Insect taxonomy can be difficult: a noctuid moth (Agaristinae: Aletopus imperialis) and a geometrid moth (Sterrhinae: Cartaletis dargei) combined into a cryptic species complex in eastern Africa (Lepidoptera)

Taxon sampling and material repositories

Materials from the following collections were studied: ANIC—Australian National Insect Collection (CSIRO), Canberra, Australia; FMNH—Finnish Museum of Natural History, Helsinki, Finland; HSS—Research Collection of Hermann Staude, Magaliesburg, South Africa; NHMUK—Natural History Museum, London, United Kingdom; Ochse—Research Collection of Michael Ochse, Weisenheim am Berg, Germany; UOZM—University of Oulu Zoological Museum; ZSM—Zoologische Staatssammlung München (SNSB), Germany. In addition, literature and online sources on Lepidoptera were screened extensively. Table 1 summarizes the new molecular (genes) and morphological data (dissections) of the study, including the examined type material.

Molecular techniques

We attempted to sequence all eight molecular markers as in Zahiri et al. (2013) for Cartaletis dargei, but only two gene regions amplified successfully from three specimens (collected in 2010, 2013, 2015, see Table 1), the first half of COI (the DNA barcode region) and wingless. In addition, one further specimen yielded the DNA barcode (collected in 2003, see Table 1). Primers and protocols were taken from Wahlberg & Wheat (2008). The successful PCR products were Sanger-sequenced by Macrogen Europe (Amsterdam). Chromatograms were checked with BioEdit (Hall, 1999), and aligned sequences were submitted to NCBI GenBank and are maintained in the VoSeq database (Peña & Malm, 2012).

Alignment and phylogenetic analyses

The Cartaletis dargei sequences were initially analysed in a dataset of eight genes (CAD, COI, EF-1a, GAPDH, IDH, MDH, RpS5 and wingless) where all major lineages of Macroheterocera were represented (taken from Rajaei et al., 2015). Based on the results (Cartaletis dargei fell within Noctuidae), we analysed the new sequences in the eight gene dataset from Zahiri et al. (2013), which has an aligned concatenated length of up to 6400 bp for 78 Noctuoidea species. In addition, we included DNA barcode sequences (Hebert et al., 2003) from six species of Agaristinae from Africa and Australia (see Table 1). Further, we analysed the same taxa using up to two genes only (COI and wingless), which had an aligned concatenated length of up to 1,876 bp. The aim was to compare how up to two genes versus up to eight genes included in the phylogenetic analysis may affect the systematic position of Cartaletis dargei.

Phylogenetic analyses were carried out using IQ-TREE 1.6.10 (Nguyen, Von Haeseler & Minh, 2015) in a maximum likelihood framework. The data were partitioned by gene and analysed with the partition finding (Chernomor, von Haeseler & Minh, 2016) and model finding (Kalyaanamoorthy et al., 2017) algorithms of IQ-TREE (using the command MFP+MERGE). Multiple runs were completed, and within one run 100 independent searches were made. Robustness of the results were assessed using UFBoot2 (Hoang et al., 2018) and a SH-like approximate likelihood ratio test (Guindon et al., 2010), each with 1000 replicates. Analyses were run on the CIPRES server (Miller, Pfeiffer & Schwartz, 2010).

Morphological analyses

Genitalia and abdomens were prepared following standard methods (for instance (Hardwick, 1950)). The male aedeagus is shown both with uneverted vesica, to allow comparison with older literature, and with everted vesica. The vesica was everted via the caecum that was cut open by placing the aedeagus inside a hypodermic syringe (Sihvonen, 2001). Some structures were photographed during dissection in situ using The Fixator (Wanke et al., 2019), to allow an optimal angle for observing and illustrating certain structures. The wings were descaled using the method described in Sihvonen (2005). All structures except wings were stained with Chlorazol Black. Numerous dissected structures shown in the plates were photographed in two to six images at different depths of focus, using a Leica DM1000 microscope and Leica DFC295 camera, and combined into single images using image-stacking software in Adobe Photoshop CC v.20.0. Larger structures were photographed using Canon EOS 5D digital camera with MP-E 65 mm EF 100 mm macro lense. Photos were taken with StackShot automated macro rail and focus stacked in Image Manager Software (Zerene Stacker). Scanning electron microscopy was performed with FEI Quanta 250 FEG (Thermo Fisher, Oregon, USA) using ETD detector of secondary electrons. As the specimen was uncoated, the low accelerating voltage of 1 kV and spot size 3 were used to minimize charging. Original images were cleaned and edited in Adobe Photoshop and compiled into plates with CorelDRAW 2020. Genitalia terminology follows Klots (1970) and Sibatini (1972), wing venation follows Wootton (1979). In ambiguous cases, descriptive terms were used and were accompanied by illustrations.

