Ring around a rosella

The rich scarlet and blue plumage of the crimson rosella (Platycercus elegans) is a familiar sight to Australians living within about 200km of the coast in a 2200km arc extending south-east Queensland to Kingston SE, in South Australia.

But not all members of the species are crimson. Around Adelaide, in the western part of its range, P. elegans is orange, while Murray Valley populations are predominantly yellow.

Noting the discontinuous ranges of the of the three colour forms in 1955, the late British evolutionary biologist Professor Arthur Cain proposed that P. elegans represented a rare example of a ring species.

A ring species typically occupies an elongated, sometimes discontinuous range, roughly in the shape of a chain, ring or a spiral. Ring species represent a special case of a cline – in a typical cline, a single species occupies a more or less continuous geographic or altitudinal range, and gene flow can occur between adjacent populations along its length.

Over time, geographic or climatic barriers may fragment the cline into separate populations, with limited gene flow between neighbouring populations.

As the races adapt to environment or climate change, the original cline can bend back on itself, forming a loop. Where the loop comes together, the populations may have accumulated sufficient genetic, physical or behavioural differences to prevent them interbreeding. Their reproductive isolation effectively makes them distinct species.

A recent collaborative study involving CSIRO Sustainable Ecosystems, Deakin University, the South Australian Museum and the University of Bristol has used a combination of modern analytical techniques to tested Cain’s hypothesis, including analysing the birds’ mitochondrial and nuclear DNA.

The team published its results in the Proceedings of the Royal Society B in June. Their conclusion: the P. elegans complex does not conform to the classical concept of a ring species – although it may have formed a ring in the prehistoric past.

The study has cast new light on the forces that created the disjunct ranges of the three colour forms of P. elegans.

Dr Gaynor Dolman of CSIRO’s Australian National Wildlife Collection says the crimson rosella is an iconic Australian bird, and as a putative ring species, offers a rare opportunity to study how species diverge from a common ancestor.

Ring species are uncommon. Examples include the Ensatina salamander species complex in Oregon and Washington, on the US west coast; Larus gulls, which form a complex that form a circumpolar ring around the land masses of the sub-Arctic region; and sub-species of the greenish warbler, Phylloscopus trochiloides, which form a ring encompassing north-east Europe and northern Asia.

---PB--- Flying distance

The crimson rosella sub-species complex forms a “d’” shape; the stem extends down the Great Dividing range from far north Queensland, then curves west through Victoria and South Australia to the Flinders Ranges, before looping back on itself through the Murray Valley to the western slops of the Great Divide.

Two disjunct populations of the typical crimson colour form occur between Mackay and Townsville, and between Ingham and Cairns, and the crimson’s range becomes more or less continuous from south-east Queensland to Kingston SE, in South Australia. Another isolated population of crimsons populates Kangaroo Island, 200km north-west of Kingston.

On the Fleurieu Peninsula, a 15km sea gap from Kangaroo Island, P. elegans has orange plumage. The orange Adelaide form extends north through Adelaide and on into the Flinders Ranges, on the edge of the arid zone.

East of the Adelaide Hills, a third predominantly yellow race occupies a range extending from around Waikerie, in the South Australian Riverland, and follows the Murray Valley east the western slopes of the Great Dividing Range. The yellow race also extends some distance into the riparian woodlands of the Murray’s two main tributaries, the Darling and Murrumbidgee rivers.

At the easterly limit of its range, in the headwaters of the Murray, the inland form of P. elegans comes within flying distance of the Great Divide’s resident crimson rosellas. The yellows and crimsons are separated by a small region populated by orange rosellas, very similar to the Adelaide form.

Dolman says the three colour forms in south-eastern Australia appear to have diverged from each other relatively recently, during the Pleistocene (1.8 million to 10,000 years ago). They appear to have arisen through a combination of genetic drift during geographic isolation, and environmental selection.

“It’s pretty amazing stuff, and we have been shaking our heads trying to work out what happened,” she says. “We believe our work disproves the hypothesis that the species’ distribution today represents a ring, although it’s difficult to be certain because of the recent influence of climate change.”

However, Dolman says the study could not rule out the existence of a ring complex in the prehistoric past, or the possibility that the orange rosellas are the product of relatively recent, secondary contact between adjacent but long-separated populations. The P. elegans complex could simply be a product of allopatric speciation.

This occurs when a species range becomes fragmented, restricting gene flow as the isolated populations undergo genetic drift and are subjected to local selection pressures, including sexual selection.

The study provided evidence that the orange Adelaide population of P. elegans arose through secondary contact between the crimson and yellow populations. The orange race appears to have been replicated through contact between the yellow and crimson races in the Great Divide, around the headwaters of the Murray River.

The study concluded, largely on genetic evidence, that P. elegans underwent a population expansion during the Pleistocene epoch that was accompanied by a range expansion in south-eastern Australia.

Then a combination of fluctuating sea levels, and landscape and habitat changes, appears to have fragmented the species’ extended range.

---PB--- Sudden geological uplift

The most significant influence in fragmenting the rosella’s range may have been an epochal geological event around 3.2 million years ago that altered the ancient course of the Murray River, creating a 33,0000 square kilometer megalake, Lake Bungunnia.

Lake Bungunnia existed until about 700,000 years ago, and covered a huge area of low-lying landscape straddling South Australia, south-western NSW and a large part of the western half of Victoria.

The modern Murray River trends roughly northwest from its headwaters in the Great Divide to Morgan, in South Australia, where it abruptly turns south and flows through Murray Bridge to its terminal lakes, Lake Albert and Lake Alexandrina, and flows out to sea near Goolwa.

A huge alluvial fan on the seabed in Spencer Gulf shows that the Murray flowed into the Gulf before the sudden geological uplift of the Pinnaroo Block stopped its course around three million years ago.

