Publication

Simulation of the development of the root system and associated microbial community of Pinus radiata

Date
1995
Type
Thesis
Abstract
A simulation scheme to model three-dimensional plant root architecture and the location, growth and interaction of associated microbial communities has been developed. The scheme expands on previous root architecture models by using the mature root system morphology observed in the field as a spatial envelope to model the system's temporal development. The three-dimensional representation allows a uniquely detailed treatment of the spatial development of microbial populations and their interactions with the root system and with each other. Root morphology and microbial populations are described by different types of "node", each node records a position in three-dimensional space and other details specific to the feature it represents. For example, "Branch" nodes indicate the start of a higher order root, and "Fungal" nodes increase or decrease the level of a fungal population on a root's surface. A large number of probability density functions are required to produce the list of nodes that describe each root. A four-dimensional matrix is used as a convenient abstraction for these functions, the dimensions represent tree type, root order, a "feature" such as branching or microbial infection, and an "attribute" such as length or lifespan. Algorithms for the generation and manipulation of the node lists are given. Specific implementation issues, such as rapid location of roots within the area potentially affected by disease lesion, and visualisation of the simulated root system, are also addressed. Validation of models of complex, irregular entities such as plant root systems is difficult, some techniques, including an objective root distribution index, have been developed and applied in this research. The simulation software includes extensions to allow input and editing of observed root system architectures to simplify the extraction of relevant parameters from such observations. The Pinus radiata / armillaria root-rot pathosystem has been used extensively as a test case for the scheme. The spread of the disease armillaria between roots of seedlings and the control of that spread by antagonistic strains of the soil fungus genus Trichoderma have been successfully modelled. Simulated root architectures are consistent with those observed in the field, and patterns of fungal spread through simulated stands are very similar to those reported in the literature. While this research was initially prompted by the Pinus radiata / armillaria pathosystem, the problems solved, namely representation and manipulation of three-dimensional root architecture and root/microbial spatial interaction, are essentially the same for most plant root ecosystems. The scheme should be a useful tool for the development of biological control programmes, both in formulating an understanding of the dynamics of the ecosystem to be modified, and in optimising the timing and position of biological control agent application. Further development of the scheme should include a generalised root ecosystem description syntax, and a re-implementation of the simulation's internal structure in an object oriented language. This should result in a versatile framework for detailed simulation of plant root microbial ecosystems.