A Nonparametric Model for Analysis of Flowering Patterns of Herbaceous Multi-flowered Monocarpic Shoots

One of the approaches to plant development description involves phenological curves, which represent the time variations of certain traits. Most models applied to various plant taxa and life forms describe their phenology, including flowering, at the population level, and insufficient attention is paid to the modeling at the individual one. Individual modeling is more complicated than populational one owing to the multilevel structure of a phenotype. And as a result, it is accompanied by a significant increase in the number of model parameters, many of which are both interdependent and non-transferable between species. Here, we present a simple structural-dynamic data-driven model to describe the flowering patterns of individual monocarpic shoots of herbaceous plants with multi-flowered inflorescences. This model is nonparametric, thus being convenient and applicable to various plant species. Our results showed the operability of the proposed model and its potential in the detection of hidden trends in the input data. This is illustrated by the example of the herbaceous polycarpic Campanula sarmatica, Campanulaceae family, whose inflorescences demonstrate a wide variability in the structure and dynamics of flowering. For example, using our model, we found that the shape of the flowering curve of C. sarmatica shoots as a whole is determined by only two factors, the ratio of the number of flowers on the main and lateral axes, and the time of the flowering shift from the main axis to the lateral axes. The proposed model accurately mimics the individual flowering pattern with due natural variability. It can be used to simulate the flowering of a group of individuals in cenopopulational studies or practical tasks, e.g., in landscape design. Flowering patterns that characterize inflorescence development at the individual level can serve as informative phenotypic traits in anthoecological and developmental studies, and in inflorescence diagnostics.