Tirumala hamata, Robinson & Vane-Wright, 2018

Robinson, Jeanne & Vane-Wright, Richard I., 2018, A specimen of Tirumala hamata hamata (Macleay, 1826) (Lepidoptera: Danainae) from Captain Cook’s first voyage, Journal of Natural History 52 (11 - 12), pp. 687-712 : 709-711

Populations of Tirumala hamata that could have been sampled during the three Cook voyages across the Pacific, with special reference to the separation of Javan

and Australian specimens by means of biometrical data

The following list indicates Cook voyages landfalls where specimens of T. hamata could have been collected, together with Ambon – the only other likely source of T. hamata material during the mid–late eighteenth century (Vane-Wright and Hughes 2005).

Fiji. Cook made brief landfall in the Lau Group during his second voyage. Watkins (1923) does not list any butterflies from Fiji in the Banks Collection. Talbot (1943) mentions a specimen of T. hamata from ‘Fiji’ from Banks, but the provenance seems very doubtful (see mendica, Appendix 1). Fiji is populated by subsp. neptunia , a small race that is dimorphic in both sexes on Viti Levu and Ovalau (Robinson 1975). Figure 4f illustrates f. protoneptunia, the form most similar to T. hamata hamata . Based on major differences in phenotype, the Dale and Hunter T. hamata cannot have come from Fiji.

Tonga. These islands were visited during the second and third Cook voyages. Watkins (1923) lists some butterflies from Tonga in the Banks Collection. The relatively small Tongan race, subsp. angustata ( Figure 4h ), to which Talbot (1943) referred the Banks ‘Fiji’ specimen (see mendica, Appendix 1), is quite unlike the Dale and Hunter material.

New Zealand. All three Cook voyages visited New Zealand, and Watkins (1923) indicates the presence of several New Zealand butterflies in the Banks Collection. Although T. hamata hamata does occur on NZ, it is a rare visitor that does not establish, with the first known record given as 1940 (Hudson 1950; see main text). That the Dale and/or Hunter material could have come from NZ seems very unlikely.

Norfolk Island. T. hamata hamata was first recorded from Norfolk Island by Smithers (1995) where, as in New Zealand, it has vagrant status. The second Cook voyage made landfall on Norfolk, but Watkins (1923) does not list any butterflies from the island in the Banks Collection, and it seems very unlikely that the Dale and/or Hunter material could have originated there.

Vanuatu. The second Cook voyage visited the New Hebrides (now Vanuatu), where the resident population belongs to subsp. moderata ( Figure 4g ). This phenotype does not match the Dale or Hunter material.

New Caledonia. Visited on the second Cook voyage, New Caledonia is also populated by subsp. moderata , where it occurs in forest and woodland, but is ‘nowhere common’ (Holloway and Peters 1976, p. 296).

New Guinea. Brief landfall is said to have been made on the southern New Guinea coast during the first voyage, after Endeavour left Australia. Most New Guinea hamata populations have the ground colour of the outer margin of the hind wing distinctly paler than the discal area (this reaches an extreme in the population found in the Jimi Valley). However, in parts of coastal Milne Bay and inland, including the Aroa River, the resident T. hamata are very similar to the Australian race. Talbot (1943) described this population as T. hamata subnubila , but D’ Abrera (1971) followed by Parsons (1998) suggested the differences between subnubila and subsp. hamata were trivial. Figure 4e suggests that, although the differences are indeed small, the two are not identical. This could be pursued, but for present purposes there is no evidence that any butterflies were collected from New Guinea during the Cook voyages. Material from southern New Guinea was very rare during the mid-late eighteenth century and, as already noted, other than in one restricted area, the New Guinea phenotypes are distinctive from both Australian and Javan material.

Savu. According to Vane-Wright and Gaonkar (2006), although Endeavour made brief landfall on the island of Savu (Sawu), south-west of Timor, there is no evidence that any butterflies were collected. Ackery and Vane-Wright (1984) found no record of T. hamata for this island, although it seems likely that subsp. paryadres occurs there.

