Agnocoris rubicundus (Fallen, 1807)

Figs 1B; 2A–D; 3F–J, M; 4E–H; 5A,C; 6F–J.

Lygaeus rubicundus Fallen 1807: 84 (original description).

Agnocoris rubicundus Moore 1956: 37 (new combination).

Slater 1950: 29 (description, key to species); Wagner & Slater 1952: 276 (description); Kelton 1955: 532 (description, key to species); Moore 1955: 180 (description, key to species); Carvalho 1959: 24 (catalogue); Kerzhner & Jaczewski 1964: 723 (description, key to species); Wagner & Weber 1964: 219 (description, key to species); Wagner 1974: 428 (description, key to species); Schuh 1995: 697 (catalogue); Vinokurov & Kanyukova 1995: 93 (key to species); Vinokurov 1979: 96 (key to species); Kerzhner & Josifov 1999: 59 (catalogue); Vinokurov et al. 2010: 77 (catalogue); Wolski & Skora 2012: 6–11 (description, key to species); Lock 2018: 146–148 (description); Vinokurov et al. 2024: 244 (catalogue).

For more references and synonyms, see Carvalho (1959); Kerzhner & Josifov (1999); Schuh (1995), Vinokurov (2010), Wolski & Skora (2012), Vinokurov et al. (2024).

Diagnosis. Body length in male 4.6–5.5, in female 4.9–5.5 (Fig. 2); vertex width / eye diameter in dorsal view ratio 1.15–1.20 in male and 1.4–1.55 in female; antennal segment II / head width ratio 0.9–1 in male and 0.84–0.98 in female (Table 1); in dorsal view, apical process gradually tapering towards apex; left paramere with apical process twice as long as sensory lobe width, sensory lobe ca. 2.4× as long as wide; in posterior view, spike on the apical part of apical process moved towards posterior margin; in left view, right paramere ca. 3× as long as wide; its body ca. 1.5× as wide as apical process, apical process triangular, forming 70° angle with paramere body (Fig. 3F–J); spicule in vesica straight, as wide as ductus seminis, surpassing secondary gonopore at distance subequal to third of spicule length; posterior sclerite only slightly surpassing secondary gonopore, with spines covered only half of its posterior side; secondary gonopore subequal to half of spicule length; plate-like sclerite above secondary gonopore distinctly sclerotized (Fig. 4E–H); distance between sclerotized rings on dorsal labiate plate subequal to half of ring width; sclerotized ring height / length ratio 0.7; sclerite under sclerotized ring oval, its length subequal to half of sclerotized ring width; (Fig. 5A, C), posterior wall of bursa copulatrix with interramal lobes entirely covered with lateral lobe.

Distribution. Agnocoris rubicundus is distributed from Western Europe to East Asia. It inhabits polar regions at least in Europe, in the south its distribution spans to North Africa (Morocco), Near East (Iran), Central Asia and Mongolia. In East Asia, A. rubicundus is known from the Russian Far East, but it was not recorded from Kuril Islands. It was also recorded from the northern regions of China and Japan (Kerzhner 1988; Linnavuori 1992; Kerzhner & Josifov 1999; Linnavuori 2009; Gorczyca & Wolski 2011; Kment & Banar 2012; Kondorosy 2011; Wolski & Skora 2012; Shamsi et al. 2014; Lock 2018; Zamani & Hosseini 2020; Vinokurov et al. 2024). Here we record this species for Iran for the first time. Agnocoris rubicundus is also known from the USA and Canada (Knight 1917, Moore 1956, Wheeler & Henry 1992, Scudder 1997, Hebert et al. 2016, Sikes et al. 2017, Dewaard et al. 2019).

Material examined (see details in the Material examined Data SI2): Afghanistan, Armenia, Azerbaijan, Belarus, Bulgaria, Canada, Estonia, Finland, Georgia, Germany, Iran, Japan, Kazakhstan, Kyrgyzstan, Lithuania, Moldova, Mongolia, Morocco, Poland, Russian Federation (Central, North and South European regions, West Siberia, East Siberia, Far East), Serbia, Tajikistan, Ukraine, Uzbekistan.

Genetic distances. Among the studied markers, COI shows the highest variation. Overall, Agnocoris rubicundus demonstrates higher intraspecific variability in the COI and 16S rRNA markers compared to Agnocoris reclairei.

