Insights into phylogenetic relationships between Trioxys Haliday, 1833 and Binodoxys Mackauer, 1960 (Hymenoptera, Braconidae, Aphidiinae), with a description of a new species of the genus Trioxys

ABSTRACT Despite extensive research on the taxonomy and phylogeny of the subfamily Aphidiinae Haliday, 1833, certain questions about the relationships between genera remain unresolved. Genera Trioxys Haliday, 1833 and Binodoxys Mackauer, 1960 are considered closely related, based on morphological and molecular analyses. However, recent studies suggest there is a need for a taxonomic revision of the two genera, since molecular data does not support monophyly of the two groups when a larger number of species is used in the analysis. We examine those relationships using molecular data and including a new species we describe in the present study. Trioxys ulmi Čkrkić & Tomanović, n. sp. is a parasitoid of the Japanese elm aphid (Tinocallis takachihoensis Higuchi, 1972) on elm hybrids (Ulmus x hollandica Mill.). Despite its probable Asian origin, this species has gone undescribed until its accidental introduction to Europe, highlighting the importance of continued research efforts.


INTRODUCTION
Trioxys Haliday, 1833 is one of the largest genera within the subfamily Aphidiinae Haliday, 1833, with currently over 70 species described worldwide (Yu et al. 2016). Classified in the subtribe Trioxina Mackauer, 1961 based on morphology, its main diagnostic characters are accessory prongs on the last abdominal sternite, as well as the absence of secondary tubercles on the petiole (Mackauer 1961). The second character, secondary tubercles on the petiole, is traditionally used to separate Trioxys and the closely related genus Binodoxys Mackauer, 1960. Recent studies employing morphological and molecular data show there is no clear distinction between the two genera, and that there is a need to revise the current status of Trioxys and Binodoxys, or at least some of the species within these two genera (Čkrkić et al. 2019).
Japanese elm aphid, Tinocallis takachihoensis Higuchi, 1972 is an aphid of Asian origin, commonly found on Ulmus spp. in its native area, as well as in Europe, North Africa, and North America, where it has been introduced (Blackman & Eastop 1994). Other than Ulmus spp., it can be found on Hemiptelea davidii (Hance) Planch. in China and eastern Siberia, and it is relatively common on bonsai Zelkova serrata (Thunb.) Makino in England (Blackman & Eastop 1994). The introduction of T. takachihoensis to Europe has been accidental, and it is probable that it came via bonsai plant trading (Petrović-Obradović et al. 2018). While this species is not considered a serious pest, large amount of honeydew deposits can inflict damage on young trees (Petrović-Obradović et al. 2018).
There have been reports of parasitoids from the genus Trioxys parasitizing Tinocallis species (Starý 1978(Starý , 1987(Starý , 1988Lumbierres et al. 2005). However, only one parasitoid (Aphidius sp.) of T. takachihoensis has been recorded so far from Algeria (Hemidi et al. 2013). Here we report a new species of the genus Trioxys, found parasitizing T. takachihoensis on Ulmus x hollandica, diagnosed by morphological differences and molecular data, using the mitochondrial cytochrome c oxidase subunit I (COI) gene. We also discuss the relationship of the new species with other members of Trioxys and with the related genus Binodoxys.