DNA barcodes, genus- and species-level taxonomy

The genetic data, together with morphology and other available evidence such as distribution, were used to draw conclusions on the exact systematic position of C. dargei within Agaristinae. DNA barcodes (658 bp region near the 5′ terminus of the COI mitochondrial gene) of seven specimens (Table 1) were studied using the analytical tools on BOLD, including BIN (Barcode Index Number) and barcode gap analysis (Ratnasingham & Hebert, 2007; Ratnasingham & Hebert, 2013). Genetic divergences between sequences were calculated using the number of base differences between sequences as implemented in MEGA X (Kumar et al., 2018), and they are reported as percentages. The taxonomic and collection data, voucher image, COI sequence and other metadata are available on the BOLD database https://v4.boldsystems.org through the public dataset DS-AGARIST, doi: dx.doi.org/10.5883/DS-AGARIST “Agaristinae - dargei and imperialis”.

Phylogenetic position of “Cartaletis” dargei

A phylogenetic analysis of 78 Noctuoidea species (up to eight molecular markers/species) with five Cartaletis dargei samples (up to two molecular markers/specimen) and six Agaristinae species (all eight molecular markers for Periscepta, COI barcode only for the other five species) recovered C. dargei in a well-supported Noctuidae: Agaristinae position (SH-like/UFBoot2 = 99.9/99) (Fig. 2). African C. dargei was recovered as sister to all other analysed Agaristinae. African (Schausia coryndoni, Heraclia africana, Agoma trimenii, Aegocera tigrina) and Australian species (Periscepta polysticta, Agarista agricola) grouped together, each in separate lineages. The analysis of up to two molecular markers/species (COI and wingless) recovered C. dargei in exactly the same position within Noctuidae: Agaristinae (SH-like/UFBoot2 = 100/100) (File S1).

Morphology supports a position within Agaristinae also: wing venation trifine (Fig. 3); hindwing vein M2 obsolescent (Fig. 3); male with trifine brush-organs complete of levers, hair-pencils and pleuro-sternal pockets (Fig. 4); counter-tympanal membrane greatly enlarged and associated with reduced counter-tympanal hood (not figured); paired vesicle (bulla abdominalis) on A1 (Fig. 4); female without signum on corpus bursae (Figs. 5–6); forewing often with pale spots or bands on black background (characters after Holloway, 2001); cone-shaped prominence on frons of head (Fig. 7) (character after Becker, 2010). Several Agaristinae species have white fringes on forewing apex, which is also present in our study species. Larva and pupa of Cartaletis dargei are unknown, therefore the diagnostic characters (for instance Kitching & Rawlins, 1999 and references therein) could not be evaluated. We illustrate the characteristic cone-shaped prominence on the frons of the head (Fig. 7), which we have found across Agaristinae studied from Australia, Thailand, Africa, and which is mentioned by Becker (2010) to occur in the Neotropical species. The apex is crater-shaped with tiny sensilla (Fig. 7). See Discussion for additional information.

After the Agaristinae position of Cartaletis dargei was established, an extensive screening of Agaristinae materials in collections and examination of literature revealed that taxon dargei should be classified in the monotypic genus Aletopus (Jordan, 1926). The genus combination was supported by DNA barcodes and morphology, but the species-level taxonomy turned out to be complex. Because Aletopus has not been subject to taxonomic revision, we decided to provide such. Taxonomic conclusions are presented below.

