The type material of the available name Euploea dorippus Klug, 1845, originated from northern Sudan, an area that lies beyond the core zone of the semispecies or subspecies currently known as Danaus chrysippus dorippus, and the description did not include examples of the phenotype currently referred to as Danaus chrysippus f. ‘dorippus’. Possible consequences for nomenclature of the infraspecific species group taxa and form names of Danaus chrysippus are discussed.

“Mr. Butler has pointed out to me that Klug’s type of L. dorippus has white hind wings, a fact which appears to have been entirely overlooked ...”

Swinhoe 1885 page 126.

“None of these specimens have [sic] any white on the secondaries, and agree best with fig. 5 of Klug’s plate.”

Godman 1885 page 537.

“ab. infumata n. nom. — … dorippus var. Klug, Symb. Phys., t. 48 f. 5.”

Aurivillius 1899 page 33

“dorippus Klug … FW unpatterned, orange; HW orange.”

Smith et al. 1997 page 57, table 2

During recent work addressing Danaus chrysippus (Linnaeus, 1758) in the eastern Mediterranean and Egypt (John and Vane-Wright 2019; Vane-Wright and John 2019), a taxonomic problem came to light concerning use and application of the available binomen Euploea dorippus Klug, 1845. As with some other available names subsumed within Danaus chrysippus sensu lato (e.g. Papilio alcippus Cramer, 1777; Danaida chrysippus var. orientis Aurivillius, 1909), the epithet dorippus is used today in two fundamentally different ways – first, as the trinomen for a semispecies or subspecies considered to have its ‘heartland’, in the sense of Smith et al. (2019), in the Horn of Africa; second, as the name of a form (phenotype) of Danaus chrysippus found more widely in Africa, and also in parts of the Indian region, including Sri Lanka (Donahue 1962).

There are at least two different problems relating to these two uses. The first, with respect to the application of dorippus to a species group taxon in the sense of the International Code of Zoological Nomenclature (ICZN 1999), depends on whether subspecies are considered geographically exclusive (non-overlapping, allopatric) or not – and, even if geographical segregation is insisted on for subspecies, and is usually expected for allospecies (Amadon 1966; Short 1969; Helbig et al. 2002), is this necessarily the case with respect to semispecies, the main populations of which are usually connected by a hybrid zone? The second problem relates to the fact that the ICZN regulations simply do not apply to form names, or any other taxon or appellation considered to be infra-subspecific (Masters 1972; Vinarski 2015b).

The interaction of these two uses of available names, as contended by John & Vane-Wright (2019), creates a further problem – that of confusion in communication. Thus in this case, if I use the name Danaus dorippus, am I referring to a taxon of the species group (species, superspecies, allospecies, semispecies, subspecies – Helbig et al. 2002), or to a phenotype that can transcend the boundaries of geographically circumscribed taxa, and thus be regarded as a form found in two or more species group taxa due to gene flow, shared descent (trans-species polymorphism in the sense of Ségurel et al. 2012), or convergent evolution? Moreover, does it matter if the type locality falls outside the area to which any such species group taxon is considered to be restricted or characteristic? – in other words, beyond its core range, or ‘heartland’. And does it matter if the phenotype currently referred to when the name is used as a ‘form’ does not correspond to any of the phenotypes originally described under that name? Before endeavouring to address these problems, it is first necessary to assess the material basis on which Klug’s name rests – and then explore the various ways in which it has been used in the past, and how it is being used now.

Throughout this paper I use, as normal, italic font for generic and species group names (e.g. Anosia, Danaus chrysippus alcippus), but non-italic font within single quotes when referring to named phenotypes (e.g. Danaus chrysippus ‘albinus’). This convention, derived from Masters (1972), was employed by Vane-Wright (1975) and in many publications since, including Ackery & Vane-Wright (1984). The reasons for doing this are re-stated by John & Vane-Wright (2019, page 142/43) – where, in addition, the phenotype names are all prefixed with f. (for form) – thus Danaus chrysippus f. ‘albinus’. Although the problems discussed in this paper go far deeper than mere typographical distinction, this convention if applied consistently does remove one source of confusion.

Before presenting synoptic information on the original combination, type material and subsequent combinations and uses for Klug’s dorippus, it is helpful to note the various generic names that, since 1845, have been applied to butterflies now placed in the genus Danaus Kluk, 1780, sensu Ackery & Vane-Wright (1984) and Smith (2014). These include four objective synonyms of Danaus, one of which is also a homonym: Danaida Latreille, 1804; Danais Latreille, 1807; Danaus Latreille, 1809; and Festivus Crotch, 1872. Subjective synonyms at generic level are Limnas Hübner, 1806 [this name, first published in Hübner’s Tentamen, is nomenclatorially invalid]; Anosia Hübner, 1816; Salatura Moore, 1880; Nasuma Moore, 1883; Tasitia Moore, 1883; Danaomorpha Kremky, 1925; Panlymnas Bryk, 1937; and Diogas d’Almeida, 1938. For completeness, it should be noted that at the time of Klug’s description it was frequent practice to include species now assigned to Danaus within Euploea Fabricius, 1807, due to a lack of agreement regarding the type species of the latter – a problem not finally resolved until publication of Opinion 163 (ICZN 1945) – see below.

