Developmental variation in sensitivity of Vitis vinifera L. (Shiraz) berries to soil water deficit

A large-scale, long-term irrigation experiment was established near Waikerie in the South Australian Riverland to investigate the feasibility of controlling berry size and ripening at the vineyard scale with modern irrigation systems. Irrigation treatments were devised to impose water stress, by withholding irrigation, during four periods of berry development after flowering of Vitis vinifera (variety Shiraz). Varying water deficits were achieved during each period and between the four seasons, which were climatically diverse. In one season water deficit during the period after flowering resulted in the greatest reduction in berry weight compared with that of well-watered vines, however, in another, water deficit during this period had no effect on berry weight. By comparison, berries appeared to be insensitive to water deficit during the month before harvest in all four seasons. A soil water deficit index was derived to compare the varying levels of water deficit between treatments and seasons on berry development. Deficit effects on berry development were assessed using either comparative growth rate or berry weight near harvest. Regression analysis of berry development against soil water deficit indicated that berries were most sensitive to water stress during the post flowering period.     

Accumulation of potassium and calcium by ripening berries on field vines of Vitis vinifera (L) cv. Shiraz

Ripening berries of Vitis vinifera (L) cv. Shiraz can show pre-harvest weight loss at sub-optimal sugar content (shrinkage). This later-age decline in berry weight implies that water loss from mature berries has begun to exceed water inflow from the parent grapevine. Such decrease in net inflow has been attributed to a cessation of xylem flow subsequent to veraison, followed by a cessation of phloem flow into berries during later stages of ripening. We address this issue in this present paper, and show a continuing increase in berry content of both potassium and calcium throughout ripening. We measured changes in berry fresh weight and berry content of potassium (phloem mobile) and calcium (phloem immobile) in fruit on field vines sampled from set to harvest. Berry fresh weight reached a plateau between 81 and 95 days after flowering, then declined to 75% of maximum fresh weight by 115 days. Dry weight maximum occurred 14 days after the onset of the fresh weight plateau. Potassium accumulation was slow pre-veraison, increased 3.5-fold post-veraison, and continued during berry shrinkage. Calcium content per berry also showed a linear increase throughout fruit enlargement and ripening phases. Assuming both potassium and calcium were entering berries via vascular conduits, our results imply a continuing connection between parent grapevine and ripening berries. Moreover, an abrupt change (increase) in the ratio of potassium/calcium content per berry subsequent to veraison implies that phloem inflow has increased relative to xylem inflow during post-veraison enlargement.     

The effect of bunch shading on berry development and flavonoid accumulation in Shiraz grapes

Opaque boxes were applied to bunches of Shiraz grapes prior to flowering to determine the effect of sunlight on berry development and accumulation of flavonoids. The boxes were designed to maintain airflow while excluding light and thus to minimise changes in temperature and humidity. There was no significant effect of shading on sugar accumulation and in two of the three seasons studied there was no effect on berry weight. Chlorophyll concentration was much lower in the shaded fruit, which appeared pale yellow until veraison. The fruit coloured normally in the shaded bunches and in two of the three seasons there was no significant change in anthocyanin content. Expression of the gene encoding UDP-glucose flavonoid-3-O-glucosyl transferase (UFGT), a key gene in anthocyanin synthesis, increased after veraison and was similar in both shaded and exposed fruit. Anthocyanin composition was altered in the shaded fruit, which had a greater proportion of the dioxygenated anthocyanins, the glucosides of cyanidin and peonidin. Shading had no significant effect on the levels of condensed tannins in the skin or seeds of ripe fruit. Shading significantly reduced the levels of flavonols in the grape skin. In the exposed fruit, flavonol concentration was highest around flowering then declined as the berries grew, but there was an increase in flavonols per berry during ripening. When the boxes were applied before flowering, shaded fruit had much lower levels of flavonols throughout berry development and at harvest the level of flavonols were less than 10% of that in exposed fruit. A gene encoding flavonol synthase (FLS) was expressed at flowering and during ripening in exposed grapes but its expression was greatly reduced in shaded fruit. The results indicate that shading had little effect on berry development and ripening, including accumulation of anthocyanins and tannins, but significantly decreased flavonol synthesis.     

