Given that these two species have indistinguishable early stages and are found on the same food plants in the same area, that there are no observations regarding differences in behaviour which might lead to segregated mate location (e.g. differences in adult activity with regard to season, habitat, height above ground or time of day), and the adult differences are based primarily on the bimodal distribution of a ratio for one spot size, it is justified to ask, are these two valid species or two interbreeding forms of one species?
One source of guidance in similar situations has been to compare their barcodes (e.g. Janzen et al. 2009, 2011). Libert (2014) states that barcodes are of little use to interpret the galenus clade. Nevertheless we compared his barcodes in BOLD (http://www.boldsystems.org/) for these two species from Kakamega Forest: MLIB-0033, MLIB-0034, MLIB-0044, and MLIB-0046 for A. opalinus and MLIB-0039 and MLIB-0043 for A. kakamegae. We found that the two species in Kakamega Forest have identical or almost identical barcodes, which strongly suggests that they are an interbreeding population. Furthermore, specimens of A. rwandae Libert from Rwanda (MLIB-0045 and MLIB-0407) also have barcodes almost identical to those of A. opalinus and A. kakamegae from Kakamega Forest. Barcode Index Numbers (BINs) have been introduced to provide a permanent numbering system for barcode clusters which in a high percentage of cases correspond to known taxonomic species and can also help flag species complexes or clusters needing taxonomic research (Ratnasingham & Hebert 2013, Miller et al. 2016). They provide a useful surrogate for species in ecological diversity estimates but there are many known cases of BIN sharing between apparently good morphological species and different BINs representing deep splits where no morphological differences have been found (e.g. Hausmann et al. 2013, Zahiri et al. 2017).
These specimens from Kakamega Forest and Rwanda have a single BIN: BOLD:ACE5474, and appear to represent one species based purely on their barcodes. However, barcoded individuals of all three species from other regions have barcodes that match other species, often those co-occurring in those locations. Accordingly, it might not be wise to assume that A. opalinus from east of the Rift Valley (type locality Kikuyu), for which no barcodes are available, has identical barcodes to the material barcoded from west of the Rift Valley. Libert (2014) was aware of these types of problems with the barcodes of the galenus clade and considered many uninterpretable when presenting his conclusions based on wing markings and male genitalia. The galenus glade will require further work to clarify what is going on in terms of population dynamics and speciation to reconcile with the barcode patterns. In the meantime, the situation in Kakamega Forest could be made clearer for A. opalinus and A. kakamegae, by (1) the documentation of mating pairs, (2) rearing broods from individual females to see if they breed true or can produce both species, and (3) crossing the two species and documenting the progeny.
FIGURE 18. Final instar caterpillar of Apallaga opalinus, collected on Justicia flava, Kenya. 1, dorsolateral view; collected 26 Oct 1987 on Hypoestes aristata, North Muguga Forest, Kenya; photographed 19 Nov; parasitized; MJWC 87 / 31. 2 – 3, details of head, collected 17 Sep 1989 on H. aristata, Ngong Forest, Nairobi; moulted to final instar 17 Oct; photographed 30 Oct; pupated 2 Nov; MJWC 89 / 56.
FIGURE 19. Final instar caterpillar of Apallaga kakamegae, collected 1 Jul 1990 on Justicia flava, Kakamega Forest, Kenya; moulted to final instar 29 Jul, photographed 10 Aug; pupated 14 – 20 Aug; 24 mm; MJWC 90 / 68.
FIGURE 20. Heads of final instar caterpillars of Platylesches spp. [TCEC] 1 – 2, P. panga, collected on Parinari curatellifolia, Mutinondo Wilderness, eastern Zambia; 1, photographed in shelter, 2 Feb 2008; 2, photographed 5 Feb 2008; 3, Platylesches rasta collected on Parinari curatellifolia, Mulanje Pond, Malawi, 20 Mar 2014.