Previous PageTable Of ContentsNext Page

Harvestability and potential seed production of Cullen australasicum

Eric Kobelt, Alan Humphries, Trevor Rowe and Steve Hughes

South Australian Research and Development Institute, Email


Cullen australasicum is a promising native perennial pasture legume targeted towards reducing soil water recharge and drought proofing cropping and grazing systems within the low rainfall wheatbelt of Australia. There is potential for this native plant to be sown over very large areas, providing seed production and other agronomy targets can be met. A major constraint with the species is that large seed losses can occur through pod dehiscence and uneven ripening of the seed. Seed production experiments over 2 seasons at Urrbrae, South Australia have resulted in 10-250kg/ha of seed harvested, with 40-410 kg/ha of dehisced seed falling on the ground. Whilst windrowing has shown to be more successful than using a chemical desiccant to aid mechanical harvesting, the major challenge with improving the seed production potential of this species relates to improving synchrony of seed production and timing of the harvest, or alternatively by improving pod retention through plant breeding and selection.

Key Words

forage, perennial, pasture, legume


Cullen australasicum is a native Australian plant that is widely distributed throughout central Australia, extending in South Australia to the Fleuireu Peninsula, to the Western regions of NSW and the south-east coast of Queensland. It was identified as a potentially valuable forage plant by Skerman (1957), Britten and De Lacy (1979), Bennett (2006) and Dear et al. (2007). Cullen australasicum was recently elevated to a priority species by the SARDI Genetic Resources Centre, which now holds 55 accessions from targeted collection missions (Hughes et al. 2008). The species has been found to be very productive and persistent in 20 field evaluation trials across southern Australia, despite initial problems with establishment (Dear et al. 2008).

For the species to be successful we need to ensure that a viable seed industry can be developed. A major constraint with the species is that large seed losses can occur through pod dehiscence and uneven ripening of the seed. In a RIRDC supported project SARDI is investigating the harvestability and seed production potential of this species. Different treatments such as harvesting technique (windrowing vs direct heading), harvest time, flowering time and plant density will be employed. The importance of pollination by bees on seed production will also be investigated. This paper is a preliminary report on these investigations.


Cullen australasicum accession (SA 4966) was sown in spring 2006 at Urrbrae, SA in 40 5 m2 plots. It was irrigated weekly and moderately, the same as for the adjacent lucerne also grown for seed production. Honey bee hives were present nearby to augment pollination of both seed crops. Two treatment factors investigated were 1) harvest method (windrowing versus desiccation), and 2) harvest time (March 13 and 26). There were four plot replicates for each treatment factor. Seed was harvested by direct heading in 2007 and 2008 using a small plot harvester. The amount of dehisced seed was also measured in plots using plastic trays at ground level. In 2008, seed content of these trays were collected at weekly intervals to assess losses during the full seed production period.

Results and Discussion

Harvested seed yields ranged from a low of 10 kg/ha (desiccated, late harvest in 2008) to a maximum of 188 kg/ha for a windrowed early harvest in 2007. In each season flowering began in early December and reached a peak in January. Bees were observed to be actively visiting and working flowers of both Cullen and the adjacent lucerne throughout the flowering period of both crops. Seed began to set in mid-December shortly after first flowering and continued right through to harvest. No damaging insect populations (thrips, seed wasps and mirids are all present at the site) were observed on the Cullen.

Harvest time and seed dehiscence

The timing of harvest had a large effect on harvested seed yield relative to the amount of seed pod falling on the ground (Figure 1). In 2007 up to 47 % of the total seed produced was harvested, whereas in 2008 only 14% of the seed produced was harvested with the small plot harvester. By the late harvest of 2007, over 390 kg/ha of seed had dehisced from the vine. The distribution of seed dehiscence over time is shown for 2008 in Figure 2. The amount of pod dehiscence peaked in the week ending February 18, 2 weeks before the first harvest. This corresponded with a period of 5 days of 37C, following a cooler period. A second, very hot spell of 15 days above 35 C between the 4th and 17th March also appeared to result in a small increase of dehiscence, but most pod was already on the ground at this stage.

Figure 1. Seed Production of Cullen australasicum at Urrbrae, South Australia that was machine harvested and collected from the ground at 2 harvest times in February and March in (a) 2007 and (b) 2008.

Figure 2. Distribution of seed dehiscence from Cullen australasicum experiments at Urrbrae, South Australia in 2008 (solid line), with average weekly maximum temperature (dashed line)

Harvest Method

The windrow pre-harvest treatment produced significantly higher yields than the pre-harvest desiccation treatment (Figure 3), though harvested seed yield in 2008 was poor. In 2007 the average early harvest yields for the windrow method was 189 kg/ ha and for the desiccation treatment, 95 kg/ha.

Figure 3. The effect of a chemical desiccant or windrow pre-harvest treatment on seed yield of Cullen australasicum in 2007 and 2008 at Urrbrae, South Australia.


The seed production potential of Cullen australasicum is adequate for commercial production providing improvements can be made in reducing the level of pod dehiscence from the vine. Tools need to be developed to determine the best timing of harvest, as the continuous, indeterminate pattern of flowering makes it difficult to determine the best time for harvest. Monitoring seed dehiscence may assist with harvest time recommendations. This research has investigated the seed production of a single accession, and further increases in harvestable seed production are considered likely when a more diverse group of germplasm is evaluated.


Bennett R, Colmer T, Real D and Ryan M (2006) Hardy Australians: ecogeography of Cullen suggests perennial legumes for low rainfall pastures. In 'Ground breaking stuff. Proceedings of the 13th Australian Agronomy Conference'. Perth Western Australia. Accessed May 2008.

Britten E and De Lacy I (1979) Assessment of the genetic potential for pasture purposes of the Psorelea eriantha-patens complex, a native legume of the semi-arid zone. Australian Journal of Experimental Agriculture and Animal Husbandry 19, 53-58.

Dear BS, Li GD, Hayes RC, Hughes SJ, Charman N and Ballard RA (2007) Cullen australasicum (Syn. Psoralea australasica): a review and some preliminary studies related to its potential as a low rainfall perennial pasture legume. The Rangeland Journal 29, 121-132.

Dear BS, Reed KFM and Craig AD (2008) Outcomes of the search for new perennial and salt tolerant pasture plants for southern Australia. Australian Journal of Experimental Agriculture 48, 578-588.

Hughes SJ, Snowball R, Reed KFM, Cohen B, Gajda K, Williams AR and Groeneweg SL (2008) The systematic collection and characterisation of herbaceous forage species for recharge and discharge environments in Southern Australia. Australian Journal of Experimental Agriculture 48, 397-408.

Skerman P (1957) Bullamoon lucerne (Psorelea eriantha Benth.) - a plant worth watching. Journal of the Australian Institute of Agricultural Science 23, 337-339.

Previous PageTop Of PageNext Page