Dynamics of grape berry growth and physiology of ripening

Data from two experiments on development of grape berries is re-examined with emphasis on partitioning of berry weight into non-solutes per berry (largely water) and solutes per berry (largely sugar), using weight times juice °Brix. This approach is based on the thought that, since xylem flow is blocked after veraison, time curves of solutes per berry indicate the activity of phloem transport into the berry during ripening growth. Experiment 1: Measurements of Muscat Gordo Blanco berries from inflorescences with a spread of flowering times showed typical double-sigmoid volume/time curves but with divergent rates and amounts of volume increase. Despite this divergence, °Brix curves after veraison were almost coincident because, in each case, the rate of increase in solutes per berry was proportional to that of berry volume. These results indicate that sugar and water increments after veraison are linked and depend on the same source, namely, phloem sap. Experiment 2: An irrigation experiment on cv. Shiraz also showed divergent berry weight curves between treatments and years but with the difference that all berries shrank after a maximum berry weight was attained at 91 days after flowering (at about 20 °Brix). At this point, the curves of solutes per berry slowed then plateaued, indicating that inflow of phloem sap had become impeded. Prior to shrinkage these berries accumulated primary metabolites (mainly phloem sugar) but, during shrinkage, when berries were apparently isolated from vascular transport, non-anthocyanin glycosides accumulated. These results have implications for the study of berry flavour buildup and berry composition, and also for the understanding of sink competition within the vine, fresh and dried yield, and juice °Brix levels.     

Direct measurement of hydraulic properties in developing berries of Vitis vinifera L. cv Shiraz and Chardonnay

Berries of Vitis vinifera L. cv Shiraz can undergo weight loss during later stages of ripening. Existing published views on how weight loss occurs are based on changes in capacity of the vascular system to import water during development (McCarthy and Coombe, Australian Journal of Grape and Wine Research, 5, 17–21, 1999). One important element of these views is the proposed cessation of water flow through the xylem after veraison. We have now measured the water flow into berries of Shiraz and Chardonnay as they develop using the pressure probe and the high pressure flow meter (HPFM). The pressure probe connected to the pedicel of individual berries provided measurements of single berry hydraulic conductance. By systematic excision of tissue segments of the berry and pedicel we determined where in the pathway hydraulic conductance changed during development. The HPFM was used on whole bunches showing that berries (including pedicels) represent parallel high hydraulic resistances and that the hydraulic resistance of the bunch axis was rather small. The hydraulic conductance per berry could be determined from excision experiments. There was close agreement between the pressure probe and HPFM measurements. Both showed a ten‐fold reduction in hydraulic conductance of whole berries from veraison to full ripeness. Shiraz had hydraulic conductances that were 2‐ to 5‐fold higher than those for Chardonnay. Shiraz maintained a higher hydraulic conductance past 90 days after flowering than Chardonnay. The decrease in hydraulic conductance occurred in both the distal and proximal parts of the berry for both varieties. The pressure probe also provided measurements of the xylem pressure that non‐transpiring berries could develop. These pressures were –0.2 to –0.1 MPa until veraison and increased to zero when the juice osmotic potential reached about –3 MPa in Chardonnay and –4 MPa in Shiraz. The results suggest values of the reflection coefficient of the osmotic barrier around the xylem vessels of about 0.1–0.2 at veraison decreasing to 0 at harvest. It is suggested that in addition to changes in xylem anatomy, aquaporins in berry membranes may play a role in regulating hydraulic conductance. Water movement from the berry back to the parent vine via the xylem (backflow) may be an important component of berry weight loss in Shiraz, particularly if the phloem ceases functioning at high osmotic potentials near maximum weight. Backflow could account for a weight loss of 43 mg per day in Shiraz berries for a relatively small gradient of 0.1 MPa.     