Specimen collection and morphological analySiS
Specimens were collected in New Belgrade (Belgrade, Serbia; GPS coordinates: 44.48003420, 020.21529524), in June 2017. Plant leaves infested with aphids were collected in plastic boxes covered with mesh to allow for ventilation. Samples were kept under laboratory conditions for 3-4 weeks, or until parasitoid emergence. Adult parasitoids were dissected and slide mounted for detailed examination of external morphology, using a ZEISS ® Discovery V8 stereomicroscope (Carl Zeiss MicroImaging GmbH, Göttingen, Germany). Photographs of the slide mounted specimens were obtained using a Leica DM LS phase contrast microscope (Leica Microsystems GmbH, Wetzlar, Germany), and used to measure relevant morphological characters with ImageJ software (Schneider et al. 2012). Morphological terminology follows Sharkey & Wharton (1997). molecular analySiS DNA was extracted from three individual adult parasitoids using the QIAGEN Dneasy® Blood & Tissue Kit (Qiagen Inc., Valencia, CA, USA) following the manufacturer's instructions. The barcode region of the COI gene was amplified using the universal primers LCO1490 and HCO2198 (Folmer et al. 1994. DNA amplification was performed in a 20 μl volume, containing 1 μl of DNA, 11.8 μl of H2O, 2 μl of High Yield Reaction Buffer A with 1 × Mg, 1.8 μl of MgCl2 (2.25 mM), 1.2 μl of dNTP (0.6 mM), 1 μl of each primer (0.5 μM) and 0.2 μl of KAPATaq DNA polymerase (0.05U/μl) (Kapa Biosystems Inc., Boston, USA). PCR amplification was performed in an Eppendorf Mastercycler® (Hamburg, Germany), using the following thermal profile: initial denaturation at 95° C for 5 min, followed by 35 cycles of 94° C for 60 s, 54° C for 60 s, 72° C for 90 s and a final extension at 72° C for 10 min. Purification of PCR products and DNA sequencing in both direction was performed by Macrogen Inc. (Seoul, Korea).
Sequences were edited with FinchTV ver. 1.4.0 (www.geospiza.com). CLUSTAL W algorithm integrated in MEGA X software (Kumar et al. 2018) was used to align sequences. Sequences were trimmed to a length of 638 bp. Newly acquired sequences in this study are deposited in GenBank.
Two additional sequences, identical to those of the new species described here, were acquired from the BOLD database (http://www.boldsystems.org; BIN AAU8586  (Gahan, 1926) andB. centaureae (Haliday, 1833) were used to place the newly discovered species within the genus Trioxys and its closely related genus Binodoxys (Table 1).
Average genetic distances were calculated using Kimura's two-parameter method of base substitution (K2P, Kimura 1980) integrated in MEGA X. Phylogenetic relationships between the new species and related species were explored using Bayesian inference of phylogeny. A phylogenetic tree was constructed in MrBayes 3. diStribution. -The current known distribution of the new species is Serbia, although we suspect a much broader distribution of this   diagnoSiS. -Morphologically most similar to T. complanatus (Tomanović & Kavallieratos 2002). Transverse carinae present on dorsal surface of propodeum (Fig. 2E), irregular postmedian carinae and the beginning of a closed central areola present in some specimens (in T. complanatus transverse carinae sometimes present, but discontinuous; no signs of a central pentagonal areola). Petiole with a slight constriction behind spiracular tubercles (almost parallelsided in T. complanatus).
remark Since the main diagnostic character, the sculpturing of the dorsal surface of the propodeum, varies to some extent in T. ulmi Čkrkić & Tomanović, n. sp., it is advisable to take aphid hosts (T. ulmi Čkrkić & Tomanović, n. sp. is a parasitoid of Tinocallis takachihoensis while T. complanatus mainly parasitizes aphids from the genus Therioaphis Walker, 1870 on legumes) and DNA data into account when identifying this new species.
Fore wing (Fig. 2F). Wing length 1 mm, width 0.4 mm. Stigma triangular, 2.6 times as long as wide and 1.7 times as long as distal abscissa of R1. Wing venation reduced, fused r and RS (r&RS) visible, reaching distally as far as R1 or shorter.
Fore wing (Fig. 3F). Wing length 1 mm, width 0.4 mm. Stigma triangular, 2.6 times as long as wide and 2.1 times as long as distal abscissa of R1. Wing venation reduced, fused r and RS (r&RS) visible, reaching distally as far as R1 or shorter.

Metasoma.
Petiole with prominent spiracles, 1.5 times as long as wide at spiracles. Dorsal disc smooth, with 2 long setae on each side (Fig. 3G).

