Three new species and one new subspecies of Depressariinae (Lepidoptera) from Europe

Abstract The species Depressaria albarracinella Corley, sp. n., Agonopterix carduncelli Corley, sp. n. and Agonopterix pseudoferulae Buchner & Junnilainen, sp. n. and the subspecies Depressaria saharae Gastón & Vives ssp. tabelli Buchner, ssp. n. are described. Depressaria albarracinella was first found in Spain in 1969 and recognised as apparently new but the specimens in NHMUK have remained undescribed. Additional Spanish material has been located in ZMUC and other collections and three specimens have been found from Greece. Agonopterix carduncelli. A single male of an unidentified Agonopterix of the pallorella group was found in Algarve, Portugal in 2010. A search for larvae in March 2011 was successful and one male and one female were reared from Carthamus caeruleus. Additional specimens of the new species have been located in collections from Spain, Greece and Morocco. Agonopterix pseudoferulae. A specimen from Greece with the name Agonopterix ferulae (Zeller, 1847) found in the Klimesch collection in ZSM had forewing markings which suggested that it might be a different species. Further specimens from Italy and Greece have been examined, among them two reared from Elaeoselinum asclepium (Apiaceae). Both genitalia and barcode show that this is an undescribed species. Depressaria saharae Gastón & Vives, 2017 was described very recently (Gastón and Vives 2017) from northern Spain with a brief description, and figures of two males and male genitalia. Here the new species is redescribed, and additional data on distribution and relationships of the new species added. The opportunity is also taken to show that Canary Islands specimens with the same male genitalia should be treated as a new subspecies D. saharae ssp. tabelli Buchner, ssp. n.


Introduction
Preparatory work for a proposed volume on Depressariinae in the series Microlepidoptera of Europe has revealed a number of taxonomically challenging species groups in the genera Agonopterix Hübner [1825] and Depressaria Haworth, 1811. This was not unexpected but there are more such groups than we had initially expected. However, in addition to the problem groups, some undescribed species have also been discovered which can be described without the necessity to resolve complex taxonomic issues. Two such species are described here in Agonopterix and one species and one subspecies in Depressaria.

Material and methods
Material has been examined from NHMUK, NHMV, TLMF, ZMUC, ZSM and additionally specimens from many private collectors including those of the authors have been checked, here only listed if the material was of particular importance for this paper: Michael Dale (England), Gabriele Fiumi (Italy), Knud Larsen (Denmark), Toni Mayr (Austria), Willibald Schmitz (Germany), Peter Sonderegger (Switzerland), Lubomír Srnka (Slovakia), Jan Šumpich (Czech Republic) and Joachim Viehmann (Germany). Apart from one exception given in the description, each species includes both reared and light-trapped specimens.
Morphological examination and photographic documentation. Genitalia preparations followed standard techniques (Robinson 1976). Male preparations were stained with mercurochrome and females with chlorazol. The placement of holotypes is given under each species. Photographic documentation: Apart from two exceptions given in the descriptions, photos of specimens in total view were taken with Canon EOS 5D Mark III and Canon lens EF 100mm 2.8 L IS USM at 1:1., specimens were illuminated with two diffused flashes, using a third flash for setting the background whiteness. Detailed photos of specimens were taken with a Canon lens MP-E 65 at 2:1, using ring flash. Genitalia photos were taken with microscope (Wild Heerbrugg) using a 10x objective and a 2.5x ocular. Photos were edited using the software Helicon Focus 4.80 and Adobe Photoshop 6.0. For creating the black and white photos, based on the used stain, the G alpha channel of the RGB originals was used in males and the Y alpha channel of the CMYK originals in females. Genitalia examination and photos by P. Buchner, if not specified.
The key below includes only the European species. D. pentheri Rebel, 1904 is omitted due to insufficient knowledge of this taxon. The North African D. deverrella Chrétien, 1915, has sometimes been listed as present in France, but we can find no evidence for this. Description. Adult (Figs 2-4). Wingspan 23-26 mm. Head cinnamon-brown on neck and crown; face light brownish buff. Labial palp with segment 3 two-thirds length of segment 2, segment 2 buff with tufted scales on ventral side cinnamon; segment 3 cinnamon with dark grey ring beyond middle, tip cinnamon-buff. Antenna light grey-brown, narrowly ringed dark brown. Thorax light brownish buff, rarely darker. Forewing light brown, often with slight cinnamon tinge, often very weakly marked but sometimes with more or less faint grey-brown interrupted streaks in cell, in fold, beyond cell, between veins to costa and between veins to termen; occasionally a faint brown spot is present at base of dorsum; equally indistinct grey-brown spots between vein-ends at termen; cilia light brown, without obvious cilia line. Hindwing light grey, slightly darker posteriorly, with narrow grey-brown line around terminal and dorsal margins; cilia light grey-brown at apex to almost white at dorsal base, with a fine darker cilia line. Abdomen light grey-brown.

