Phenotyping the kinematics of leaf development in flowering plants: recommendations and pitfalls

Genetic architecture of variation in Arabidopsis thaliana rosettes

Rosette morphology across Arabidopsis accessions exhibits considerable variation. Here we report a high-throughput phenotyping approach based on automatic image analysis to quantify rosette shape and dissect the underlying genetic architecture. Shape measurements of the rosettes in a core set of Recombinant Inbred Lines from an advanced mapping population (Multiparent Advanced Generation Inter-Cross or MAGIC) derived from inter-crossing 19 natural accessions. Image acquisition and analysis was scaled to extract geometric descriptors from time stamped images of growing rosettes. Shape analyses revealed heritable morphological variation at early juvenile stages and QTL mapping resulted in over 116 chromosomal regions associated with trait variation within the population. Many QTL linked to variation in shape were located near genes related to hormonal signalling and signal transduction pathways while others are involved in shade avoidance and transition to flowering. Our results suggest rosette shape arises from modular integration of sub-organ morphologies and can be considered a functional trait subjected to selective pressures of subsequent morphological traits. On an applied aspect, QTLs found will be candidates for further research on plant architecture.

High-throughput phenotyping analysis of maize at the seedling stage using end-to-end segmentation network

Image processing technologies are available for high-throughput acquisition and analysis of phenotypes for crop populations, which is of great significance for crop growth monitoring, evaluation of seedling condition, and cultivation management. However, existing methods rely on empirical segmentation thresholds, thus can have insufficient accuracy of extracted phenotypes. Taking maize as an example crop, we propose a phenotype extraction approach from top-view images at the seedling stage. An end-to-end segmentation network, named PlantU-net, which uses a small amount of training data, was explored to realize automatic segmentation of top-view images of a maize population at the seedling stage. Morphological and color related phenotypes were automatic extracted, including maize shoot coverage, circumscribed radius, aspect ratio, and plant azimuth plane angle. The results show that the approach can segment the shoots at the seedling stage from top-view images, obtained either from the UAV or tractor-based high-throughput phenotyping platform. The average segmentation accuracy, recall rate, and F1 score are 0.96, 0.98, and 0.97, respectively. The extracted phenotypes, including maize shoot coverage, circumscribed radius, aspect ratio, and plant azimuth plane angle, are highly correlated with manual measurements (R2 = 0.96-0.99). This approach requires less training data and thus has better expansibility. It provides practical means for high-throughput phenotyping analysis of early growth stage crop populations.

All roads lead to growth: imaging-based and biochemical methods to measure plant growth

Plant growth is a highly complex biological process that involves innumerable interconnected biochemical and signalling pathways. Many different techniques have been developed to measure growth, unravel the various processes that contribute to plant growth, and understand how a complex interaction between genotype and environment determines the growth phenotype. Despite this complexity, the term 'growth' is often simplified by researchers; depending on the method used for quantification, growth is viewed as an increase in plant or organ size, a change in cell architecture, or an increase in structural biomass. In this review, we summarise the cellular and molecular mechanisms underlying plant growth, highlight state-of-the-art imaging and non-imaging-based techniques to quantitatively measure growth, including a discussion of their advantages and drawbacks, and suggest a terminology for growth rates depending on the type of technique used.

Image-based methods for phenotyping growth dynamics and fitness components in Arabidopsis thaliana

BACKGROUND

The model species Arabidopsis thaliana has extensive resources to investigate intraspecific trait variability and the genetic bases of ecologically relevant traits. However, the cost of equipment and software required for high-throughput phenotyping is often a bottleneck for large-scale studies, such as mutant screening or quantitative genetics analyses. Simple tools are needed for the measurement of fitness-related traits, like relative growth rate and fruit production, without investment in expensive infrastructures. Here, we describe methods that enable the estimation of biomass accumulation and fruit number from the analysis of rosette and inflorescence images taken with a regular camera.

