Continuing a 70-year enduring commitment to narrow-leafed lupin improvement, a team at the University of Western Australia is revisiting germplasm to mine it for flowering diversity.

Australian narrow-leafed lupins (NLL), a significant grain legume in Western Australia’s acid sandy soil cropping systems, saw a rapid expansion in production in the 1980s and 1990s. Challenges such as weeds and low prices led many growers to opt for canola or fallow in more recent years, but renewed interest now stems from the demand for naturally fixed nitrogen, increased crop diversity and growing markets for plant protein.

Historically, NLL were bred to flower early, resulting in limited genetic diversity in breeding programs for flowering time. This makes current lupin varieties unsuitable for early sowing and longer growing seasons in higher-rainfall zones.

A new project supported by GRDC, the Council of Grain Grower Organisations (COGGO) and Australian Grain Technologies (AGT) aims to address these issues under the guidance of veteran plant breeder Professor Wallace Cowling and Associate Director of The University of Western Australia (UWA) Institute of Agriculture.

“During domestication the germplasm pool often narrows. While centuries of breeding have alleviated this for cereals, the 70-year history of NLL breeding, with lower investment due to its production size, has limited progress,” Professor Cowling says.

The improvement of Australian NLL can be traced back to Dr John Gladstones’ seminal work developing the sweet lupin industry in the 1970s, followed by the efforts of Professor Cowling, Dr Jon Clements and Dr Bevan Buirchell. This breeding effort has been underpinned with fundamental research at CSIRO by Professor Karam Singh and Dr Lars Kamphuis, and Drs Matthew Nelson and Candy Taylor during their tenures at UWA.

This new investment will build on the train of discovery for NLL.

Professor Cowling explained the background to the new project during a recent interview.

Why were lupins historically bred to flower early?

The original naturalised lupins thrived as weeds in long-season environments, such as the Swan Valley and southern WA where some bitter types were cultivated for green manure or soil improvement. These lupins shared the late-flowering allele ku.

Dr John Gladstones bred the first sweet lupin, Uniharvest, which had the ku flowering allele from a wild type, but it was too late-flowering for the wheatbelt. Discovering a mutant with the early flowering Ku allele, he released the first early flowering Ku variety Unicrop – it was well-adapted to the wheatbelt in terms of flowering time, but mostly too early for southern high-rainfall or eastern states.

Of course, history tells the story – lupins with Ku allele grew in popularity and were bred for disease resistance such as Phomopsis, brown spot and anthracnose. But this left long-season environments without well-adapted lupins for cropping – nobody was interested in breeding ku varieties since they were just too late-flowering for most environments. Until now, there have only been two options – either Ku (early, no vernalisation requirement) or ku (late, strong vernalisation requirement).

The quest for an intermediate flowering type – a compromise between the Ku and ku alleles – has been the ongoing research focus at UWA with Dr Matthew Nelson, Dr Candy Taylor and me, plus colleagues at CSIRO and Curtin University and international colleagues for the past two decades.

Will it use data from the sequenced NLL genome?

We achieved a lot with the original sequencing of NLL at CSIRO by Karam Singh and Lars Kamphuis. In Candy Taylor’s PhD at UWA, co-supported by GRDC, we used the existing sequence from this CSIRO work to locate a new allele (let’s call it the P allele since we found it in a wild lupin with a P number, P22660). Then we ran the first backcross into NLL with assistance of a COGGO-funded project in 2019–21.

Now GRDC and COGGO are co-supporting this project, where we will backcross this P allele into elite germplasm with the aid of molecular markers developed in the previous COGGO project. So yes, we used sequence from the NLL genome, just not the latest version.

Why increase harvest height?

If you are in the wheatbelt and come across a drought-affected lupin crop, you will see how short and low to the ground it is – harvesting is a headache as harvesters pick up stones and dirt, damaging machines.

The Ku varieties are short due to quick flowering (16 to 18 nodes to the first flower). If you delayed flowering to nodes 23 or 24 or about one week, then both height and biomass would increase and the crop would be easier to harvest, free of stones and dirt – and potentially the higher biomass would set the foundation for higher yield.

The problem is, we have never had a useful alternative allele to test this idea – but now we have the P allele and we hope to determine which agronomic regimes it will be most useful to. The only alternative allele we had until now was the ku allele, which delayed flowering by 15 to 20 nodes or four weeks in WA – too late to be useful.

Where does the wild lupin variety P22660 come from- why are its genes ‘novel’?

P22660, a wild lupin collected in Israel in 1973 by Dr BJ Quinlivan of the WA Department of Agriculture, shares a latitude with Perth (32° North). We suspect now that other ‘early’ flowering wild types may have been found in the eastern Mediterranean, but this is the first to be described, genetically analysed, located on the genome, sequenced, and proven to be a new distinct allele at the Ku locus.

It is quite an exciting discovery, because all we thought we had in lupins was Ku and ku alleles – not so! Ku is completely dominant for flowering time, meaning that very little other variation in flowering time can be expressed or observed by breeders.

Now that we have the P allele, we hope to expose other previously unrecognised effectors of flowering time that were hidden by the Ku allele for the past 70 years of lupin breeding.

What new technology will be applied in this project?

Growth lights are being used to hasten flowering and reduce the backcrossing period – ‘speed backcrossing’, if you will. We are using reliable molecular markers developed via Polymerase Chain Reaction (PCR) techniques to identify alternative forms of the alleles – this must be accurate for backcrossing to proceed, because if you fail in the markers, you fail in the backcrossing.

We cannot afford to use untested techniques as we would fail in rapid backcrossing. This small and efficient project demonstrates the value of having experienced people in lupins to do this work. We have Karen Nelson, who has worked on lupins and legumes for the past 10 to 15 years, and Dr Aneeta Pradhan, who ran the markers for the previous COGGO lupin project in 2020. I was lupin breeder at the WA Department of Agriculture from 1982 to 1999, and I continue to publish in lupins from ongoing UWA research.

When could growers expect earlier flowering NLL varieties in the field?

We intend to backcross the new alleles into elite lines for four generations and will select seedlings that have copies of P alleles from both parents at this stage – we will hand these to AGT for further bulking and testing before they consider releasing new varieties.

AGT’s lupin breeder Dr Matt Aubert says this is important research to tease apart the potential maturity groups for NLL and expand its fit in cropping systems, together with its area of production.

“Given the alleles are being introduced from a wild relative, it takes time to introduce it into elite lines,” Dr Aubert says.

“As a large national breeding company, AGT has the capacity to validate the importance of this new NLL allele in a range of yield trials in different environments, while UWA has expertise in this pre-breeding area. This research alliance with UWA will enable us to get new NLL lines into growers’ hands faster.”

More information: Professor Wallace Cowling, wallace.cowling@uwa.edu.au