The influence of site, crop load and cluster light exposure on Pinot noir fruit composition in Canterbury

The aim of this study was to identify the relative importance of site, crop load and cluster exposure on various aspects of fruit composition in the Vitis vinifera cultivar Pinot noir in the Canterbury region of New Zealand. Both reductions in crop level and increases in cluster exposure were found to increase berry skin phenol levels. Crop reduction was further found to advance maturity at all sites. A field trial was established at six vineyard sites in January of 1996. Sites varied in soil type, mesoclimate and training system. However, at all sites vines were of similar ages, planted at similar densities, were own-rooted and of the same clone of Pinot noir (AM 10/5). Also at all sites viticultural practises were identical after the start of the trial with no leaf-plucking, shoot-thinning or trimming of the canopy sides. In order to investigate the effect of crop level on fruit maturation a 2 x 5 randomised block design was established at each site in early February. This was achieved by removing 50% of the clusters in 5 of the 10 experimental bays at each site. At all sites crop-thinning was performed between 19-40 days before 50% veraison was reached. Fruit maturation was assessed weekly from veraison onwards at each site. This involved the systematic removal of a 60-berry sample. These were subsequently analysed for soluble solids, malic acid concentration, titratable acidity (TA) and pH. Fruit was harvested from each of the two crop treatments at each site when the mean soluble solids level of the weekly berry samples reached 19.5°Brix. Crop removal considerably advanced the date of harvest. On average the low crop treatment reached the harvest maturity point of 19.5°Brix 15 days before the high crop treatment. The rate of soluble solids accumulation and malic/TA degradation was nearly identical between crop loads at each site. Veraison was not identified separately for each crop treatment, but differences in fruit composition at 50% veraison over both crop treatments were maintained until harvest. This therefore suggested that differences between crop treatments in the date of harvest were primarily attributable to changes in the date of veraison for each crop treatment, rather than a difference in the rate of soluble solids accumulation after veraison. At veraison a second treatment was also implemented in order to investigate the influence of cluster light exposure on berry skin phenol levels at harvest. This involved tagging six bunches per bay. Two of the most exposed bunches on the canopy exterior were designated the 'exposed' treatment. Two bunches from the canopy interior were designated the 'partial exposure' treatment, while two bunches from the canopy interior were covered with brown paper bags and designated 'shaded'. This trial was harvested at the same time as the crop treatments and bunches were frozen for later analysis. After harvest was completed, these clusters were thawed and berry skin discs were removed and extracted in acidulated ethanol. Extracts were analysed individually by spectrophotometer at 290nm (for total phenols), 360nm (for flavonols) and 540nm (for anthocyanins). In order to assess the validity of measuring absorbance at 360nm to estimate flavonol content, pooled samples from each exposure treatment were also analysed at 360nm by high performance liquid chromatography (HPLC). Absorbance at 360nm provided a reliable estimate of skin disc flavonol levels with analysis by HPLC showing that flavonols were responsible for an average of 80% of the total absorbance at this wavelength. Quercetin glycoside was the main flavonol present and this compound exhibited a closer response to changes in cluster exposure levels, than did overall flavonol levels. This suggests that quercetin may be the best bioindicator of the cluster light environment. Increases in cluster exposure increased the concentration of skin disc berry phenols (anthocyanins, flavonols and total phenols) as did crop reduction. Furthermore, there was an overall relationship between yield per linear metre of canopy and anthocyanin, flavonol, and total phenol levels in the exposed cluster treatment. This was consistent across all sites, and independent of the specific site effects (eg. soil type) thus suggesting that site is relatively unimportant in determining berry phenol levels. Highest skin phenol levels occurred at yields of 1.5-2.5 kg.m⁻¹. This corresponded to overall yields of 5-8.2 tonnes/ha at the trial sites, taking into account differences in planting density. Shortly after veraison the canopies at each site were characterised using the Point Quadrat technique (Smart & Robinson, 1991). From this % canopy gaps, % interior leaves and clusters along with leaf layer number was calculated. During the winter following the experiment all vines included in the study were pruned and pruning weight, shoots/m, average shoot length and internode length were measured. From this the leaf area to fruit weight (LA:FW) ratio and yield/pruning weight ratio were calculated. Sites closest to the canopy ideotype (as defined by Smart & Robinson, 1991) and with optimum values for LA:FW ratio and yield/pruning weight reached 19.5°Brix faster than did sites with sub-optimum canopies. Pruning weight per linear metre of canopy showed a significant positive correlation with both pH and malic acid concentration at harvest. Similarly, LA:FW ratio was significantly positively correlated with malic acid concentration and TA at harvest. Yield/pruning weight was significantly positively correlated with days from veraison to harvest and accumulated mean daily temperature (AMDT). There was also a significant negative correlation between both days and AMDT from veraison and malic acid concentration.