At their greatest extent, the shallow waters of Lake Bungunnia extended from the Chowilla Plain east to Robinvale and Sea Lake – Lake Tyrrell is a remnant of the megalake – and as far south as the Grampians range.

Freshwater sediments on the bed of the lake formed a thick, impermeable layer of clay, known as the Blanchetown Clay, which today confines shallow, hypersaline aquifers that are a threat to agriculture in the region.

Eventually, some 700,000 years ago, a stream running along the base of the eastern face of the Pinnaroo Block cut through the dam and cut a pathway to the ocean at Goolwa, allowing Lake Bungunnia’s waters to drain away, and removing the geographic barrier to contact between the crimson and yellow P. elegans populations.

Lake Bungunnia’s location overlaps with the discontinuity that separates the southerly crimson race of P. elegans from the inland, yellow race, suggesting it could have been responsible for fragmenting the ancient range of P. elegans, while drainage of the lake 700,000 years ago may have allowed the yellow and crimson races to re-establish genetic contact and form a hybrid race – the orange-hued Adelaide race.

The presence of orange-hued rosellas around the headwaters the Murray and Murrumbidgee rivers in the Great Dividing range points to similar contact and hybridisation between the yellow and crimson races at the eastern margins of the yellow rosella’s range.

---PB--- DNA support

Microsatellite studies of nuclear and mitochondrial DNA (MtDNA) tell an intriguing tale that broadly supports these scenarios, but also show that geographic gaps in the modern range of the three races do not necessarily coincide with genetic discontinuities, suggesting the ranges of the crimson, yellow and orange forms have contracted and expanded in response to climate change during the past 700,000 years.

Microsatellites circulate in populations like alleles of genes; patterns of microsatellites at multiple loci distinguish populations that have been genetically isolated over long time periods, and also reveal genetic contact between different populations. MtDNA is maternally transmitted, and hybridisation events can be inferred from the two types of data.

Dolman and her colleagues have shown that, genetically speaking, what you see is not necessarily what you get. A phenotypically yellow rosella may carry microstallite sequences from a crimson population, or vice versa.

For example, at Griffith, in the NSW Riverina, phenotypically yellow rosellas have a matching yellow genotype, as revealed by microsatellite DNA. But a short flight to the south-east, the phenotypically yellow rosellas carry some microsatellite DNA sequences characteristic of the eastern population of crimson rosellas that inhabit the Great Divide.

In the higher country to the east, around Gundagai, on the Murrumbidgee River, and around the Hume Dam, on the Murray, the rosellas have an orange phenotype, but carry microsatellite sequences from eastern crimsons.

The genetic data reveal a complicated pattern of contact and hybridisation, and three distinct genetic discontinuities, whereas the ring hypothesis predicts only one region of discontinuity where the ring has closed, on the western slopes of the Great Divide, where the genetically incompatible populations should coexist without hybridising.

Instead, there is a genetic discontinuity further west, in the yellow population; on the western slopes the phenotypically orange population has a microsatellite signature intermediate between yellow and crimson, confirming hybridisation has occurred between them.

There is another discontinuity between the orange Adelaide rosellas and the most westerly population of yellows, in the SA Riverland. The third discontinuity occurs at the 15km gap between Kangaroo Island’s disjunct population of crimsons, and the orange rosellas of the Fleurieu Peninsula.

---PB--- If not a ring?

If not a ring, then what? The researchers suggest several alternative hypotheses, based on mtDNA and microsatellite data.

One is incomplete ring speciation. Ancestral southern crimson and northern yellow populations underwent genetic differentiation in different habitats, before both spread westwards during the last glacial period – crimsons in the south, yellows in the north.

Mainland crimsons and Kangaroo Island crimsons hybridised when low sea levels (120m below present) reconnected Kangaroo Island with the mainland. Secondary contact at the western margin of the yellow rosella’s range then gave rise to the orange Adelaide phenotype, which has Kangaroo Island microsatellite DNA.

The second hypothesis concerns allopatric differentiation with secondary hybridisation, which predicts that the Adelaide and Western Slopes orange phenotypes should be genetic admixtures of the crimson and yellow races. MtDNA shows the Western Slopes oranges fit this pattern, but the microsatellite DNA shows the oranges are genetically continuous with the Great Divide crimsons.

In South Australia, the orange plumage of the Adelaide rosella suggests hybridisation between crimsons and yellows, but both MtDNA and microsatellite evidence shows the Adelaide form has more in common genetically with crimsons than yellows. The authors speculate that the range now occupied by the oranges may never have been occupied by either crimsons or yellows – early hybrids may have established their own environmental niche.

The final hypothesis is in situ vicariant differentiation, which predicts that all of today’s Adelaide, crimson, yellow and Western Slopes populations evolved in allopatry – in geographic isolation – and should be genetically independent.

Under this scenario, the orange Adelaide race evolved from an ancestral crimson population through genetic drift in isolation, and thus retains its mtDNA signature, but also shows evidence of occasional introgression by the yellow race. But this does not explain why the mtDNA of the Adelaide race clusters with the Kangaroo Island race.

Nor does it explain the phenotypic similarity of the orange Adelaide and Western Slopes races, or why the Kangaroo Island mtDNA haplotypes cluster with the Adelaide and yellow in different areas. The hypothesis also fails to explain why yellow, crimson, Adelaide and Western Slopes rosellas are not genetically independent.

The Deakin researchers studying the complex have recently predicted that the onset of more arid conditions with global warming could see the yellow colour form expand its range, at the expense of the crimson form. Rainfall over the southern Murray Darling Basin has already declined by around 20 per cent since the 1970s and is predicted to fall by another 20 per cent by 2030.