Key to males of European species of
Variation: The forewing markings vary from almost completely obsolete to present but faint compared with most other Depressaria species. The specimen from Huesca, Spain (Fig. 4) is the most strongly marked that we have seen. Sometimes a faint Vshaped pale fascia is visible beyond end of cell.
Neighbour-joining analysis (Fig. 7) shows Depressaria eryngiella as the nearest neighbour with 2.45% p-distance. Intraspecific variability, based on present knowledge, 0% within the Spanish population and 1.08% between Spanish and Greek populations.
For Maximum Likelihood analysis, see Fig. 70. Etymology. The species name is an adjective derived from Albarracin in Spain, an area where two of the paratypes were taken.
Distribution. Spain: Mountain areas of Eastern Spain from Sierra Nevada and Sierra de Los Filabres northwards, in the provinces of Granada, Almería, Castellón, Teruel, Zaragosa and Huesca. Greece: Parnassos Mountains in Central Greece, Lesbos.
Bionomics. Larva and food-plant unknown, but the latter is likely to belong to Apiaceae. Adult moths have been taken in May, June, July and October. It is probable that overwintering takes place in the adult stage, but less clear when the larvae would be feeding.
Remarks. The genus Depressaria Haworth, 1811 includes around 125 species (Wikipedia 2016) with the greatest number in the Palaearctic region. The majority of the species are rather similar externally, but for the most part, the male genitalia give clear differences between species. Indeed there is such diversity in genital morphology within the genus that it is difficult to characterise the genus using genitalia characters. Within this great diversity there are some clearly defined groups, with a number of species sharing a suite of genitalia characters. One such group is the veneficella group (Hannemann, 1953), currently with 13 species described from the Palaearctic region extending from western Europe and North Africa through the Middle East to central Asia, with the most eastern records from north-east China, Mongolia and the Altai region of Siberia. Lvovsky (1996) when describing D. erzurumella Lvovsky, 1996 from Turkey, provided a key based on male genitalia to 11 species, omitting D. pentheri Rebel, 1904, which was known only from the female. Subsequently he described D. kailai Lvovsky, 2009(Lvovsky 2009). The new species, D. albarracinella, belongs to this group.
The veneficella group is characterised by rather long wings, forewings brown with pattern usually consisting of blackish streaks between the veins, but the pattern very reduced in some species. Male genitalia have elongate gnathos (nearly globose only in altaica Zeller, 1854 and kailai), valvae incurved at apex, costal margin sometimes with median bulge, sacculus widely crossing the valva with two (rarely three) processes on the posterior edge, the outer reaching close to the costal margin of valva or exceeding it, saccus often elongate, aedeagus slender, long with a single cornutus. Female genitalia with long ductus bursae. Species identification most often rests on the male genitalia, where the shape of the incurved apex of the valva, the length of the saccus and the relative proportions of the various parts provide diagnostic characters, in particular the length of the cornutus relative to the aedeagus and the length of the aedeagus relative to the length of the valva. Those species with known food-plants all feed on Apiaceae.
The presence of an undescribed species of this group in Spain was recognised by Klaus Sattler after he and David Carter collected several specimens in Sierra Nevada in 1969. These have remained unnamed in NHMUK since that date. Further specimens were later collected in the same area and elsewhere in Spain, most of these deposited in ZMUC. It was these that first came to the notice of M. Corley in 2004. Recently the species has been found in additional localities in Spain and in Greece. It is described here as D. albarracinella Corley sp. n.
