Anagrus (Anagrus) atomus ( Linnaeus, 1767 )

Triapitsyn, Serguei V., Rugman-Jones, Paul F., Tretiakov, Peter S., Daane, Kent M. & Wilson, Houston, 2020, Reassessment of molecular and morphological variation within the Anagrus atomus species complex (Hymenoptera: Mymaridae): egg parasitoids of leafhoppers (Hemiptera: Cicadellidae) in Europe and North America, Journal of Natural History 54 (27 - 28), pp. 1735-1758 : 1743-1752

Anagrus (Anagrus) atomus ( Linnaeus, 1767)

Ichneumon atomos Linné [Linnaeus] 1767: 941. Neotype female, designated by Chiappini and Triapitsyn 2007: 2. Type locality (of the neotype): Hågadalen , Uppsala, Sweden.

Anagrus ustulatus Haliday 1833: 346 . Syn. n.

Anagrus atomus (L.): Haliday 1833: 347.

Anagrus spiritus Girault 1911: 209–210 . Synonymy by Triapitsyn 2015: 41.

Anagrus bartheli Tullgren 1916: 8–11 . Questionable synonymy by Graham 1982: 199–200, confirmed by Chiappini 1987: 95.

Anagrus minimus Menozzi 1947: 38–39 . Synonymy by Chiappini 1987: 94.

Anagrus tullgreni Heqvist 1954: 272–273 . Synonymy by Graham 1982: 199–200.

Anagrus hundsheimensis Soyka 1956: 24 . Synonymy by Chiappini 1989: 102–103.

Anagrus lemonicolor Soyka 1956: 24 . Synonymy by Chiappini 1989: 102–103.

Anagrus parvus Soyka 1956: 24 . Syn. n.

Anagrus stammeri 1956: 24 . Synonymy by Chiappini 1989: 102–103.

Anagrus varius Soyka 1956: 24 . Synonymy by Chiappini 1989: 102–103.

Anagrus devius Soyka 1956: 25 . Synonymy by Chiappini 1989: 102–103.

Anagrus gabitzi Soyka 1956: 25 . Synonymy by Chiappini 1989: 102–103.

Anagrus kressbachi Soyka 1956: 25 . Synonymy by Chiappini 1989: 102–104.

Anagrus levis Soyka 1956: 25 . Synonymy by Chiappini 1989: 102–103.

Anagrus debilis Foerster [sensu Soyka]: Soyka [1955] 1956: 25 (misidentification); Chiappini 1989: 104 (= A. (Anagrus) ustulatus Haliday ).

Anagrus proscassellatii Viggiani and Jesu 1995: 95 . Synonymy by Triapitsyn 2015: 41.

Anagrus parvus Soyka : Viggiani 2014: 399 (de facto resurrected as a valid species from the previous synonymy with A. ustulatus by Chiappini 1989, p. 104).

Anagrus (Anagrus) atomus (L.): Chiappini 1989: 102 (redescription, comments).

Anagrus (Anagrus) ustulatus Haliday : Chiappini 1989: 104 (redescription [at least in part], comments); Viggiani 2014: 399 ( A. (Anagrus) ustulatus Haliday sensu Chiappini 1989 ).

Anagrus erythroneurae S. Trjapitzin and Chiappini 1994: 137–140 . Syn. n.

Anagrus lindberginae Nugnes and Viggiani 2014: 66–71 . Syn. n.

Anagrus ustulatus Haliday : Viggiani 2014: 400 (nomen dubium).

Anagrus nepetellae Viggiani and Nugnes in Nugnes et al. 2017: 591–594. Nomen nudum.

Anagrus nepetellae Viggiani and Nugnes 2018: 4 . Syn. n.

