Green Warbler Phylloscopus nitidus ringed at Blåvand: molecular confirmation of a Danish first and European eighth vagrant record

Green Warbler Phylloscopus nitidus breeds from northern Turkey eastwards to Iran, and winters in India and southeast Asia (del Hoyo et al. 2006). It reaches marginally into Europe in southern Russia and northern Georgia (del Hoyo et al. 2006). East of its breeding distribution the Greenish Warbler Ph. trochiloides and the Two-barred Warbler Ph. plumbeitarsus display a continuous distribution along a broken ring around the Tibetan Plateau, with several discernible subspecies/populations (Irwin et al. 2001). The relationship between Green Warbler, Greenish Warbler, and Two-barred Warbler is complex (Irwin et al. 2001) and the Green Warbler has traditionally been regarded as a subspecies of Greenish Warbler. It was raised to species status by the British Ornithologists’ Union in 2008 (Knox et al. 2008, BOU 2009), and by the Clements Checklist of Birds of the World in 2009 (Clements et al. 2009), while the International Ornithological Union regarded it as a species already in its first edition (Gill & Wright 2006). The Two-barred Warbler was given species status later (Gill & Wright 2006, Clements et al. 2014). ORNIS SVECICA 27: 121–131, 2017


Introduction
Green Warbler Phylloscopus nitidus breeds from northern Turkey eastwards to Iran, and winters in India and southeast Asia (del Hoyo et al. 2006).It reaches marginally into Europe in southern Russia and northern Georgia (del Hoyo et al. 2006).East of its breeding distribution the Greenish Warbler Ph. trochiloides and the Two-barred Warbler Ph. plumbeitarsus display a continuous distribution along a broken ring around the Tibetan Plateau, with several discernible subspecies/populations (Irwin et al. 2001).
The relationship between Green Warbler, Greenish Warbler, and Two-barred Warbler is complex (Irwin et al. 2001) and the Green Warbler has traditionally been regarded as a subspecies of Greenish Warbler.It was raised to species status by the British Ornithologists' Union in 2008 (Knox et al. 2008, BOU 2009), and by the Clements Checklist of Birds of the World in 2009 (Clements et al. 2009), while the International Ornithological Union regarded it as a species already in its first edition (Gill & Wright 2006).The Two-barred Warbler was given species status later (Gill & Wright 2006, Clements et al. 2014).
Before this first Danish record of Green Warbler there were two records from Fennoscandia and another five from remaining Europe.Two new records were added in 2016, making the total until and including 2016 ten records outside the restricted European breeding range (Table 1).Furthermore, there are seven accepted records in Israel (Slack 2009, IRDC 2016).

Description
On Wednesday 27 May 2015, the daily standardized ringing at Blåvand Bird Observatory could not be done because of strong wind (NW 12 m/s) and rain.Around noon, the wind decreased to 5 m/s the rain stopped, and the sun began to shine.Some of the mist nets were opened and after about one hour, one bird was caught.Surprisingly, this turned out to be a Greenish Warbler type.However, this bird clearly differed from the Greenish Warblers that are caught almost every year at Blåvand.The date was also quite early for Greenish Warbler at Blåvand, as most records of Greenish Warbler at Blåvand are from June.Because of the colouration, the bird was https://doi.org/10.34080/os.v27.19565Table 1.European records outside the peripheral breeding grounds of Green Warbler Phylloscopus nitidus.Europeiska fynd utanför det perifera häckningsområdet av kaukasisk lundsångare Phylloscopus nitidus.*Treatment/decision by national rarities committee.Behandling/beslut av nationell raritetskommitté.**Another record, made 2000-09-27 in Greece, was listed by Hudson (2010), but this has been rejected by the Hellenic Rarities Committee (Nikos Probonas in litt.).Ytterligare ett fynd i Grekland 2000-09-27 rapporterades av Hudson (2010), men detta fynd har underkänts av den grekiska raritetskommittén (Nikos Probonas in litt.).immediately suspected of being a probable Green Warbler.
The bird was measured and photographed (Figure 1).The wing length (maximum chord ;Svensson 1992) was recorded at 61.5 mm, primary 3-4 formed the wing tip, and the tip of the second primary fell at the tip of primary 7.During the process, the bird shed two downs which were collected for DNA typing, as the species determination was not conclusive.After five hours, the bird was relocated and seen clearly in the field.The next day, however, the bird could not be found.
The colouration of the bird was quite different compared to a Greenish Warbler.The upperparts were bright green without the greyish wash, which is typical of Greenish Warbler.The supercilium was yellow, not white, as in Greenish Warbler.The throat and upper breast were yellow contrasting to the white lower breast and belly.The wing bar on the tip of the greater secondary coverts was prominent and broader compared to Greenish Warblers seen in the spring (Figure 1).In the hand, the size and shape of the bird was not perceived as different from Greenish Warbler, but on some photos the head looks larger and more robust than that of Greenish Warbler.The bird was not heard, but when a recording of Green Warbler song was played, the bird duly appeared.Regrettably, no recording of Greenish Warbler was played, so the potential response to that species is unknown.