RESULTS

We developed two models to predict plant dry mass and fruit number from the parameters extracted with the analysis of rosette and inflorescence images. Predictive models were trained by sacrificing growing individuals for dry mass estimation, and manually measuring a fraction of individuals for fruit number at maturity. Using a cross-validation approach, we showed that quantitative parameters extracted from image analysis predicts more 90% of both plant dry mass and fruit number. When used on 451 natural accessions, the method allowed modeling growth dynamics, including relative growth rate, throughout the life cycle of various ecotypes. Estimated growth-related traits had high heritability (0.65 < H2 < 0.93), as well as estimated fruit number (H2 = 0.68). In addition, we validated the method for estimating fruit number with rev5, a mutant with increased flower abortion.

CONCLUSIONS

The method we propose here is an application of automated computerization of plant images with ImageJ, and subsequent statistical modeling in R. It allows plant biologists to measure growth dynamics and fruit number in hundreds of individuals with simple computing steps that can be repeated and adjusted to a wide range of laboratory conditions. It is thus a flexible toolkit for the measurement of fitness-related traits in large populations of a model species.

Image-based methods for phenotyping growth dynamics and fitness components in Arabidopsis thaliana

BACKGROUND

The model species Arabidopsis thaliana has extensive resources to investigate intraspecific trait variability and the genetic bases of ecologically relevant traits. However, the cost of equipment and software required for high-throughput phenotyping is often a bottleneck for large-scale studies, such as mutant screening or quantitative genetics analyses. Simple tools are needed for the measurement of fitness-related traits, like relative growth rate and fruit production, without investment in expensive infrastructures. Here, we describe methods that enable the estimation of biomass accumulation and fruit number from the analysis of rosette and inflorescence images taken with a regular camera.

RESULTS

We developed two models to predict plant dry mass and fruit number from the parameters extracted with the analysis of rosette and inflorescence images. Predictive models were trained by sacrificing growing individuals for dry mass estimation, and manually measuring a fraction of individuals for fruit number at maturity. Using a cross-validation approach, we showed that quantitative parameters extracted from image analysis predicts more 90% of both plant dry mass and fruit number. When used on 451 natural accessions, the method allowed modeling growth dynamics, including relative growth rate, throughout the life cycle of various ecotypes. Estimated growth-related traits had high heritability (0.65 < H² < 0.93), as well as estimated fruit number (H² = 0.68). In addition, we validated the method for estimating fruit number with rev5, a mutant with increased flower abortion.

CONCLUSIONS

The method we propose here is an application of automated computerization of plant images with ImageJ, and subsequent statistical modeling in R. It allows plant biologists to measure growth dynamics and fruit number in hundreds of individuals with simple computing steps that can be repeated and adjusted to a wide range of laboratory conditions. It is thus a flexible toolkit for the measurement of fitness-related traits in large populations of a model species.

Identifying developmental phases in the Arabidopsis thaliana rosette using integrative segmentation models

The change in leaf size and shape during ontogeny associated with heteroblastic development is a composite trait for which extensive spatiotemporal data can be acquired using phenotyping platforms. However, only part of the information contained in such data is exploited, and developmental phases are usually defined using a selected organ trait. We here introduce new methods for identifying developmental phases in the Arabidopsis rosette using various traits and minimum a priori assumptions. A pipeline of analysis was developed combining image analysis and statistical models to integrate morphological, shape, dimensional and expansion dynamics traits for the successive leaves of the Arabidopsis rosette. Dedicated segmentation models called semi-Markov switching models were built for selected genotypes in order to identify rosette developmental phases. Four successive developmental phases referred to as seedling, juvenile, transition and adult were identified for the different genotypes. We show that the degree of covering of the leaf abaxial surface with trichomes is insufficient to define these developmental phases. Using our pipeline of analysis, we were able to identify the supplementary seedling phase and to uncover the structuring role of various leaf traits. This enabled us to compare on a more objective basis the vegetative development of Arabidopsis mutants.