In recent decades Anosia (type species Papilio gilippus Cramer) and Salatura (type species Papilio genutia Cramer) have been used by a few authors for genera separate from Danaus (type species Papilio plexippus Linnaeus), or more frequently as subgenera of Danaus. In rejecting all subgeneric divisions of Danaus, Smith (2014: 120) stated that Salatura and Anosia were “resurrected” by Ackery & Vane-Wright (1984) as subgenera. Their three-fold division was however identical to that recognised by D’Almeida (1939): Diogas (subjective synonym of Danaus s.s.), “Danaus” (equivalent to Salatura, based on a misidentified type species in the current sense), and Anosia. Under the influence of Takashi Shirôzu (e.g. 1960) the same division into three full genera, Danaus, Salatura and Anosia, was for a period employed by Japanese lepidopterists (e.g. Kawazoé and Wakabayashi 1979). Brower et al. (2010) also recognised all three, but as subgenera, with the hierarchical or cladistic relationship ((Danaus) (Anosia + Salatura)) – as proposed by Ackery & Vane-Wright (1984). Others (e.g. Morishita 1981) placed Salatura as a subjective synonym of Anosia, recognising the latter together with Danaus as two separate genera. Morishita’s arrangement is supported by data derived from cardenolide resistant Na+/K+-ATPases (Petschenka et al. 2013). However, the most recent statement remains that of Smith (2014) who, based in part on evidence of conflict among haplotypes, rejected all subgeneric divisions of Danaus.

Euploea dorippus Klug, 1845

Euploea dorippus Klug, 1845: text (signature h[iii]), pl. 48, figs 1–5. Male and female syntypes, Sudan: [New] Dongola, Dunqulah, 19°10’N, 30°29’E, and Ambikol, 18°03’N, 31°31’E, leg. C.G. Ehrenberg & W.F. Hemprich (Baker 1997: 194) (material apparently lost – not in Museum für Naturkunde, Berlin: Olivier and Nekrutenko 2000).

Limnas dorippus (Klug): Butler, 1886: 758. Butler acted as first reviser, restricting application of the name Euploea dorippus to Klug 1845: pl. 48, figs 1–4 (see also Swinhoe 1885: 126). [In this paper Butler excluded the specimen represented by Klug 1845: pl. 48, fig. 5, to become part of the type series of a new nominal taxon, Limnas klugii (Butler, 1886: 758 – type locality currently accepted to be Somalia: inland south of Berbera).

Type material

Klug’s (1845) description of Euploea dorippus was based on at least two males and one female collected by Ehrenberg and Hemprich at two localities on the banks of the Nile in northern Sudan, ca 150 km apart, in May and July 1822 (“Habitat ad Dongalae novae castra Aethiopiae in Echii floribus Maio; ad Ambukohl Iulio.”). The precise locations, as noted above, have been given by Baker (1997); a very helpful map is presented by Chester Bradley (1968). Their approximate location is plotted on Fig. 1.

Figure 1. 

Outline distribution map for the four subspecies or semispecies of Danaus chrysippus in mainland Africa, as recognised by David Smith and co-workers (e.g. Smith et al. 2010 fig.1). This map is based on Lushai et al. 2003 fig. 1 (for which the original caption reads: ‘geographical distribution of the subspecies of D. chrysippus in the Afrotropical region, showing the approximate boundaries of the hybrid zone … ’). Note this extensive ‘hybrid zone’ in eastern Africa. The large five-pointed star in northern Sudan is centred on the type localities of D. dorippus (Klug), in an area supposedly occupied by subspecies D. c. chrysippus or semispecies D. [c.] chrysippus. Not one of the three phenotypes exhibited by the type material of D. dorippus (Fig. 2) corresponds to the phenotype regarded as typical of the Horn of Africa subspecies D. c. dorippus or semispecies D. [c.] dorippus (as currently named). The small five-pointed star in northern Somalia is centred on the approximate type locality of D. klugii (Butler) (See text).

Klug’s original description was in Latin: “Euploea alis testaceis, nigro-limbatis, limbo, subtus praesertim, albo-punctato, posticis maculis, subtus albo-notatis, in disco nigris. Mas. Fem. …. Magnitudine E. Chrysippi, cui valde affinis. Caput et thorax nigra, albo-punctata. Antennae nigrae. Alae fulvo-testaceae, supra margine omni, sparsim albo-punctato, nigro; anticae macula insuper triangulari costali media, posticae maculis discoidalibus quatuor (in femina tribus) nigris; subtus alae basi fulvae, limbo maculisque discoidalibus nigris ubique maculis albis ornatis. Pedes nigri. Abdomen fulvo-testaceum, subtus album. Variat uterque sexus alis posticis medio albis.”