The specimens from Greece have not been included in the type series. Although there is no reason to doubt the identification, the p-distance of over 1% between the barcodes of Spanish and Greek specimens suggests that caution is not out of place. Diagnosis. The characteristic shape of segment 2 of the labial palp and the absence of a posterior crest on the thorax are features shared with a few other species mostly with similar coloration. A. straminella (Staudinger, 1870) is most similar with black dot at base of dorsum and black terminal dots together with paler hindwing, but lacks cell dots. Forms of A. carduncelli sp. n. without evident cell dots require genitalia examination to distinguish them from A. straminella. Other related species have better developed cell dots. In the male genitalia, A. carduncelli sp. n. is recognisable by the longer curved cuiller and broader valva in comparison with related species. The female is unique among European Agonopterix in the absence of a signum.

Agonopterix carduncelli
Description. Adult (Figs 8-9). Wingspan 19.5-21 mm. Head dull ochreous-buff, face creamy buff. Labial palp segment 2 with only the distal half rough-scaled and furrowed, pale buff with scattered light brown scales, segment 3 pale buff or ochreousbuff. Antenna with scape dull ochreous-brown, proximal part of flagellum ochreousbuff, ringed grey-brown, distally grey-brown. Thorax dull ochreous-buff often with darker median line, without posterior crest. Forewing pale ochreous-buff with faint pinkish tinge when fresh, with a variable amount of scattered light brown and blackish scales, particularly along veins towards termen and sometimes also in cell and between dorsum and fold; a black or brown dot at base of dorsum, a small dot in cell at twofifths and usually another at end of cell; terminal spots dark grey-brown; a faint greybrown stripe stretching through subdorsal area ending in a wider patch below end of cell; cilia pale ochreous-buff with weak cilia line. Hindwing light grey, darker in outer half; cilia light greyish ochreous with indistinct cilia lines. Legs pale ochreous-buff, foreleg blackish on upper side of tibia and part of tarsus. Abdomen light greyish buff.
Variation: Some specimens have many more scattered dark scales than others. The subdorsal spot can be distinct or dull pale brown; the cell dots may be obsolete, or if developed may still be indistinct due to the abundance of scattered scales; the development of the subdorsal streak is variable.
Male genitalia. (Fig. 12). Similar to related species, but gnathos almost exceeding socii by its own length, valva broader and smooth-sided, slightly curved, tapered to rounded apex; cuiller curving outwards at middle, parallel-sided, round and slightly wrinkled at apex, crossing four-fifths width of valva, longer than in related species due to broader valva. Fig. 13 shows male genitalia of the other six pallorella group species for comparison. Female genitalia. (Fig. 14). Anterior margin of sternite VIII nearly straight, not bulging, ostium just beyond middle of plate; ductus bursae smooth, gradually expanding to corpus bursae; signum absent.
Differences in DNA barcodes arise over time through chance mutations. Such stochastic events sometimes lead to fairly unrelated species appearing as nearest neighbours. This is evidently the case with A. carduncelli sp. n. and A. multiplicella. The latter species has none of the characters of the pallorella group.
Etymology. The species name, a noun in genitive case, is derived from the larval food-plant Carthamus [=Carduncellus] caeruleus (Asteraceae).
Distribution. Currently known only from Portugal, Spain, Greece and Morocco, but potentially more widespread around the Mediterranean with its food-plant.
Bionomics  male genitalia are rather similar throughout the genus. Easily defined groups within the genus are less obvious than in Depressaria, but there are some such groups. One of these is the pallorella group which includes several closely related species all feeding on Asteraceae tribe Cynareae and sharing similar pale ochreous coloration, forewings without defined basal patch, oblique pair of dots reduced to one or absent, labial palp characteristic with appressed scales on underside of segment 2 in proximal half, distal half with forward projecting scales making a triangular tuft, thorax without posterior crest. The species of this group are A. pallorella (Zeller, 1839), A. kaekeritziana (Linnaeus, 1767), A. bipunctosa (Curtis, 1850), A. broennoeensis Strand, 1920, A. straminella (Staudinger, 1870 and A. squamosa (Mann, 1864). A few other described taxa are synonyms of the above mentioned species. However one species has been previously overlooked and is described here as A. carduncelli sp. n.  (Busck, 1902) (BIOUG07080-A07, BOLD:ACF9151). The evolutionary history was inferred by using the Maximum Likelihood method based on the Tamura-Nei model. The tree with the highest log likelihood (-4818.52) is shown. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The proportion of sites where at least 1 unambiguous base is present in at least 1 sequence for each descendent clade is shown next to each internal node in the tree. The analysis involved 54 nucleotide sequences from selected species of Agonopterix and Depressaria. Only the Agonopterix part of the tree is shown here. Codon positions included were 1st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 657 positions in the final dataset. Evolutionary analyses were conducted in MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets (Kumar, Stecher and Tamura 2015). The result is shown in radiation graphic, because in this view the evolutionary aspect is visualized better than in rectangular tree. The existence of an Agonopterix feeding on Carthamnus in Algarve, Portugal was suspected from the late 1990s when empty spinnings were found by M. Corley on the plant in late April. After Michael Dale found an adult of an undescribed species in 2010, a visit to Algarve in March 2011 by M. Corley targeting larvae on this plant was successful, resulting in two reared adults (see paratypes).