The mean distance for COI between A. reclairei and A. rubicundus is 1.4%, which is the highest among the markers studied. The mean distances between those two species for 16S rRNA and Ca-ATPase equal 0.4%, and for ITS1 it equals 0.7% (Table SI3). The intraspecies genetic distances within A. rubicundus are 1.5%, 0.5%, 0.12% for COI, 16S rRNA and Ca-ATPase respectively. Specimens of A. rubicundus have identical ITS1 sequences. Agnocoris reclairei shows lower intraspecific variability across most markers, and it equals 0.36% for COI, 0.11% for Ca-ATPase, and all sequences of 16S rRNA and ITS1 are identical within this species. Mean p-distance between Palearctic and Nearctic Agnocoris species for COI is 7.89%.

Phylogenetic relationships between Palearctic Agnocoris species. The topologies are shown in Figs 7–8 and Figs SI1–SI6. Phylogenies based on the mitochondrial markers only are consistent with each other. Phylogenies based on nuclear markers only also show consistent topologies. However, mitochondrial markers only show that A. reclairei do not form a clade, whereas in the phylogeny based on ITS1 and ITS1 + Ca-ATPase this species is monophyletic with the highest supports. Agnocoris rubicundus and A. reclairei are monophyletic in STARBEAST 2 and STARBEAST 3 analyzes based on the dataset with all markers.

СOI + Genbank (Fig. 7). The clade comprising the Palearctic species A. reclairei and A. rubicundus, and the clade with the Nearctic species, are sister groups with high supports at least in the Bayesian analyses (for Nearctic clade: PPSB2 = 1, PPMB=1, BS = 98, for Palearctic clade: PPSB2 = 1, PPMB=1). The Palearctic clade comprises two subclades. One of them includes most A. rubicundus specimens from the central and south European regions, some specimens from the European north and all specimens of A. reclairei (PPSB2 = 1, PPMB = 0.99). All representatives of A. rubicundus from Asia, some specimens from the European north, a specimen from Dagestan and representatives from Nearctic form a second clade (PPSB2 = 0.96, PPMB = 0.87). Within this clade there are two main groups. One of them comprises all specimens from Nearctic, some specimens from the Russian northern regions (Murmansk Province) and some specimens from Asia (Kamchatka Territory Khanty-Mansi Autonomous District) (PPSB2 = 1, PPMB = 0.96, BS = 64). The second group is formed by a single specimen from Dagestan, single specimen from Western Siberia (Khanty-Mansi Autonomous District) and two specimens from European north (PPSB2 =1, PPMB = 0.82). Our results show that the specimens of A. pulverulentus and A. utahensis are not monophyletic, however, these conclusions rely solely on GenBank sequences derived from barcoding studies (e.g., Park et al. 2011; DeWaard et al. 2019) and voucher verification was not performed.

16S rRNA (Fig. SI1). The relationships between the most specimens are unresolved. The specimens from Asia and most specimens from the European north, and specimens from Dagestan form a clade (PPMB = 0.67).

COI + 16S rRNA (Fig. SI2). All A. reclairei and A. rubicundus from Bulgaria and Russian Central, Southern and North part form a clade in STARBEAST3 and ML analyzes (PPSB3 = 1, BS = 81). The specimens from Asia and most specimens from the Russian north, and specimen from Dagestan form a clade (PPSB3 = 0.89, PPMB = 0.96).

ITS1 (Fig. SI4). Agnocoris rubicundus and A. reclairei are monophyletic (PPSB2 = 1, PPMB = 0.99, BS = 100) and (PPSB2 = 1, PPMB = 1, BS = 100 respectively), and there is no any variability within those two clades.

Ca-ATPase (Fig. SI3). Agnocoris reclairei is monophyletic (PPMB = 0.98, BS = 100) and A. rubicundus is not monophyletic. However, the most specimens except for the one from Volgograd Province (G6) form a clade (PPMB = 0.67, BS = 85).

Ca-ATPase + ITS1 (Fig. 8). Agnocoris rubicundus (PPSB3 = 1, PPMB = 0.99, BS = 0.99) and A. reclairei (PPSB3 = 1, PPMB = 1) are monophyletic. Agnocoris reclairei does not form a clade in the RAxML analysis.

