Dioxys atlanticus Saunders, 1904

2. Dioxys atlanticus Saunders, 1904 View in CoL

Figs 5 A – F View Figure 5 , 6 A – F View Figure 6

Dioxys atlantica Saunders, 1904: 232, ♀ ♂ [ Spain: Tenerife, NHMUK, examined]. View in CoL

Material examined.

Spain • 1 ♂, 1 ♀; Santa Cruz [Tenerife]; 4 Apr. 1904; A. E. Eaton leg.; NHMUK (syntypes) (Figs 5 A – F View Figure 5 , 6 A – F View Figure 6 ) • 1 ♀; Gran Canaria, Santa Lucia [Santa Lucía de Tirajana]; 800 m a. s. l.; 15 Jan. 2001; H. & I. v. Oorschot leg.; RMNH; ZMA.INS.5142848 • 1 ♀; Gran Canaria, Bco. Tasartea [ Barranco de Tasarte ]; 21 Mar. 1987; F. de la Roche leg.; B. Tkalců det.; OÖLM .

Remarks.

Discussion of the species and its distribution is required. In addition to Tenerife and Gran Canaria, Bogusch (2023) mentioned also Lanzarote citing Hohmann et al. (1993), Egypt citing Warncke (1977), and newly recorded the species from Sardinia based on novel records. Each of these points requires dissection.

The record (s) from Lanzarote are difficult to interpret because of the subsequent description of D. lanzarotensis (see Section 8). It is possible that specimens of the then unknown D. lanzarotensis were unwittingly determined as D. atlanticus , since the two are similar in size and colouration; the type specimen of D. lanzarotensis was captured on 3 March 1987 ( Tkalců 2001: 49), it may have formed the basis of the report of Hohmann et al. (1993). The records from southern Egypt (Luxor, Abydos Baliana), as written, do not make ecological or morphological sense. Warncke (1977: 278) says that the morphological differences between D. atlanticus and North African D. cinctus ( Jurine, 1807) are extremely slight, citing slightly finer punctation of T 3 – T 6, that the female S 6 is more rounded, and the male has similarly long pointed teeth medially on the margin of S 4. Examination of the two syntypes of D. atlanticus raise the question as to whether Warncke ever examined them, because the shape of T 6 in D. atlanticus is clearly substantially more rounded than in D. cinctus (compare Fig. 5 F View Figure 5 with Fig. 1 A View Figure 1 ), as is the margin of S 6, whereas North African (and indeed, all) D. cinctus have S 6 rectangular in outline. The male of D. atlanticus does indeed have long pointed teeth medially on S 4 (Fig. 6 F View Figure 6 ), but this places it much closer to D. rotundatus Pérez, 1884 (Fig. 4 F View Figure 4 ; this name is re-established as the senior synonym of D. moestus Costa, 1884 , see Section 11), as D. cinctus males only have slight bumps on the margin of S 4 (Fig. 4 D View Figure 4 ). I concur with Bogusch (2023) in rejecting the combination of D. cinctus ssp. atlanticus proposed by Warncke (1977). The morphology of these specimens is discussed below.

Finally, the new records of D. atlanticus from Sardinia are doubly questionable. The first is because there is not a single insect species which has a distribution of the Canary Islands and Sardinia, without records from North Africa, and more pertinently because of the original description of D. moestus which has a locus typicus of Sardinia ( Costa 1884). The description is short, in Latin and Italian, and it is worth reproducing in full here:

“ Dioxys moesta . – D. nigra unicolor, cinerea pubescens, abdominis segmentis primis quatuor vel quinque postice fasciola e pilis stratis albis cinctis. – Long. mill. 4. ”

Molto diversa dalla pyrenaica non solo per avere egualmente neri tutti gli anelli addominali, ma anche per le dimensioni minori ed il sesto anello addominale più semicircolare ”.

[ Dioxys entirely black, pubescence grey, the first four to five abdominal segments posteriorly covered with layers of white hair.

Very different from pyrenaica not only for having all of the abdominal segments entirely black, but also for its smaller size and the sixth abdominal segment more semi-circular.].