This has kindly been translated by Tony Galsworthy as follows: “A Euploea [now Danaus] with orange wings, bordered with black, the border, particularly below, spotted with white, hind wings black on the disc marked below with white. Male. Female. …. Size as E. chrysippus, to which it is certainly related. Head and thorax black, spotted with white. Antennae black. Wings fulvo-testaceous, upperside with the margin completely black, sparsely spotted with white; fore wing upperside with a triangular costal marking, hind wings with four black discal marks (three in female); underside wings fulvous at base, with black border and discal marks, everywhere ornamented with white markings. Legs black. Abdomen fulvo-testaceous, white below. Both sexes vary in the white in the middle of the hind wings.”

Given the three different phenotypes originally illustrated by Klug (Fig. 2), it might be considered desirable to select a lectotype. This, however, is not possible in terms of an actual specimen as it appears all the original material is lost (Olivier and Nekrutenko 2000). Moreover, with respect to the taxonomic application of the name, unless we insist that Klug was working from a mixed series of two or more species (possibly the case based on the conclusions of Smith et al. 2005), the type localities are accurately known and sufficiently close together (ca 150 km) that any application dependent on geo-location is not likely to be equivocal. Thus I consider making a lectotype (or neotype) designation unnecessary (even undesirable) at the present time with respect to application of Euploea dorippus as an available name (see Discussion).

Before considering these matters further, the various ways in which the name dorippus has been applied over the past 175 years are enumerated and exemplified.

Figure 2. 

The five images published by Klug (1845, plate 48) to illustrate his original description of Euploea dorippus. a, b – pl. 48 figs 1,2 (“E. dorippus mas.”); c, d – pl. 48 figs 3, 4 (“E. dorippus fem.”); e – pl. 48 fig. 5 (“E. dorippus mar. var.” – this “variety” was later included as part of Limnas klugii Butler, 1886, and subsequently given the infra-subspecific name Danais dorippus ab. ‘infumata’ (Aurivillius 1899)). See text.

Previous applications of the name Euploea dorippus

Ignoring differing generic combinations, there are at least nine ways in which the available name Euploea dorippus has been employed since it was first proposed as a new species: 1) separate monotypic species (Danaus dorippus); 2) separate polytypic species; 3) form of a polymorphic D. dorippus; 4) subspecies of a polytypic D. dorippus; 5) semispecies of superspecies Danaus [chrysippus]; 6) subspecies of species D. chrysippus; 7) form of a monotypic but polymorphic D. chrysippus; 8) form of D. chrysippus subspecies chrysippus; and 9), form of D. chrysippus subspecies aegyptius (Schreber, 1759). It could also have been used as a form of subspecies D. chrysippus dorippus or of semispecies D. [chrysippus] dorippus, but I have not seen this done explicitly – and insofar as David Smith and his co-workers have regarded the ‘dorippus’ forewing pattern to be fixed in semispecies or subspecies dorippus in its core area (Smith 2014, page 188; Smith et al. 2016 – but see Discussion), it would be an unlikely previous usage. Instances of all nine uses that I have found are listed sequentially below (ciphers in square brackets refer to the usages listed above).

Literary synonymy of Euploea dorippus

Euploea dorippus – Klug 1845: text (signature h[iii]), pl. 48, figs 1–5. [1]

Danais dorippus – Peters 1862: 371; Godman 1885: 537; Marshall & de Nicéville 1882: 52; Pagenstecher 1902: 129. [1]

Limnas dorippus – Moore 1883: 238; Swinhoe 1885: 126. [1]

Danaida dorippus – Aurivillius 1910a: 2, 1910b: 72 [1]

Danaus dorippus – Smith et al. 2005. [1]

Danaus dorippus bataviana (Moore, 1883) – Smith et al. 2005. [2]

Danais dorippus var. klugii – Holt White 1894: 49. [2 or 3]

Danaus dorippus f. dorippus – Suffert 1900: 116. [3]

Danaus dorippus dorippus – Smith et al 2005. [4]

Danaus [chrysippus] dorippus – Smith et al. 2010; Smith 2014; Gordon et al. 2014, 2015. [Note – square brackets around chrysippus used here indicate that dorippus is a member of a superspecies of which Danaus chrysippus is the first-named taxon, in accordance with the proposals of Amadon (1966; see also Helbig et al. 2002).] [5]

Danaus chrysippus dorippus – Bryk, 1937: 66; ?Talbot 1943: 121; Gordon 1984; Smith et al. 1997, 1998, 2002, 2016, 2019; Lushai et al. 2003a, 2003b, 2005; Herren et al. 2007; Traut et al. 2017; Williams 2018; Duplouy and Hornett 2018; Martin et al. 2019. [6]

