1. Proposed subspecies
   1. Olsson and colleagues’ findings
2. Results
   1. Clade 1, blythi
      1. GAPS IN KNOWLEDGE, RESEARCH GOING FORWARD
   2. Clade 2, halimodendri
   3. Clade 3, margelanica
   4. Clade 4, althaea
   5. Clade 6, minula
3. Votier and colleagues’ findings – no hybrid zone between halimodendri and blythi?
4. Conclusions

Desert Whitethroat (Curruca minula), Hume’s Whitethroat (C. althaea) and Lesser Whitethroat (C. curruca) have been lumped by the IOC just years after their initial split. What new research, if any, supports this decision? And what research justified their initial split in the first place?

This time last year, a sleep deprived me found himself in the central Amir Temur Square of Tashkent, Uzbekistan. I distinctly remember a pair of ‘Lesser Whitethroats’ chattering in the tree above me, but exhaustion from a 14 hour journey from London meant I gave them little notice. Two weeks later, I remember hiking the hills south of Samarkand. Excited by the Eastern Rock Nuthatches just 8 metres in front of me, I hardly registered the ‘Lesser Whitethroats’ in the bush next to them. At the time, these ‘subspecies’ were of no interest to me. Ebird, which hadn’t accepted the split in the first place, didn’t distinguish them from the regular-old Lesser Whitethroat I was used to seeing in England, and I wasn’t aware of the species status that the IOC awarded them at the time. It was only after reading the literature surrounding this complex that I realised that those two sightings could have represented two different new species for me.

The Lesser Whitethroat complex is made up of a number of populations which extend from the British Isles through to Eastern Mongolia and Tibet. Numerous relationships within this complex pose interesting case studies, the morphological similarities and supposed clinal zones between populations ensure that discerning them will be no easy feat. This post aims to synthesise all research surrounding the taxonomy of this group, and subsequent posts will aim to provide insight into identification and morphological comparisons. EDIT: IDENTIFICATION GUIDE TO THE WHOLE COMPLEX OUT NOW: LINK HERE.

Proposed subspecies

The cryptic nature of the complex meant 13 taxa have been proposed prior to phylogenetic analysis:

Taxa	Type Locality	Range (as understood before phylogenetic analysis)	Taxonomy (as understood before phylogenetic analysis)
curruca	Sweden	Europe, Turkey, European Russia east to c. Urals	Thought to have clinal (hybrid) zone with blythi and halimodendri.
minula	Pakistan (on migration)	W China (Xinjiang), Lowland Turkmenistan, Uzbekistan and N Iran	Thought to mix with margelanica at the Qaidam depression, China
althaea	India (on migration)	Mountainous areas in central Asia and possibly E Iran.	Understood to be reproductively isolated from parapatric halimodendri.
margelanica	Uzbekistan (on migration)	Gansu and Qinghai provinces of China	Thought to mix with minula at the Qaidam depression, China
halimodendri	Kazakhstan (lower Tugay river)	Thought to inhabit steppe from Astrakhan to Uzbekistan and east to W Mongolia	Though to clinally transition into blythi and curruca
caucasica	Georgia	Caucasus and adjacent mountains (NW Iran and E Turkey)	considered by some a junior synonym of curruca
zagrossiensis	Zagros mountains, Iran	Zagros mountains, Iran	Considered by some to be a junior synonym of curruca or caucasica. Others considered it a subspecies / junior synonym of althaea
telengitica	Chuya steppe, Russia	Altai mountains, W Mongolia, E Kazakhstan.	Considered by some a junior synonym of halimodendri
turkmenica / snigirewskii	Turkmenistan	Karakum desert area (Lowland central Asia)	(turkmenica was renamed to snigirewskii) Considered by some synonymous with: jaxartica, halimodendri or minula
chuancheica	Qinghai	Qinghai	considered by most a junior synonym of margelanica
jaxartica	Syr’darya river, Kazakhstan	Kazakh and Uzbek steppe	considered by some to be a junior synonym of halimodendri or minula.
blythi	India (on migration)	Siberia	Thought to clinally transition into curruca and halimodendri, considered by some a synonym of curruca
monticola	Tajikistan	Tajikistan	Thought to be a junior synonym of althaea

Hopefully the main take-away from this table is just how confusing, complicated and contradictory the literature was. Practically all taxa had some kind of supposed clinal transition into another, and was proposed to be a synonym of something else. In 2013, Olsson and colleagues aimed to fix this mess through phylogenetic analysis.

Olsson and colleagues’ findings

232 individuals were sampled, many of which were on migration or in wintering grounds. The holotypes for althaea, blythi, minula and margelanica were sampled. For the taxa whose holotypes couldn’t be sampled, samples were taken at the type localities. No samples were found around the type locality of snigirewskii, more on that later.

