On Afromantispa and Mantispa (Insecta, Neuroptera, Mantispidae): elucidating generic boundaries

Abstract The genus Afromantispa Snyman & Ohl, 2012 was recently synonymised with Mantispa Illiger, 1798 by Monserrat (2014). Here morphological evidence is presented in support of restoring the genus Afromantispa stat. rev. to its previous status as a valid and morphologically distinct genus. Twelve new combinations (comb. n.) are proposed as species of Afromantispa including three new synonyms.


Introduction
Mantispidae (Leach, 1815) is a small cosmopolitan family in the very diverse order Neuroptera. The former is characterised by an elongated prothorax, elongated procoxa protruding from the anterior pronotal margin and conspicuous raptorial forelegs. Recently, one of the genera, Mantispa Illiger, 1798 has been the focus of taxonomic studies (Snyman et al. 2012;Monserrat 2014). Mantispa was originally described by Illiger (1978) and quickly became the most speciose genus with a cosmopolitan distribution. Studies by Lambkin (1986a;b), Hoffman (2002), Machado and Rafael (2010) and Snyman et al. (2012) excluded Mantispa's distribution from much of the world and consequently Mantispa is no longer thought to occur in the Neotropics, Nearctic, Afro-tropics or Australasia. Mantispa, according to the morphology of the type species M. pagana (Fabricius, 1775) (synonymised with M. styriaca (Poda, 1761)), is thus probably a small genus from the Palearctic. As the previously mentioned studies focused on the fauna elsewhere the generic boundaries between Mantispa and other similar groups have remained poorly understood.
In their study on the Afrotropical mantispid genera, Snyman et al. (2012) proposed that the majority of the Mantispa-like species from the Afrotropics and south western parts of Europe can be defined as a separate genus and consequently described Afromantispa. The authors unfortunately did not provide a list of species belonging to the newly erected genus claiming it would be best left until a full revision of the genus could be launched. It appears that this might have caused some additional confusion. Monserrat (2014) synonymised Afromantispa with Mantispa in a study only focussing on the local fauna of the Iberian Peninsula and Balearic Islands. A new species Mantispa incorrupta was also described. Additionally, the author synonymised Sagittalata Handschin, 1959 with Mantispa. The status of Sagittalata is currently still in dispute and not well understood.
Afromantispa, Mantispa and Sagittalata are quite difficult to distinguish, but several morphological characters do support separation of the genera. Adding to the difficulty is that there is a distribution overlap between species from both genera in southern and western Europe. The antennae, prothorax, mesothorax, pterostigma, and fifth tergite are morphologically different between members of the genera (Table 1). Mantispa are represented by only two species that can confidently be placed in the genus (supplementary files: Appendix II). The status of Afromantispa (Snyman & Ohl, 2012) is hereby restored as a genus morphologically distinct from Mantispa, and a list of the species that belong to Afromantispa is provided. This study thus aims to elucidate the boundaries between these two genera.

Material and methods
The specimens used in this study are housed at the following institutions: Photos were taken with either a Canon 500D equipped with a 100mm Canon macro lens or with a Leica Z16 APOA camera setup.

AMG
All type specimens that are not housed at MRAC and ZMB were studied using high resolution photographs provided by ZMB (supplementary files: Appendix IV). Adult morphological terminology follows that of Lambkin (1986a;b).

Morphological overview
Head: The flagella of the genera are quite similar in appearance but all species of Afromantispa have a pale band in the distal third of the flagella, this character is not shared by the species of Mantispa. The band is even distinct in Afromantispa species with light yellowish flagella. In the latter, the band is then presented by a few dark antennules prior to the band so it remains visible ( Fig. 1b; d) (Monserrat 2014: fig. 43). In Mantispa, the occiput is covered by short stout setae dorsolaterally (Fig. 1f ); this feature is not present in the Afromantispa species studied. The rest of the head capsule is very similar between the taxa. The prothorax dorsum of Afromantispa is always covered in granules and setae, where the Mantispa species lack granules, even if small pigmentation "dots" are visible at the origin of the setae on M. styriaca (Fig. 1a-f ) (Monserrat 2014: fig. 23-24). Peculiarly, a region in the lateral mid-zone of the prothorax of Afromantispa always lacks granules ( Fig. 2e; f ) (Monserrat 2014: fig: 45). Mantispa in turn have short stout setae on the dorsum of the mesothorax, which is lacking in Afromantispa (Fig. 1f ). The wing venation of both genera is very similar in structure except for features pertaining to the pterostigma (Fig. 2a-d). The costal space in Mantispa seems slightly larger than in Afromantispa, but it can vary. The pterostigma in Mantispa however, is different. The subcosta and radius of Afromantispa is always pale/yellowish in colouration up to or just distal to midway of radial cell 2. Thereafter, the pterostigma commences. The proximal end of the pterostigma is the same pale colouration of the subcosta and radius veins. The centre of the large distal half is always reddish or dark in colouration flanked by a thin yellowish margin until meeting the veins. The pterostigma of some species might be slightly truncated and anteriorly rounded (Fig. 2b). Mantispa in turn always have a reddish monocoloured pterostigma (Fig. 2c-d). The terminalia of both genera are similar ( Fig. 3a-b) in structure where variation on ectoproct size and length is common in Afromantispa. Both genera have an extrusible gland present between tergite V and VI (Eltringham 1932). Tergite V of Afromantispa is conspicuously enlarged, especially in fresh specimens (Fig. 3c). From various photos of live M. styriaca and M. aphavexelte Aspöck & Aspöck, 1994, including those of Monserrat (2014), it was determined that this tergite is not as prominent in Mantispa.
The elongated line on the procoxa of Sagittalata suggested by Snyman et al. (2012) was considered as a weak character by the authors. Sagittalata lacks the greatly enlarged gland present between the V and VI tergites. In addition, species from Sagittalata have  (Handschin, 1959) b Afromantispa moucheti (Navás, 1925) c Afromantispa nana (Erichson, 1839) d Afromantispa tenella (Erichson, 1839) e Mantispa styriaca f Mantispa aphavexelte (photo credits: a-d Johan Saayman).  fig. 3 H, L)). The cusp on the anterior margin of the pronotum lacks short stout setae which are present in Mantispa. The pronotal dorsum lacks short stout setae that are present in Mantispa, but may have a few sparsely distributed setae. The mesothorax completely lacks prominent setae and is either glaborous or pubescent (velvet appearance) which is also different in Mantispa. Sagittalata might be forming part of the previously synonymised genus Mantispilla, thus, changing the taxonomic status of the genus in this paper only

Discussion
These two genera are possibly quite closely related and therefore present several confusing morphological characteristics. The suggestion by Monserrat (2014) to synonymise the genera has cascading effects on the taxonomy of Mantispinae. Even though western Europe is not specifically rich in mantispid species, these genera are not confined to that area. The suggested synonymy by Monserrat (2014) means that Mantispa will again include 144 species spanning Africa, Europe, Asia and some Australasian islands (numbers from Ohl 2004). This might be possible, but should be approached with caution and include a substantially larger number of species than what was included by the author. Monserrat (2014) further only considered species formally recorded from Spain, where a much larger scope should have been included.
This study confidently presents enough data for the separation of Afromantispa and Mantispa. Current integrative studies including the authors of this study are ongoing focussing on the elucidation of the world genera of the mantispines. the photos used in this publication. Mervyn Mansell is thanked for his willingness to discuss and for his meaningful advice. We are also grateful for all the staff of SANC, MRAC and ZMB who helped with the curation and lending of specimens. Lastly, we would like to acknowledge the NRF as the primary funding body of this study.