Is weight loss in ripening grape berries cv. Shiraz caused by impeded phloem transport?

Berry shrinkage in ripening grapes cv. Shiraz is systemic within a given grapevine and coincident between grapevines in any given season. In this present study on weight loss in ripening berries, ripening curves of non-solutes per berry (largely water) were similar to curves for berry weight (as a function of time). Both sets of curves were equivalent with respect to timing of maximum weights and subsequent rates of weight loss. However, curves of solutes per berry (largely sugar) increased steeply up to the time of maximum berry weight, then slowed and plateau-ed. We suggest that phloem sap is the sole source for water and solutes that enter grape berries subsequent to veraison, and accumulate until maximum berry weight. We further suggest that phloem flow becomes impeded at maximum berry weight.
As berry ripening proceeds, continuation of berry transpiration leads to berry shrinkage and a concentration of solutes; i.e. any increase in juice Brix depends on shrinkage. One implication is that assimilates enter a berry up to the onset of shrinkage, whereafter accumulation of non-anthocyanin glycosides (including glycosides of flavour compounds) depends upon their synthesis in situ.     

Vascular function in berries of Vitis vinifera (L) cv. Shiraz

Anatomical studies on the movement of a xylem tracer dye were combined with functional studies on changes in grape berry volume during final stages of berry ripening to gauge xylem effectiveness. Movement of a xylem tracer dye into pre-veraison fruit was compared with movement into post-veraison fruit by feeding a solution of acid fuchsin to excised shoots with bunches still attached, and then sectioning fruit for photo-microscopy. Those comparisons confirmed published studies showing an apparent blockage to dye movement along major vessels within the brush tissue of post-veraison fruit. However, our functional approach yielded a different impression of vascular activity. A continuation of xylem transport in ripening fruit was inferred from comparisons of berry volume where pedicels were either girdled (phloem interrupted, but xylem intact) or excised (both phloem and xylem interrupted). Volume changes in manipulated berries were compared with immediately adjacent intact control berries within the same bunch. Control fruit lost volume subsequent to 78 days after flowering (DAF) while manipulated fruit lost volume from the first day of treatment at 67 DAF. By harvest time at 95 DAF, both control fruit and girdled fruit had fallen to 91% and excised fruit to 46% of maximum volumes recorded around 78 DAF. Berry volume loss in girdled fruit was further enhanced by deficit irrigation. We conclude that xylem flow into those Shiraz berries must have continued beyond veraison despite dye evidence of a vessel blockage within the brush region of analogous post-veraison fruit.     

Synthesis of flavonols and expression of flavonol synthase genes in the developing grape berries of Shiraz and Chardonnay (Vitis vinifera L.)

Flavonols were determined in Shiraz and Chardonnay grapes throughout berry development. The predominant flavonols were quercetin-3-glycosides with trace amounts of kaempferol-3-glycosides detected in Shiraz flowers but not in developing berries. Flavonols were present in the skin of ripening grapes but were not detected in seeds or flesh. Flavonols were also present in buds, tendrils, inflorescences, anthers and leaves. The concentration of flavonols in flowers (mg/g fresh weight) was high and decreased between flowering and berry set then remained relatively constant through berry development. The total amount of flavonols in berries (mg/berry) was low until pre-veraison then increased during berry development, particularly before veraison, the onset of ripening, in Chardonnay and during ripening in Shiraz. Two cDNA fragments with homology to genes encoding the enzyme flavonol synthase (FLS) were isolated from Shiraz flowers. In the overlapping region of the two cDNAs, they had 80% sequence identity at the nucleotide level and both had high homology to FLS genes from other plants. VvFLS1 was expressed in leaves, tendrils, pedicels, buds and inflorescences as well as in developing grapes. Expression was highest between flowering and fruit set then declined, increasing again during ripening coincident with the increase in flavonols per berry. Expression of VvFLS2 was much lower than for VvFLS1 and did not change during berry development. The results indicate that two distinct periods of flavonol synthesis occur in grapes, the first around flowering and the second during ripening of the developing berries.     