Taxonomy of genus Aletopus (Lepidoptera, Noctuidae: Agaristinae)

Aletopus, as reclassified here, contains three described species, but it could contain five or even more species. We had access to seven DNA barcoded specimens, whose morphology was studied also, in addition to other material (Table 1). Both genetic and morphological data are compatible with a number of taxonomic scenarios, but because material on both sexes was rather limited, we took a conservative view and formally recognise with certainty only two species. See Discussion for alternative classifications.

External characters and abdomen (Figs. 5, 8 and 9): Wingspan: males 36–42 mm (n = 7), females 42–44 mm (n = 2). Wings rounded. Basal part of wings orange to red, distal part blackish brown, traversed by white band or spots on forewings, by row of small white dots on hindwings. Terminal line black, except apex white on both wings. Wings below as above. Frons with four lateral white spots, medial area black with cone-shaped projection. Eyes large, lined ventrally by long and narrow yellow or orange scales. Proboscis well-developed. Basal parts of labial palps white, otherwise structures black. Medial parts of tegulae white, otherwise structures black. Antenna filiform in both sexes. Mesothorax behind tegulae with long and narrow yellow or orange scales. Legs mostly black, with white scales on proximal parts of segments. Spur formula 2–2–4 in both sexes. Abdomen slender, dorsally orange-brown banded; laterally black with two rows of white dots; ventrally black with one row of white dots on segments 1–3. Counter-tympanal membrane greatly enlarged, associated with reduced counter-tympanal hood. Paired vesicle (bulla abdominalis) on A1. Male abdomen with trifine brush-organs on A2 and tergite A8 weakly horse-shoe shaped (Fig. 4), other abdominal sternites and tergites of both sexes undifferentiated. Variation: Aletopus imperialis is slightly sexually dimorphic; the males are orange, while the females show redder colouration, particularly on the hindwings. Reddish-orange males do exist, but normally the colour is not as deep as in the females. Basal part of the female forewing can also be slightly reddish-brown. Occasionally, veins are weakly visible as narrow white lines on black areas.

Venation (Fig. 3): Typical trifine venation of Noctuidae. Forewing R veins form an elongated areole. Hindwing vein M2 obsolescent. Both wings with one anal vein (1A + 2A), A2 short and reduced. Jordan (1926, Fig. 1) illustrates part of the hindwing venation, matching with the venation illustrated here.

Male genitalia (Fig. 9, lower part): Uncus long, setose, expanded medially. Tegumen long and narrow, ventral margins slightly triangular, covered with long setae, dorsal sclerotisations very narrow, y-shaped. Valva long, narrow at base, widest subapically, ventral margin weakly concave, apical half covered with setae. Valva with hook-shaped process (harpe) and sclerotised ridge in middle (sacculus). Saccus narrow, elongated. Aedeagus long, narrow, curved ventrally, dorsal part of shaft broadly open at junction with ductus ejaculatorius for about half of aedeagus length, caecum short. Basal extension of vesica short, blunt-ended, covered with minute spines. Vesica long, tubular, covered with microcornuti, with ventrally recurved tip after ductus ejaculatorius.

Female genitalia (Fig. 5): Papillae anales prominent, covered with long setae. Apophyses posteriores slightly shorter than apophyses anteriores. 8th segment with membranous pouches laterally, on both sides of ostium bursae. Ventral margin of ostium bursae smooth. Lamella postvaginalis membranous. Ductus bursae narrow, sclerotised. Holotype female (Noctuidae Brit. Mus. slide No. 8298 lacks corpus bursae (dissection artefact), therefore we exclude description of the corpus bursae (see. A. dargei).

Distribution, habitat, phenology, biology: Exact distribution is not known, requiring dissection and DNA barcoding of additional material. Therefore the records on the distribution map are not more detailed as shown. Aletopus species are known from eastern Africa, from Tanzania to Malawi (Fig. 10). Most records are from forest habitats between 1800–2070 m, one specimen is from 418 m. The specimens have been recorded between September and March, most specimens are from December, potentially hinting to one generation per year. It is uncertain whether Aletopus species are nocturnal or diurnal. The material in coll. ZSM was collected by Philippe Darge, and the specimens were among artificial light-collected material when those arrived to museum, therefore indicative of being collected at light. The material in coll. HSS was collected during day as bycatch by butterfly collectors, who were doing fieldwork for the African Butterfly Research Institute (ABRI, Kenya). Otherwise, biology and immature stages are unknown.

