Morphometrics, characterization and seedling emergence studies of a collection of Russell lupin (Lupinus polyphyllus x Lupinus arboreus) genotypes in Canterbury

In a field experiment, a germplasm collection of 18 Russell lupin accessions, obtained from the USSR (5), New Zealand (5), Poland (3), Germany (3), Portugal (1), and the United Kingdom (1), was grown on a Wakanui silt loam soil at Lincoln (43º 38' S.) in Canterbury. The morphology, growth, performance, and flowering characteristics of the accessions were studied. Canonical variate analysis was used to characterize the germ plasm collection on morphological and agronomic attributes. Seedling emergence of the Russell lupin accession (Connie), which has considerable agronomic potential for sheep grazing in the hill and high country of New Zealand, was evaluated for response to scarification method, depth of sowing, and temperature in a series of laboratory and glasshouse experiments. The scatterplot of the canonical variate means of accessions facilitated the identification of four distinct groups of accessions in the germ plasm collection. Characterization of the genotypes using univariate statistics, or by simple visual observation was extremely difficult. The groups and their main characteristics are described and discussed. In terms of overall relationships, the Russell lupin should be regarded as a plant which is highly variable in its morphology, growth, performance, and flowering characteristics. A latitude of origin pattern of variation was represented along the first canonical axis, which accounted for 35.2% of the between-accession variation. Along this axis, all, but one, accessions from New Zealand and the accession from Portugal were grouped together as opposed to the group of accessions from northern Europe. Flowering before exposure to cold and/or short days, days to first flower, dry matter yield in autumn and leaflet dimension appear to be the most discriminatory attributes in separating accessions along the first canonical axis. Along the second canonical axis, which accounted for a further 16.4% of the between-accession variation, there was wide variation in plant size and spring dry matter yield. Two New Zealand accessions (HN & ON), with a higher dry matter yield than the mean of the 18 accessions were identified, although these accessions show different flowering responses, perhaps, to cool conditions along the first canonical axis. Conversely, all accessions from the USSR were vernalization requiring, high latitude types, and had a low dry matter yield in both autumn and spring. Connie lupin (CN), used as the control, was an early flowering, non-vernalization requiring low latitude type, and was of medium plant height and had an intermediate dry matter yield. The wide diversity within and among the groups should provide good opportunity for future breeding and agronomic work on Russell lupin in New Zealand. A high proportion of freshly harvested Russell lupin seed is hard. Laboratory and glasshouse studies showed that the germination and emergence of freshly harvested Russell lupin seed can be improved by scarification using chipping or concentrated sulphuric acid (36N). However, a large proportion of abnormal seedlings (36% of the seeds sown) was produced by chipping. Acid scarification for 30 or 45 minutes produced more than 75% normal seedlings. In a glasshouse study, depth of sowing affected both emergence and seedling vigour of Russell lupin. Emergence was best (92%) when the seed was sown at 1 cm below the soil surface. The best seedling growth (seedling height, root length, oven dry weight of tops and roots) was also obtained by sowing at 1 cm. However, sowing at 2 cm gave results which were not statistically different (p