Material examined

Anagrus atomus sensu stricto (F4 of female antenna with 1 mps on both antennae; all on slides, UCRC; measured specimens marked by an asterisk). France, Yvelines, Béhoust, 6. i.1987, K. H. Chen, on hornbeam, Carpinus sp. , infested with Typhlocyba sp. [1 female * ( UCRC _ENT 00533569)]. Italy, Friuli-Venezia Giulia, Udine Province, Buttrio, 46°01’15.5’N 13°21’35.9’E, 91 m, F. Pavan, emerged 29.ix–10.x. 2014 in laboratory (Udine) from grape leaves, Vitis vinifera , from a vineyard infested with Empoasca vitis (Göthe) and Zygina rhamni Ferrari [3 females *, molecular vouchers PR15-030, PR15-041, and PR15-043 ( UCRC _ENT 00533572, UCRC _ENT 00533570, and UCRC _ENT 00533571, respectively)]. UK, England: Hampshire Co.: Awbridge, 51°01’25.3”N 1°32’23.2”W, 41 m, 30.viii.2014, S. V. Triapitsyn, sweeping [1 female *, molecular voucher PR15-036 ( UCRC _ENT 00533568)]. Hollom Down, 51°06’52.5”N 1°37’02.3”W, 86 m, roadside of A30 Highway, 30.viii.2014, S. V. Triapitsyn, sweeping [1 female, molecular voucher PR15-040 ( UCRC _ENT 00533560)]. Kent Co., Sevenoaks, viii.2014, A. Polaszek, yellow pan traps in garden [1 female *, molecular voucher PR15-025 ( UCRC _ENT 00533567)]. USA, California, Napa Co.: Near Napa, 38°19’44”N 122°16’06”W, 21 m, collected 16.i.2013, emerged 24.i.2013, H. Wilson, on blackberry, Rubus sp. [1 female *, molecular voucher PR14-015 ( UCRC _ENT 00463763)]. Near St. Helena, River Ranch Farm Workers Housing, 38°30’11”N 122°26’17”W, 57 m, collected 19.iv.2012, emerged 24.iv.2012, H. Wilson, on California buckeye, Aesculus californica [1 female *, molecular voucher PR13-548 ( UCRC _ENT 00463341)]. Near Zinfandel, Quintessa, 38°29’21”N 122°24’53”W, 66 m, collected 16.i.2013, emerged 18. i.2013, H. Wilson, on rose, Rosa sp. [1 female *, molecular voucher PR13-562 ( UCRC _ENT 00463697)]. Long Meadow Ranch Winery, 38°27’58”N 122°27’52”W, 263 m, collected 16. ii.2012, emerged 26.ii.2012, H. Wilson, on Rubus sp. [1 female *, molecular voucher PR13- 545 ( UCRC _ENT 00463058)].

