INTRODUCTION

Metaparagia doddi was described by Meade-Waldo (1911) from a single female collected in “Cairns, N. Queensland” by F. P. Dodd. The type is in the Natural History Museum, London, and has remained the only known specimen of this species until now. We here report the rediscovery of this species, provide the first host plant record, and describe the male.

NEW RECORDS AND HOST PLANT RECORD

On February 27–28, 2009, and March 4, 2009, we collected four females and two males of this species at Mt. Carbine Caravan Park, 16°31′44″S 145°08′14″E and 351 m. elevation in Far North Queensland. The caravan park comprises 13.5 acres adjoining the old mining town of Brooklyn Village. All the specimens were taken in ironbark scrub on a rocky jasper hill overlooking the buildings of the park and village (fig. 1). Three of the females were taken at flowers of Tephrosia brachyodon (Fabaceae), which was common on the hill (fig. 2). The other female and the males were taken in flight near the legume.

Figure 1

Habitat where M. doddi was collected.

Figure 2

Collecting site with host plant.

The senior author has seen the type of M. doddi, and these specimens correspond to it in details of both morphology and color pattern. The scape is somewhat thick, the occiput in dorsal view is posteriorly truncate and nearly straight across mesally, the clypeus is entirely yellow and metasomal Tergum II has a spot at the posterior margin (see the key of Richards, 1962: 59). The females and one of the males have a posterior yellow spot on the scutum, mentioned by Meade-Waldo (1911), and the mandibles are yellow except for the teeth. The size of the scutal spot varies, and one of the males lacks it entirely.

We have also seen a specimen in the Australian National Insect Collection, a female labeled “15.39S 144.31E / Split Rock QLD / 18 Feb–25 Apr 1993 / Malaise Trap / P. Zborowski,” to which the senior author has attached a determination label. This is slightly blacker than the other specimens, with no pale color laterally on the scutum, and metasomal Tergum VI anteriorly and Sternum VI laterally black. Due to being collected into ethanol the glossa is entirely retracted (see Concluding Remarks).

DESCRIPTION OF MALE

Key morphological features are identical to the female. The association of sexes is therefore not in doubt. The pale markings are yellower than in the female, which has them tinged with orange. The posterior mesal spot on the second metasomal tergum is more triangular than square. The antennal apex is not modified, being conical like that in the female, unlike the case in some other species of Metaparagia (cf. figs. 9–11 in Snelling, 1986). The foretarsal segments are symmetrical in both sexes, and the forefemur is rounded, not angular.

Aside from the usual sexual dimorphism, the males are notably different from the female in several characters. The foretrochanter is produced, narrow, and slightly emarginate at the apex (fig. 3). This process is quite different from that in other species of Metaparagia (cf. figs. 7, 16–18 in Snelling, 1986). The last visible metasomal sternum has the seventh and eighth (eighth and ninth abdominal) sternites partly differentiated, not completely fused, and the seventh is cleft apically, while the eighth has a basal process that projects into the cleft. This condition is seen in other species of Metaparagia (cf. fig. 61 in Richards, 1962), as well as other masarines. The apex of the eighth sternite is produced into five points, two lateral, two dorsolateral above, and a medial one above the plane of the dorsolateral ones (fig. 3). A similar condition is seen in some species of Paragia (cf. fig. 30 in Snelling, 1986).

Figure 3

Characters of male M. doddi. Above, apex of foretrochanter process. Below, apex of last visible metasomal sternum. Scale bar  =  1 mm.

In the genitalia, the parameral spines are rather broad, curved, and with the apices produced into minute recurved tips. The parameres are laterally sharply pointed. The aedeagus is broadly rounded apically, and the volsellae are short, not reaching the apex of the aedeagus. These features are reminiscent of the condition in M. maculata, but the recurved tips to the parameral spines are much smaller than in that species. Metaparagia doddi differs from M. maculata in that the digitus is not fingerlike; rather it is broadly triangular.

