The value of germplasm nurseries in selecting species for field evaluation

This paper documents observations made in irrigated seed multiplication nurseries of novel pasture germplasm to aid in the selection of species for field trials. For the first time it also compares these observations to results from past and concurrent field trials to ascertain the value of observations made in germplasm nurseries in predicting performance in the field. Flowering abundance, seed yield and plant survival over a 2-year period was assessed at four consecutive germplasm nurseries used for seed multiplication between 2002 and 2007. Germplasm included 373 accessions from 67 species and 28 genera of exotic and native perennial forage species, mostly legumes with potential in recharge environments. Classification and regression tree analysis was used to allocate accessions within a species to fewer groups of accessions using the germplasm characteristics. Plant survival and ratings for herbage production in germplasm nurseries were compared to results from field trials for the same species and for the same accessions in order to identify which plant characteristics gave good concordance. Observations from germplasm nurseries were, on the whole, useful in predicting survival of species and accessions in field trials. Ratings for herbage production in germplasm nurseries corresponded poorly with measurements or ratings of herbage production infield trials but, nevertheless, were useful in eliminating some poor-performing germplasm. Several species were identified that had superior survival and seed production in germplasm nurseries, some of which also performed well in field trials, including Dorycnium hirsutum, Kennedia prostrata, Lespedeza juncea var. sericea, Lotononis bainesii and Sanguisorba minor. Western Australian native species had mixed performance in germplasm nurseries but showed promise in field trials, and should be further investigated for their potential in farming systems in southern Australia.