Limnas chrysippus var. dorippus – Butler 1897: 923. [6 or 7]

Danais chrysippus var. dorippus – Kirby 1871: 7; Trimen 1887: 53. [6 or 7]

Salatura chrysippus var. dorippus – Ormiston 1918: 5. [6 or 7]

Danaida chrysippus var. dorippus – Ormiston 1924: 3. [6 or 7]

Danaida chrysippus f. dorippus – Fruhstorfer 1910: 194; Hulstaert 1931: 27. [7]

Danais chrysippus f. dorippus – Manders 1912; Woodhouse and Henry 1942: 39, pl. 2 fig. 3. [7]

Danaus chrysippus f. dorippus – Woodhouse 1952; Donahue 1962; Gifford 1965; Owen 1971; Dickson and Kroon 1978; Kielland 1990; Idris and Hassan 2012; Idris 2013; Hassan et al. 2013. [7]

Danaus chrysippus chrysippus f. dorippus – Carcasson 1963: 21; Larsen 1990, 1991; Gillett 1998; ?Braby et al. 2015; van der Poorten & van der Poorten 2016. [8]

Danaus chrysippus aegyptius f. dorippus – ?Talbot 1943: 120; Pierre 1974; Rothschild et al. 1975; D’Abrera 1980; Ackery & Vane-Wright 1984; Smith et al. 1988; Ackery et al. 1995; d’Abrera 1997. [9]

Danaus dorippus

This usage reflects its original, species-level status – having first been introduced by Klug (1845) as a species of Euploea Fabricius, 1807. Fabricius included three nominal species in Euploea, one being Papilio plexippus L., the type species of Danaus Kluk, 1780. The current usage of Euploea was not finally stabilised until publication of Opinion 163 (ICZN 1945), which fixed Papilio corus Fabricius, 1793, as its type species. Papilio corus is now used as the name for a subspecies of the butterfly currently known as Euploea phaenareta (Schaller, 1785).

Species level status for dorippus was maintained by Aurivillius in several publications, notably the very influential ‘Seitz’ (Aurivillius 1911, as Danaida dorippus). In that work he states that the forewing apex is brown-yellow without any subapical white band, and “the hindwing is not white [sic]”. He does acknowledge that it is often only regarded as a form of D. chrysippus, “but of this there is no sufficient evidence” (Aurivillius, 1911: 12). From Aurivillius (1899: 32/33) it is evident that he regarded Butler’s Limnas klugii as a synonym of dorippus, but recognised Danais dorippus ab. ‘albinus’ Lanz, 1896, as a variety, and gave Klug’s 1845: pl. 48 fig. 5 “var.” the infra-subspecific name Danais dorippus ab. ‘infumata’. This suggests that Aurivillius (see also Suffert 1900) conflated Klug’s dorippus with the all-orange ground colour morph that also lacks all trace of the forewing pre-apical transverse white spots (‘transiens’) as well as hindwing discal white – a morph which, as shown above, Klug did not describe or illustrate. Thus regardless of Swinhoe’s (1885, p. 126) earlier stricture “Klug’s type of L. dorippus has white hind wings, a fact which appears to have been entirely overlooked”, this misapplication of Klug’s name to the all-orange phenotype has continued ever since.

Despite Aurivillius’s great authority, during most of the rest of the 20^th century dorippus was generally regarded as a form and/or subspecies of D. chrysippus, not a separate species. Particularly important in this shift was Poulton’s (1924) short paper entitled “Danaida chrysippus L. and D. dorippus Klug, proved by breeding to be forms of the same species”. However, separate species status was accorded once more by Smith et al. (2005). Following a series of complex analyses based on mitochondrial 12S rRNA and COI sequences, the authors concluded that “dorippus … is the basal clade of the genus and is reinstated as the species D. dorippus” (Smith et al. 2005: 1291). Furthermore, the authors regarded it as composed of two subspecies: D. dorippus dorippus (range: Somalia, Kenya, Tanzania, Uganda, Sudan, Ethiopia, Arabia, Iran, Pakistan (Baluchistan), India (Sind, Kutch), and D. dorippus bataviana (Moore, 1883) from SE Asia.

Compared with all other interpretations of this group, these were unexpected and radical findings. Five years later Smith et al. (2010) announced “we wish to renounce our decision to restore dorippus to full species status, based solely on the evidence of mtDNA-based phylogenies …. and designate D. c. dorippus (stat. rev.) as its fourth African semispecies”. This is so despite the fact that fig. 3.3. in Smith (2014, p. 129, p. 153) shows D. chrysippus dorippus twice – once (dorippus-NB haplotype) as the sister clade to all other taxa of Danaus (including D. plexippus), and second (dorippus-DP haplotype) as sister to Danaus chrysippus bataviana. In contrast, as pointed out by Brower (2016) “A recent molecular systematic paper by Braby et al. (2015) does not support Smith’s results and shows a monophyletic D. chrysippus including African and Asian forms with very low levels of sequence divergence.”

