A geotechnical approach to root anchorage modelling in wheat, oats and oilseed rape

Rouse, Amelia (2019). A geotechnical approach to root anchorage modelling in wheat, oats and oilseed rape. University of Birmingham. Ph.D.

Preview

Rouse2019PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (13MB) | Preview

Abstract

Lodging is a naturally occurring phenomenon that reduces the yield and quality of cereal and oilseed rape crops, costing farmers millions of pounds in losses. Lodging is the permanent displacement of a plant stem from the vertical position and can occur due to stem failure or root anchorage failure. Understanding mechanisms behind risks of each of these types of lodging could reduce losses to UK and Irish farmers. Mathematical models have been developed to predict the risk of lodging failure in stems and root systems. However, the root anchorage failure model may not be accurately predicting root system failure (root failure moments). The analysis and computation of the root failure of wheat and oilseed rape crops are currently based on the theory of lateral resistance of rigid piles modified by Crook and Ennos, (1993) and Goodman et al., (2001). However, there was uncertainty about the accuracy of the model, the failure mechanism and modifications to the rigid pile analysis. Also, oat root systems had not been included in root anchorage studies.

This research reviewed and evaluated root anchorage models across different plant types and developed methodologies for collecting a dataset to test the models and compared the models for accuracy. It was found that the Crook and Ennos, (1993) and Goodman et al., (2001) models over-predicted root anchorage. These models incorporated soil shear strength, root plate diameter, root diameter and root length. Models by Fourcaud et al., (2008) and Coutts et al., (1999) were more accurate for the wheat and oats dataset collected. These models were suggested for trees, incorporating parameters such as the root plate mass and depth of the roots as well as root plate diameter. In oilseed rape, the results showed a model suggested by Niklas, (1992) was more accurate than the Goodman et al., (2001) model.

A new methodology was developed for laboratory testing of flexible and rigid wheat and oilseed root models in coarse and fine-grained soils. Load-deflection curves for root failure were recorded using a specially created lodging machine. It was found that in coarse-grained soil the root failure moment increased with increased water content. In fine-grained soils, the root anchorage was 4-5 times stronger and closer to the values measured in the field. When the water content was increased in these soils the root failure moment decreases rapidly.

Finally, a model calculating the capacity for under-reamed pile foundations was adapted for root systems of wheat, oats and oilseed rape. The models had options for both fine-grained and coarse-grained soils. It was found that for wheat, the coarse-grained model and an adjusted fine-grained model had the highest accuracy. The coarse-grained model had a normalised root mean squared error (NRMSE) of 0.464 and the fine-grained model had an NRMSE of 0.399 compared to 21.5 for the past model by Crook and Ennos, (1993) . For oats, the same models as wheat had the highest accuracy with the coarse-grained model and fine-grained models having an NRMSE of 0.430 and 0.354 respectively, compared to 63.5 for the model by Crook and Ennos, (1993) . For oilseed rape, it was found that the Niklas, (1992) model was the most accurate model having an NRMSE of 0.282 compared to 10.3 for the past model by Goodman et al., (2001) . A sensitivity analysis found that the new models for wheat and oats were sensitive to the adhesion factor; the root length and diameter were important factors. For oilseed rape root length was still the most important factor. Field and laboratory data were used to validate the proposed model. It was found that the new models could reliably predict root anchorage.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):