On the true identity of Scolopendra aztecorum Verhoeff , 1934 ( Chilopoda : Scolopendromorpha : Scolopendridae )

The genus Scolopendra Linnaeus, 1758 currently comprises 13 valid species in Mexico. One of the species, Scolopendra aztecorum Verhoeff, 1934 has been considered a potential synonym of S. viridis Say, 1821, S. polymorpha Wood, 1861, or Arthrorhabdus pygmaeus (Pocock, 1895). To shed light on its true status, five of the six syntype specimens were examined, redescribed and illustrated for the first time. Scolopendra aztecorum is diagnosable from all other Mexican species and remains in the genus Scolopendra as a valid species.


Introduction
published a comprehensive catalog of the chilopods from the Neotropical region.In his work, he regarded as doubtful the new forms of the viridis group from Mexico described by Verhoeff (1934), with the exceptions of Scolopendra octodentata and "perhaps" S. aztecorum.Bücherl (1942) was the first author to question the validity of Scolopendra aztecorum, noticing its closeness to Scolopendra polymorpha Wood, 1861 (as Scolopendra viridis polymorpha sensu Attems (1930)) and the lack of a comparison in the original treatment by Verhoeff (1934).
In consequence, Bücherl (1942) tried to provide a basic comparison between S. aztecorum and S. viridis, specifically with S. v. polymorpha sensu Attems (1930).While doing so, he failed to find the most important character separating these two species: the number of tarsal spurs on the locomotory legs.Bücherl (1942) correctly mentioned S. aztecorum as having two tarsal spurs on legs 1-20 but stated that Attems (1930) did not establish the spur number in S. v. polymorpha.He overlooked that Attems (1930) described Scolopendra viridis Say, 1821 (including S. v. polymorpha) as having two tarsal spurs on leg pair 1 and one tarsal spur on leg pairs 2 to 19 or 20.Bücherl (1974) did not mention S. aztecorum in his second catalog of Neotropical Scolopendromorpha.Shelley (2006) justly considered clarifying Neotropical species of Scolopendra as a daunting challenge.For the benefit of all subsequent workers, he took the initial step of consolidating the speciesgroup names and literature on New World Scolopendra species.This author regarded 13 Mexican species as valid, a figure followed by Cupul-Magaña (2013).However, the Mexican endemic Scolopendra aztecorum Verhoeff, 1934 was considered by Shelley (2006) a potential synonym of S. viridis Say, 1821, S. polymorpha Wood, 1861, or Arthrorhabdus pygmaeus (Pocock, 1895).
To solve this taxonomical problem, five of the six syntypes of Scolopendra aztecorum were located, examined and confirmed as conspecific.The relevant taxonomic literature is consolidated and comments on nomenclatural and curatorial issues are presented.

Material and Methods
Anatomical terminology follows the recommendations of Lewis et al. (2005) and Bonato et al. (2010).The term "accessory spine/spines" recommended by Bonato et al. (2010) is substituted here by one of its synonyms, "accessory spur/spurs" (Crabill, 1958;Lewis et al., 2005).This increases the consistency of the terminology because the "accessory claws" are "actually spurs ankylosed to base of pretarsus", as noted by Crabill (1960).The change will allow using without confusion expressions such as "legs 1-20 without spines".As the terms for the homologous structures on the second maxillae are not listed by Bonato et al. (2010) we use the ones recommended here.
Standardized label information is proposed for syntypes, following recommendations of Wheeler et al. (2001) and ICZN (1999).Georeference for La Paz was directly taken from Google Maps.Georeference for Los Inocentes is not accurate in Google Maps but it was step by step centered in the only country house that remains, where old roads meet, and then retrieved.Free map data at 1:110m scale were obtained from Natural Earth (http://www.naturalearthdata.com/) and processed using TileMill v.0.10.1.The specimen deposited in Prague was photographed using an Olympus SZX12 stereomicroscope equipped with an Olympus E-510 camera.Type material.Verhoeff (1934) described S. aztecorum based on six syntypes.
Type locality.Mexico, Baja California Sur, near La Paz and near Los Inocentes (Fig. 1).See remarks.Distribution.Known only from type localities.Endemic to Mexico.
Original rank.Species.20) against two spurs only present on tarsi 1 of legs 1.