David Smith’s research group has not been consistent with application of semispecies/subspecies rank. In recent publications (e.g. Smith et al. 2016, 2019; Traut et al. 2017; Martin 2019) they have reverted to calling dorippus a subspecies of D. chrysippus.

Danaus [chrysippus] semispecies dorippus

Smith and his associates (Smith et al. 2010, summarised in Smith 2014: fig. 2.34) developed a classification of Danaus chrysippus in the Afrotropical Region based on a hypothesis of past incipient geographical speciation in Africa during the late Pleistocene. They recognised a superspecies with four semispecies in the Afrotropics, for which they applied the following names (without the use of square brackets, a convention adopted here based on Amadon 1966):

• D. [c.] chrysippus in the north-east, to include North Africa, Egypt, northern Sudan (including the type localities for Euploea dorippus), the Canary Islands and Mediterranean eastwards to China etc.);

• D. [c.] alcippus (Cramer) in West Africa north of the Equator and south of the Sahara;

• D. [c.] dorippus in the Horn of Africa region (including the type localities for Limnas klugii); and

• D. [c.] orientis (Aurivillius) in most of Africa south of the Equator, including the Malagasy region.

However, according to the Smith scheme, all these populations are now in contact within a very extensive east African ‘Hybrid Zone’ covering much of Uganda, Tanzania, Kenya, Ethiopia and north of the Horn through Eritrea to Yemen. Although genetically compatible only to varying degrees involving several complex factors (e.g. localised infections with male-killing bacteria – Jiggins et al. 2000), due to the migratory nature of these butterflies, extensive gene flow nonetheless occurs. As a result, all of the various phenotypes can be found well beyond their core zones, including those of the other semispecies (within Africa, most notably the case with semispecies orientis – Smith et al. 2019, fig. 2). Within the Hybrid Zone, Smith (2014) has recognised many forms and genotypes as the result of hybridization between the semispecies, where they apparently coexist, in some sense, as sympatric taxa. Thus although the core area for supposedly ‘pure’ D. [chrysippus] semispecies dorippus is Somalia, eastern Ethiopia and north-eastern Kenya, Smith (2014: fig. 4.1) describes the polymorphism at Dar es Salaam, some 600 km south of this core area, as the result of hybridization between all four semispecies.

Danaus chrysippus subspecies dorippus

Talbot (1943: 121) stated “The form dorippus appears dominant in Abyssinia, Somaliland, Kenya and Tanganyika Territory; it may be considered almost as an eastern race.” Subsequently dorippus has quite often been used for the name of a subspecies of D. chrysippus centred on the Horn region (e.g. Gordon 1984; Herren et al. 2007; Lushai et al. 2003a; Smith et al. 1997, 2002). As noted above, Smith et al. (2016, 2019) have recently reverted to this usage.

During the first two weeks of February 1909 the indefatigable traveller, naturalist and ‘butterfly-hunter’ George Longstaff was in Khartoum. He reported finding “typical chrysippus”, chrysippus “with the veins of the hind-wing dusted with white” “f. alcippoides”, “f. alcippus”, and “f. dorippus, Klug, var. albinus”. He also noted that “Of the total specimens seen, I estimated at the time that at least three-fourths were either alcippus or alcippoides” (Longstaff 1912, page 406).

Under the superpecies scheme, this strongly suggests that in 1912 at least three of the semispecies were present: D. [c.] chrysippus, D. [c.] alcippus, and D. [c.] dorippus – in other words, Khartoum was within the hybrid zone. Some 90 years earlier, it would seem on the evidence of the Hemprich and Ehrenberg material described by Klug as Euploea dorippus, that the hybrid zone extended further north, perhaps as far as the Egyptian border. If we were to accept Longstaff’s (1912, page 395) suggestion that the butterflies depicted on a tomb at Thebes are ‘mid-way in colouring between the type [D. chrysippus] and the variety [var. dorippus]’, then we could speculate that at ca 1400 BCE the hybrid zone extended as far north as Luxor. And why not? The ‘dorippus’ forewing phenotype extends east to India and south to Sri Lanka (van der Poorten & van der Poorten 2016) – where, on the face of it, and making the assumption that the same alleles express as the same phenotypes, then the same three semispecies may also co-occur (this is in fact David Smith’s current view: in litt., August 2019).

The limitation is that we can only observe the phenotypes of the vast majority of specimens on which these observations are based – we have far less knowledge of the supposedly fundamental genetic distinctions between them. Observed differences in nuclear gene Elongation Factor 1-α and haplotypes for mitochondrial genes 12S rDNA and COI were summarised by Smith et al. (2010, table 1). These are:

• D. [c.] chrysippus EF1-α–1, haplotype ST1;

• D. [c.] orientis EF1-α–2, haplotype ST2;

• D. [c.] alcippus EF1-α unstated, haplotype GH;

• D. [c.] dorippus EF1-α–3, haplotype NB/DP.