Phenotyping oilseed rape growth-related traits and their responses to water deficit: the disturbing pot size effect

Following the recent development of high-throughput phenotyping platforms for plant research, the number of individual plants grown together in a same experiment has raised, sometimes at the expense of pot size. However, root restriction in excessively small pots affects plant growth and carbon partitioning, and may interact with other stresses targeted in these experiments. In work reported here, we investigated the interactive effects of pot size and soil water deficit on multiple growth-related traits from the cellular to the whole-plant scale in oilseed rape (Brassica napus L.). The effects of pot size on responses to water deficit and allometric relationships revealed strong, multilevel interactions between pot size and watering regime. Notably, water deficit increased the root:shoot ratio in large pots, but not in small pots. At the cellular scale, water deficit decreased epidermal leaf cell area in large pots, but not in small pots. These results were consistent with changes in the level of endoreduplication factor in leaf cells. Our study illustrates the disturbing interaction of pot size with water deficit and raises the need to carefully consider this factor in the frame of the current development of high-throughput phenotyping experiments.

Novel imaging-based phenotyping strategies for dissecting crosstalk in plant development

In an era of genomics, proteomics, and metabolomics a large number of mutants are available. The discovery of their phenotypes is fast becoming the bottleneck of molecular plant physiology. This crisis can be overcome by imaging-based phenotyping, an emerging, rapidly developing and innovative approach integrating plant and computer science. A tremendous amount of digital image data are automatically analysed using techniques of 'machine vision'. This minireview will shed light on the available imaging strategies and discuss standard methods for the automated analysis of images to give the non-bioinformatic reader an idea how the new technology works. A number of successful platforms will be described and the prospects that image-based phenomics may offer for elucidating hormonal cross-talk and molecular growth physiology will be discussed.

Modelling the coordination of the controls of stomatal aperture, transpiration, leaf growth, and abscisic acid: update and extension of the Tardieu-Davies model

Stomatal aperture, transpiration, leaf growth, hydraulic conductance, and concentration of abscisic acid in the xylem sap ([ABA]xyl) vary rapidly with time of day. They follow deterministic relations with environmental conditions and interact in such a way that a change in any one of them affects all the others. Hence, approaches based on measurements of one variable at a given time or on paired correlations are prone to a confusion of effects, in particular for studying their genetic variability. A dynamic model allows the simulation of environmental effects on the variables, and of multiple feedbacks between them at varying time resolutions. This paper reviews the control of water movement through the plant, stomatal aperture and growth, and translates them into equations in a model. It includes recent progress in understanding the intrinsic and environmental controls of tissue hydraulic conductance as a function of transpiration rate, circadian rhythms, and [ABA]xyl. Measured leaf water potential is considered as the water potential of a capacitance representing mature tissues, which reacts more slowly to environmental cues than xylem water potential and expansive growth. Combined with equations for water and ABA fluxes, it results in a dynamic model able to simulate variables with genotype-specific parameters. It allows adaptive roles for hydraulic processes to be proposed, in particular the circadian oscillation of root hydraulic conductance. The script of the model, in the R language, is included together with appropriate documentation and examples.

Phenotyping and beyond: modelling the relationships between traits

Plant phenotyping technology has become more advanced with the capacity to measure many morphological and physiological traits on a given individual. With increasing automation, getting access to various traits on a high number of genotypes over time raises the need to develop systems for data storage and analyses, all congregating into plant phenotyping pipelines. In this review, we highlight several studies that illustrate the latest advances in plant multi-trait phenotyping and discuss future needs to ensure the best use of all these quantitative data. We assert that the next challenge is to disentangle how plant traits are embedded in networks of dependencies (and independencies) by modelling the relationships between them and how these are affected by genetics and environment.

Cell to whole-plant phenotyping: the best is yet to come

Imaging and image processing have revolutionized plant phenotyping and are now a major tool for phenotypic trait measurement. Here we review plant phenotyping systems by examining three important characteristics: throughput, dimensionality, and resolution. First, whole-plant phenotyping systems are highlighted together with advances in automation that enable significant throughput increases. Organ and cellular level phenotyping and its tools, often operating at a lower throughput, are then discussed as a means to obtain high-dimensional phenotypic data at elevated spatial and temporal resolution. The significance of recent developments in sensor technologies that give access to plant morphology and physiology-related traits is shown. Overall, attention is focused on spatial and temporal resolution because these are crucial aspects of imaging procedures in plant phenotyping systems.