Anagrus erythroneurae (F4 of female antenna without mps; all on slides, UCRC; measured specimens marked by an asterisk). USA, California: Mendocino Co., Fetzer- Hooper Ranch, 38°57’52”N 123°02’14”W, 256 m: collected 15.viii.2014, emerged 20. viii.2014, H. Wilson, on wild grape, Vitis sp. [1 female *, molecular voucher PR15-018 ( UCRC _ENT 00466084)]; collected 15.viii.2014, emerged 22.viii.2014, H. Wilson, on wild grape, Vitis sp. [1 female *, molecular voucher PR15-013 ( UCRC _ENT 00466233)]. Napa Co.: American Canyon, Grgich Hills Estate American Canyon Vineyard, 38°12’34”N 122° 14’27”W, 46 m: collected 16.i.2013, emerged 23.i.2013, H. Wilson, on blackberry, Rubus sp. [1 female, molecular voucher PR13-563 ( UCRC _ENT 00463728)]; collected 3.iii.2013, emerged 8.iii.2013, H. Wilson, on Ceanothus sp. [1 female, molecular voucher PR14-020 ( UCRC _ENT 00464098)]; collected 2.iv.2013, emerged 11.iv.2013, H. Wilson, on mint, Mentha sp. [1 female, molecular voucher PR14-022 ( UCRC _ENT 00464332)]; collected 3. xi.2013, emerged 7.xi.2013, H. Wilson, on Ceanothus sp. [1 female *, molecular voucher PR15-017 ( UCRC _ENT 00465150)]. Near Napa, 38°19’44”N 122°16’06”W, 21 m, collected 16. i.2013, emerged 1.ii.2013, H. Wilson, on Rubus sp. [1 female, molecular voucher PR14-016 ( UCRC _ENT 00463854)]. Near Oakville, Robert Mondavi Winery To Kalon Vineyard, 38° 25’40”N 122°24’03”W, 47 m: collected 4.xi.2012, emerged 7.xi.2012, H. Wilson, on coyote brush, Baccharis pilularis [1 female *, molecular voucher PR13-557 ( UCRC _ENT 00463640)]; collected 18.vi.2013, emerged 20.vi.2013, H. Wilson, on Mentha sp. [1 female *, molecular voucher PR13-565 ( UCRC _ENT 00464586)]. Near Rutherford, Joseph Phelps – Banca Dorada Vineyard, 38°26’59”N 122°25’26”W, 56 m, collected 11.vi.2012, emerged 15. vi.2012, H. Wilson, on Vitis vinifera [1 female *, molecular voucher PR13-552 ( UCRC _ENT 00463438)]. Near St. Helena: 38°30’07”N 122°25’44”W, 64 m, collected 19.iv.2012, emerged 22.iv.2012, H. Wilson, on Baccharis pilularis [1 female * ( UCRC _ENT 00463325)]. River Ranch Farm Workers Housing, 38°30’11”N 122°26’17”W, 57 m: collected 2.ii.2012, emerged 12. ii.2012, H. Wilson, on Rubus sp. [2 females, molecular vouchers PR13-538 and PR13-539 ( UCRC _ENT 00462866 and UCRC _ENT 00462883, respectively)]; collected 16.ii.2012, emerged 20.ii.2012, H. Wilson, on Baccharis pilularis [1 female, molecular voucher PR13- 542 ( UCRC _ENT 00462962)]; collected 4.xi.2012, emerged 10.xi.2012, H. Wilson, on wild grape, Vitis sp. [1 female, molecular voucher PR13-558 ( UCRC _ENT 00463656)]. Near Yountville, Wappo Hill, 38°23’53”N 122°19’37”W, 27 m, collected 14.vi.2012, emerged 24. vi.2012, H. Wilson, on wine grape, Vitis vinifera [1 female *, molecular voucher PR14-014 ( UCRC _ENT 00463611)]. Riverside Co., Temecula, Bella Vista Winery, 33°31’29”N 117° 03’49”W, 428–430 m, 12.ix.2014, S. V. Triapitsyn and H.-y. Hu, from grape leaves (in organic vineyard, Vitis vinifera ) infested with Erasmoneura variabilis (Beamer) : emerged 15.ix.2014 [1 female *, molecular voucher PR15-028 ( UCRC _ENT 00533573)]; emerged 16.ix.2014 [1 female *, molecular voucher PR15-019 ( UCRC _ENT 00533574)]; sweeping [1 female, molecular voucher PR15-038 ( UCRC _ENT 00533566)].

Anagrus lindberginae (on slides, UCRC). Italy, Campania, Napoli Prov., Portici, G. Viggiani, from eggs of Lindbergina aurovittata (Douglas) on Quercus ilex : 21.ii.2006 [1 female, molecular voucher PR13-574 ( UCRC _ENT 00533558)]; 23.ii.2006 [1 female, molecular voucher PR13-575 ( UCRC _ENT 00533559)].

Anagrus parvus (F4 of female antenna without mps; all on slides, UCRC; measured specimens marked by an asterisk). Italy, Campania: Avellino Prov., Domicella, 26.iii.2004, A. Di Luca, from leafhopper egg on blackberry, Rubus sp. [1 female * ( UCRC _ENT 00533575)]. Napoli Prov.: Palma Campania, 40°51’08”N 14°33’26”E, 80 m, collected 2. ix.2014, emerged in laboratory at Portici 4.ix.2014, S. V. Triapitsyn, from wine grape, Vitis vinifera , leaves infested with Empoasca vitis (Göthe) [1 female, molecular voucher PR15- 033 ( UCRC _ENT 00533562)]. Portici, A. Di Luca: 17.ii.2006, from leafhopper egg on Rubus sp. [1 female *, molecular voucher PR13-576 ( UCRC _ENT 00533579)]; 22.ii.2006, from leafhopper egg on Rubus sp. [1 female *, molecular voucher PR13-577 ( UCRC _ENT 00533578)]; 12.ii.2007, from egg of Ficocyba ficaria (Horváth) on Lonicera sp. [1 female *, molecular voucher PR13-572 ( UCRC _ENT 00533576)]; 17.iv.2007, from leafhopper egg on Rubus sp. [1 female *, molecular voucher PR13-579 ( UCRC _ENT 00533577)]. San Giorgio a Cremano, 40°50’27”N 14°21’08”E, 88 m, collected 2.ix.2014, emerged in laboratory at Portici from wine grape, Vitis vinifera , leaves infested with Empoasca vitis (Göthe) , S. V. Triapitsyn: 3.ix.2014 [1 female *, molecular voucher PR15-024 ( UCRC _ENT 00533580)]; 4.ix.2014 [1 female *, molecular voucher PR15-022 ( UCRC _ENT 00533581)]. UK, England, Kent Co., Sevenoaks, viii.2014, A. Polaszek, yellow pan traps in garden [3 females *, molecular vouchers PR15-029 ( UCRC _ENT 00536184), PR15-032 ( UCRC _ENT 00533582), and PR15-034 ( UCRC _ENT 00533583)].