It would be very interesting to have comparable data for the three forms of D. chrysippus found on Sri Lanka. In this context it may be relevant to note that Singh et al. (2018) have recently reported six haplotypes from 30 individuals of D. chrysippus chrysippus obtained from six sites in the western Himalayan state of Uttarakhand – and, by analysing 54 additional D. chrysippus sequences obtained from the National Center for Biotechnology Information database (samples spanning Africa to Korea), they arrived at a total of 24 D. chrysippus haplotypes.

According to the map of Smith et al. (1998, fig. 3), and those later derived from it (e.g. Smith 2014, fig. 1.13.1; Smith et al. 2019, fig. 2a), Khartoum lies within the range of D. c. chrysippus, beyond that of D. c. dorippus. However, Idris & Hassan (2012) report that the same morphs noted by Longstaff (2012) co-occur at Khartoum today (with ‘albinus’ and ‘dorippus’ still relatively rare) – and they conclude, in contrast to Smith et al. (1998), that Khartoum lies within the hybrid zone.

Butler (1886, page 758) based his new species Limnas klugii – which he distinguished from Limnas dorippus (Klug) – on the following material:

• The “variety” of Euploea dorippus from northern Sudan represented by Klug (1845, pl. 48, fig. 5) – reproduced here as Fig. 2e;

• A male collected by J.W. Yerbury “on the Somali coast” (Butler 1886, page 756), “2^nd April, 1884.” [on evidence of specimen label, lapsus calami for 22.iv.1884] – Fig. 3a;

• One or more males and females collected by J.G. Thrupp “southwards into the interior of Somali from Berbera … Inland south of Berbera” – Fig. 3b–d).

As already noted above, the Ehrenberg/Hemprich “variety” of Euploea dorippus from northern Sudan illustrated by Klug (1845, pl. 48, fig. 5) was subsequently named Danais dorippus ab. ‘infumata’ by Aurivillius (1899). Butler’s (1886) paper is devoted to butterflies of Somalia and, in his description of L. klugii, he makes no mention of northern Sudan (or Nubia as this region was then referred to), despite his inclusion of Klug’s “var.”. Somalia south of Berbera is currently accepted as the (restricted) type locality for L. klugii (Ackery et al. 1995).

The Natural History Museum London (NHMUK) has two male and two female specimens from Somalia identified in 1971 by Phillip Ackery as syntypes of E. klugii Butler. All four are illustrated in Fig. 3. Applying current terminology, three of them (Fig. 3a–c) would be referred to f. ‘dorippus’ (cf Smith 2014 fig. 4.1.2/3); the fourth (Fig. 3d) to f. ‘transiens’ (cf Smith 2014 fig. 3.20.9).

Figure 3. 

Four syntypes of Limnas klugii Butler, 1886, from northern Somalia (NHMUK, London). Upperside left; halved underside right. a – male “Somali 84.83 22.4.84” [Somali coast, J.W. Yerbury; NHMUK010241714]; b – female “Somali 85.38 Dec. 23” [inland, south of Berbera, J.G. Thrupp; NHMUK010241716]; c – female “Somali 85.38” [inland, south of Berbera, J.G. Thrupp; NHMUK010241717]; note – this specimen appears to be have been ‘repaired’]; d – male “Somali 85.38” [inland, south of Berbera, J.G. Thrupp; NHMUK010241715].

Vane-Wright & John (2019, page 152) stated: “In passing, we note that … there is a nomenclatural problem regarding application of the name Euploea dorippus Klug, 1845, to a semispecies or subspecies of Danaus chrysippus restricted to the Horn of Africa.” In this paper the focus is on the nomenclature affecting dorippus; there are also issues of classification that require an extensive overview of the whole species complex, including discussion of the superspecies and subspecies concepts in both theory and application (see e.g. Helbig et al 2002; Vinarski 2015a,b), which I am unable to undertake at this time. Even so, some account of classificatory differences cannot be entirely avoided.

With respect to the classification of Danaus chrysippus in mainland Africa, Vane-Wright & John (2019, page 147) stated “There are now two different, rival systems … the ‘Larsen classification’, which includes all populations in a greatly expanded ssp. chrysippus, and the ‘Smith classification’, which divides African D. chrysippus into four named semispecies together with a large ‘hybrid zone’.” However, there is a third system, also due to Smith and his co-workers, and currently apparently favoured by them (e.g. Smith et al. 2019). This corresponds to the ‘Smith classification’, but treats the four segregates as subspecies. The difference, even if at first sight merely semantic, or simply an issue of assigned rank, can have consequences for nomenclature. Below I will refer to these three systems for D. chrysippus in Africa as:

• C1 (one polymorphic subspecies – Larsen 1991)

• C2 (four subspecies – Smith et al. 1998)

• C3 (a superspecies comprising four semispecies – Smith et al. 2010).