Ambon. None of the Cook voyages visited the Spice Islands (Maluku) . Ambon (Amboina) was then a garrison and headquarters for the extensive Dutch trading in the Far East, and one of the main sources of South-east Asian insect material during the eighteenth century . However, as discussed by Vane-Wright and Hughes (2005), at that time most of the material from Ambon ended up, understandably, in Dutch and other continental cabinets; little reached England . Central Maluku is inhabited by T . hamata nigra ( Figure 4d ) – a slightly darker race that can be differentiated from T. hamata hamata and T. h. neomelissa .

Java. As recounted by Banks (in Beaglehole 1962) and O’ Brian (1987), during the return of the first Cook voyage, Endeavour was forced to stop at Java for several weeks. Even though the crew were ill and found themselves in terrible circumstances, some insect material was obtained – although it was poorly documented then, or subsequently (Vane-Wright and Gaonkar 2006) . The endemic T . hamata population belongs to subsp. neomelissa ( Figure 4c ). Although Javanese hamata are very similar to T . hamata hamata, most individuals can be separated by ‘jizz’ – which we have made operational by biometric comparison of certain fore wing spots ( Table 1).

Australia. Only the first voyage spent time in Australia in those areas (e.g. Queensland coast) where T. hamata is common. We know from Banks’ s diary that near Rockhampton he and other members of the Endeavour crew encountered and apparently collected multiple specimens of a species which he likened to another ‘blue tiger’, Ideopsis similis (Linnaeus) . Among the Australian fauna T. hamata is the only species that fits such a comparison. Two examples of T. hamata hamata are illustrated ( Figure 4a,b ), to give some of idea of the range of variation. Of the hamata populations that Banks and the Cook voyages could have encountered, working from individual specimens without

Hunter female

Fore wing l. 44 mm ≈ Average for h. hamata , = max. observed for h. neomelissa

M1 8.4 mm> Max. observed for h. hamata ,> max. observed for h. neomelissa M2 6.5 mm ≈ Mean for h. hamata ,> max. observed for h. neomelissa

fwl/M1 + M2 2.95 Within range for h. hamata , outside range for h. neomelissa

Dale male

Fore wing l. 39 mm Within observed range for both h. hamata and h. neomelissa

M1 5.9 mm Within observed range for both h. hamata and h. neomelissa

M2 5.9 mm Within observed range for h. hamata ,> max for h. neomelissa fwl/M1 + M2 3.31 Within range for h. hamata , outside range for h. neomelissa

Dale female

Fore wing l. 44 mm ≈ Average for h. hamata , = max. observed for h. neomelissa

M1 7.3 mm <Max. observed for h. hamata , = max observed for h. neomelissa M2 6.2 mm> Min. observed for h. hamata ,> max. observed for h. neomelissa fwl/M1 + M2 3.30 Within range for h. hamata , outside range for h. neomelissa

provenance the most difficult to separate are those from Java and Australia. However, the biometric data presented in Table 1 suggest that, having eliminated all possibilities except Java and Australia, the putative Cook voyage material discussed in this paper almost certainly came from Australia, not Java.

Figures

Figure 4. Examples of seven of the 20+ currently recognised subspecies of Tirumala hamata (all specimens in NHMUK; all females; not to scale): (a) T. hamata hamata Australia, Queensland, Westwood, nr Rockhampton, 8 March 1924, GH Wilkins [NHMUK 010242199]; (b) T. hamata hamata Australia, Port Darwin, June 1890, JJ Walker [NHMUK 010242198]; (c) T. hamata neomelissa Indonesia, Java, 1904, ex. J Waterstradt, ex. Oberthür [NHMUK 010242197]; (d) T. hamata nigra Indonesia, Ambon, ex. J Waterstradt, ex. Oberthür [NHMUK 010242196]; (e) T. hamata subnubila, paratype, Papua New Guinea, Aroa River, AS Meek [NHMUK 010242200]; (f) T. hamata neptunia f. protoneptunia Fiji, Mango, Tairuni, September 1882, Woodford [NHMUK 010242201]; (g) T. hamata moderata Vanuatu, Espiritu Santo, Woodford [NHMUK 010242203]; (h) T.hamata angustata Tonga, Tongatabu [NHMUK 010242202].