Results

Before I dive into the findings, it’s worth looking at this map, which reflects both the findings of Olsson and colleagues (2013) as well as Abdilzadeh and colleagues (2020). Votier and colleagues (2016) will also be discussed but does not affect the map per se. All colours and decisions on the map will be justified and explained, but a quick look now may help to picture the rough situation.

6 major clades were identified based upon mtDNA. It is important to note that nuclear loci examined DID NOT discriminate populations. This is important as dates on apparent divergence DOES NOT mean that no gene flow has occurred since. MtDNA is inherited from our mother and different factors can mean that our mtDNA and Y chromosomes (inherited from our father) can have wildly different histories, the same is true for birds.

Firstly, clades 5 and 6 or curruca (pink) and minula (purple) respectively were found to have diverged from the other 4 clades roughly 5.7 million years ago, and to have diverged from each other over 3 million years ago. While, as previously mentioned, this does not mean no gene flow has occurred since, this is still a long period of apparent separation. A 2 million year period of separation is often considered sufficient for bird species splits, hence, disregarding morphology (which will be considered in the next blog, stay tuned), a minimum of three species seems likely so far: Curruca curruca (Clade 5), Curruca minula (Clade 6) and Curruca althaea (Clades 1-4)… more on this later.

Clade 1, blythi

Clade 1, or blythi was found to be closest related to Clade 4 or althaea. This was very surprising considering its practically identical morphology to curruca, as well as its supposed (more on this later) hybrid zone with halimodendri. This suggests that blythi colonised Siberia from the Central Asian mountains and came into contact with curruca at the Urals, as opposed to previously postulated gradual cline between the two. All samples during breeding periods were found in typical boreal forest habitat. However, a few birds caught on migration or in wintering grounds were nestled within a subclade, 1b. The breeding range for this taxa(?) is unknown, although the only bird sampled in summer belonging to this clade was found in Mongolia, as shown by the light blue question marks on the map.

GAPS IN KNOWLEDGE, RESEARCH GOING FORWARD

The newfound phylogenetic distance from curruca means that a comprehensive study into the supposed ‘hybrid-zone’ hypothesised in the Urals is needed. It is unknown where it occurs, and to what degree, if at all, there is gene flow. It is therefore entirely possible that the practically identical morphologies are a result of similar selective pressures either side of the Urals, rather than recent ancestry. More research is certainly needed, but a mtDNA divergence of 5+ million years shows that a hidden cryptic species is possible.

The relationship between blythi and halimodendri was also considered to be clinal, this was adressed by Votier and colleagues, and will be discussed soon.

Clade 2, halimodendri

The wealth of supposed subspecies in the Central Asian steppes and desert was not reflected in the genetic analysis. Samples at the type locality of jaxartica were firmly nestled within the halimodendri clade, whereas no Lesser Whitethroats were found during breeding season at the type locality of snigirewskii. Multiple trips were made to suitable desert habitat in the Karakum desert but nothing was found. Olsson therefore suggests that these collected specimens were migrating birds, and that Lesser Whitethroat don’t actually breed in Central Asian deserts. Staff at a reserve in the area, Reptek, claimed not to be aware of Lesser Whitethroat breeding in the area, nor have there been confirmed breeding reports in suitable habitat in NE Iran. Hence, I have tentatively marked Uzbekistan and Turkmenistan with question marks on the map.

Another lineage within the halimodendri clade, clade 2b, was identified from mtDNA of wintering and migrating birds. The breeding location of this clade is unknown (perhaps it’s in the central Asian desert), although strangely, a bird was recovered from Gansu in China during summer, within the breeding range of margelanica. This is probably(?) a migrant bird, although more surveying in the region should be done to confirm no sympatric breeding with margelanica.

Clade 3, margelanica

Margelanica was found to inhabit much the same range as was previously thought, although a sample from the type-locality of telengitica was firmly within clade 3. It is therefore likely that telengitica is a junior synonym of margelanica. This means that the range of margelanica is larger than previously thought. A bird sampled in eastern Mongolia also belonged to this clade, suggesting that the range possibly extends into Mongolia.

Birds from the type locality of chuancheica were nestled within margelanica as expected, considering their range was within that of margelanica.

Clade 4, althaea

This was the only clade that didn’t change. The distribution matches the published range, and all summer samples at high altitudes fitted within the clade. Although Tajikistan was not sampled, samples from mountains in Afghanistan, Pakistan, Kyrgyzstan and later Iran (by Abdilzadeh and colleagues), surrounding Tajikistan were all part of the same clade. It is therefore highly unlikely that Tajik birds ‘monticola’ constitute a separate population.

