Files
Abstract
Plant roots display a wide range of architectural types, each with unique spatial arrangement and shape characteristics. Traditional theoretical models don't fully capture this root architectural variation, often attributing it to genetic (G) and environmental (E) variation or the G by E interaction. Addressing this, we developed Dirt-Pop, a computational pipeline, designed to cluster the root architectural variations of a single genotype into multiple root architecture types. This pipeline employs the DS-curve as a shape descriptor, integrates the K-means++ clustering algorithm, an outlier removal strategy, and the Fréchet similarity metric. Applying this pipeline, three common bean (Phaseolus vulgaris L.) genotypes (DOR364, L8857 and SEQ7) exhibited five distinct root architecture types, with their composition varying under different water conditions. Validating DOR364 and SEQ7 in a mesocosm system, where water distribution was monitored by soil moisture sensors, both genotypes repeatedly displayed these five root architecture types. Moreover, the composition of SEQ7's five root architecture types changed across developmental stages and water conditions. By linking these root architecture types to published simulation models, each root architecture type observed can be assumed with a specific function in water and nutrients (Phosphorus and Nitrogen) uptake. We further investigated how root architecture types and biomass allocation impact fitness outcome in both monoculture and mixture of SEQ7 and DOR364 under water-limited (WL) and non-limiting conditions to explore plant-plant interactions through the lens of resource partitioning and kin selection theories. The study reveals that mixtures of these two genotypes exhibited greater population fitness than monocultures. SEQ7 showed a significant increase in population fitness in mixtures, attributed to its tendency to maintain or reduce root biomass allocation, especially under WL conditions, and a strategic shift towards more deep root architecture types, enhancing water acquisition. In contrast, DOR364 increased root allocation for belowground resources acquisition in mixtures, but this did not confer a fitness benefit. These results underline the complexity of plant interactions, showing that neither kin selection nor niche partitioning theories fully explain the observed trait expression and fitness outcome.