Pisidia longimana (Risso, 1816), a junior synonym of P. bluteli (Risso, 1816) (Crustacea: Decapoda: Anomura: Porcellanidae) and a species distinct from P. longicornis (Linnaeus, 1767)

Pisidia longimana (Risso, 1816) and P. bluteli (Risso, 1816), both described from Nice, France, have been considered each other’s synonyms or have been validated depending on successive taxonomic opinions. The validity of both in respect to P. longicornis (Linnaeus, 1767) has also been contradicted a number of times. The current lack of clarity in the use of the names P. longicornis, P. longimana and P. bluteli has resulted in nomenclatural instability, but also in unreliability and miscommunication as regards the available ecological and distributional information. The validity of P. bluteli and P. longimana is revisited herein based on a large number of specimens (241 males, 190 females and 33 juveniles) from many different localities. The latter species is confirmed as a junior synonym of the former, whereas P. bluteli and P.  longicornis are herein considered separate species. Diagnostic characters and morphological variations are discussed and illustrated. Key-Words. Biodiversity; Eastern Atlantic; Mediterranean; Porcelain crabs.


INTRODUCTION
The worldwide genus Pisidia Leach, 1820, currently consists of 15 species distributed across the tropical and subtropical zones of the eastern Pacific, western and eastern Atlantic and Indo-West Pacific Oceans, and Mediterranean and Black Seas (Haig, 1978;Osawa & McLaughlin, 2010;Dong & Li, 2014;WoRMS, 2020).
Three species are currently considered valid from the eastern Atlantic and Mediterranean and Black Seas basins: Pisidia longicornis (Linnaeus, 1767), P. bluteli (Risso, 1816) and P. longimana (Risso, 1816) (Osawa & McLaughlin, 2010). However, P. longimana and P. bluteli have been lumped and split alternately with each successive taxonomic opinion, and the validity of both in respect to P. longicornis has also been contradicted a number of times (Zariquiey-Álvarez, 1951;Holthuis, 1961;García-Raso, 1987;d'Udekem d'Acoz, 1995, 1999Koukouras et al., 2002;Osawa & McLaughlin, 2010). Thus, the current lack of clarity in the use of the names P. longicornis, P. longimana and P. bluteli has resulted in nomenclatural instability, but also in unreliability and miscommunica-tion of the available ecological and distributional information.
An ongoing phylogenetic analysis of the genus Pisidia and the examination of 241 males, 190 females and 33 juveniles from many different localities in the collections of the National Museum of Natural History, Smithsonian Institution (USNM) and Museu de Zoologia, Universidade de São Paulo (MZUSP), prompted us to revisit the validity of P. bluteli and P. longimana. The latter species is confirmed as a junior synonym of the former, whereas P. bluteli and P. longicornis are herein considered two separate species. Diagnostic characters and morphological variations are discussed and illustrated.

