Influence of flowers per cluster, fruit-set and berry weight on cluster weight in Verdejo grapevine (<em>Vitis vinifera</em> L.)

Estimation of yield has always been one of the greatest concerns in the wine-making industry, which is why researchers have attempted to create models and methods that make accurate predictions. However, the variability of the yield components, which depends on the environmental factors, like light and temperature, and on cultural and natural factors which affect vigor, water status, photosynthetic and hormonal activity and the accumulation of reserve substances, make the prediction of the future harvest very difficult to calculate. In certain situations, fertility expressed as the number of bunches, tends to display very stable values from one harvest to the next. The weight of the cluster is the performance factor that has the greatest variability and mostly causes the variations in production. In this report, we analyze the influence of the number of flowers per cluster, the fruit-set percentage, the number of berries per cluster and the berry weight in the variability of the cluster weight of the Verdejo variety (Vitis vinifera L.). The cluster weight and its components have been determined along canes of 6 and 10 buds, in the harvests of 1999, 2000 and 2001. The yield component upon which cluster weight variation fundamentally depends has been the number of berries it contains, whereas the berry weight has had much less of an effect. The number of berries per cluster is mainly determined by the number of flowers per inflorescence, a yield component which has gone so far as to explain, by itself, an average of 75%of the variability in the number of berries per cluster and 70 % of the variability of the cluster weight. Despite the possible influence that the fruit-set percentage may have on the variability in the number of berries per cluster, measuring the number of flowers per inflorescence would make it possible to know, to a reasonable level of accuracy, the future number of berries per cluster long before the harvest date. At the same time, if a yield estimate is not needed far in advance, prediction of the future cluster weight can be determined by measuring the number of berries per cluster 15 or 20 days after fruit-set time, when the falling of small fruits is right minimal. In any case, whether by using the number of flowers per inflorescence or the number of berries per cluster, the berry weight on the harvest date would be the only yield component that would require estimation, which is usually much easier to achieve due to the lesser variability seen in this parameter.