Methods
The bases of the collected downs were cut off and digested for three hours at 56 °C in 100 μl lysis buffer (0.1 M Tris, 0.005 EDTA, 0.2% SDS, 0.2 M NaCl, pH 8.5) with 1.5 μl proteinase K (10 mg/ml) and then precipitated with ethanol and eluted in 20 μl ddH 2 O.For typing of the mitochondrial cytochrome b (cytb) gene, we used the Qiagen Multiplex PCR Kit (Qiagen Inc.), with amplification reactions containing 5 μl Qiagen Multiplex PCR Master Mix, 0.2 μl each of 10 μM primers ND5-Syl (Stervander et al. 2015) and mtF-NP (Fregin et al. 2009), 2 μl template DNA (non-diluted elution), and 2.6 μl water.We ran the PCR reactions for activation at 95 °C for 15 min.Then followed 40 three-step cycles with denaturation at 94 °C for 30 s; annealing for 20 cycles in a touchdown profile decreasing by 0.5°C per cycle from 55°C, followed by 20 cycles at 45°C for 90 s; and extension at 72 °C for 90 s.Finally, we allowed extension at 72 °C for 10 min.PCR products were checked on a 1% agarose gel, precipitated with NH4Ac and ethanol, and then dissolved in 25 μl water.We used 2 μl for sequencing with the internal sequencing primer Cytb_seq_H15541 (Stervander et al.2015) with the BigDye sequencing kit (Applied Biosystems) in an ABI Prism 3100 capillary sequencer (Applied Biosystems).
Substitution models were evaluated with jMo-delTest v. 2.1.4(Guindon & Gascuel 2003, Darriba et al. 2012), selecting from 88 available models allowing for rate heterogeneity according to four gamma categories and for a proportion of invariable sites.Model selection was performed per the Bayesian Information Criterion (BIC; Schwarz 1978).
Cytb gene trees were computed within a Bayesian inference (BI) framework with BEAST v. 2.4.4 (Bouckaert et al. 2014), using a Yule tree prior, and a strict molecular clock with a rate of 0.0105 substitution/site/lineage/million years, based on overall cytb substitution rates for a wide range of avian species (Weir & Schluter 2008).We applied four discrete categories over the gamma distribution, and estimated the frequency of invariant sites as well as the base frequencies.We performed two replicate runs, and sampled trees every 1,000 generations, over 30 million generations, of which the first 10% were discarded as burn-in.The results were inspected using Tracer v. 1.6 (Rambaut et al. 2013), ensuring stationarity and effective sample sizes (ESS) of >200, and ascertaining congruence between replicate runs.

Results
The trimmed sequence from the Blåvand bird was 681 bp, of which 633 bp covered parts of the cytb gene from its start, preceded by the last 40 bp of the ND5 gene and 8 intergenic bp.The cytb sequence aligned with no gaps to cytb sequences of related species, and is available at Genbank with the accession number MF188243.
The best estimated substitution model was HKY (Hasegawa et al. 1985) with rate variation following a discrete gamma distribution with four rate categories (G) and with an estimated fraction of invariant sites (I).
The computed cytb phylogeny groups the Blåvand bird with full statistical support (posterior probability 1.0) together with two sequences of Green Warblers in a nitidus clade, that is sister to the viridanus clade (Figure 2).Those two clades, in turn, make up as a sister clade to the clade containing the sisters trochiloides and plumbeitarsus (Figure 2).

