Project Aims:

1. To develop non-destructive and high throughput Near Infrared Reflectance (NIR) assays for the selection of single sweet lupins (Lupinus angustifolius) seeds with increased protein content.
2. To establish the merit of this assay as a routine early generation screening test for improved protein content.
3. To use these assays to screen F2 and F3 seeds generated from selected crosses made by Dr Bevan Buirchell and the germplasm from the National Lupin Breeding Program (NLBP) and select seeds with increased protein content.

A non-destructive procedure for determining the protein concentration in single L. angustifolius seeds has been developed using near-infrared (NIR) spectrometry. The R squared value for the calibration curve was 0.83 which, for a determination on a single seed, is excellent. More than 4,000 L. angustifolius single seeds were analysed using the procedure. The procedure has been used to study the inheritance of protein in the progeny from a single parent seed. The results have shown that the protein of the parent is transferred to the progeny. However, there are some instances where the agreement is not that good and in these cases, environmental factors may be a cause of the variation. The protein concentration of the single seeds from the various pods on the lateral branches is similar to the results on the main stem and there is little variation in the protein concentration of the single seeds within an individual pod. The variation is within experimental error. In determining the protein concentration of a single plant it is recommended from this initial work that the third or fourth pod on the main stem be taken and the protein concentration of the single seeds in the pod be determined and the average result for protein concentration for seed in the pod represents the seed on the whole plant. A number of seeds from some crosses were examined and the protein concentrations ranged from 29 to 49% protein. These seeds were returned to the breeder for replanting.

It is recommended that the non-destructive single seed protein analytical procedure be incorporated into the L. angustifolius breeding program. Initially the procedure will be used to determine the protein level of single seeds from new crosses. However, the system has the potential to examine larger seed populations looking for a high protein seed that may result in a new lupin variety. As other parameters are measured and determined on a single seed, then they can be added to the single seed calibration.

Background

Lupin production in Australia is expected to increase to two million tonnes per annum by the year 2005 and production in Western Australia (WA) alone will exceed 1.5mt. It is anticipated that lupins will continue to be a 'price taker' in the international feed market and, as a consequence, may struggle to maintain both market share and price. Lupin grain is sold primarily on the basis of its relatively high protein content. However, there are a number of quality deficiencies, which diminish its nutritional value and hence price to monogastrics and ruminants. In the case of monogastrics, lupins have a relatively low metabolisable energy (ME) content. In the case of ruminants, the key nutritional constraint is protein quality, especially the relatively poor protein bypass quality of lupin protein compared to other protein sources. Australian sweet lupins are considered to be an excellent source of dietary protein, although deficient in sulphur (S), for monogastrics, especially pigs and chickens. This deficiency is normally remedied through fortification of the feed with methionine.

The genetically modified (GM) debate in Europe has increased interest in lupins as an alternative to soybean meal. Although sweet lupins contain approximately 32% protein, there is considerable potential to increase this level, given the high hull (25%) and non-starch polysacharides (30%) and oligosaccharides (10%) contents of the kernel. Indeed, a recent review of historical data on protein content from the NLBP by Dr Mark Sweetingham indicated that the protein content (based on 50g milled samples) ranged by as high as 37%. The Grain Pool of WA also offers $2/t for every 1% protein content increase from a base of 32% protein. In the longer term it is estimated that each 1% increase in protein content over the current average of 32% could increase the grain value by $8/t. Accordingly, there is now more interest in developing new varieties with increased protein content. This project aimed to develop a non-destructive technique for the determination of protein levels in individual seeds. Dr Buirchell has made a series of selected lupin crosses, in addition to the 300 he currently makes each year. The key advantages of this approach are that the tests are very rapid and non-destructive and, unlike current techniques, are applicable to single seeds rather than the destructive milling of samples representing more than 200 seeds.

The NIR non-destructive single seed protein concentration determination was calibrated and validated against standard instrumental techniques, in particular Leco determination (combustion technique). Once this was achieved, the information was conveyed to Dr Bevan Buirchell at the Department of Agriculture and Food Western Australia (DAFWA), who then arranged for plants to be grown and seeds to be forwarded to the laboratory for testing. The results of the project were presented to the national lupin breeders at the annual Australian Co-ordinated Lupin Improvement Program (ACLIP) meeting held in Wagga Wagga. The results from this project will be presented to the wider scientific community by publishing the results in the appropriate scientific journals.

The analytical procedure was used to determine the inheritance of the protein concentration from the parent to the progeny. It was found that, in most cases, the progeny had protein concentrations similar to the parent seed. In one germplasm line (22841), the protein concentration of the single seeds of the progeny was approximately two percentage points lower than the parent seeds. The single seeds from pods on single plants were analysed and the results demonstrated that the protein level on the whole plant does not vary greatly. When analysing a single plant, the individual seeds from the third and fourth pod on the main stem were found to be representative of the whole plant. This information has been given to the L. angustifolius breeders and present data will be published in the scientific literature.

Additional Supporting Information

Table 1: The protein concentration of a single parent seed and the average and range of protein concentration of the progeny seed.

At present, the only calibration that is available is for L. angustifolius single seed. It is possible that calibrations for L. albus and L. luteus could be developed and then incorporated as an analytical tool into the appropriate breeding programs.

The effect of environment on the protein concentrations in a single seed needs to be examined. Little is understood about the effect environment has on protein content in lupins.

Additional Supporting Information

Table 1: The protein concentration of a single parent seed and the average and range of protein concentration of the progeny seed.

Figure 1: Calibration curve for the determination of protein concentration of single L. angustifolius seed.