Effects of environment and floral intensity on fruit set behaviour and annual flowering in apple

• Published in: Scientia horticulturae Vol. 210; pp. 258 - 267
• Main Authors: Breen, K.C., Tustin, D.S., Palmer, J.W., Hedderley, D.I., Close, D.C.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 10-10-2016
• Subjects: Artificial spur extinction; Biennial bearing; Crop load; Thinning; Model; Crop load; Thinning; Artificial spur extinction; Biennial bearing
• ISSN: 0304-4238; 1879-1018
• DOI: 10.1016/j.scienta.2016.07.025

•On unmodified trees flowering varied widely among trees, sites and seasons.•On unmodified trees fruit set was variable and unpredictable.•Artificial spur extinction (ASE) halved variability in flowering and reduced floral bud density.•ASE improved reliability of fruit set and allowed predictive fruit set models to be developed.
Natural variability in flowering and fruit set of commercial apple trees (Malus x domestica Borkh.) must be effectively managed to achieve optimised yield and quality. Early removal of excess flowers and fruit improves fruit size and quality. However, physiological responses to environment and flowering intensity, and their interaction with chemical thinning are not well understood, causing unpredictable fruit set responses. We altered flowering intensity on ‘Gala’ trees on five sites through New Zealand and Australia over four consecutive seasons using artificial spur extinction (ASE) at bud densities of 2–6budscm−2 branch basal cross-sectional area. We compared this with natural flowering and fruit set on trees thinned by hand after final fruit set (Control). ASE uses hand-thinning of dormant buds to specific densities in late winter, in order to reach targeted floral bud densities on every branch on whole trees the following spring. Naturally occurring seasonal differences in weather and orchard environment across five regions (Hawke’s Bay and Nelson in New Zealand and Queensland, Tasmania and Victoria in Australia) allowed us to investigate the effect of these factors on fruit set and annual flowering behaviour. Typical of trees in commercial production, flowering and fruit set in Control trees differed widely among trees within sites, among sites, and between seasons. Natural fruit set in these trees, especially at high floral bud densities, was highly variable and largely unresponsive to natural differences in floral bud density, but often highly responsive to differences in weather and other environmental conditions. If only one fruit was set per bud, the number of flowering buds was always sufficient to produce a commercial crop, but in some cases four times more floral buds were present than required. Systematic removal of dormant buds using ASE reduced floral bud density to within 20–30% of the bud density set during dormancy, and reduced variability in floral bud density among trees to about half of that in Control trees. In ASE trees, reduced floral bud density resulted in reduced number of fruit set per cm2 branch cross-sectional area, but increased fruit set per floral bud because lower proportions of floral buds set zero fruit and greater proportions set two-or-more fruit, compared with Control trees. The ASE treatments allowed models to be fitted for each site which described relationships between fruit set and floral bud density, and identified years where weather or environment altered fruit set responses. Consequently, use of ASE in ‘Gala’ provided a method of very early crop load adjustment which produced reliable and predictable outcomes to fruit set over a number of sites and seasons.