QUALITY ATTRIBUTES, PECTOLYTIC ENZYME ACTIVITIES AND PHYSIOLOGICAL CHANGES DURING POSTHARVEST RIPENING OF NECTARINE

The postharvest ripening at 20C and 90–95% RH for 10 days of ‘Armking’ nectarine grown in a greenhouse was investigated over two seasons. Firmness, titratable acidity, ascorbic acid, pH and maturity index were all adequate to stablish the rate of ripening. However, soluble solids content and reducing and non-reducing sugars showed no significant changes. It took about 10 days for very early ripening fruit (100 g weight and 82 N firmness) and 6 days for normal early ripening fruits (115 g weight and 46 N firmness), both harvested at preclimacteric stage, to become eating ripe (near 20 N). Total weight loss and decay after 10 days was approximately 11%. During ripening there was a temporal coincidence among higher rates of ethyleneproduction, higher pectinmethylesterase (PME) and polygalacturonase (PG) activities, lower firmness and acidity and higher maturity index. PME and PG activities increased during ripening, with a highly negative linear correlation between activities of the two enzymes and firmness. In very early ripening fruit, PME activity was more closely related to softening than PG, whereas in normal early ripeningfruit, PG activity was slightly more closely linked to the loss offirmness than PME.     

Effect of postharvest dehydration on the composition of pinot noir grapes (Vitis vinifera L.) and wine

This study was conducted in order to improve our understanding of how phenolics and aroma compounds change in wine grapes during postharvest dehydration. Pinot noir grapes grown in the Willamette Valley of Oregon were harvested at 22.0 and 24.0 °Brix. Grapes harvested at 22.0 °Brix were divided into three equal lots with one lot immediately used for wine production, and the remaining two lots placed inside an air tunnel with an air speed of 1.0–1.8 m s⁻¹, 38% relative humidity and a temperature of 22 °C. The soluble solids content and weight loss were measured daily and wines were made from grapes when they reached 24.8 and 26.7 °Brix. The soluble solids of grapes increased about 1 °Brix per day; therefore, on the third and fourth day the berries reached the desired concentration; weight loss was 14 and 16%, respectively. Results from berry phenolic analysis indicated that per berry anthocyanin amount remained unchanged during dehydration. The composition of proanthocyanidins isolated from berries changed during dehydration. Volatile compounds in wines made from dehydrated grapes contained more terpenes and norisoprenoids (β-ionone, β-damascenone) when compared to wine made from the original fruit. Wines made from increasingly dehydrated grapes tended to resemble the composition and flavour profile of wines made from grapes left on the vine (i.e. with extended ripening). The results of this study suggest that postharvest flavour changes consistent with changes during fruit ripening can occur in grapes when harvested early and allowed to dehydrate under controlled conditions prior to fermentation.     

Relationships between different berry components in Tempranillo (Vitis vinifera L) grapes from irrigated and non‐irrigated vines during ripening

The impact of water availability on Vitis vinifera L cv Tempranillo grape yields and juice composition over a 3 year period was studied. Grape juice composition during the different stages of berry growth was compared. The analytical data collected were used to investigate the relationships between some of the different components studied in these berries during the ripening period. The object was to determine whether the relationships could be used to derive mathematical equations for use as indicators of the changes taking place in the different parameters considered and thus of the optimum time for harvesting the grapes for a given wine‐making process. Very similar equations were derived for this grape variety irrespective of the culture conditions (use or non‐use of irrigation) or the different ripening conditions employed. A high degree of correlation between some of the components considered was observed. In any case, cause‐and‐effect relationships could only be established for certain of the parameter pairs considered (total acidity/pH, pH/potassium, °Brix/glucose + fructose, total acidity/tartaric acid and malic acid). For other parameter pairs (berry weight/°Brix, °Brix/pH, °Brix/total acidity) the relationship found was attributable to accumulation and breakdown processes which took place concurrently or overlapped in time. In two cases only there was uncertainty as to whether the relationship observed between the parameters was a cause‐and‐effect relationship or whether the parameters were so closely related that they acted as mutually limiting factors (°Brix/proline, °Brix/total anthocyanins). © 2002 Society of Chemical Industry     