Potential consequences for form names

Based on the nomenclature of Smith (2014), the three phenotypes illustrated by Klug (1845 pl. 84 – see Fig. 2 a–e) are:

• male f. ‘albinus’ (with ‘transiens’) (Klug 1845 pl. 84 fig. 2: 1,2; cf Smith 2014: fig. 4.1.11);

• female f. ‘semialbinus’ (with ‘transiens’) (Klug 1845 pl. 84 fig. 2: 3,4; cf Smith 2014: fig. 4.1.12);

• male f. ‘klugii’ (‘transiens’ not apparent) (Klug 1845 pl. 84 fig. 2: 5; cf Smith 2014: fig. 4.1.1).

Smith et al. (2019, table 1) consider all three to be of hybrid origin: f. ‘albinus’ = alcippus × dorippus F2 hybrid; f. ‘semialbinus’ = alcippus × dorippus F1 hybrid; and f. ‘klugii’ = orientis × dorippus F1 hybrid.

The ground colour of the first two specimens represented in Klug (1845: pl. 48, figs 1–4) appears to be plain orange (Fig. 2a–d). Butler (1886), acting as first reviser, restricted application of the name Euploea dorippus to these ‘albinus’ / ‘semialbinus’ morphs. The third specimen, the “variety” (Klug 1845: pl. 48, fig. 5), which appears to be brown/orange, he included in the type material of his new taxon, Limnas klugii.

The primary problem is then that not one of the phenotypes described and illustrated in the original description of Euploea dorippus corresponds to the all-orange-without-‘transiens’ phenotype currently designated in almost all relevant literature on Danaus chrysippus as f. ‘dorippus’ (including Smith’s 2014 summary volume). Moreover, were we to apply the rules of nomenclature to form names as if we were dealing with available names (and both E. dorippus and L. klugii are available), then we would have to consider treating either f. ‘albinus’ or f. ‘semialbinus’ as a junior synonym of f. ‘dorippus’ (Table 1). A lectotype has yet to be designated for Butler’s Limnas klugii – three of the Somalia syntypes appear to be the “dorippus” phenotype (Fig. 3a–c) as that name is currently applied.

Figure 4. 

Four specimens of Danaus chrysippus collected during 1929 in NW Somalia by M. Portal Hyatt (NHMUK, London, BM 1929-534). Upperside left; halved underside right. a – female “10.10N, 45.45E, ca 1500 ft, Jan.” [NHMUK010241712]; b – male, same data as (a) [NHMUK010241710]; c – female, same data as (a) [NHMUK010241713]; d – female, “9.30N, 44.30E, 4000 ft, May.” [NHMUK010241711].

Despite the various issues presented above, I would not lightly recommend implementation of any changes that might appear to be called for at the present time. There are several reasons:

Despite the evident regret expressed by Vinarski (2015b), unlike available names, form names are currently unregulated. I can call a black butterfly f. ‘alba’ if I wish; I can call an all-orange phenotype of Danaus chrysippus f. ‘dorippus’ if I wish, even though that form was not among those included by Klug under the available name Euploea dorippus. Of course, naming a black butterfly ‘white’ in Latin would be perverse – and the same could be said, as Swinhoe (1885) was surely trying to indicate, of the application of dorippus to a phenotype that Klug did not describe. But both can be done – form names are not regulated.

With respect to the three infraspecific classifications applied to African Danaus chrysippus, there are at least two issues: 1) much work on the butterfly in Africa and elsewhere is in progress, including increasingly complex molecular studies; and 2) there are many other issues affecting the classification of D. chrysippus as a whole, throughout the Old World tropics and subtropics (see e.g. Braby et al. 2015). To give just one example, as discussed above, what do the three forms of chrysippus found on the island of Sri Lanka really represent? Also, from a more theoretical perspective, I have avoided discussion here of different species concepts. A curiosity is that, following the initial criticisms of Wilson & Brown (1953; see also e.g. Zink 2006; Vinarski 2015a, 2015b), the subspecies category remains so widely used in butterfly systematics when in most other animal groups, apart from birds and mammals, it has largely been abandoned. But lepidopterists need not necessarily repent. To quote Gippoliti & Amori (2007) “while the taxonomic validity of many described subspecies is regularly questioned … there is mounting evidence that an astounding amount of taxonomic diversity among mammals has been overlooked; many recognised subspecies may deserve to be considered full species according to evolutionary species concepts … and many clearly distinct taxa remain undescribed”. This often seems to be the case for butterflies too. However, in-depth genomic investigations can reveal excessive taxonomic splitting, as recently shown for the hawkmoth Hyles euphorbiae (L.) by Hundsdoerfer et al. (2019). Each case has to be taken on its own merits.

