De la fleur au fruit : étude et modélisation de la floraison, de la fécondation-fructification et de la croissance du fruit chez le goyavier-fraise (Psidium cattleianum)
Strawberry guava (Psidium cattleianum) domestication is studied within a project of crop diversification in humid highlands o f Reunion Island. The number of fruit per plant and their individual weight determine the commercial yield. These characteristics are highly variable and do not seem to be linked to the genotype. An agronomic approach o f domestication is proposed for this species. It consists in identifying how are built-up these characteristics and their variability through the different steps of the production cycle and under the effect of environmental, physiological and genetic factors. The objective is to act on some of these factors using appropriate cultivation techniques. Main features of this approach are the genetic diversity and the lack of references on the species, as confirmed by a bibliographical analysis. Strawberry guava phenological cycle was studied on feral stands in Reunion Island. In cultivation, nitrogen modifies the natural phenological cycle by triggering a new production cycle. The setting of the study is a natural production cycle in an orchard, without interaction with triggered cycles. A model representing the process of fruit number and individual weight build-up is proposed, structured around five steps: number of flowers build-up, flowering, pollination, fecundation-fruit set, and fruit growth. The model outputs are the individual fruit weight and harvest dates distributions at the orchard level. Its functions are simulation and prediction in order to assist farmers in their technical choices. This work deals with the flowering, fecundation-fruit set and fruit growth steps. The plant genetic diversity induces a flowering asynchrony which is not taken into account by flowering model conceived for a monovarietal orchard. A general stochastic flowering model is then built. It describes flowering at different levels and considers the flowering asynchrony both within and between plants. A controlled pollination experiment showed that strawberry guava is partially self-compatible, and that self-compatibility level varies among plants. The distribution of seed number per fruit depends on the individual, on the pollen genetic origin and on the pollination conditions. The modelling of pollination and fecundation in a genetically heterogeneous population appears complex and further research is needed. A general approach to model the effect of growth factors on individual fruit growth curve is presented. It is then applied to strawberry guava fruit growth, with the seed number per fruit and the plant leaf: fruit ratio as growth factors. Statistical errors are associated to the prediction model in order to estimate the output variability. The distribution of harvest dates at the plant level is estimated from the plant leaf: fruit ratio and the mean number of seeds per fruit. The general approach used in this study and the results are discussed, and further research is outlined. Practical consequences of this work are underlined. Finally, the generalisation of the agronomic approach of domestication and of the models is discussed.
| Summary: | Strawberry guava (Psidium cattleianum) domestication is studied within a project of crop diversification in humid highlands o f Reunion Island. The number of fruit per plant and their individual weight determine the commercial yield. These characteristics are highly variable and do not seem to be linked to the genotype. An agronomic approach o f domestication is proposed for this species. It consists in identifying how are built-up these characteristics and their variability through the different steps of the production cycle and under the effect of environmental, physiological and genetic factors. The objective is to act on some of these factors using appropriate cultivation techniques. Main features of this approach are the genetic diversity and the lack of references on the species, as confirmed by a bibliographical analysis. Strawberry guava phenological cycle was studied on feral stands in Reunion Island. In cultivation, nitrogen modifies the natural phenological cycle by triggering a new production cycle. The setting of the study is a natural production cycle in an orchard, without interaction with triggered cycles. A model representing the process of fruit number and individual weight build-up is proposed, structured around five steps: number of flowers build-up, flowering, pollination, fecundation-fruit set, and fruit growth. The model outputs are the individual fruit weight and harvest dates distributions at the orchard level. Its functions are simulation and prediction in order to assist farmers in their technical choices. This work deals with the flowering, fecundation-fruit set and fruit growth steps. The plant genetic diversity induces a flowering asynchrony which is not taken into account by flowering model conceived for a monovarietal orchard. A general stochastic flowering model is then built. It describes flowering at different levels and considers the flowering asynchrony both within and between plants. A controlled pollination experiment showed that strawberry guava is partially self-compatible, and that self-compatibility level varies among plants. The distribution of seed number per fruit depends on the individual, on the pollen genetic origin and on the pollination conditions. The modelling of pollination and fecundation in a genetically heterogeneous population appears complex and further research is needed. A general approach to model the effect of growth factors on individual fruit growth curve is presented. It is then applied to strawberry guava fruit growth, with the seed number per fruit and the plant leaf: fruit ratio as growth factors. Statistical errors are associated to the prediction model in order to estimate the output variability. The distribution of harvest dates at the plant level is estimated from the plant leaf: fruit ratio and the mean number of seeds per fruit. The general approach used in this study and the results are discussed, and further research is outlined. Practical consequences of this work are underlined. Finally, the generalisation of the agronomic approach of domestication and of the models is discussed. |
|---|