Effects of manipulated grape berry transpiration on post-veraison sugar accumulation

A series of experiments were conducted on container-grown Pinot Noir and Sangiovese grapevines ( Vitis vinifera L.) to investigate whether changes in berry water loss at veraison influence the pattern of sugar accumulation. Berry transpiration was induced to vary either by changing the vapour pressure deficit (VPD) around bunches through temperature or relative humidity (RH) manipulations, or by applying drying-accelerating emulsion or a hydrophobic coating (vaseline) over the berry skin. In all the experiments, every berry of every bunch was rated prior to treatment as to softness and colour intensity. Transpiration rates were derived either from measurements of attached bunches using a custom-built open gas-exchange system or from weight loss calculated for single excised berries. Berry development and ripening were monitored throughout each experiment as deformability, fresh weight, and sugar concentration and content. Berries either did not respond to VPD-enhancement or showed reduced water loss when bunches were subjected to high temperature. Low berry transpiration in the latter treatment led to lower sugar content per berry up to harvest, and berry transpiration and net sugar intake were linearly correlated up to 0.20–0.25 mmol/m².s. A similar correlation was found also in the coating experiment within the first five days after treatments. When berry transpiration was restricted by applying vaseline, sugar accumulation was retarded. Low values of VPD, induced by raising the RH around the bunches, lowered sugar concentration but not sugar content per berry. The present study provides preliminary evidence that sugar accumulation in grapevine berries responds to changes in the evaporative demand around the bunch at veraison.     

Effect of short-term temperature and shading on fruit-set, seed and berry development in model vines of V. vinifera, cvs Chardonnay and Shiraz

Small Chardonnay and Shiraz vines were grown under controlled environment conditions at 25/20°C and 400 μE/(m².s) radiant energy. These conditions were interrupted for some plants for one week, at two periods near flowering, by imposing reduced temperatures (17°/14°C or 12°/9°C) and shading (8, 40 or 72% light reduction), in factorial combinations. Fruit-set was reduced by exposure to the lower temperature regime in both cultivars (confirming the results of a previous experiment) and tended also to decrease with increasing intensity of shading. Seed characteristics were examined only in Chardonnay where total number of seeds per berry was not affected by any of the treatments. However, pre-flowering exposure to 12°/9°C resulted in fewer berries with one or two sinker seeds and more berries with floater seeds and, overall, in a greater proportions of seeds being floaters. Shading had little effect on seed development. Berries on vines exposed to 12°/9°C, at both stages of development, had lower pericarp and seed weights compared with those treated with 17°/14°C or 25°/20°C. At harvest, pericarp weight and seed weight per berry were positively correlated; the linear regression for berries exposed to the higher temperature regimes had a significantly steeper slope than that for berries exposed to the lowest temperature regime. It is concluded that differences in pericarp weight were due to differing cell numbers. These differences may have developed firstly in the ovary due to a direct effect of exposure to varying temperature regimes, and secondly in the developing berry after fruit-set because of variation in the stimulatory activity of the seeds, caused by residual effects of pre-flowering temperature conditions on ovule and seed development.     

Effects of site, season and viticultural practices on grape seed development

The effects of vineyard site, season, clone and viticultural management practices were examined as possible sources of variability in seed development between populations of grape berries, cvs Traminer, Riesling and Pinot Noir. Mean values for number and weight of seeds per berry were similar in most seasons and sites, but some were significantly different. In some but not all cases, smaller means for seed number per berry were associated with larger mean weight per seed. Trellising, training, pruning level, canopy manipulation and regulation of bunch number per vine had little effect on seed complement per berry except that mean weight per seed was smaller with minimal pruning. Reduction in water supply imposed after flowering did not alter seed number but reduced seed weight. Berries of boron-deficient vines had fewer and smaller seeds. The implications of seed development for berry development are discussed.     

