Mechanism of rootstock effects on flowering in kiwifruit

Abstract

Rootstocks selected from different Actinidia species have marked effects on flower production of 'Hayward' kiwifruit. The objective of this project was to investigate the physiological mechanism underlying these rootstock effects. Five clonal rootstocks selected from A. hemsleyana, A. eriantha, A. rufa, A. deliciosa and A. chinensis respectively were studied. Flower production of 'Hayward' kiwifruit on these rootstocks, listed in the order above, averaged 6.2, 5.5, 3.8, 3.5 and 2.7 flowers per shoot. None of these rootstocks had any particular influence on the vigour of the vine. Field examinations showed that the rootstocks significantly affected the budburst on the scion. Vines on the flower-promoting rootstocks had higher and more synchronous budburst than vines on the flower-retarding rootstocks. Dissection of kiwifruit buds in early spring, i.e., the early stage of flower development, revealed that the number of flower primordia initiated within each bud was similar irrespective of the rootstock used. This indicates that rootstocks affect flower production by influencing the level of floral abortion during flower development. Seasonal changes in concentrations of carbohydrates in the shoots and roots of the five rootstock-scion combinations were monitored. The concentrations of carbohydrates in the shoots were similar in all five rootstock-scion combinations throughout the year, except for a higher concentration of starch and sugars in vines on the A. eriantha rootstock. The concentrations of carbohydrates in the root tissues showed marked variation among rootstocks, with the greatest variation occurring with starch and polysaccharide mucilage. Starch concentrations in root wood tissues were generally higher in the flower-promoting rootstocks than in the flower-retarding rootstocks. The correlation was not perfect, however, because the highest level occurred in A. eriantha rather than in the best rootstock A. hemsleyana. Starch concentrations in the root bark tissues were similar in winter among rootstocks, but the proportion of starch that was mobilised in spring varied markedly. There was a close relationship between the extent of starch mobilisation and the abundance of flower production. The extent of starch mobilisation in the flower-promoting rootstocks was greater than that in the flower-retarding rootstocks. Mucilage was not mobilised in spring, and did not therefore appear to function as a carbohydrate reserve. In general, flower-promoting rootstocks contained more mucilage than flower-retarding rootstocks. In A. hemsleyana, mucilage in root bark tissue accounted for up to 27% of dry matter. Root anatomy of the five rootstock clones was compared. In all these rootstocks, the cortex and endodermis were retained and developed during secondary root thickening. Groups of endodermal cells (meriphytes) derived from cell divisions were enclosed in suberin lamellae. Idioblasts containing raphide crystals and mucilage were present in the cortex and phloem. Differentiation of these idioblasts involved at least two stages: formation of raphide crystals of calcium oxalate, and the production of mucilage. Increases in endoplasmic reticulum and in the density of cytoplasm were associated with crystal formation. The size of the idioblast cells increased greatly as mucilage accumulated. The flower-promoting rootstock A. hemsleyana had much more numerous and larger idioblasts in the cortex and phloem than other rootstocks. Total cross-sectional areas of xylem vessels and the abundance of starch grains in parenchyma of the cortex and stele were also generally greater in the flower-promoting rootstocks. The results indicate that flower development in 'Hayward' kiwifruit is limited by the availability of reserve carbohydrates in spring. Flower-promoting rootstocks decrease the level of floral abortion by encouraging the buds of the scion to use greater proportions of the reserve carbohydrates for flower development. The possible cause for the differences in the extent of carbohydrate mobilisation among rootstock-scion combinations was discussed, and the criteria for future selection of superior kiwifruit rootstocks were proposed.