Anagrus ? atomus (F4 of female antenna with 1 mps on one antenna only but not on the other; all on slides, UCRC). Italy, Campania (locality unknown), 2006 or 2007, G . Matteucig , from leafhopper egg on Silene sp . [1 female, molecular voucher PR13-570 ( UCRC _ ENT 00533561 )] . USA, California, Napa Co .: American Canyon, Grgich Hills Estate American Canyon Vineyard , 38°12’34”N 122°14’27”W, 46 m: collected 16 .i .2013, emerged 2 .ii .2013, H. Wilson , on mint, Mentha sp . [1 female, molecular voucher PR14-017 ( UCRC _ ENT 00463865 )]; collected 16 .i .2013, emerged 19 .i .2013, H. Wilson , on catnip, Nepeta sp . [1 female, molecular voucher PR13-561 ( UCRC _ ENT 00463705 )]; collected 16 . i .2013, emerged 30 .i .2013, H. Wilson , on Nepeta sp . [1 female, molecular voucher PR13-547 ( UCRC _ ENT 00463840 )]; collected 9 .ii .2013, emerged 24 .ii .2013, H. Wilson , on Nepeta sp . [1 female, molecular voucher PR14-018 ( UCRC _ ENT 00463990 )] . Near St . Helena, River Ranch Farm Workers Housing , 38°30’11”N 122°26’17”W, 57 m, collected 7 .v .2012, emerged 11 . v .2012, H. Wilson, on California buckeye, Aesculus californica [1 female, molecular voucher PR13-551 ( UCRC _ ENT 00463402 )] .

Anagrus ? erythroneurae (F4 of female antenna without mps; all on slides, UCRC). USA, California: Mendocino Co ., near Old Hopland, Campovida , 38°58’07”N 123°05’54”W, 151 m: collected 4 .x .2013, emerged 5 .x .2013, H. Wilson , on rose, Rosa sp . [1 female, molecular voucher PR14-025 ( UCRC _ ENT 00465010 )]; collected 8 .xi .2013, emerged 12 . xi .2013, H. Wilson , on Rosa sp . [1 female, molecular voucher PR15-002 ( UCRC _ ENT 00465215 )]; collected 10 .xii .2013, emerged 18 .xii .2013, H. Wilson , on Rosa sp . [1 female, molecular voucher PR15-005 ( UCRC _ ENT 00465281 )]; collected 17 .ii .2014, emerged 26 . ii .2014 and 28.ii.2014, H. Wilson , on Rosa sp . [2 females, molecular vouchers PR15-012 and PR15-006 ( UCRC _ ENT 00465377 and UCRC _ ENT 00465392 , respectively)] . Napa Co., American Canyon, Grgich Hills Estate American Canyon Vineyard , 38°12’34”N 122° 14’27”W, 46 m, collected 16 .i .2013, emerged 17 .i .2013, H. Wilson , on catnip, Nepeta sp . [1 female, molecular voucher PR13-560 ( UCRC _ ENT 00463694 )] . Sonoma Co.: near Healdsburg, Simi Chalk Hill , 38°38’05”N 122°45’51”W, 57 m: collected 1 .iii .2012, emerged 8 .iii .2012, H. Wilson , on alder, Alnus sp . [1 female, molecular voucher PR13-544 ( UCRC _ ENT 00464116 )]; collected 2 .iii .2013, emerged 9 .iii .2012, H. Wilson , on Alnus sp . [1 female, molecular voucher PR14-021 ( UCRC _ ENT 00464116 )] . Medlock-Ames Vineyards , 38° 36’11”N 122°45’11”W, 95 m, collected 19 .iv .2012, emerged 23 .iv .2012, H. Wilson , on coyote brush, Baccharis pilularis [1 female, molecular voucher PR13-546 ( UCRC _ ENT 00463334 )] .

