Floral and Fruit Phenology in Three Young Olive (Olea europaea L.) Cultivars Grown in Central Northern Morocco. Part III: Correlation and Regression Study of Phenological Traits and Agroclimatic Requirements According to the Specific BBCH Scale
摘要
Knowledge of the relationship between agronomical traits and climatic variables is crucial for plant breeding and olive grove establishment and management. Here, we examined the relationship between the main reproductive phenological phases of olive (‘Arbequina’, ‘Arbosana’, and ‘Picholine Languedoc’), their agroclimatic (chill and heat) requirements, chlorophylls as well as flower and fruit density according the BBCH scale. Such relationships were investigated through Pearson correlations, simple (SRM) and multiple regression models (MRM), and principal component analysis (PCA). Our outcomes reveal that PCA separated between the Spanish cultivars (‘Arbequina’ and ‘Arbosana’) and the French cultivar (‘Picholine Languedoc’) on one hand and the cardinal points into two groups (south-east and north-west) on the other hand. Such separation was according the first two components with a total variance exceeding 80%. Following Pearson correlations, in general, there were significant positive (r up to 0.943***) correlations between inflorescence events BBCH51–55, among flowering times BBCH61, 65 and 69, and between ripening times (BBCH80–89). BBCH51 was positively associated with the available chill (r up to 0.959***). Duration and onset of inflorescence and flowering were negatively associated (r up to 0.606*). Timing of inflorescence onset, full flowering, and ripening were strongly dependent on their respective heat requirements (growing degree days [GDD] and growing degree hours [GDH], r exceeded 0.999***). Full flowering (BBCH65) was positively associated with GDD and GDH accumulated during inflorescence (r up to 0.930***). Moreover, GDD and GDH required for fruit development and ripening were negatively correlated with the durations of these events. Fruit set was positively correlated with flower bud density and density of early and fully ripe fruits (r 0.692–0.773*). Based on the correlation matrix, several SRM were developed for timing of phenological events (BBCH51–65–80–89), ripening duration, flower density, fruit set and density (R2 up 0.9998). MRM were developed for timing of phenological events (BBCH51–55–60–61–65–69–80–89), their durations, fruit set and density (R2 up 0.9995). In conclusion, SRM and MRM could provide important insights into the timing and duration of some phenophases in the studied cultivars for better programming of agronomic practices.
Graphic abstract