Wheat physiology predictor: predicting physiological traits in wheat from hyperspectral reflectance measurements using deep learning.

BACKGROUND: The need for rapid in-field measurement of key traits contributing to yield over many thousands of genotypes is a major roadblock in crop breeding. Recently, leaf hyperspectral reflectance data has been used to train machine learning models using partial least squares regression (PLSR) to rapidly predict genetic variation in photosynthetic and leaf traits across wheat populations, among other species. However, the application of published PLSR spectral models is limited by a fixed spectral wavelength range as input and the requirement of separate custom-built models for each trait and wavelength range. In addition, the use of reflectance spectra from the short-wave infrared region requires expensive multiple detector spectrometers. The ability to train a model that can accommodate input from different spectral ranges would potentially make such models extensible to more affordable sensors. Here we compare the accuracy of prediction of PLSR with various deep learning approaches and an ensemble model, each trained and tested using previously published data sets. RESULTS: We demonstrate that the accuracy of PLSR to predict photosynthetic and related leaf traits in wheat can be improved with deep learning-based and ensemble models without overfitting. Additionally, these models can be flexibly applied across spectral ranges without significantly compromising accuracy. CONCLUSION: The method reported provides an improved prediction of wheat leaf and photosynthetic traits from leaf hyperspectral reflectance and do not require a full range, high cost leaf spectrometer. We provide a web service for deploying these algorithms to predict physiological traits in wheat from a variety of spectral data sets, with important implications for wheat yield prediction and crop breeding.

Measuring CO2 and HCO3- permeabilities of isolated chloroplasts using a MIMS-18O approach.

To support photosynthetic CO2 fixation by Rubisco, the chloroplast must be fed with inorganic carbon in the form of CO2 or bicarbonate. However, the mechanisms allowing the rapid passage of this gas and this charged molecule through the bounding membranes of the chloroplast envelope are not yet completely elucidated. We describe here a method allowing us to measure the permeability of these two molecules through the chloroplast envelope using a membrane inlet mass spectrometer and 18O-labelled inorganic carbon. We established that the internal stromal carbonic anhydrase activity is not limiting for this technique, and precisely measured the chloroplast surface area and permeability values for CO2 and bicarbonate. This was performed on chloroplasts from several plant species, with values ranging from 2.3 × 10-4 m s-1 to 8 × 10-4 m s-1 permeability for CO2 and 1 × 10-8 m s-1 for bicarbonate. We were able to apply our method to chloroplasts from an Arabidopsis aquaporin mutant, and this showed that CO2 permeability was reduced 50% in the mutant compared with the wild-type reference.

Digital imaging approaches for phenotyping whole plant nitrogen and phosphorus response in Brachypodium distachyon.

This work evaluates the phenotypic response of the model grass (Brachypodium distachyon (L.) P. Beauv.) to nitrogen and phosphorus nutrition using a combination of imaging techniques and destructive harvest of shoots and roots. Reference line Bd21-3 was grown in pots using 11 phosphorus and 11 nitrogen concentrations to establish a dose-response curve. Shoot biovolume and biomass, root length and biomass, and tissue phosphorus and nitrogen concentrations increased with nutrient concentration. Shoot biovolume, estimated by imaging, was highly correlated with dry weight (R(2) > 0.92) and both biovolume and growth rate responded strongly to nutrient availability. Higher nutrient supply increased nodal root length more than other root types. Photochemical efficiency was strongly reduced by low phosphorus concentrations as early as 1 week after germination, suggesting that this measurement may be suitable for high throughput screening of phosphorus response. In contrast, nitrogen concentration had little effect on photochemical efficiency. Changes in biovolume over time were used to compare growth rates of four accessions in response to nitrogen and phosphorus supply. We demonstrate that a time series image-based approach coupled with mathematical modeling provides higher resolution of genotypic response to nutrient supply than traditional destructive techniques and shows promise for high throughput screening and determination of genomic regions associated with superior nutrient use efficiency.

Diel time-courses of leaf growth in monocot and dicot species: endogenous rhythms and temperature effects.

Diel (24 h) leaf growth patterns were differently affected by temperature variations and the circadian clock in several plant species. In the monocotyledon Zea mays, leaf elongation rate closely followed changes in temperature. In the dicotyledons Nicotiana tabacum, Ricinus communis, and Flaveria bidentis, the effect of temperature regimes was less obvious and leaf growth exhibited a clear circadian oscillation. These differences were related neither to primary metabolism nor to altered carbohydrate availability for growth. The effect of endogenous rhythms on leaf growth was analysed under continuous light in Arabidopsis thaliana, Ricinus communis, Zea mays, and Oryza sativa. No rhythmic growth was observed under continuous light in the two monocotyledons, while growth rhythmicity persisted in the two dicotyledons. Based on model simulations it is concluded that diel leaf growth patterns in mono- and dicotyledons result from the additive effects of both circadian-clock-controlled processes and responses to environmental changes such as temperature and evaporative demand. Apparently very distinct diel leaf growth behaviour of monocotyledons and dicotyledons can thus be explained by the different degrees to which diel temperature variations affect leaf growth in the two groups of species which, in turn, depends on the extent of the leaf growth control by internal clocks.

Root cooling strongly affects diel leaf growth dynamics, water and carbohydrate relations in Ricinus communis.

In laboratory and greenhouse experiments with potted plants, shoots and roots are exposed to temperature regimes throughout a 24 h (diel) cycle that can differ strongly from the regime under which these plants have evolved. In the field, roots are often exposed to lower temperatures than shoots. When the root-zone temperature in Ricinus communis was decreased below a threshold value, leaf growth occurred preferentially at night and was strongly inhibited during the day. Overall, leaf expansion, shoot biomass growth, root elongation and ramification decreased rapidly, carbon fluxes from shoot to root were diminished and carbohydrate contents of both root and shoot increased. Further, transpiration rate was not affected, yet hydrostatic tensions in shoot xylem increased. When root temperature was increased again, xylem tension reduced, leaf growth recovered rapidly, carbon fluxes from shoot to root increased, and carbohydrate pools were depleted. We hypothesize that the decreased uptake of water in cool roots diminishes the growth potential of the entire plant - especially diurnally, when the growing leaf loses water via transpiration. As a consequence, leaf growth and metabolite concentrations can vary enormously, depending on root-zone temperature and its heterogeneity inside pots.