Anagrus ? erythroneurae (F4 of female antenna with 1 mps on one antenna only but not on the other). USA, California, near St . Helena, River Ranch Farm Workers Housing , 38° 30’11”N 122°26’17”W, 57 m, collected 7 .x .2012, emerged 11 .x .2012, H. Wilson , on wild grape, Vitis sp . [1 female, molecular voucher PR13-556 ( UCRC _ ENT 00463630 )] .

Anagrus ? parvus (F4 of female antenna without mps; all on slides, UCRC). Italy, Campania: Avellino Prov., Domicella, A. Di Luca: 15.iii.2004, from leafhopper egg on Rubus sp. [1 male, molecular voucher PR13-578 ( UCRC _ENT 00533565)]; 4.iv.2004, from leafhopper egg ( Edwardsiana sp. or Ribautiana sp. ) on Rubus ulmifolius [1 male, molecular voucher PR13-568 ( UCRC _ENT 00533564)]. Benevento Prov., Torrecuso, 13.ix.2005, G. Viggiani, from leafhopper egg on Vitis vinifera [1 female, molecular voucher PR13-580 ( UCRC _ENT 00533563)]. USA, California, Napa Co., American Canyon, Grgich Hills Estate American Canyon Vineyard, 38°12’34”N 122°14’27”W, 46 m, collected 16.i.2013, emerged 25.i.2013, H. Wilson, on catnip, Nepeta sp. [1 female, molecular voucher PR13-564 ( UCRC _ENT 00463786)].

Anagrus ? parvus or? erythroneurae (F4 of female antenna without mps). USA, California, Napa Co ., American Canyon, Grgich Hills Estate American Canyon Vineyard , 38°12’34”N 122°14’27”W, 46 m, collected 2 .iii .2013, emerged 7 .iii .2013, H. Wilson , on mint, Mentha sp . [1 female, molecular voucher PR14-019 ( UCRC _ ENT 00464064 )] .

Anagrus sp. ( atomus species complex). USA, California, Sonoma Co ., near Healdsburg, Simi Chalk Hill , 38°38’05”N 122°45’51”W, 57 m, collected 1 .iii .2012, emerged 6 .iii .2012, H. Wilson , on alder, Alnus sp . [1 male, molecular voucher PR13-543 ( UCRC _ ENT 00463159 )] .

Other material examined

All on points [ UCRC]. Anagrus ? atomus : Italy, Campania, Napoli Prov., Portici, A. Di Luca, from eggs of Ficocyba ficaria (Horváth) on Lonicera sp. : 12.ii.2007 [1 female]; 17.iv.2007 [5 females]. Anagrus lindberginae : Italy, Campania, Napoli Prov., Portici, G. Viggiani, from

eggs of Lindbergina aurovittata (Douglas) on Quercus ilex : 21.ii.2006 [1 female]; 27.ii.2006 [1 female]. Anagrus parvus : Italy, Campania: Avellino Prov., Domicella, 23.ii.2004, A. Di Luca, from leafhopper eggs on blackberry, Rubus sp. [1 female, 1 male]. Napoli Prov., Portici, 17.iv.2007, A. Di Luca, from leafhopper eggs on Rubus sp. [2 males]. Anagrus ? parvus : Italy, Campania, Napoli Prov., Portici, A. Di Luca, from eggs of Ficocyba ficaria on Lonicera sp. : 12.ii.2007 [4 females]; 17.iv.2007 [4 females]. Anagrus spp. (either A.? atomus or A.? parvus ): Italy, Campania, Napoli Prov., Portici, 12.ii.2007, A. Di Luca, from eggs of Ficocyba ficaria on Lonicera sp. [9 males].

