Cholovocera maderae Wollaston 1854

Cholovocera maderae Wollaston, 1854

This taxon has a convoluted history regarding both its generic placement and its original geographic distribution. It was described from a single specimen, collected in the island of Madeira, in the North Atlantic Ocean. In the original description, Wollaston (1854: 180) wrote that it differed from the other two members of Cholovocera by its larger size (“Long. corp. lin. 7/8”), no punctuation, a more rounded outline, and described the eyes as “prominent”, “being composed of merely few large facets, set widely apart upon a convex surface”. Firstly, the given body length would be equivalent to 1.86 mm, a much longer measurement that the longest species of Cholovocera (see above); secondly, the multifaceted eye is not a generic feature of Cholovocera, which has only one facet, protected by a lateral rim (Fig. 24E).

“ Cholovocera ” maderae has been recorded from several localities around the world, always associated with the longhorn crazy ant, Paratrechina longicornis (Latreille, 1802), a pantropical invasive species, which appears to be responsible for the wide distribution of the beetle, but both have an unknown geographic origin.

Dajoz (1975: 200) transferred Cholovocera maderae to the genus Displotera Reitter, 1887, erected for the species Displotera simoni Reitter, 1887 from Addah, in the old British Gold Coast (present Ivory Coast) in western Africa. Also, Dajoz (1975) regarded D. simoni as a junior synonym of D. maderae and included two other species in Displotera: D. beloni (Wasmann, 1899) from north-eastern India, and D. grandis Dajoz, 1975 from Sri Lanka, because these three species had six facets on each ocular area. Thus, Dajoz (1975) implied that this genus had an Asiatic origin, but with a cosmopolitan member. Although Reitter (1887) did not mention any association of D. simoni with ants, Paratrechina longicornis does occur in Ivory Coast.

According to Wasmann (1905: 385), most of the localities where the species P. longicornis and “ Ch.” maderae had been recorded together (Madeira, Trinidad and Pará in northern Brazil) were anthropogenic, on the sea coast, and connected with commerce. In contrast, the only places where this ant and beetle have been collected together in natural environments were located in India and Southeast Asia. Over one hundred years later, Wetterer (2008: 142) referred to further records of P. longicornis with “ Coluocera ” maderae in anthropogenic habitats of Myanmar (as Burma), Taiwan, Haiti, Galápagos Islands and Hawaii, which agreed with Wasmann’s (1905) hypothesis that both the host-ant and the beetle have been dispersed by human activity. Wetterer (2008: 138) also referred to the ant cricket Myrmecophilus americanus Saussure, 1877, another insect closely associated with P. longicornis, as widespread in India and Southeast Asia, but known from other parts of the world only in coastal areas along major old trade routes. The conclusion was that the most likely origin of the ant-host and its two symbionts was the Indomalayan Region (Wetterer & Hugel 2008, 2014; Wetterer 2015).

Kistner (1982: 124) studied the association between P. longicornis and “ Ch.” maderae in Hawaii, observing that if the ant nest was disturbed, the workers moved not only their larvae to a secure place, but the beetles as well. Furthermore, Kistner (1982) referred to a groove in the elytral epipleura of the beetles, where the ants’ mandibles would fit to facilitate carrying them. No such intimate ant-beetle association is found between any species of Cholovocera and their ant hosts.

Shockley et al. (2009b: 65) placed Displotera as a new junior synonym of “ Cholovocerida” Belon, 1884b (an unjustified emendation of Colovocerida Belon, 1884b), including “ Ch.” maderae together with two Neotropical species: “ Cholovocerida ” unicurva Belon, 1884b from Chile and “ Cholovocerida ” ecitonis (Wasmann, 1890) from Brazil. We disagree with Shockley et al. (2009b: 65) in both actions: in our opinion Displotera is not a junior synonym of Colovocerida but a valid genus of Indomalayan distribution, and “ Ch.” maderae does not belong in Colovocerida, because this genus appears restricted to the Neotropics.

In conclusion, from the morphological, distributional and biological evidence given above, we agree with Dajoz (1975: 200) that “ Cholovocera ” maderae should be placed in the genus Displotera, and referred to as Displotera maderae (Wollaston, 1854).