**4. Discussion**

Flame suckering reduced the number of suckers observed before the start of the experiment each year (Table 2). The date of flaming influenced the final number of suckers. In fact, in 2017 and 2018, the suckers removed with flaming on 21 April 2017 and 2 May 2018 led to a significant resprouting of suckers compared to the flaming two weeks later on 5 May 2017 and 16 May 2018 (or applied on both dates), which resulted in a nonsignificant reduction in the initial number of suckers (Table 2). Flame suckering was more effective if not carried out too early, because although the suckers were more developed (15–16 BBCH growth stage *vs* 12–13 BBCH growth stage), and therefore would seem more difficult to devitalize, sucker resprouting is more difficult due to the more developed main grape productive shoots (18-19 BBCH growth stage *vs* 15–16 BBCH growth stage).

In fact, a vine is a 'acrotonic branching' plant due to the distal position of shoots, irrespectively of its size relative to the parent stem [26]. More developed shoots exploit their higher sink strength, and absorb more nutrients [27,28]. Apical dominance of *Vitis vinifera L.* increases with increasing temperature [29]. This explains why, over time, there was a significant decrease in the resprouting of basal buds. The amount of resprouting, in the first two weeks after the start of the experiment, was the same as the resprouting in the hand-suckered plots, where suckers were completely removed during the operation. However, two weeks later the number of suckers was no longer zero, but on average there were two suckers per plant in 2016 and one sucker per plant in 2017 and 2018. This amount, added to approximately one sucker per plant which flaming did not devitalize, led to the nonsignificant decrease in the number of suckers compared to the initial number when FlamingA was applied in 2017 and 2018. There was no resprouting in the control plants because new buds only sprout when buds (e.g., the suckers) have been removed from the plant [4].

In the three-year experiment, the yield components and grape composition were similar among treatments, suggesting that flaming did not damage the grapevine plants, or modify the quality of the grapes. Byrne and Howell [3] found that sucker removal increased the yield per vine. In our experiment, the low plant potential fertility due to the low number of buds per vine, the tendency to an excess of total leaf surface per plant in relation to the quantity of grape production, and the few suckers per plant (a maximum of around four suckers per control plant in 2016), although statistically higher compared to the flame- and hand-suckering, was probably not enough to negatively affect the production of grapes.

Although the number of suckers per plant did not influence the yield, and suckers were manually removed during the winter pruning (as per general vineyard procedure), suckering during the spring is nevertheless required. This is because the presence of suckers during the vine growing season can cause problems during vineyard managemen<sup>t</sup> such as weed, insect, and disease control, and mechanical harvest [3]. Moreover, postponing the suckering of a high number of suckers until winter pruning would significantly increase the cost of manually removing the excessively developed and lignified shoots, and most of all, would increase the development of basal buds capable of resprouting [4]. Finally, the majority of vine disease pathogens infect via winter wounds, because winter is associated with high rainfall, which increases inoculum availability, increasing the risk of sucker wounds acting as portals for grapevine trunk pathogen infections [30,31].

The reduction in yield from 2016 to 2018 cannot be attributed to the suckering, because the reduction was found in all treatments indiscriminately. It was probably the result of the dry year in 2017 (Figure 1), which led to a lower potential fertility for 2018, resulting in a lower number of clusters. The higher average berry weight estimated in 2018 was probably because of the lower number of clusters and the higher rainfall occurring during the 2018 growing season compared to 2016 and 2017 (Figure 1). Concerning grape composition, as the total soluble solids increase in the berries, the juice pH rises and the tartaric acid declines.