Pre‐budburst temperature influences the inner and outer arm morphology,phenology, flower number,fruitset, TSS accumulation and variability of Vitis vinifera L. Sauvignon Blanc bunches

Background and Aims: Inflorescence morphology, flower formation and subsequent fruit development of grapevines are influenced by genetic, environmental and cultural practices. While the effect of temperature on inflorescence primordia number per bud is documented, its effect during dormancy and budburst (BB) on floral development is less clear. In our study, winter dormant buds were passively heated for different periods of time from mid‐winter (July) to BB (October). Methods and Results: Individual canes were heated pre‐BB for different periods in polyethylene tunnels. Heating during winter dormancy had no effect on the number or position of inflorescences on the shoot, or on the type of structure occurring at the outer arm position of the inflorescence. Heating buds for either July or August to BB advanced the date of BB and the start date of flowering by 12–14 and 14 days, respectively, compared with that of the unheated Control treatment. There was a significant (P < 0.05) negative correlation between the mean temperature 12 days pre‐BB and flower number per shoot. The time between 50% flowering and the fruit TSS reaching 14°Brix was not influenced by winter heating or the date of flowering; the difference in this time interval, however, between the inner and outer arm components increased as shoot node number increased. Conclusions: Flower number per inflorescence had the greatest influence on the proportion of fruitset. The date of flowering of the outer arm was later at higher inflorescence positions on the shoot and later than that of the inner arm. Likewise, the outer arms of bunches higher up the shoot took longer from flowering to accumulate a fruit TSS of 14°Brix. Significance of the Study: Much of the within‐vine variability in berry composition at harvest can be attributed to the position of bunches on the shoot and the presence of an outer arm. The outer arm on apical bunches took significantly longer to progress from flowering to a fruit TSS of 14°Brix, suggesting that removal would significantly reduce within‐vine fruit variability.     

Relationships between seed and berry development of Vitis Vinifera L. cv Shiraz: Developmental changes in seed morphology and phenolic composition

Berries were collected regularly from fruit set to berry maturity from irrigated (Shiraz) grapevines in a Barossa valley vineyard. Seeds were removed for detailed study of physical attributes (weight, moisture, colour) and phenolic composition (seed tannins). Three phases of seed growth and development were discerned: (1) a phase of seed growth characterised by a steady increase in both fresh weight and dry weight, biosynthesis and accumulation of flavan‐3‐ols and tannins, and green appearance; (2) a transition phase where seed fresh weight and dry weight reached a maximum, but with continuing enlargement of the basal end. Accumulation of flavan‐3‐ols and seed tannins also reached a maximum during phase 2, and was accompanied by an onset of tannin oxidation, and yellow appearance; and finally, (3) a phase of seed drying and maturation defined by a decrease in fresh weight due to water export, a sustained oxidation of tannins, and overall brown appearance. These phases in seed development correspond to particular stages in berry development. Seeds reached maximum fresh seed weight and full size at the beginning of berry colouring (veraison), while maximum dry seed weight coincided with maximum berry weight. Changes in seed phenolics were linked to berry development and maturation. Changes in seed coat colour were also related to developmental changes in berry anthocyanins and total skin phenolics, indicating that the external appearance and colour of the seed coat may be used as an additional indicator of overall berry ripeness. A graduated colour chart was developed to provide an objective index of seed coat colour and thus developmental status of seeds and berry.     