DISCUSSION
Genera Trioxys and Binodoxys are considered closely related, and were initially considered to be one genus, Trioxys (Mackauer 1959). Mackauer (1960) separated Binodoxys as a subgenus of Trioxys based on the presence of secondary tubercles on the petiole. Since then, Binodoxys has been considered as a separate genus (Mackauer 1961) or as a subgenus of Trioxys (Mackauer 1960). Morphological data seem to support the separation of two groups (Lazarević et al. 2017). Molecular analyses of the phylogeny of Aphidiinae recovered Trioxys and Binodoxys as closely related, but usually with a small number of species used in the analysis (Belshaw & Quicke 1997;Smith et al. 1999;Kambhampati et al. 2000;Sanchis et al. 2000;Shi & Chen 2005;Derocles et al. 2012). Recent studies that focused solely on the two genera, rather than the whole sub- Čkrkić J. et al. family, showed that molecular data is not in concordance with morphological analyses, implying that Binodoxys and Trioxys are not monophyletic groups (Čkrkić et al. 2019;Rakhshani et al. 2020). The results of this study confirm the ambiguities shown by molecular results -while all Binodoxys species cluster together, the positioning of Trioxys species used in the analysis does not support monophyly of this genus. Genetic distances between species of the two genera based on COI are relatively high for Aphidiinae (Table 2), a situation that is becoming increasingly common with new studies of various Aphidiinae genera (Čkrkić et al. 2019, 2020Kocić et al. 2019). In addition, the divergence rates between congeners are not significantly different from those between species designated to different genera, as previously shown (Čkrkić et al. 2019). The grouping of some Trioxys species does not reflect ecological similarities either. Trioxys ulmi Čkrkić & Tomanović, n. sp. and T. pallidus, parasitoids of arboricolous aphids, cluster with T. complanatus, a parasitoid of the spotted alfalfa aphid, Therioaphis trifolii on Medicago and Trifolium (Tomanović & Kavallieratos 2002). These results suggest that a comprehensive revision is needed, where all available species of both genera will be included and subject to an integrative approach. Combining morphology, several molecular markers and biology and ecology of analysed species with thorough sampling for greater resolution will allow for a clearer picture of the phylogenetic relationships between Trioxys and Binodoxys species.
Taxonomic studies of Aphidiinae wasps throughout Europe have been ongoing for decades (Starý 1966a, b;Kavallieratos et al. 2004Kavallieratos et al. , 2016Kos et al. 2012;Žikić et al. 2012). Despite continuous research efforts, new species for science or for European fauna are reported very often (e.g. Tomanović et al. 2009;Petrović et al. 2013Petrović et al. , 2017Žikić et al. 2015), including some Trioxys species (Rakhshani et al. 2017(Rakhshani et al. , 2020Čkrkić et al. 2019). Aphidiinae are often accidentally introduced into new areas due to their small size and intercontinental traffic and trade. Based on the origin of its aphid host, it is probable that T. ulmi Čkrkić & Tomanović, n. sp. also has an Asian origin, and that it was accidentally introduced with T. takachihoensis and bonsai plant material. A European origin is unlikely, given the numerous taxonomic studies conducted in Europe lately, where this species would probably have been recorded by now (Kavallieratos et al. 2004;Žikić et al. 2012;Rakhshani et al. 2017Rakhshani et al. , 2020. Its currently known host, T. takachihoensis is not considered a serious pest, and T. ulmi Čkrkić & Tomanović, n. sp. most likely has little significance in terms of biological control of aphids at this moment. However, there are some concerns that T. takachihoensis may become a pest of ornamental plants in non-native areas, especially in the light of global warming, and that it may adapt to more severe climatic conditions (Kanturski et al. 2018). On the other hand, as a suspected alien species, T. ulmi Čkrkić & Tomanović, n. sp. should be carefully monitored, because there is always a possibility of adaptation to new hosts in non-native areas (Roy et al. 2011;Petrović et al. 2013). Although Hemidi et al. (2013) listed Aphidius sp. as a parasitoid of T. takachihoensis, this finding should be considered with caution, since all known parasitoids of Tinocallis aphids belong to the genera Betuloxys Mackauer, 1960and Trioxys (Starý 1978, 1987, 1988Lumbierres et al. 2005), with the exception of Praon flavinode (Haliday, 1833) (Starý 1987). This study represents the first record of a Trioxys species parasitizing T. takachihoensis. Other Tinocallis species in Europe are parasitized by three Trioxys species -T. curvicaudus Mackauer, 1967, T. pallidus and T. tenuicaudus Starý, 1978(Starý 1988.
These findings highlight the importance of ongoing taxonomic research efforts, since most studies continue to yield new data about species distributions and tritrophic associations with aphids and host plants. Given