Redescription
Color: According to Verhoeff (1934), body and legs pale yellow, antennae, head and T1 dark bluish, sometimes also tergites 2 and 3, forcipular coxosternite and forcipules reddish-brown, tarsungula black.Only one adult with head and T1 more yellowish-brown.At present, former bluish color on the head and T1 is faded.
Antennae: 27-33 antennal articles, 15-17 basal sparsely hirsute.The transition between sparsely hirsute and hirsute antennal articles is gradual and difficult to see.This may be due to the setae being very short and pale.Setae not arranged in longitudinal rows.ZSM-A-20051044 with antennae broken, only 15 left and 26 right articles remaining.Bücherl (1942) incorrectly transcribed the count from Verhoeff (1934) as only "27-32 artículos antenais".
Mandibles and maxillary complex: Specimen ZSM-A-20051044 with right mandible, 1st maxillae and 2nd maxillae dissected.Right mandible with five tricuspid teeth, condyle of lamina condilifera well developed.2nd maxillae articles not ridged or pointed ventrally, article 2 with a lateral spur.According to Verhoeff (1934), the spur ("Dorn") may be present or absent, but non dissected specimens were not checked for this character to avoid manipulation damage.Article 3 with conspicuous dorsal brush.Claw with 2 accessory spurs.
Forcipular segment: Coxosternite with a short anteromedian longitudinal suture (aprox.1/8 to less than 1/3 of the coxosternite's length) (in specimen ZSM-A-20051045 double and reticulate in the middle), without transversal suture.Tooth-plates wider than long, slightly separated.Typical single short strong seta in the middle of tooth-plates, not seen by Verhoeff (1934): "Die Zahnplatten innen mit Grube, aber ohne borstentragenden tuberkel".Tooth-plates with 1+1+2 teeth, the lateral tooth more isolated and sharp, the median ones more blunt.Process of trochanteroprefemur large, generally with 2 tubercles on its internal margin (in specimen ZSM-A-20051045 left process without noticeable tubercles, right with two insinuated bumps, in specimen ZSM-A-20051044 tubercles well developed).Trochanteroprefemoral process surpassing the anterior border of the tooth-plates.Interior surface of tarsungula with a single sharp longitudinal ridge, smooth, not serrate.
In tergites anterior to T7 there is not a definite oblique suture but a reticulum.Margination starting from T7-T11 to T21 (in ZMB 13378 seen only from T16 to T21), in anterior tergites incomplete and less marked, generally marked and complete from T8-T14 onwards.T21 as wide as long or slightly wider than long, posterior margin rounded.
Legs: Tarsi 1 longer than tarsi 2. Legs 1-20 without spines.Leg pair 1 with 1 prefemoral, 1 femoral, 1 tibial, 2 tarsal (on tarsi 1) and two pretarsal accessory spurs.Legs 2-20 without tibial spurs, legs 2-19(20) with two tarsal spurs on tarsi 1, sometimes only one tarsal spur on leg pair 20, legs 2-20 with pretarsal accessory spurs.Coxopleura: Coxopleural surface without setae.Inner half of the coxopleuron not notably warped, in contrast to species like S. angulata Newport, 1844 and S. gigantea Linnaeus, 1758.Extensive pore-field, covering most of the ventral surface of the coxopleuron, including the posterior margin, the inner margin and part of the coxopleural process, but leaving a short and narrow poreless stripe anterior to the coxopleural process.Coxal pores numerous, small, grouped (generally separated by not more than 1-2 pore diameters).
Posterior margin of coxopleuron slightly oblique, with one side spine.Coxopleural process short, conical, with 4-7 spines, from which two are apical and the rest subdistal in ventral, lateral or dorsal position, the generally 4 (3-5) biggest spines organized in a row.Verhoeff (1934) considered the coxopleural process as strongly developed and slender, moderately long, with 3-5 outer spines.(2015), where the authors wrote "Scolopendra pomacea differs from the other described species of the genus Scolopendra in Mexico in the following combination of characteristics (Attems 1930): …tibia of legs 1-19 with one spur and tarsus 1 with two spurs; legs 20 with tibia usually without spur...".This is clearly an erroneous translation of Attems (1930) description, the correct translation being: "…tarsi 1 of legs 1-19 with one spur and pretarsi with two accessory spurs; legs 20 with tarsus 1 usually without spur...".

