Comparing Imperial mandarin and Silverhill satsuma mandarin as seed parents in a breeding program aimed at developing new seedless citrus cultivars for Australia

Abstract

In a breeding program aimed at developing new seedless citrus cultivars, crosses were conducted between Silverhill satsuma and Imperial mandarin as seed parents, and sweet orange and pummelo cultivars as pollen parents. Poncirus trifoliata pollen was also used to provide a morphological marker to compare each seed parent's ability to produce hybrid seedlings. Both Silverhill and Imperial can produce seedless fruits. Methods used to produce and identify hybrids highlighted the relative ease with which either can be used as a seed parent. Imperial mandarin produced more seeds per fruit (mean >8) than Silverhill satsuma (mean <4), thus requiring fewer controlled pollinations to produce desired numbers of hybrids. Germination rates of Imperial-cross seeds were generally greater than for Silverhill-cross seeds. A single seedling emerged from every Imperial-cross seed that germinated, demonstrating the monoembryonic nature of Imperial seeds. All Imperial x P. trifoliata seedlings were hybrids, confirmed by their trifoliate leaves. From this, all seedlings from other Imperial-cross seeds were assumed to be hybrids. By contrast, embryo rescue after in vitro germination demonstrated the polyembryonic nature of the Silverhill-cross seeds. Mean embryo number per Silverhill seed varied from 3.6 to 12.0 between crosses, and from 2 to 30 between seeds within crosses. About 50% of all embryos from Silverhill-cross seeds were rescued and retained for hybrid identification. Silverhill hybrids were identified visually and/or by isozyme banding patterns. Silverhill x pumrnelo hybrids were identified or confirmed using glutamate oxaloacetate transaminase and phosphoglucomutase isozymes. Silverhill x Sweet orange hybrids were identified using isocitrate dehydrogenase, phosphoglucoisomerase or shikimic acid dehydrogenase (SKD) isozymes, although the banding patterns for SKD isozymes were complex and difficult to interpret genetically. Thirty-three Silverhill satsuma hybrids were identified. This represented a hybrid recovery rate of 11.15% compared with >90% for every Imperial-cross combination.