Irrigation of grapevines with saline water at different growth stages: Effects on leaf,wood and juice composition

Background and Aims: This study reports on the effects that timing of saline irrigation has on leaf and wood tissue concentrations of Na⁺ and Cl^– and on juice composition. Methods and Results: Colombard vines on Ramsey rootstock were drip irrigated with saline water during any one of four annual growth stages: pre‐flowering, berry formation, berry ripening and postharvest. At other times, vines were irrigated with non‐saline water as was the control. Salts were annually flushed from the rootzone. Over six seasons, saline irrigation caused five‐ and sevenfold rises, respectively, in the leaf and juice Na⁺ concentrations and two‐ and fourfold rises in respective Cl^‐ concentrations. Saline irrigation raised juice pH and this was associated with a rise in juice Na⁺. Normalising responses for inter‐treatment differences in the seasonal salt load to isolate the effects of timing showed that juice Cl^‐ concentration was most sensitive to saline irrigation during berry formation and juice malate concentration most sensitive to saline irrigation pre‐flowering. Conclusion: Cl^‐ uptake was greatest when saline irrigation was applied early in organ formation, whereas Na⁺ uptake reflected seasonal salt load in irrigation water. Significance of the Study: In vines on the chloride excluding rootstock Ramsey, yield loss under saline irrigation was associated with high concentrations of sodium in the leaf.     

Analysis of tannins in seeds and skins of Shiraz grapes throughout berry development

The flavan-3-ol and proanthocyanidin composition of both seeds and skin of Vitis vinifera L. cv. Shiraz grapes was determined by reversed-phase HPLC after acetone extraction and acid-catalysis in the presence of excess phloroglucinol. Samples were taken at weekly intervals from fruit-set until commercial harvest. The main period of proanthocyanidin accumulation in grape seeds occurred immediately after fruit-set with maximum levels observed around veraison. Over two seasons there was variation in both the timing and content of proanthocyanidins in seeds. In skin, proanthocyanidin accumulation occurred from fruit set until 1–2 weeks after veraison. Proanthocyanidin subunit composition was different in seeds and skin and changed during berry development but the mean degree of polymerisation of the tannin polymers in skins was higher than in the seeds at all stages of berry development. Proanthocyanidin levels in both seeds and skin decreased between veraison and harvest. Additional proanthocyanidin subunits were released when the residues remaining after acetone extraction were subjected to direct acid-catalysis in the presence of phloroglucinol. In the seeds, these accounted for much of the post-veraison decrease, but not in grape skin. At harvest, 75% of extractable berry proanthocyanidin was in the seeds. Accumulation of proanthocyanidins in the seeds appears to be independent of that in the skins, but in both tissues synthesis occurs early in berry development and maximum levels are reached around veraison.     

Berry size and vine water deficits as factors in winegrape composition: Anthocyanins and tannins

Soluble solids, seed tannin, skin tannin, and skin anthocyanin were measured in fruit from Cabernet Sauvignon vines that had experienced either High, Control or Low water status during ripening. Berries from each treatment were segregated into 6 size categories at harvest in order to test independently for relationships due to size compared with those due to water deficits. Berry content of all solutes increased approximately in proportion to the increase in berry size. Deviations from proportionality caused Brix and anthocyanin concentration (mg per unit berry fresh mass) to decrease, and the concentration of skin tannin to remain unchanged or decrease slightly with increasing berry size. The concentration of seed tannin did not decrease and appeared to increase with berry size in multiple-seeded berries. In comparison with skin tannin or anthocyanin content, seed tannin content varied more with berry size and less with vine water status. In addition to decreasing berry size, water deficits increased the amount of skin tannin and anthocyanin per berry and the concentrations of skin tannin and anthocyanins, but did not significantly affect the content or concentration of seed tannin. The results show that there are effects of vine water status on fruit composition that arise independently of the resultant differences in fruit size. The effect of vine water status on the concentration of skin tannin and anthocyanin was greater than the effect of fruit size on those same variables. However, the increases in skin tannin and anthocyanin that accompanied water deficits appear to result more from differential growth sensitivity of inner mesocarp and exocarp than direct effects on phenolic biosynthesis.