All markers (Fig. SI6). Agnocoris reclairei (PPSB3 = 0.99, PPSB2 = 0.79) and A. rubicundus (PPSB3 = 0.79, PPSB2 = 0.52) are monophyletic in STARBEAST2 and STARBEAST3 analyzes, these clades were not recovered in either the MrBayes or RAxML reconstructions.

Wolbachia. We tested 20 samples of Agnocoris for the Wolbachia infection, performing two independent PCR screens for 16S rRNA and Wolbachia surface protein gene (wsp). No products of length 438 bp and 555- to 560-bp-long, corresponding to those markers, were found.

Haplotype networks. The results for the nuclear and mitochondrial markers are different (Fig. 9).

COI (Fig. 9): The haplotype network for COI shows three main groups. The first group includes both A. reclairei and A. rubicundus. Agnocoris reclairei comprises three haplotypes (haplotypes 11, 12, 13). Haplotype 13 is formed by both A. reclairei (Bryansk Province, G18) and A. rubicundus from European countries, as well as northern, central and southern Russia. The second group includes A. rubicundus from the USA, Canada, the Russian Far East, northern Russia. The third group is represented by A. rubicundus from Norway KM287225 (haplotype 1); Khanty-Mansi Autonomous District G16 and Dagestan Republic G8 (haplotype 2).

COI with Nearctic species (Fig. SI7): The haplotype network for COI with Nearctic species shows three main groups. The first and second groups include Nearctic species. The first group (haplotypes 1–6) includes A. pulverulentus and A. utahensis from Canada. The second group (haplotypes 7–10) includes A. pulverulentus from Canada and the USA. The third group (haplotypes 11–22) includes Palearctic species: A. rubicundus from Palearctic and Nearctic and A. reclairei.

16S rRNA (Fig. 9): The haplotype network for 16S rRNA consists of two groups. Agnocoris reclairei have the same haplotype as some A. rubicundus specimens. The first group includes A. rubicundus and A. reclairei from southern, central and northern parts of Russia and Southern Europe (haplotype 1), as well as specimens from Altai Province and Dagestan Republic (haplotype 2) The second group consists of haplotypes 3 and 4, which include A. rubicundus from the Far East and northern Russia (haplotype 3) and exclusively northern Russia (haplotype 4).

ITS1 (Fig. 9): The haplotype network for ITS1 shows two haplotypes, each of the corresponding to either A. rubicundus or A. reclairei.

Ca-ATPase (Fig. 9): Haplotype 1 includes A. rubicundus from Dagestan (G8). Haplotype 2 includes A. rubicundus from northern Russia and the Far East. Haplotype 3 includes all A. reclairei.

Species delimitation. SPEEDEMON analysis for all markers and ITS1 + Ca-ATPase datasets, ABGD, bGMYC, bPTP analysis based on ITS1 and bGMYC analyses based on ITS1 + Ca-ATPase retrieved A. rubicundus and A. reclairei as separate species each (Fig. 8, Fig. SI4). PTP and bPTP analyzes for all markers dataset delineated each specimens as separate species. The bGMYC analysis for the dataset with all markers and PTP analysis for the ITS1 did not split the specimens into species. The PTP and bPTP analyzes based on ITS1 + Ca-ATPase dataset placed all A. reclairei into a single species, while all A. rubicundus specimens were also assigned to one species, except for sample G6, which was placed into a separate group.

Divergence dates. The analysis with fossils showed that Agnocoris and Liocoris diverged ~27.1 Mya (95% highest posterior density (HPD) 11.68–40.87 Mya) (Fig. SI5). This date was used to calibrate the root for the Agnocoris datasets. Analysis of COI dataset suggests that the Nearctic species (A. pulverulentus, A. utahensis) diverged from the A. rubicundus and A. reclairei in Miocene, around 6.8 Mya (95% HPD 4.27–9.6 Mya). According to the analysis based on ITS1 A. rubicundus and A. reclairei diverged in Pleistocene (Calabrian), 0.97 Mya (95% HPD 0.28–1.77 Mya). Based on COI data, these clades diverged approximately 0.52 Mya (Chibanian). According to the analysis based on All markers A. rubicundus and A. reclairei diverged in Late Pleistocene (Chibanian), 0.63 Mya (95% HPD 0.38–0.91 Mya).