CHARACTER CODING

The discovery of the male of M. doddi allows us to fill in some of the missing data entries in the matrix published by Carpenter (1997). This is done in table 1. A state has been added to character 27, male foretrochanter, to denote the unique process in M. doddi (state 5, emarginate). We also take this opportunity to correct an erroneous entry in the previous publication: character 30, female foretarsal lobes, was incorrectly scored as state 1 (strongly asymmetrical) in M. maculata; it should have been scored with state 0 (symmetrical), as in table 1.

Table 1

Revised data matrix for species of Metaparagia Outgroups are the masarine tribe Gayellini, and the genera Ceramius and Paragia. Multistate variables are treated as additive except where noted in the character list. A question mark denotes missing values. An asterisk indicates a polymorphism that includes the entire range of values for a given variable. A $ is used for a subset polymorphism (in the sole case, the included values are 0 and 1).

Analysis of the revised matrix using the program TNT (Goloboff et al., 2008), using implicit enumeration under either equal or implied weights results in the same cladogram as published by Carpenter (1997). It is presented here with revised character plotting as fig. 4.

Figure 4

Cladogram of Metaparagia species resulting from present analysis. The length is 65 steps, with consistency index  =  0.65 and retention index  =  0.73. Characters have been optimized and the cladogram plotted with the program Winclada (Nixon, 2002). Character numbers are above the hash marks; state changes are shown below, with the respective primitive and derived conditions separated by a caret, >. Filled hash marks indicate an uncontroverted step, while open hash marks indicate homoplastic change. Only unambiguous optimizations are plotted.

CONCLUDING REMARKS

All the specimens when collected had their glossa retracted, and extruded it upon being placed in a cyanide tube. This is noteworthy because the generic classification of Australian Masarinae was for long based on a misapprehension that some species could not retract the glossa while other species could, and several monotypic genera were accordingly described (Richards, 1962; Snelling, 1986). Carpenter (1997) pointed out that this was a mistake, and rationalized the classification to recognize two genera for Australian Masarinae, Paragia Shuckard and Metaparagia Meade-Waldo. These were the same two genera recognized by Meade-Waldo (1911), when he described both Metaparagia and M. doddi.

Meade-Waldo included three species in Metaparagia, and species described by Richards (1962, 1968) and Snelling (1986) in other genera and transferred to Metaparagia by Carpenter (1997) bring the present total in the genus to nine species. In the cladistic analysis by Carpenter (1997), and repeated here (fig. 4), M. doddi is the sister-group of M. pictifrons (Smith), and their sister-group in turn is M. maculata (Meade-Waldo). Metaparagia pictifrons remains known only from the type and there is no host record. Hosts for M. maculata are known: Houston (1995) reported females foraging on flowers of Jacksonia sericea, J. sternbergiana, J. stricta, and J. spinosa (Fabaceae). Known hosts for other species of Metaparagia do not include Fabaceae; rather they are Goodeniaceae and Rubiaceae (Richards, 1968; Snelling, 1986; Gess, 1995; Gess et al., 1995). The finding of a host in Fabaceae for M. doddi thus may have phylogenetic significance, possibly being apomorphic for the clade comprising M. doddi, M. pictifrons, and M. maculata.

Finally, Richards (1962: 60) stated of Paragia, the other Australian masarine genus: “I suspect that the species were much commoner in the early days of the Australian colonies but that they have become extinct or rare with the destruction of the native vegetation.” While M. doddi perhaps no longer occurs in the type locality, Cairns, due to development, when rediscovered more specimens were seen than ever previously known. The fact that it was found at two localities in savannah country may indicate that it is widespread in such habitat. The seeming rarity of many Australian pollen wasps may instead reflect limited and local abundance tied to appearance of host plants, and corresponding lack of collecting. Clearly there is much more entomological exploration to be done with the Australian fauna.