The differences were exacerbated in cluster weight , berries/cluster , and yield when 66% of the leaf area was removed. In 2018, the effect of defoliation was evident with a 12% decrease in berry weight. As in the previous year, we saw a diminution in cluster weight, berries/cluster, and yield. However, the decline in yield in 2018 was 56% when 66% of the leaves were defoliated. Fruit removal was effective in modulating the cluster number and thus the yield in both experimental years as expected . Furthermore, we measured a strong linear decrease in cluster weight in 2017. However, the same response was not evident in 2018. Removing 66% of the cluster resulted in a 55 and 60% decrease in yield of Cabernet Sauvignon in 2017 and 2018, respectively. Surprisingly, we did not measure an interaction of defoliation and fruit removal on components of year in either of the experimental years . The carry-over effects of source-sink adjustments on components of yield in 2019 were strongly evident ; even though no defoliation or fruit removal treatments were applied. Berry weight, cluster number per vine, cluster weight, and yield per vine were all affected by the carry over effects of defoliation from the previous 2 years. They all declined linearly with the 33% defoliation treatment. Conversely, large plastic pots we did not measure a carryover effect of fruit removal in 2019 in the majority of components of yield monitored. There was an interaction of defoliation and fruit removal in 2019 on the number of berries per cluster.
The 33%L-100%F had the fewest berries per cluster compared to all other treatment combinations.Canopy area was affected by defoliation, but not with fruit removal or its interaction with defoliation during the experimental years . There was a strong linear trend, as expected with defoliation; where removing 66% of canopy area resulted in a 65% decrease of it in 2017 ; and a 58% in 2018 . The carryover effects of source-sink adjustments on the canopy area in 2019 were evident. The strong linear decrease in 33%L continued to this year, as the grapevines did not recover from the removal of 66% of their leaf area, regardless of fruit removal in previous years. In 2017, both defoliation and fruit removal affected leaf area to fruit ratio; however, the interaction amongst them was not significant . There was a strong linear trend where leaf area to fruit ratio decreased by 47% when66% of the canopy area was removed . Conversely, we measured a strong linear trend where leaf area to fruit ratio increased by 56% this time when 66% of the fruit was removed. In 2018, defoliation and fruit removal interacted to affect the leaf area to fruit ratio . Within the interaction, there was a linear trend where leaf area to fruit ratio decreased linearly as the number of clusters retained increased. We measured the greatest leaf area to fruit ratio with 100%L-33%F as well as in 66%L-33%F. The 33%L-100F and 66%L-100F had similar leaf area to fruit ratio . There was no carry over effect of source-sink adjustments on leaf area to fruit ratio in 2019 , nor was there an effect of main effects carrying over to this year . We measured an interactive effect of fruit removal and year on Ravaz Index . During both experimental years,there was a strong trend of fruit removal on Ravaz Index . In 2017, removing 66% of fruit resulted in a 56% decrease of Ravaz Index. We saw a similar response in 2018 as well. There was no effect of defoliation within the experimental years. We also did not measure an interactive effect of defoliation and fruit removal on Ravaz Index either.
There was no carryover effect of source-sink adjustments on Ravaz Index in 2019 ; and we did not measure a carry-over effect of main effects of Ravaz index either . Berry TSS was affected strongly by the defoliation treatments during the experimental years, and we did not measure an interactive effect of defoliation and fruit removal in 2017 or in 2018 . There was a linear increase in TSS as the severity of defoliation decreased in both experimental years . The effect of fruit removal on speed of ripening in 2017 was negligible. However, we also saw a strong effect of fruit removal on berry TSS where it declined linearly with the decrease in fruit removal severity in 2018 . This indicated a level of self-adjustment in previous grapevine season’s applied treatments. Conversely, under free-growth in 2019 when the grapevine was allowed to sprawl without any defoliation and fruit removal, we saw a reversal of the trends with defoliation . Although there were limited main and interactive effects, in the absence of fruit removal there was a linear decrease in TSS accumulation as the severity of defoliation treatments decreased, revealing the carry-over effect .We measured a significant year and defoliation interactive effect on AN, and gs . There was never an interactive effect of fruit removal and year on any of the leaf gas exchange variables monitored. In regards to AN during the experimental years , there was a strong linear trend where AN and gs increased linearly with the increase in the severity of defoliation . In 2017, AN and gs increased by 10 and 13%, respectively, when 66% of leaves were defoliated. Defoliating 66% of leaves resulted in a 23% increase in AN, and a 30% increase in gs integrals in 2018. In either experimental year we did not see an effect of cluster thinning on leaf gas exchange integrals or an interaction of defoliation and fruit removal. The carryover effects to source sink adjustments in 2019 were inconsistent.