Updated diagnosis

Female. Antenna with scape (including radicle) 2.4–6.0× as long as wide; mps on F4 (0 or 1, sometimes on one antenna only); F5 (usually 1, rarely 2), and F6 (usually 2, rarely 1); clava with 3 mps. Midlobe of mesoscutum without adnotaular setae. Fore wing 6.1–- 11.6× as long as wide, its disc with a distinct bare area at posterior margin. Ovipositor 1.2–2.5× length of protibia.

Distribution

Afrotropical region: Cape Verde Islands, Yemen; Australasian region: New Zealand; Nearctic region: Canada, Mexico, USA; Neotropical region: Argentina, Chile; Oriental region: China, Pakistan; Palearctic region: Austria, Belgium, Bulgaria, China, Croatia, Denmark, Egypt, Finland, France, Germany, Greece, Iran, Ireland, Israel, Italy, Kyrgyzstan, North Macedonia, Moldova, Netherlands, Poland, Republic of Korea, Romania, Russia, Serbia, Spain, Sweden, Switzerland, Turkey, Turkmenistan, UK ( Triapitsyn 2015; Noyes 2019 [also as A. erythroneurae , A. parvus and A. ustulatus ]), and Japan ( Triapitsyn et al. 2020).

Hosts

Various Cicadellidae , mainly in the subfamily Typhlocybinae , as listed by Chiappini et al. (1996), Triapitsyn (1998), Triapitsyn and Berezovskiy (2004), Matteucig and Viggiani ([2008] 2009), Nugnes and Viggiani (2014), Triapitsyn (2015), Nugnes et al. (2017), and Noyes (2019) for A. atomus and its synonyms proposed herein. Some of these records probably require confirmation (particularly those prior to the 1990s), as both the host leafhoppers and their egg parasitoids could be misidentified.

Biology

Some biological traits of A. atomus were studied in Europe by MacGill (1934) and Agboka et al. (2003, 2004)), and also in Iran by Hesami et al. (2004). This species is a solitary, proovigenic, primary egg parasitoid reproducing by arrhenotoky.

Remarks

Results of the genetic analyses had a key role in justifying the proposed new synonymies because morphometry, based on some key female diagnostic characters of the nominal species treated herein, would not separate them conclusively ( Table 1), with broad overlaps among the nominal species. See ‘Genetic analyses’ and ‘Discussion’ sections below for the justification of these synonymies. More detailed morphometric analyses of the nominal species within the Anagrus atomus complex were also given in Nugnes and Viggiani (2014) and Nugnes et al. (2017); illustrations of A. atomus were provided by Chiappini (1987).

The proposed synonymy of A. ustulatus , as treated by Viggiani (2014), with A. atomus is obvious, as the lectotype male of this nominal species could only belong to either A. parvus or A. atomus , which are considered here to be conspecific.

Interestingly, Anagrus specimens reared in northern California from leafhopper eggs on Mentha sp. and Nepeta sp. (Lamiaceae) and mostly tentatively identified morphologically as A.? atomus , A.? erythroneurae or A.? parvus (only two, PR13-565 and PR14-022, as a more or less definite A. erythroneurae ), cluster together in our ITS2 analyses ( Figures 2 and 3 ) along with one specimen of A. parvus from Sevenoaks, England, UK (PR15-034); the latter was collected in a yellow pan trap in a private garden in which members of this plant family could be present. However, in our COI analysis ( Figure 1 ) they do not come as A. nepetellae , which is associated with the leafhoppers feeding on various Lamiaceae in Italy ( Nugnes et al. 2017), but rather with the broader ‘ A. atomus ’ (sensu stricto) cluster mentioned below under ‘Genetic analyses’. In these specimens, the scape of female antenna, including radicle, is 3.6–5.3× as long as wide (mean 4.3, standard deviation 0.6, n = 9).

Genetic analyses

The present study obtained COI sequences from 59 specimens . In 15 cases the sequence contained a number of ambiguous base calls, typically 1–3. However, the two specimens identified morphologically as A. lindberginae , which were reared at the type locality from the same host leafhopper as its type series, each had 11. The DNA from all 15 specimens was re-amplified and re-sequenced confirming these ambiguous base calls (data not shown). Considering the dataset as a whole, in all but two instances, ambiguous bases were always found at polymorphic positions, characterised by the unambiguous alternatives, across the dataset.

