Predicting Genotypic Variability in the Field Using High Throughput Root Phenotyping Data

Background: Many of the strategies proposed for sustainable intensification of agriculture rely on breeding crops for greater yields with reduced use of water and fertiliser inputs (White et al. 2013, Ann. Bot. 112:207-222). This effort can be informed by mathematical models linking genetic markers to root traits and root traits to resource acquisition and crop growth in the field. Unfortunately, this cannot be achieved with existing models because they do not combine both genotypic variation in the crop and interactions of roots with the soil (de Dorlodot, et al. 2007, Trends Plant Sci. 12:474-481).

Aim: The aim of this project is to develop a predictive crop growth model that combines both genotypic variation in the crop and interactions of roots with the soil.

Objectives/Timetable: [01, months 0-5] To develop algorithms to quantify root traits (root growth rate, root growth angles, root branching) for genetic mapping populations of barley and brassica from our existing BBSRC-CIRC image databases. [02, months 5-6] To perform genetic association analyses for root traits using existing mapping data. [O3, months 6-12] To develop a Mechanistic Model describing correlations between root architectural traits, nutrient acquisition and crop growth using our existing BBSRC-CIRC datasets. [O4, months 12-18] To develop a Functional Genetic Model by combining a genetic association model with the Mechanistic Model describing the correlations between root architectural traits, nutrient acquisition and crop growth. [05, months 18-22] To incorporate the Functional Genetic Model into a standard Crop Growth Model. [O6, months 22-24] To identify ideotypes with greater resource capture and yield using the modified Crop Growth Model. [O7, months 24-36] To validate predictions of the modified Crop Growth Model in field and glasshouse experiments.

Methods/Approach: [O1] The student will develop a computational pipeline linking image datasets from our high-throughput root system phenotyping systems to a root trait database. They will utilize a large root system architecture image database collected in a BBSRC CIRC project that includes a barley association genetic mapping population (AGOUEB; n= 304 Spring barley genotypes, n=282 Winter barley genotypes) and an oilseed rape genetic mapping population (TNDH; n=204 genotypes). The computational challenges will include the development of more efficient algorithms for (a) detection of root tips, (b) root tracing, and (c) optimizing the selection of architectural traits. Thus, the student will improve our customized software previously developed for image processing and data analysis. [O2] Genetic association analyses will be performed using standard techniques (Shi et al. 2013, Ann. Bot. 112:381-389; George et al. 2014, New Phytol. 203:195-205). A composite interval mapping model (CIMM) will be implemented to associate root traits with genetic markers for incorporation in predictive Functional Genetic Models. [O3, O4] A Mechanistic Model (MM) describing correlations between root architectural traits, nutrient acquisition and crop growth will be developed based on our existing models for root system architecture and nutrient uptake (Dupuy et al. 2010, Plant Cell Environ. 33:358-369; Brown et al. 2013, Ann. Bot. 112:317-330) using our existing BBSRC-CIRC datasets. The composite Interval Mapping Model will be combined with the Mechanistic Model to produce the Functional Genetic Model (FGM). [O5, O6] A crop modelling platform, such as APSIM or DSSAT (McCown et al. 1995, Math. Comput. Simu. 39:225-231; Hoogenboom et al. 2003, Decision Support System for Agrotechnology Transfer Version 4.0. Vol. 1: Overview. University of Hawaii), will be used as the template for developing the ultimate Crop Growth Model (CGM). All models (CIMM, MM, FGM, CGM) will use the same programming language. This work will be carried out in collaboration with Davide Cammarano, who will assist with parameterisation and simulation of crop models. [O7,O8] The modified crop growth model will be used to identify ideotypes with greater resource capture and yield, and these predictions will be validated for selected genotypes of barley (n=12) and oilseed rape (n=12) in both glasshouse or field experiments in which root and shoot biomass, mineral composition and yield of plants will be determined from crop emergence to commercial maturity.