Costa clearly describes an all-black species that is smaller than D. pyrenaicus Lepeletier, 1841 [= D. cinctus ] and has the final segment of the abdomen [= T 6] more semi-circular. This description can only correspond to D. rotundatus , and corresponds to the black specimens of “ D. atlanticus ” found on Sardinia. This strongly suggests that the all-black Sardinian specimens simply are D. rotundatus as described by Costa (1884) as D. moestus . Moreover, specimens from Sardinia are not always entirely black. Warncke (1977: 275) writes that:

“ ♀ San Lussurgio / Sardinien, Lectotypus Mus. Napoli. Die Beschreibung kennzeichnet eine vollkommen schwarz gefürbte kleine Dioxys - Art. Das stimmt nicht, das 2. Tergit ist vollständig rot und die Seiten des 1. und 3. Tergits sind trübe rot!, womit das Tier auch in den übrigen Merkmalen mit Dioxys rotundata übereinstimmt! ”

[The description indicates a small and completely black Dioxys species. This is not true, as the second tergum is entirely red and the sides of the first and third terga are dull red!, which means that the animal agrees with Dioxys rotundata in other characteristics].

This raises the question as to whether or not Warncke’s “ lectotype ” is actually valid since it does not match the description (cf. Schwarz et al. 1996), but this is of secondary importance and the overall comment combined with examination of new specimens from Sardinia (kindly loaned by Petr Bogusch, see material examined in Section 11. Dioxys rotundatus ) indicates considerable colour variation. One female from Sardinia has T 2 dull red, almost black, and a second female has T 2 entirely black. The question is therefore, is D. atlanticus morphologically distinct from a widespread D. rotundatus which can be entirely black or have a metasoma marked with red, with variation between these two states?

I agree with Bogusch (2023) that D. atlanticus and D. rotundatus (referred to as D. moestus ) are very morphologically similar; this is most apparent in the males which both possess S 4 with a pair of long projecting teeth medially (Figs 4 F View Figure 4 , 6 F View Figure 6 ), together these emphasising the median emargination. Based on the small number of specimens that I have been able to examine, the only clear structural character that I can see to allow separation of these species is the sculpture of the apical tergal segments, as well as the width: length ratio of the disc of T 6. In female D. atlanticus , the punctures of T 6 are very shallow and placed closely together, making it difficult to distinguish between punctures and shiny interspaces (Fig. 5 F View Figure 5 ); in the male, the punctures of T 5 and T 6 almost disappear due to their shallowness (Fig. 6 E View Figure 6 ). In contrast, in female D. rotundatus the apical tergal segments have the punctures deeper and slightly more spaced, therefore clearly and unambiguously contrasting the shiny interspaces (Fig. 1 B View Figure 1 ); in the male, the punctures of T 6 remain clear and distinct.

The two Egyptian specimens examined by Warncke were also sent to me on loan (see material examined in Section 11. Dioxys rotundatus ), representing a female and a male. The female has T 2 entirely red-marked, and T 6 is rounded with the punctures clear and well-defined, with shiny interspaces. In this regard it clearly matches D. rotundatus . The male has S 4 with relatively short teeth flanking the medial emargination, making it somewhat intermediate between the condition found in D. cinctus and D. rotundatus / atlanticus , but the punctation of T 5 and T 6 is strong and well-defined, not disappearing into the integument. Measurement of the width: length ratio of the disc of the female T 6 (from maximum visible width and maximum visible length of the disc in dorsal view) produces values of 1.82–1.90: 1 for D. rotundatus (Figs 1 B View Figure 1 , 7 D, H View Figure 7 ) and 2.12: 1 for the syntype of D. atlanticus (Fig. 5 F View Figure 5 ). Even though the punctation of T 6 of one of the Sardinian specimens is more chaotic and begins to resemble that of D. atlanticus (Fig. 7 H View Figure 7 ), the overall width: length ratio of T 6 is 1.90, and less than that observed in D. atlanticus . This means that neither the female Sardinian specimens or the female or male Egyptian specimens are referable to D. atlanticus in the narrow sense used here.

With access to suitable material, these characters (T 6 punctation and width: length ratio) can be used consistently. Therefore, I take the position that D. atlanticus is restricted to Tenerife and Gran Canaria, as I have not seen any specimens with equally weak punctation on the apical tergal segments elsewhere in the western Mediterranean or North Africa, the distributional range of D. rotundatus (see Section 11).

Distribution.

Spain (Tenerife, Gran Canaria) ( Saunders 1904).