Plant-based controls on soil structural dynamics: elucidating the interactive roles of the genotype, phenotype and soil microbial community

Lead Research Organisation: University of Nottingham
Department Name: Sch of Biosciences

Abstract

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.

Planned Impact

The principal impacts we wish to attain with this project are:

(i) a practitioner-level awareness of the concepts and findings arising from this research, particularly in relation to the potential that maincrops could contribute to soil health through enhanced structure;

(ii) a wider understanding of the concepts and potential of such approaches by policy communities;

(iii) utilisation of the data and knowledge acquired by the environmental sciences academic community.

These will be achieved by the establishment of a project related website and periodic production of briefings which will be posted on the site and distributed via a group of 'Promulgation Partners' who have agreed to use their respective communications networks to disseminate the instigation of the project and the key findings as they are realised. This will include material on their websites and online portals, newsletters and annual member meetings and trade conferences.

Project members will also give series of talks and seminars to growers and various fora which the investigators are regularly contribute to, which will include two attendances to the key UK cereal-producer event, Cereals.

Publications

10 25 50
 
Description A statistical modelling approach has been developed to express X-ray density in images of rooted soil as a function of distance to the nearest root. This is currently being applied as a way to understand how root-based processes (mechanical or biological) influence the development of soil structure. At present the method is being deployed with the results of a large experiment in which contrasting wheat varieties have been grown in soil with the same composition. The varieties are known to have different rooting habits when grown in culture, this work will allow us to evaluate whether there are differences between the varieties expressed in real soil, and, in due course, whether these differences are likely to be of practical significance (e.g. in improving soil structure with benefits for nutrient use, reduced emissions and crop yield). Provisional exploratory results have shown that the methods developed can identify differences in soil-root interactions between different cases.

Final analyses are currently underway to show precisely how the statistical model for variation in X-ray density differs between contrasting examples (e.g. rooted and unrooted soil), and how the statistical outputs can facilitate interpretation of the imagery.
Exploitation Route The methods developed in this project could be used to screen new crop varieties to identify those with traits that improve soil-root interactions during crop growth and development.

The general statistical framework developed for the analysis of the CT imagery could be applied to other problems where CT or similar technology is used to form 3-D images of complex materials with non-uniform structures.
Sectors Agriculture, Food and Drink,Environment
 
Title R package for spatial analysis of x-ray CT imagery by parallel processing 
Description In the course of this project it has proved necessary to undertake parallel processing in order to compute spatial analyses of imagery in feasible time. The PDRA on the project has developed an R package which can be used to run processes on a HPU or similar cluster, but which does not require the user to be familiar with the parallel environment. It is intended that this package will be published and made generally available. 
Type Of Material Computer model/algorithm 
Year Produced 2020 
Provided To Others? No  
Impact This package has allowed timely processing of CT imagery in the project. 
 
Title R software for analysis of X-ray CT images 
Description Software has been developed to work with segmented X-ray CT images of soil, extracting and registering data sets from successive scans of the same material over time, extracting the soil and root components of the image from the containder, and then computing distance-to-nearest-root values for each voxel. This software has been substantially developed because of the considerable computational demands of running it on large image stacks. It has been completely rewritten to run in R in an HPC environment. Rather than simply generating scripts and functions, the software has been produced as an R package for the HPC environment. It can be run in a familiar R interface, so the user does not need to be particularly familiar with HPC conventions as these are handled by the package. The parallelization of the original code and the development of systems for the distribution of computational tasks and for visualization of large images has been a substantial task. However, it is now possible to process data from a large experiment undertaken in the project in feasible time frames. This software will be of considerable value after the project to facilitate high-thoughput processing of CT images for phenotyping studies. There are also more generic problems in the processing of large 3-D images for which it would be very useful. 
Type Of Technology Software 
Year Produced 2020 
Impact This software allows data sets from large experiments to be processed prior to statistical modelling of the CT density as a function of proximity to roots. It has enabled the central experiments of the project, comparing root lines, and ultimately could contribute to functionality for using CT scanning as part of plant phenotyping. The parallelization and development of the code for an HPC environment has substantially increased its utility and made it feasible to apply the method to large volumes of data from replicated experiments on multiple wheat lines.