These issues aside, to introduce a few name changes at this point risks adding to the confusion. The infraspecific taxonomy of Danaus chrysippus across its entire range is very much still a work in progress. Until consensus and greater stability emerge, name changes are generally better resisted unless literally unavoidable (e.g. objective synonymy). After all, unlike genus and species, subspecies is (fortunately!) not a mandatory category in animal taxonomy, and forms names are still unregulated.

Thus the problems revealed here, and the possible solutions proposed, should perhaps be ‘stored’ for appropriate action once a clearer picture has emerged. This investigation also serves to remind that for taxonomic nomenclature there is still a great need for scholarly research – taxonomists themselves, and those who have to make use of taxonomic works and acts, can never rely solely on secondary sources, as these are invariably interpretations. Over time, as our taxonomic concepts change with the growth of biological knowledge, failing to check and reinterpret primary sources can be a recipe for cumulative error and/or confusion. At the same time, when the classificatory framework is itself in flux, piecemeal and premature name changes rarely help. “There is a time to be silent, and a time to speak.” The time to make changes regarding the application of Euploea dorippus has arguably not yet arrived.

However, one of the reviewers of this paper asked that I be encouraged “to ‘stick [my] neck out’ and recommend whether C1, C2 or C3 should preferably be used.” Given the huge amount of work already lavished on Danaus chrysippus in Africa, when compared to the relative poverty of investigations to date in Asia and the Malay Archipelago, and for all the reasons given above, there is to me currently no objective basis for choosing between C1, C2 or C3. Moreover, the assumption that we should always strive for a single, preferred ‘unitary’ classification has been challenged. Thus Scoble (2004) cogently observed that “in reality multiple classifications of the same taxon exist, since different taxonomists often hold different concepts of their taxa”, and argued instead for a unifying rather than unitary approach, whereby the richness and value of maintaining “multiple taxonomies” could be acknowledged and made accessible.

Larsen’s simple system (C1), to record Mediterranean, Afrotropical and Indian subregion Queen butterflies as Danaus chrysippus chrysippus, if given together with form name, phenotype and/or genotype plus locality, is sufficient for most if not all individuals to be ‘translated’ into the current C2 or C3 classifications if those are preferred. However, having been pressed to be more assertive, I am willing to offer four provisional recommendations at this juncture regarding the use of names. Taken together these would result in making one change to species-group nomenclature (under C2 or C3; none under C1), and one change to form nomenclature:

1. Anyone wishing to signify a subordinate taxon of Danaus chrysippus that encompasses the populations of northwest Somalia should in future be encouraged to call this D. chrysippus klugii or D. [chrysippus] klugii, not (respectively) subspecies or semispecies dorippus.

2. The name Euploea dorippus Klug, under C3 arguably based on hybrids, should not be used to indicate a species-group taxon but should, instead, be included in the synonymy of Danaus chrysippus chrysippus (or D. chrysippus aegyptius, if that name is ever resurrected from synonymy).

3. Anyone wishing to use a form name for the ‘all-orange’ phenotype of Danaus chrysippus should continue to apply ‘dorippus’, but in so doing explicitly state (or accept) that they are using the name as a form in the sense of Aurivillius, not Klug. This is perfectly permissible in my view, and avoids confusion so long as dorippus Klug is not used for a taxon of the species group (recommendation 2). [While I greatly regret that butterfly workers ever started to use available names in a two-fold way, the practice is so deeply engrained that no resolution of this awkward case will wash away the many ‘sins of the past’.]

4. Anyone wishing to refer to ‘all orange’ D. chrysippus phenotypes that have the basal upperside coloration of both wings darkened (infumed) should in future be encouraged to refer to these as f. ‘infumata’, not f. ‘klugii’. This is perfectly permissible, and removes ambiguity with klugii as a species group name.

If these four recommendations are followed, then there is no current threat to other named ‘dorippus’-like forms in common use, notably ‘transiens’, ‘albinus’ and ‘semialbinus’.

The author is most grateful to the Reading Room Library Staff at the Natural History Museum London, Val McAtear (Royal Entomological Society library), and Jo McCaffrey for access to literature. Sir Anthony Galsworthy very kindly translated Klug’s original description from the Latin. Eddie John is thanked for inspiration and, together with Nancy and Michael van der Poorten, discussion. And although we do not always agree on taxonomy, I also wish to pay tribute to David Smith for his unquenchable enthusiasm for everything ‘African Queen’, and his remarkable work on the biology of this fascinating butterfly. My thanks are also due to the Editor, and to reviewers Ian Gordon, Gerardo Lamas and Mark Williams, for kindly suggesting several improvements which I have been happy to adopt.

Finalisation of this paper has been facilitated by an Emeritus Fellowship (2017–2019) awarded to the author by The Leverhulme Trust, which is gratefully acknowledged.