Remarks
Regarding an avoided potential homonymy.Verhoeff (1934) stated in a footnote "aztekorum = azteka Verh.i. litt.(non azteka Sauss.!)".This means that he first used the name S. azteka for the species in a letter or other documented correspondence, probably with a date.The existence of this letter is unknown, and the document is not published in the sense of the ICZN.Verhoeff (1934) noted that this name will fall in homonymy with S. azteka (sic!) Saussure, so he changed it to S. aztekorum.
Regarding species epithets.Verhoeff (1934) used "k" instead of "c" in the first mention of the epithet "aztekorum n. sp." in his key for Scolopendra species from the viridis group.He immediately remitted to the aforementioned footnote, and there also used "k" for the epithets aztekorum (it was the second time) and azteka (twice, Verh.and Sauss.).However, he used the binomen Scolopendra aztecorum n. sp. in the species description, immediately after the key and on the same page as the footnote.Verhoeff (1934) utilized the epithet aztecorum two more times, in the description of Scolopendra viridis storkáni [recte storkani (Shelley, 2006)].All subsequent authors except Moritz & Fischer (1979) have used the spelling Scolopendra aztecorum.The most frequently cited Scolopendra aztecorum Verhoeff, 1934 is here selected as the correct original spelling.The Scolopendra aztekorum spelling is incorrect (Art 24.2.3,ICZN) and unavailable (Art.32.4,ICZN).We are the First Revisers (Art.24.2.3,ICZN).
Regarding the designation of a lectotype.Lectotype designations made after 1999 should have the purpose of clarifying the application of the name to a taxon (ICZN, 1999, p. 11).This is not the case of Scolopendra aztecorum, so it is not necessary.On the other hand, the type series is shared among at least four institutions in three countries.To declare a lectotype will deprive all institutions but one from the possession of name bearing specimens and it will also affect published references on syntype holding.The loss or destruction of the lectotype will lead to a new nomenclatural act designating a neotype while the loss of one or few syntypes from this series will require no action.Keeping the syntypes avoids two potential nomenclatural acts and maintains stability.
Regarding the missing syntype.The sixth syntype was not found.It is not at any of the BMNH (Greg Edgecombe, pers. comm., 18.xi.2014),NHMW (Nesrine Akkari, pers. comm., 11.xii.2014),ZMUC (Henrik Enghoff, pers. comm., 20.x.2015), MHNG (John Hollier, pers. comm., 1.xi.2015) or NHRS (Karin S. Kronestedt, pers. comm., 21.i.2016).It is listed neither by Weidner (1960) nor by Rack (1974) as being at ZMUH, nor by Stoev & Beron (2000) as being at NMNHS.It is not found in the dataset of the Myriapoda and Onychophora collection of the MNHP cited by Le Bras et al. (2015).Verhoeff's type series usually consist of wet material and micropreparations of dissected parts.There might be a chance that the sixth syntype is dissected and deposited as slides in one of the European collections that stores part of Verhoeff's collection.
Regarding colleting dates.Dolejš (2015) was the first author to provide a collecting date for the specimens, the year 1930.
Regarding a standardized label.As a curatorial recommendation, an additional standardized label is proposed for all specimens, which should be labeled as syntypes following recommendation 72D (ICZN, 1999).Scolopendra aztecorum Verhoeff, 1934 Regarding specimens' size: Verhoeff (1934) twice mentioned that he examined 4 adults and 2 juveniles.The adults were described as 118-125 mm in length, with 30-33 antennal articles and margination incomplete on T7 but marked on TT8-21.The juveniles were described as 70-77 mm in length, with 27-28 antennal articles and margination marked on TT14-21.However, it seems that Verhoeff (1934) mistook the numbers.Our data (table I) shows that Verhoeff (1934) categories are not absolute (juvenile RMNH.CHIL.154has an antennal article count corresponding to the adult category) and that 2 syntypes are adults and 3 are juveniles.This makes us think that there were originally 2 adults and 4 juveniles, but without the sixth syntype this cannot be assured.
Table I.Comparison between Verhoeff's (1934) data for adults and juveniles of Scolopendra aztecorum and current character status of 5 examined syntypes.

Conclusion
This work summarizes several changes and updates with respect to the S. aztecorum synthesis provided by Shelley (2006): type specimen status, number of syntypes, confirmed repositories, type locality/distribution, anatomical illustrations and map.The taxonomical hypotheses by Bücherl (1942) and Shelley (2006) are corrected.S. aztecorum remains a valid species and the correct name spelling is properly fixed.