In this study two new species of the C. vernale species complex are described, C. montanum sp. nov. and C. orthostylum sp. nov., and C. vernale is re-defined. C. vernale and C. montanum sp. nov. are similar species, while C. orthostylum sp. nov. can be clearly separated from C. vernale and C. montanum sp. nov. based on morphology. An integrative approach (traditional morphology, genetics, geometric morphometry, ecology) was used to distinguish C. montanum sp. nov. from the similar C. vernale.
The main morphological character distinguishing C. montanum sp. nov. from C. vernale is the colour of pile on face, occiput, scutum and scutellum in males which are black in C. montanum sp. nov. (Fig. 8B) and yellow in C. vernale (Fig. 8A). In many syrphid species such as Episyrphus balteatus (De Geer, 1776) and Eupeodes corollae (Fabricius, 1794) colour differences are not a fixed character and depend on the period of adult emergence and on the temperature experienced by the pupa (Dušek and Láska, 1974). It is also known that within certain genera, spring emerging species have overall body colours darker than those of the lighter summer-emerging species (Rotheray and Gilbert, 2011). Genetic, morphometric and ecological evidence obtained here is congruent and indicates that differences between morphotypes A and B of C. vernale (sensu Nedeljković, 2011) are explained by their belonging to different taxa, C. montanum sp. nov. (B) and C. vernale (A).
The European Chrysotoxum species can be grouped according to the proportional length of the basoflagellomere: species with basoflagellomere shorter than scape and pedicel together, Chrysotoxum bicinctum (Linnaeus, 1758), Chrysotoxum elegans Loew, 1841, Chrysotoxum festivum (Linnaeus, 1758), Chrysotoxum gracile Becker, 1921, Chrysotoxum lineare (Zetterstedt, 1819), C. montanum sp. nov., Chrysotoxum octomaculatum Curtis, 1837, C. orthostylum sp. nov., Chrysotoxum parmense Rondani, 1845, Chrysotoxum tomentosum Giglio-Tos, 1890, C. vernale and Chrysotoxum verralli Collin, 1940; species with basoflagellomere longer than scape and pedicel together, Chrysotoxum arcuatum (Linnaeus, 1758), Chrysotoxum cisalpinum Rondani, 1845, Chrysotoxum cautum (Harris, 1776), Chrysotoxum fasciolatum (De Geer, 1776) and Chrysotoxum intermedium Meigen, 1822. Within the group with a relatively short basoflagellomere, Masseti et al. (2006) defined the morphological festivum group, C. bicinctum, C. festivum, C. elegans, C. octomaculatum, C. parmense and C. vernale, excluding characters such as the colour of femur and lateral margin of terga 3 and 4. Based on morphology, C. tomentosum, C. montanum sp. nov., and C. orthostylum sp. nov. are likely to belong to the festivum group (sensu Masseti et al., 2006). These three species have the yellow fasciae on terga 3 and 4 not reaching the lateral margin of the tergum, but the festivum group (sensu Masseti et al., 2006) includes some species with yellow fasciae reaching the lateral margin. Further morphological and molecular analyses are required to access the concept of a festivum group.
Regarding the genetic results of the present study C. montanum sp. nov. was grouped in a separate cluster of the ML tree (Fig. 3A). Analysis showed that COI sequences of mtDNA separated C. montanum sp. nov. and C. vernale. The species C. montanum sp. nov. clustered in a clearly independent position both in the ML tree (Fig. 3A) and MJ network (Fig. 3B). The species C. orthostylum sp. nov. also clustered in an independent position in the ML tree (Fig. 3A), and showed the greatest genetic divergence from C. vernale and C. montanum in the MJ network (Fig. 3). Only one fresh specimen of C. orthostylum sp. nov. was available for molecular study and further specimens must be analysed to determine the phylogenetic position of C. orthostylum sp. nov.
Prior to the present study, the COI region had already been shown as a useful taxonomic tool; for instance, it proved informative for separating C. cautum from C. bicinctum, and also C. parmense and C. fasciolatum, but failed in delimiting species borders between C. festivum and C. elegans (Masseti et al., 2006). Our preliminary investigation on 3’ fragment of COI gene variation in the genus Chrysotoxum (Veličković et al., 2012) showed that the implementation of COI as a molecular marker was found insufficiently precise to identify the differences within vernale complex (morphotypes C. vernale A, C. vernale B) and festivum complex (morphotypes C. festivum A and C. festivum B, and C. elegans). Following these conclusions Nedeljković et al. (2013) have used ITS2 as a marker of choice differences have been found between C. festivum and C. elegans. In the present study we have tested our second idea that the combination of both 5’ and 3’ sequences of COI gene will generate sufficient level of divergence for delimiting putative cryptic species. We have shown that COI was sufficient to separate C. montanum sp. nov. from the similar C. vernale. One of the conclusions could be that there is no a priori chosen single marker that could successfully solve taxonomic problems in Syrphidae. Results of this study show that the COI gene has justified its reputation of a stable barcoding marker, but there have been cases within the genus Chrysotoxum (Masetti et al., 2006), and even in other syrphid genera when the mtDNA COI failed to work well and the sequence divergence was relatively low, and even invariant (Milankov et al., 2008, 2009; Francuski et al., 2011, Haarto and Ståhls, 2014). Therefore an integrative taxonomy approach provides a better method to delimit closely related hoverfly species.
Significant differences in wing size and shape between C. montanum sp. nov., C. vernale and C. orthostylum were detected (Fig. 5A). Inconspicuous but statistically significant differences in surstylar shape between C. montanum sp. nov. and C. vernale were also revealed with a geometric morphometric analysis; this is the second example after that of Nedeljković et al. (2013) in which geometric morphometry of genitalia has successfully contributed to the separation of cryptic syrphid species. In addition, niche divergence analysis reveals that C. montanum sp. nov. is more specialised than C. vernale. Chrysotoxum montanum sp. nov. occurs only at high altitudes with high precipitation, while C. vernale is able to thrive in a wider range of environmental conditions of temperature, precipitation and altitude. The sympatric (e.g. in FYR Macedonia, Golema poljana) and synchronic (early May to mid-August) distributions of C. montanum sp. nov. and C. vernale also support their taxonomic separation.
Animal diversity in the Balkan Peninsula is exceptionally rich and includes many endemic and relict species (Savić, 2008). In this region, insects have experienced an extraordinary adaptive radiation as a result of glacial and inter-glacial periods, particularly in Coleoptera (Mesarosh, 1996), Lepidoptera (Jakšić, 1998) and Diptera, Syrphidae (Vujić et al., 2001). In the present study, integrative taxonomy is confirmed as a useful tool to better understand the entomological diversity in the Balkans and consequently improve the basic knowledge required for species conservation in this exceptional place.