Two reared specimens (which survived the deterioration of their larval food-plant after M. Corley returned to England), show a grey-brown tinge in place of scattered dark scales and lack the row of terminal dots, but these features are shared by some of the Moroccan specimens.
Corley (2002) mentions a specimen without signum from Setúbal, Portugal in MNHN which was considered to be a possible aberration of A. mendesi Corley, 2002. As the signum is not known to be absent in any other Agonopterix species, it is extremely probable that this specimen belongs to A. carduncelli sp. n.  (Figs 28-29) by Klimesch. At first glance, they do look similar, but a closer look shows two constant differences: the brick-red line between the proximal pair of dots and the distal dot and the diffuse dark spot which touches this line on costal side in A. pseudoferulae sp. n., which are both absent in A. ferulae (if diffuse dark spots are present, they are found in other areas). A. atomella (Denis & Schiffermüller, 1775) (Fig. 33) and A. scopariella (Heinemann, 1870) (Fig. 32), the two species closest to A. pseudoferulae sp. n. in DNA barcode, do not have the reddish elements in central forewing pattern. A. oinochroa (Turati, 1879) (Fig. 31) has reddish elements here, but they surround the dots and do not form a line. A forewing pattern similar to A. pseudoferulae sp. n. is found  (Fig. 30), but here differences are found in outline of forewing and shape of interneural dots at outer margin: apex rounded, outer margin convex, interneural dots round and diffuse in A. pseudoferulae sp. n., apex pointed, outer margin straight to concave, interneural dots narrow lines in A. cluniana.

Agonopterix pseudoferulae
Description. Adult: Wingspan 19-21 mm. Scales of head brown, tips markedly paler. Labial palp segment 2 inner side pale, outer and ventral sides medium greyish brown or rusty brown scales mixed with blackish scales; third segment bicoloured, blackish at base, shortly above middle and at extreme tip, pale between the dark areas. Antenna dark brown. Thorax with posterior crest, rather dark brown, tegulae similar. Forewing predominantly dark reddish brown, whitish and black scales interspersed in low (but variable) numbers, basal field markedly paler, gradually passing into a pale stripe which runs along costa especially in proximal half and is interrupted by irregular dark patches. The centre of the forewing has the typical basic pattern of Agonopterix (two oblique dots at about one-third, one or two dots along veins at about one-half and a diffuse black spot between the two pairs of dots but closer to the costa) but with very distinct details: the two black, oblique dots partly bordered with reddish (brick-red to ochreous) scales which may connect the two dots on their proximal margin, distal margin pale to white, the third dot at about one-half with clear white centre and sur- rounded by a few dark scales, a brick-red to ochreous line connects the oblique dots with the distal dot and exeeds it a little; the diffuse blackish spot touches the brick-red line between the dots on the costal side. Cilia concolorous with wings. Under side of forewing dark grey except costa which is predominantly yellowish with interspersed groups of dark scales. Hindwing rather dark greyish brown, moderately translucent at base, cilia concolorous with wings, base and tips darker than in between. Legs covered with a mix of dark grey and pale scales, tibia yellowish to rusty brown on outer side, especially on fore-and hindlegs. Abdomen greyish, with broad dark line laterally and two rows of indistinct dark spots on ventral side.
No gender-specific differences could be found. Variation: Little variation was found within the nine examined specimens. The number of interspersed white scales on forewing varies to some extent, and between the proximal pair of dots and the distal dot, an additional white dot may be developed or not. In one specimen the thorax (but not tegulae) is entirely black.
Male genitalia (Fig. 34): There is no single feature which separates male genitalia of A. pseudoferulae sp. n. from all other species of Agonopterix, therefore it is best to compare genitalia with each of the externally similar species individually.