We did not measure any significant effects of defoliation, fruit removal, or their interaction on AN, or gs . However, as mentioned above, the year effect on AN, and gs were significant. There was a quadratic response to years where AN and gs declined from 2017 to 2018 but then increased significantly in 2019.We destructively harvested the grapevines following the 2018 growing season and separated them into roots, trunk, and aerial organs. Trunk and cordon masses were not affected by the defoliation, fruit removal, or their interactive effects . Shoot and root masses were affected by the defoliation treatments. The shoot mass decreased by 1/3 with the 33%L treatment compared to the 100%L treatment . Root mass decreased by 1/4 with the 33%L treatment when compared to 100%L treatment . The total grapevine mass was also 20% lower in the 33%L treatment compared to 100%L treatment. Fruit removal did not affect plant biomass or biomass accumulation in plant organs, and there was no interaction of defoliation or fruit removal evident in our work. The starch accumulation in grapevine roots was affected by defoliation treatments and time of sampling . We did not measure an interactive effect of defoliation and fruit removal. The starch accumulation was affected at similar times during the experiment during each year. In 2018, the root starch content of 33%L was one-third of 100%L starting in mid-ripening until harvest . The starch content of 33%L roots, however, raspberry container equilibrated to the same content of 100%L by December 2018 when pruning was conducted. In 2019, when no treatments were applied to observe the carry over effects, we saw a reversal of this response. The starch content of root tissues of the 33%L treatments had ca. 20 and 40% more starch than 100%L in 2019 during mid-ripening through harvest. Crop level rather than defoliation affected the soluble sugars content in the roots . We did not measure the interactive effect of defoliation and fruit removal on soluble sugars at any time point during the experiment. In contrast to root starch accumulation , the soluble sugars were greater in the 33%F when compared to 100%F at treatment application in 2018. However, their content decreased to ½ of 100%F by December 2018 . In the follow-up year where grapevines grew without treatments, we saw transient differences in root soluble carbohydrates at mid-ripening where 100%F had greater content than 33%F. However, post-harvest in 2019, the soluble carbohydrate content in roots of 33%F was ca. 2× of 100%F prior to leaf fall. At pruning time, compared to the previous year; there was no difference.The day of bud break was not affected by defoliation treatments during the experimental years or the follow-up year of 2019 . However, bud break was 1 day later in 2019 in the 66%F and 100%F treatments compared to 33%F . The date of flowering was not affected by defoliation treatments in 2018 . However, it was 1 day later with 100%L in 2019. Likewise, we did not see a shift in flowering date with fruit removal in 2018, but a carry-over effect was evident with 33%F being 2 days earlier than 66%F and 100%F in 2019 . In 2017, the treatments started having an impact few weeks after they were imposed. Veraison was delayed by defoliation treatments by ca. 5 days in 2017 and 2018 .
Leaf senescence was consistently delayed the defoliation treatments . Extreme defoliation delayed leaf senescence by 9, 5, and 2 days in 2017, 2018, and 2019, respectively. Conversely, we saw a reversal of this trend where retaining more fruit delayed leaf senescence, albeit with a less strong effect than defoliation.At the moment when the treatments were imposed, the proportions between the numbers of leaves to clusters were extremely different amongst treatments. These differences were self-balanced largely by harvest. Changes in berry size explained a great part of this variation. Typically, berry size is managed by limiting the access of grapevine to water . However, in this case, defoliated plants had a most certain lower consumption of water, and therefore, better water status , but this was concomitant to a reduction in berry mass which ruled out the water status as cause. Another commonway to manipulate berry size is a delay in cluster thinning that encourages a competition of fruits early in their development, making berries smaller . Mechanistic experiments revealed that berry growth and the import of sugars may not occur in absence of the one or the other , and therefore, defoliationmay have induced lower berry size through a reduction a sugar allocation . Berries per cluster were also affected by the severity of defoliation. This was a complex result as number of flowers, percentage of fruit set, or spontaneous fruit abortion may determine the number of berries at harvest development. However, in the 1st year of treatments, berry abortion after pepper-corn stage was evident. This was in fact the case, and vines in 33%L treatment displayed berry abortion close to veraison which was associated to carbon starvation . Exogenous gibberellin applications are typically used in commercial table grape production to reduce cluster compactness through flower abscission . Endogenous gibberellin and auxin determine the number of berries set, therefore affecting the number of berries and thusly berry size . Although reductions in the number of berries per cluster were significant in the second season, berry abortion was not evident, which suggested this was an acclimation response to the treatments of the previous season. This was reported by previous studies where the number of inflorescence and flowers were reduced by defoliation during the previous season .Achieving vine balance has been remarked as key to achieve an adequate ripeness . However, balancing leaf area to fruit mass is a very precise task that involves a great amount of hand-labor or investment in a large equipment park to conduct these practices mechanically . In this pursuit, trellis, vine spacing, canopy management, and irrigation have been proposed as indirect methods to optimize LA/FM . Thus, the concept of vine balance appears to work in a practical sense, as it is used to make cluster-thinning decisions and but not defoliation. The main factor that made leaf area outweigh the value of fruit mass is that fruit is only a fraction of the total amount of carbon assimilated by leaves throughout the year is translocated to fruit . Furthermore, leaves do not only assimilate carbon , and canopy size has deeper implications than fruit mass for root growth and the transduction of ripening signals.