Genetic data (Fig. 8, Table 2): Aletopus imperialis group splits between two Barcode Index Numbers (BINs): BOLD:AAZ5982 is represented by three specimens from Tanzania (barcode length 625–658 bp), and of those three, BC_ZSM_Lep_42575 is the nearest to the BIN BOLD:AEG8871 at minimum pairwise distance of 1.83%.

Similar species: Aletopus imperialis and A. dargei are externally very similar, and so far reliable external diagnostic characters have not been found. Basal extension of vesica is short and straight in A. imperialis (digitiform and curved in A. dargei), see Figs. 8 and 9. Margin of lamella antevaginalis is smooth in A. imperialis (rough in A. dargei), see Fig. 5–Figs. 6 and 8, Fig. 9. Minimum pairwise distance between A. imperialis group and A. dargei group is 2.79% (Figs. 8 and 9, Table 2). In addition, Aletopus taxa are superficially similar to unrelated Euphaedra ruspina (Hewitson, 1865) (Nymphalidae), which occurs sympatrically in eastern Africa, and to a lesser degree to Apaegocera aurantipennis Hampson, 1912 (Agaristinae).

External characters, abdomen, venation (Figs. 1, 8, 9 and 11). Similar to A. imperialis, so far reliable external diagnostic characters have not been found. See A. imperialis for a general description, including venation.

Male genitalia (Fig. 11): Similar to A. imperialis, but with following differences. Basal extension of vesica is digitiform and curved in A. dargei (short and straight in A. imperialis). Also, valva appears narrower in A. dargei, and wider in A. imperialis, but this quantitative character must be used with caution, because the width may appear different if viewed from different angles.

Female genitalia (Fig. 6): Similar to A. imperialis, but with following difference. Margin of lamella antevaginalis is rough in A. dargei (smooth in A. imperialis). Ductus bursae is distinctly curved before opening of ductus seminalis. Corpus bursae large, elongated, very thin, with short sub-conical appendix bursae posteriorly from which ductus seminalis arises. Signum absent.

Distribution, habitat, phenology, biology: See text under A. imperialis.

Genetic data (Fig. 8, Table 2): Aletopus dargei group splits between two Barcode Index Numbers (BINs): BOLD:ABW9036 is a singleton (658 bp) from Tanzania (holotype of A. dargei), and the nearest BIN is BOLD:AAK2817 at minimum pairwise distance of 3.4%. BOLD:AAK2817 (in dargei group) is represented by two specimens (612–658 bp) from Tanzania, and the nearest BIN is BOLD:AAZ5982 (in imperialis group) at minimum pairwise distance of 2.8%. The complex genetic patterns of A. dargei and A. imperialis merit further attention and integrative taxonomic study (see Discussion).

Similar species: Aletopus imperialis, see text under A. imperialis.

The identity and generic combination of Aletopus ruspina (Aurivillius, 1909), from Republic of the Congo are uncertain. Aurivillius (1909) classified T. ruspina in Agaristidae and according to the original publication, the type is in “Museum Bruxelles” (Royal Belgian Institute of Natural Sciences), but it is lost according to Kiriakoff (1977). Therefore we reproduce the original illustration in Fig. 12. Based on the original illustrations, T. ruspina (Aurivillius, 1909, Fig. 43) and Weymeria athene (Weymer, 1892) an agaristine moth from Tanzania (Weymer, 1892; Goff, 2020), are similar. Aurivillius (1909) also mentions the similarity, but excludes the ruspina-athene relationship because ruspina has thorny hindleg tarsus (long bristles in W. athene). These both resemble in facies more Aletis than Aletopus, by being significantly larger (wingspan is over 60 mm), and the angled inner margin on the forewing blackish area near tornus is distinct, being straighter in Aletopus. However, the combination of A. ruspina and W. athene with Aletis is excluded because both these taxa have filiform antennae (bipectinate in Aletis). This is under the assumption the filiform antenna of T. ruspina are correctly depicted in the original drawing (Aurivillius, 1909, Fig. 43).