MATERIAL AND METHODS
Abbreviations used include: cl, carapace length, taken from the front to the posterior median margin of the carapace; cw, carapace width, taken at the level of its widest point; P1, cheliped (pereopod 1); P2-P4, pereopods 2 to 4; St, station.
Pisidia bluteli was generally regarded as a junior synonym of P. longicornis (Linnaeus, 1776), until Zariquiey-Álvarez (1951) provided evidence that both species were morphologically distinct. While agreeing with Zariquiey-Álvarez, Holthuis (1961) argued that not only was P. bluteli valid, but so was P. longimana, and he therefore removed the latter species from the synonymy with P. longicornis.
Holthuis' (1961) view, however, was challenged by the observations of Manning & Števčić (1982), who, without further details, commented that some specimens from the Piran Gulf (northern Adriatic Sea) showed intergradations between P. bluteli and P. longimana. García-Raso (1987) went farther and moved P. bluteli and P. longimana back into the synonymy with P. longicornis. Conversely, Koukouras et al. (2002), once again considered P. bluteli and P. longimana as being distinct from each other and from P. longicornis.
Arguments in favor of splitting P. bluteli from P. longimana are essentially those of Holthuis (1961): (1) the orbital margin shows a row of spines in P. bluteli, whereas the orbital margins are usually crenulate or minutely serrate, never spinous in P. longimana; (2) there are several distinct spines on the dorsal surface of the carapace in P. bluteli, whereas the carapace spines are smaller and in P. longimana larger specimens, hardly visible; (3) the antennal basis-ischium and merus have a distinct spine at the distal end of the mesial margin in P. bluteli, whereas the antennal merus bears no spine, although a distinct spine is present in the antennal basis-ischium in P. longimana; (4) numerous spinules, arranged in more or less distinct longitudinal rows, are found on the dorsal surface of the carpus and the palm in P. bluteli, whereas the dorsal surface of the carpus and palm are smooth, although in the juveniles they may be provided with a median longitudinal row of granules or spinules, in P. longimana; (5) a row of slender spinules usually is present along the lateral margin of the carpus in P. bluteli, whereas the lateral margin of the carpus is smooth in the adults, but may be provided with spinules in the juveniles in P. longimana; and (6) numerous strong dorsal spines are present on the merus, carpus and propodus of the walking legs in P. bluteli, whereas the carpus and merus of the walking legs do not show a row of spinules, although very few short and blunt granules or spinules may sometimes be observed on the merus in P. longimana (Holthuis, 1961). Additionally, Koukouras et al. (2002) submitted that P. bluteli and P. longimana could be further differentiated in that the branchial region, behind the cervical groove, is provided with 2 or more spines (rarely 1) in P. bluteli, whereas, in contrast, the branchial region bears 0 to1 spines (rarely 2) spines in P. longimana. However, García-Raso (1987) opined that the characters used by Holthuis (1961) do not allow for distinguishing between P. longicornis, P. longimana and P. bluteli for intermediate forms in which all possible combinations of the purportedly distinguishing characters are commonly found, sometimes even in the same specimen. Consequently, García-Raso (1987) concluded that P. bluteli and P. longimana should be sunk into the synonymy with P. longicornis (see also d'Udekem d 'Acoz, 1995'Acoz, , 1999.The  Avulsos Zool., 2020;v.60: e20206036 5/6 large number of specimens examined herein from the collections of the USNM and MZUSP lends support to the view that P. bluteli and P. longimana are synonyms. The purportedly diagnostic characters for distinguishing between P. bluteli and P. longimana actually intergrade between specimens, even from the same locality. For instance, the specimen USNM 1278011 (Fig. 1F-J) presents the "bluteli type" of carapace with epibranchial spines (Fig. 1F) and the P1 carpus bears a row of spines laterally and mesially (Fig. 1G), but also presents the "longimana type" of P1 ischium with one ventrodisto-mesial spine (Fig. 1H); antenna merus without a spine mesially (Fig. 1I); and P2-P4 dorsal spines absent (Fig. 1J). Likewise, the characters proposed by Koukouras et al. (2002) clearly overlap with each other and therefore, cannot be used to distinguish among P. longicornis, P. bluteli and P. longimana. However, three diagnostic characters differentiate P. longicornis from P. bluteli: (1) P. longicornis (as already noticed by Holthuis, 1961) presents inconspicuous or absent spinulation compared to P. bluteli, whose spines in the carapace, antenna and P1 are always well-developed (Figs. 1A-O; 2A-E); (2) P. longicornis males present the major P1 broader and swollen, whereas in P. bluteli the P1 is long, slender and slightly flattened (present study); (3) the front in P. longicornis presents a deep longitudinal groove in the median lobe (so that the median lobe seems to be divided into two), whereas P. bluteli presents three conspicuous lobes, with the longitudinal one shallow and poorly visible (present study) (Figs. 1A, F, K; 2A).
Pisidia longicornis s. str. is known from the Atlantic coast of Europe, from south Norway to Portugal, as well as from the Mediterranean Sea, where it inhabits greater depths, between 30 and 100 m (d'Udekem d'Acoz, 1999). Its record from the west African coast, from Mauritania to Angola (Chace, 1956) deserves further investigation.