A. ferulae (Fig. 35) belongs to the alpigena/selini species group, which is characterised by a two-horned process of the anellus toward transtilla (see arrow in insert of Fig. 35) in combination with transtilla significantly widened in the middle. In A. pseudoferulae sp. n. these features are not present. In A. atomella (Fig. 36), A. oinochroa (Fig. 37) and A. scopariella (Fig. 38) anellus lobes are large, nearly touching in A. atomella and A. oinochroa, overlapping in A. scopariella in standard preparation. In A. pseudoferulae sp. n. anellus lobes are narrow with a wide gap in between. In A. cluniana (Fig. 39) cuiller is rather short (about 70% of valva-width) and socii markedly narrow, gnathos far exceeding end of socii, in A. pseudoferulae sp. n. cuiller nearly reaching costa of valva (at least 90% of valva-width), socii and gnathos of average shape, compare Figs 34 and 39 (appearance of socii and gnathos may be influenced by preparation artifacts, so it is not helpful to point out a numerical ratio).  (Figs 40, 41a). Anterior margin of sternite VIII with a triangular process, which is separated from lateral parts of anterior margin by distinct steps (arrows 41a), ostium round, in the centre of sternite VIII, not reaching into the triangular process. Ductus seminalis with about 8 turns. Ductus bursae rather stout with structures common in genus Agonopterix, widening gradually in its course. Corpus bursae of average size (diameter approximately equalling width of sternite VIII in standard preparation, i.e. dorsoventrally flattened), signum narrow oval (4 times wider than long), rather large (maximum diameter about one half diameter of bursa).

Female genitalia
As in the males, all the species compared with A. pseudoferulae sp. n. also show distinct differences in female genitalia: A. ferulae (Fig. 41b) has a straight margin of sternite VIII without any fold. In A. atomella, oblique folds (arrow Fig. 41c) are developed at each side of centre of margin. In A. scopariella these folds are also present and between them, the anterior margin shows a somewhat rectangular extension (arrow Fig. 41d). In A. oinochroa it is slightly curved with a narrow transverse fold (arrow Fig. 41e), in A. cluniana it is extremely bulged (arrow Fig. 41f ), which gives a character of the female genitalia of this species which is unique within Agonopterix. Neighbour-joining analysis shows Agonopterix atomella ([Denis & Schiffermüller], 1775) (BOLD:ABZ0059) as the nearest neighbour at a minimum of 2.45% pdistance. So far there are only sequences from the Italian population available, where no intraspecific divergence had been found, but this may change when Greek specimens are sequenced.
For Maximum Likelihood analysis, see Fig. 18. Related species: Searching for the most closely related species based on a neighbour-joining tree (Fig. 42), Maximum Likelihood analysis (Fig. 18) and genitalia patterns of both sexes has not achieved a satisfactory result in A. pseudoferulae. Compared with the nearest neighbour, there are some distinct differences: A. atomella is a Fabaceae-feeder, in male genitalia anellus lobes are very different. Looking further afield at the second nearest neighbour, Agonopterix scopariella (Heinemann, 1870) with a p-distance of 2.6% is also a Fabaceae-feeder, and the differences in genitalia are at least as marked as in A. atomella. On the other hand, the two Fabaceae-feeders A. atomella and A. oinochroa have very similar male genitalia, but a barcode distance of 4.08%. This suggests that every single parameter must be handled with care. Pronounced similarity may result from being closely related, but it does not prove it, because a single distinctive feature may develop independently in different groups. The only certainty from present evidence is that A. pseudoferulae is not a cryptic species.  Etymology. The species name is a noun in genitive case. The first specimens of this new species were discovered in ZSM in the Klimesch collection under A. ferulae. This was decisive for the species name pseudoferulae, which means "the false ferulae".
Distribution. So far known from Italy and Greece. In Italy it had been collected from Mt Terminillo (Latium), Gargano (Puglia), Madonie, Piano Battaglia (Sicily) and Laconi (Sardinia) and in Greece from Chelmos (Peloponnese).
Bionomics. Peter Sonderegger reared it from larvae collected on Elaeoselinum asclepium (L.) Bertol. (Apiaceae) from Gargano, Italy. Unfortunately he was not expecting anything of great interest, so no photo or larval description was obtained. Larvae were collected on 4 April, while the moth emerged in late April. Moths in good condition have been caught in June and July, and a worn specimen has been caught in October. It remains unclear in which stage the species survives winter.
