Agronomic potential of genetically diverse narrow-leafed lupins (Lupinus angustifolius L.) with restricted branching

Abstract

The indeterminate growth habit of narrow-leafed lupin appears to cause a suboptimal pattern of grain filling in the Mediterranean-type environment of south-western Australia. Development of cultivars with genetically restricted branching (RB) has been proposed to overcome the problem. However, restriction of branching causes profound phenological and architectural changes, and it may be necessary to compensate for these by incorporating RB into a genetic background that confers high shoot mass. In order to make a robust assessment of the value of RB in a range of backgrounds, the trait was incorporated from 5 donor parents into the genetic background of 10 recurrent parents by 2 rounds of back-crossing followed by self-fertilisation of the progeny for 4 generations to produce BC2S4 lines. Thirty-two of these lines were obtained with highly RB or mildly RB, a range of flowering times from 68 to 118 days after sowing, and 16–34 leaf nodes on the main stem. They were tested with their parents in replicated field trials at 3 sites in Western Australia at latitudes from 28°S to 33°S. The RB genotypes generally gave higher grain yield than the normal-branching genotypes at the high-latitude, high-rainfall, long growing season, high shoot mass producing site of Esperance, and the 2 types gave approximately equal yield in the low-latitude, low-rainfall, short-season, low shoot mass site of Mullewa. Only at the intermediate site of Wongan Hills did the normal-branching genotypes have a clear advantage. RB genotypes had higher harvest index than corresponding normal-branching genotypes, particularly at Esperance, and tended to produce more pods but slightly fewer seeds per pod and lighter seeds. There was no consistent difference in performance between highly and mildly RB genotypes, contrary to an expectation that the highly RB type would produce insufficient shoot mass. There was a tentative indication that, within RB lines, a large number of leaf nodes on the main stem conferred more reliably high grain yield in the environments of Esperance and Wongan Hills. Overall, these results provide ample justification for the development and further evaluation of RB cultivars. However, this conclusion comes with 2�caveats: that a different background development pattern should be adopted to that used in normal branching lupins, and that RB cultivars should be evaluated in the target environments where the character confers an advantage.