The COI and ITS2 sequences provided little evidence to validate the current set of nominal species within the atomus species complex. Average genetic divergence among 61 ‘ingroup’ COI haplotypes was less than 3% (uncorrected p-distance from averaging over all sequence pairs was 0.026 ± 0.005). In the ML analysis, the COI haplotypes formed three loose genealogical clusters (or clades) ( Figure 1 ). Respectively, the three clusters contained the majority of specimens identified morphologically as A. erythroneurae , A. parvus , and A. atomus . However, a number of specimens identified morphologically as each of these species were also grouped with the ‘wrong’ cluster ( Figure 1 ). Furthermore, support for the ML tree was generally very poor, although the ‘ A. atomus ’ (sensu stricto) cluster was well-supported, and>3% divergent from the other clusters ( Table 2). Anagrus nepetellae clustered with ‘ atomus ’, but average divergence within that cluster was only 1.7% ( Table 2). The previously published A. lindberginae haplotype ( KX691520 ) also fell into the ‘ atomus ’ clade but two additional specimens included in this study clustered with ‘ erythroneurae ’. The ABGD method was even less convincing. Running the software with the default gap-width setting (X = 1.5) resulted in a partition in which all 63 haplotypes (including the outgroup) were considered as a single group. Changing X to values between 0.5 and 1.0 consistently yielded one (everything), two (outgroup and ingroup), four (see below), or improbable 35 groups, according to the descending levels of prior maximal intraspecific divergence tested by the software. Four groups seemed to be a realistic number, but again this turned out to be largely uninformative. After the barcode gap, corresponding to a prior intraspecific divergence of 0.0046, the 4 groups recognised by ABGD were: (1) the outgroup A. japonicus ; (2) a single A. atomus haplotype, corresponding to KM 677236 in Figure 1 ; (3) another A. atomus haplotype, corresponding to KX691548 , KX691549 , KX691550 , KX 691551 in Figure 1 ; and (4) everything else.

Sequences of the nuclear ribosomal ITS2 proved equally vague at reflecting (morphological) species boundaries. Almost half of the 56 ‘ atomus ’ group specimens examined (excluding outgroups) required cloning in order to ‘extract’ their ITS2 sequences. An NJ analysis of these cloned sequences, alongside the direct-sequenced individuals, again generally ‘scattered’ the morphologically identified specimens across multiple clusters ( Figure 2 ). The different cloned sequences from several individuals were also often ‘scattered’ indicating extensive intraspecific variation in ITS2 (e.g., PR15-005, PR15-032, and PR15-040; Figure 2 ). Interestingly, the two specimens for which species identity could not be resolved based on morphology displayed the extremes of this intraspecific variation. One, PR14-019, was successfully direct-sequenced suggesting zero intraspecific variation in the ITS2. The other, PR15-012, yielded three very diverse clones, spread across at least two clades ( Figure 2 ). Furthermore, there was little consensus in the grouping of our specimens between the COI and ITS2 sequences ( Figure 3 ).

Figures

Figure 1. Relationships among species of the Anagrus atomus complex, inferred from ML analysis of a 345bp sequence of the COI gene. Analyses were conducted in RAxML with the GTR + Γ + I model and data partitioning by third codon position. Support for the major branches, expressed as a percentage, was assessed with 1000 rapid bootstrap replicates. Scale bar represents substitutions per nucleotide site. PR numbers relate to GenBank accessions MW012433–MW012491.

Figure 2. Inter- and intragenomic relationships among the ITS2 rRNA sequences of 56 Anagrus ‘atomus’ complex specimens. The analysis involved 29 direct-sequenced specimens and 69 cloned sequences from a further 27 specimens. Circular mid-point rooted neighbour-joining tree inferred from pairwise uncorrected p-distances calculated in MEGA v.6.06. Branch support>50% (assessed with 1000 bootstrap replicates) is shown next to the branches. Intragenomic variation in the placement of just three individuals morphologically identified respectively as A. atomus, A. erythroneurae, and A. parvus are highlighted with dashed lines (3 cloned sequences for each specimen). Two morphologically unresolved specimens are further highlighted in bold. PR numbers relate to GenBank accessions MW012950–MW013047.

Figure 3. Comparison of the clustering of A. ‘atomus’ complex specimens using ITS2 (tree) and COI (colouring).