Quantifying the dynamics of sugar concentration in berries of Vitis vinifera cv. Shiraz: a novel approach based on allometric analysis

Abstract

The definitive version is available at www.blackwell-synergy.comConcentrations of key compounds (e.g. sugar) in berries are the net result of relative changes in the amount of compound per berry and berry size. The complex nature of concentrations is widely recognised, but the widespread use of chronological scales for comparisons implies that ontogenetic drift or size-dependent effects are often overlooked. This paper presents an allometric analysis of sugar concentration in berries of cv. Shiraz as a way to formally account for ontogenetic drift. Our starting point is the double-sigmoid growth pattern of a grape berry where we distinguish Phase 1, from flowering to veraison; Phase 2, from veraison to peak berry fresh mass, and Phase 3, after peak fresh mass. Phase 3 explicitly accounts for the late season shrinkage typical of Shiraz berries. We advance an allometric model of sugar per berry with berry fresh mass, rather than time, as descriptor. The condition for an increase in sugar concentration in Phase 2 is that the relative rate of sugar accumulation per berry (RSB) exceeds the relative rate of berry net accumulation of fresh mass (RFM). This is equivalent to an allometric coefficient, calculated as the slope of the regression between amount of sugar per berry and berry mass in a log-log scale, being greater than 1. For Phase 3, the condition for increase of sugar concentration is that a large reduction in berry mass offsets any putative change of sugar per berry, yielding an allometric coefficient < 1. Such an allometric model was tested against measured data from sixteen contrasting crops resulting from the combination of eight water regimes and two seasons. Berry mass peaked between 96 and 105 days after anthesis, and these dates were used to separate Phases 2 and 3. In Phase 2, the relative rate of increase in sugar per berry varied from 0.01 to 0.02 d-1 in comparison to the relative rate of increase in berry fresh mass that varied from 0.0038 to 0.0066 d-1. Sugar per berry thus increased 2.4–3.3 times faster than berry mass, with allometric coefficients between 1.98 and 2.91 accounting for 78% of the variation in the relative rate of change of sugar concentration. In Phase 3, the relative rate of change in sugar per berry was not different from zero (P 0.05) in most cases, whereas the rate of change in berry size ranged from –0.0013 to –0.0035 d-1 and was significant (P < 0.05) in 14 out 16 cases. The small changes in sugar per berry and the net loss of berry material yielded allometric coefficients between 0.17 and 1.11, which accounted for 72% of the variation in the relative rate of change in sugar concentration. We conclude that a model, which pivots around peak berry mass, with allometric coefficients above 1 in Phase 2 and below 1 in Phase 3, is suitable to quantitatively account for ontogenetic drift in the dynamics of sugar concentration in berries of Shiraz. This allometric approach demonstrated that sugar per berry during the stage of berry shrinkage is a plastic trait under significant environmental influence. For the same genotype, environmental conditions could determine either, a putative backflow of water accounting for net loss of berry fresh mass (RFM < 0) that could also carry some sugar from berries back to the parent vine (RSB < 0) or a small gain of sugar (RSB 0) closely coupled with a net loss of berry fresh mass (P = 0.003).Victor O. Sadras, Michael G. McCarth