Remarks. In ZSM under the name Agonopterix ferulae P. Buchner found a specimen collected by Josef Klimesch in Greece, which had a red mark in the discal cell which is not present in A. ferulae. Genitalia examination showed that it was distinct from A. ferulae and subsequently, when more recent specimens were found it was possible to obtain barcodes. Both genitalia and barcodes show this to be a new species not closely related to A. ferulae. It is described here as A. pseudoferulae.
Holotype. ♂, Spain, Canary Islands, Tenerife, Guimar, 6.iii. Bupleurum aciphyllum [Bupleurum salicifolium ssp. aciphyllum], ex. 16.iv.1907, Wlsm. 99748 | Walsingham Collection 1910-427 | B.M. ♂ Genitalia Slide No. 23304, NHMUK010305296, coll. NHMUK. Paratypes. 1 ♀, Spain, Canary Islands, Tenerife, Guimar, La Ladera, 800 m, 23.iv.1998 Introductory note. It may be considered unusual to give a detailed description of the nominate subspecies before the description of a new subspecies, but in this case the original Spanish description is not detailed enough to serve as the basis for a comparison of the two subspecies. The original description is completely without information on genetic data and has little on relationships of the new species.It is therefore necessary to include such information on the nominate subspecies in this investigation.
Diagnosis. The wing pattern of both subspecies of D. saharae belongs to one of the basic patterns in the genus Depressaria which can also be found e.g. in D. ultimella Stainton, 1849 andD. daucella (Denis &Schiffermüller, 1775), with which this species was confused by Walsingham (published by him as Depressaria apiella (Hübner, 1796)). A situation which is often found in Depressaria is a combination of high intraspecific variability and near identical basic wing patterns used by several species, which makes it very difficult to determine specimens externally. When intraspecific variability is larger than the mean difference between the species, identification may become impossible. On the other hand, most species of Depressaria have distinctive genitalia in both sexes. This is the case in D. saharae, where diagnosis must be based on genitalia: see relevant paragraphs below.
Description. Depressaria saharae ssp. saharae specimens (only males) from mainland Spain (Figs 43-47): Wingspan 18-23 mm. Head greyish brown, tips of the scales markedly paler than the rest. Labial palp second segment with long, forward projecting scales which are dark grey with a narrow whitish distal margin, third segment medium grey with flesh-coloured tinge, only at base with some blackish scales. Antenna with scape blackish, flagellum blackish on dorsal side and medium yellowish grey on ventral side. Thorax and tegulae medium greyish brown, thorax with 3 dark longitudinal streaks, one in the middle and one at each side. Forewing ground colour grey, with distinct blackish longitudinal streaks, especially in outer one-third; whitish scales are interspersed in low numbers over the whole surface, also forming an acute angled transverse line at about two-thirds, angle about 50°, and a longitudinal, somewhat interrupted line in the middle from about one-fifth to one-half; in older specimens the patterns formed by the whitish scales soon become invisible, but the longitudinal blackish streaks remain visible even in rather worn specimens; cilia dark grey, without distinct contrast from wings. Hindwing moderately translucent at base, becoming increasingly opaque toward distal part, medium greyish brown, veins darker; cilia concolorous with wings, basal one-third markedly darker than the rest in fresh specimens. Legs and abdomen without distinct patterns, covered with a mixture of light grey and blackish scales.
Depressaria saharae ssp. tabelli ssp. n. : Wingspan 22-24 mm. Head warm yellowish brown, tips of the scales only slightly paler than the rest. Labial palp second segment with long, forward projecting scales which are dark warm brown with a narrow whitish distal margin, third segment yellowish at the very tip, rest of distal half predominatly black, basal half with varying proportions of blackish and pale scales. Antenna as in nominate ssp. Thorax and tegulae warm medium brown, thorax without black longitudinal streaks, only a slightly darker shadow may be visible. The most striking differences are colour and patterns of forewings: ground colour warm medium brown in costal half, becoming darker in dorsal half, but without sharp borderline between these areas, longitudinal streaks reduced, much less prominent than in nominate ssp., in central part of costal half almost completely absent; interspersed whitish scales and acute angled transverse line as in nominate ssp., cilia following the general tendency more warm brown, no remarkable difference in hindwings, legs and abdomen.