Discussion
Green Warbler may be an overlooked species in western and northern Europe because it is rather similar to Greenish Warbler.Even with good views the two species may be difficult to distinguish, and good views may be hard to achieve owing to the restless nature of the species and their habit of foraging high above the ground.However, very few Green Warblers have been caught in Europe compared to the number of Greenish Warblers, so the species is probably a quite rare vagrant in Europe.
The phylogenetic analysis clearly confirms that the Blåvand bird is a Green Warbler (Figure 2).It further highlights the special case that the Green/ Greenish/Two-barred Warbler complex constitutes.Under a phylogenetic species concept, a species must be monophyletic, i.e. all of its subspecies/ populations/individuals must be most closely related to each other, and together form a clade that contains no other taxa.However, the cytb phylogeny shows that the two Greenish Warbler clades (trochiloides and viridanus; Figure 2) are not each other's closest relatives, but instead make up sister clades with Two-barred Warbler (plumbeitarsus) and Green Warbler (nitidus), respectively.Thus, judging from the cytb phylogeny alone, Greenish Warbler is paraphyletic, and either all taxa within the complex represent one species, or all should be regarded separate species.
However, while mitochondrial markers such as cytb are appropriate for taxon identification, taxonomic conclusions should not rest solely on such information.Mitochondrial genetic markers differ from nuclear genetic markers because of a faster molecular substitution rate, and because of maternal inheritance.When populations diverge, mitochondrial and nuclear genetic markers may leave different phylogenetic signatures because they sort at different rates.Also, if diverged populations later come into secondary contact, mitochondrial and nuclear genetic markers will be differentially prone to introgression depending on the stage of divergence (Rheindt & Edwards 2011).For that reason, interpretations of phylogenetic relationship should best be based on multi-locus analyses including nuclear genetic markers (e.g.Rheindt & Edwards 2011, Stervander et al. 2015).
Further, the Greenish/Green/Two-barred Warbler clade is unusually well studied, and illustrates some of the problems highlighted above.The Greenish and Two-barred Warbler occur in a broken ring around the Tibetan plateau.The Two-barred Warbler occupies the north-eastern part of the ring, and is isolated with a break in the ring to the south, while it meets the Greenish Warbler in a contact zone to the west.Despite the contact, the two species remain their integrity.The distribution of the Greenish warbler continues counter-clockwise around the Tibetan plateau until the gap the occurs east of the plateau, before reaching the distribution of the Two-barred Warbler.When Irwin at el. (2001) analysed divergence in song and mitochondrial DNA along the broken ring, they concluded that the complex had likely evolved along the ring.Neighbouring populations were only slightly differentiated, but the differentiation progressed on a gradient along the ring, and reproductive isolation was complete in the two end-points, represented by the Siberian populations of Greenish and Two-barred Warbler.However, when Alcaide et al. ( 2014) made use of "next generation sequencing" of massive amounts of nuclear DNA, they revealed that this interpretation was not correct.Instead they uncovered that there have been multiple historical breaks in gene flow, when populations have been isolated at several locations along the ring.Nevertheless, while this species complex may not represent a rare case of a "ring species", it illustrates a complex evolutionary history resulting in varying levels of genetic differentiation between subspecies and species.
Finally, it is worth to mention that the two genera Phylloscopus and Seicercus, which presently constitute the family Phylloscopidae, are paraphyletic (Figure 2; Olsson et al. 2004).The family will therefore likely be split into multiple genera that better reflect their evolutionary relationships (P.Alström pers.comm., Gill & Donsker 2017).

Appendix 1
Taxa and GenBank accession number for cytochrome b sequences included in the phylogenetic analysis.Taxa och GenBank-nummer för de cytokrom b-sekvenser som inkluderats i den fylogenetiska analysen.