The closest relative of this clade was blythi, according to mtDNA. This poses a taxonomic problem as a split of althaea but not blythi would result in a paraspecies. While paraspecies are not per se a bad thing, and are considered a normal part of peripatric speciation, they are typically avoided when it comes to splitting. This is further complicated when we consider that althaea is probably reproductively isolated from halimodendri whereas blythi is thought to hybridise with halimodendri. Hopefully Votier’s research will clear this up…

Clade 5, curruca

As predicted, clade 5 matched the published range of curruca. However, the Caucasian population, caucasica formed a subclade nestled within clade 5. This could be a very early stage of peripatric speciation, although the mtDNA divergence is small and probably doesn’t constitute a legitimate subspecies.

Abdilzadeh and colleagues (2020) sampled birds in Iran, in order to test whether ‘zagrossiensis‘ was a valid taxon. They found that it was nestled within clade 5 and it was likely a junior synonym like caucasica. NE Iran was also searched for halimodendri type birds, none were found, which corroborates suggestions by Olsson (see clade 2).

Clade 6, minula

Clade 6 found that the range of minula, previously considered to extend through Tajikistan into Turkmenistan, is likely restricted to NW China (Xinjiang). Some individuals were found in the Qaidam depression (within the published range of margelanica) and hybridisation is therefore possible. Research is therefore needed to see whether this is the result of sympatric breeding or gene flow.

Votier and colleagues’ findings – no hybrid zone between halimodendri and blythi?

It was assumed that blythi birds gradually getting more halimodendri-like and vice-versa was the result of gene flow. Hence, the mtDNA results from Olsson and colleagues were surprising and warranted a closer look at the relationship between the two populations. It was reported that blythi types are found above 55°N (roughly a line from Yekaterinburg to Novosibirsk) , halimodendri are found below 50°N (south of Astana), and it is believed that the zone inbetween is populated by intermediate birds. Votier and colleagues aimed to verify the existence of this supposed gene flow.

Birds were collected at wintering and migration sights, the morphology was noted and the birds were assigned to either clade (1 or 2) based upon mtDNA. It is worth noting that the actual breeding sights were not visited as they were inaccessible. I would suggest that follow up studies are certainly possible given perfectly accessible cities such as Chelyabinsk, Kostanay and Pavlodar lie within this supposed hybrid zone. Regardless, worn feathers (which would have grown while in the breeding season) were tested for ratios of isotopes: 15N:14N, 13C:12C, 2H:1H and 18O:16O. Essentially, differences in niche, food-sources and environment can lend themselves to different atomic composition. For example, a bird that feeds on herbivorous insects (such as caterpillars) would differ in carbon and nitrogen isotope ratios from a bird that feeds on predatory insects (such as dragonflies). The idea was that birds breeding in the hybrid zone would have intermediate isotope ratios between the two populations if gene flow was indeed occurring, and that isotope signatures from birds in the transition area would not match up to / correlate with their mtDNA haplotype (clade).

H and O isotope signatures are typically markers of environment (precipitation, altitude etc.), but C and N markers can and do often vary between sympatric species given that they have different niches. Interestingly, isotope signatures between the two populations did not overlap. Even blythi birds with morphology more consistent with halimodendri birds had isotope signatures consistent with typical blythi-type birds. The two birds not only likely have different niches, but also niche does not correlate with external morphology. Hence, halimodendri-type morphology may be selected for in intermediate habitat between the boreal forests in the north and steppe habitat in the south. Sympatric breeding between the two populations is therefore possible.

There is therefore no evidence of intermediate birds, and the evidence suggests the two birds occupy different niches. Of course, a lack of sampling in breeding areas means that it is entirely possible that no birds in the hybrid zone were sampled on wintering grounds or migration by chance. However, the fact that many blythi haplotype birds with halimodendri-type plumage were sampled without any sign of intermediate isotope signatures is very promising.

Another interesting finding was a lack of isotope signature difference between minula and margelanica, hence, a comprehensive study into possible hybridisation between the two, particularly in the Qaidam depression, is needed. This could also mean that the two have similar niches and therefore don’t breed sympatrically.

Conclusions

Well. I hope this article has shown just how ambiguous taxonomy can be. What do you think? Are there, 1, 2, 3, 4, 5 or 6 species of ‘Lesser Whitethroat’? One clear omission from this article is morphology and vocalisations which will come in my subsequent blog posts, so stay tuned. Regardless, I hope you now understand why I am now repeatedly kicking myself for not bothering to look at those ‘plain-old Lesser Whitethroats’ in Uzbekistan. Were they migrant blythi, halimodendri or margelanica in Tashkent? Perhaps althaea in the hills south of Samarkand? Or perhaps mystery clade 1b or 2b… I guess I’ll never know.