No gender-associated differences could be found in the specimens from Canary Islands.
Male genitalia. Male genitalia of D. saharae  are really similar only to those of D. bupleurella Heinemann, 1870. The most distinctive difference is the width of the excavation in the costa of valva: narrow (less than half of the basal diameter of the bulges at each side of the excavation) in D. bupleurella (Fig. 60), wide (about equalling the basal diameter of these bulges) in D. saharae. Apart from the species pair D. bupleurella / saharae, the genitalia of D. radiella (Goeze, 1789) show some similarity, but with differences in many details; see comparison in Figs 58-61.
Female genitalia. Female genitalia (Figs 62-63 + 66) are also most similar to D. bupleurella (Figs 64 + 67) with nearly the same shape of ostium and an expansion in the middle of the long and narrow ductus bursae. The best feature to separate the species is the shape of the expansion: an asymmetrical swelling without longitudinal streaks estimation. The close relatedness of D. saharae and D. bupleurella is also supported by biology with both species (so far only known from ssp. tabelli) feeding on Bupleurum.
Etymology. The subspecies name, a noun in the genitive case, honours Jukka Tabell, the Finnish lepidopterologist, who collected D. saharae -at this time still an undescribed species -in 2015 from the Spanish mainland, and sent specimens to Peter Buchner for study. They were essential to understanding this species and led to a search for females, which were found in collections from the Canary Islands and which are here treated as a separate subspecies.
Distribution. So far known only from Spain: Canary Islands (Tenerife). Bionomics. Walsingham reared one moth from larvae collected on Bupleurum aciphyllum (Bupleurum salicifolium ssp. aciphyllum (Webb & Berthel.) Sunding & G. Kunkel) from Canary Islands, Tenerife, Guimar. This plant is an endemic species of Macaronesia. The food-plant of D. saharae ssp. saharae is unknown, but is likely to be another species of Bupleurum.
Remarks. The first encounter with male genitalia of D. saharae was a simple drawing in literature: Klimesch (1985) reports on a letter from Klaus Sattler regarding Walsingham's bred male from Tenerife which Walsingham had referred to D. apiella: "….According to Dr. Sattler, NHMUK London, this specimen belongs to a species near D. bupleurella or to a form of D. bupleurella. Dissection showed differences in costa of valva and in cuiller. It must be left to a later revision of this group to decide on the final status" [translated from German]. Some males from Teruel, dissected by P. Buchner, showed this distinctive genitalia feature also. DNA barcoding supported the view that it was not a form of D. bupleurella, but a distinct species. As at this stage females were unknown, it remained undescribed.
In the large collection of Knud Larson, two females from Tenerife were found, which were both in external appearance and in genitalia patterns close to D. bupleurella, but showed a 6.36% p-distance in DNA-barcode, while barcodes show a 2% difference compared to D. saharae from Teruel, separating into two reciprocally monophyletic clusters. This suggested they were at least closely related, but left open the question of conspecificity. A male reared by Walsingham from Tenerife in 1907 was the key to this so far unanswered question: it has genitalia like D. saharae from Teruel, but in external appearance is like the females from Tenerife. The lack of genitalic separation suggests that the Teruel and Tenerife specimens are conspecific, in spite of their different external appearance. The different external appearance of the Canary Island population, in combination with the corresponding external features of both sexes of the Canary Island population justify the treatment as two separate subspecies.  . The evolutionary history was inferred by using the Maximum Likelihood method based on the Tamura-Nei model. The tree with the highest log likelihood (-4818.52) is shown. Initial trees for the heuristic search were obtained automatically by applying Neighbor-Join and BioNJ algorithms to a matrix of pairwise distances estimated using the Maximum Composite Likelihood (MCL) approach, and then selecting the topology with superior log likelihood value. The tree is drawn to scale, with branch lengths measured in the number of substitutions per site. The proportion of sites where at least 1 unambiguous base is present in at least 1 sequence for each descendent clade is shown next to each internal node in the tree. The analysis involved 54 nucleotide sequences from selected species of Agonopterix and Depressaria. Only the Depressaria-part of the tree is shown here. Codon positions included were 1st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 657 positions in the final dataset. Evolutionary analyses were conducted in MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets (Kumar, Stecher and Tamura 2015). The result is shown in radiation graphic, because in this view the evolutionary aspect is visualized better than in traditional tree.