The effect of competition on rooting patterns of plants in semi-arid systems: an analysis of a unified model of soil type, climate and leaf area index

Submitted by charlotte.schilt on Thu, 10/11/2018 - 14:33
Abstract

In most environments trees and grasses do not co-occur, but in semi-arid system they do. Walters two-layer hypothesis suggests that rooting patterns are an important factor to explaining this phenomena. However these patterns do not seem to be consistent within a plant type (e.g. tree or grass). The roots a plants roots are competing with might be an important factor in determining whether it is best for a plant to position its roots deep or shallow.

Maximization of transpiration is considered to be the driving factor in plant rooting patterns. In this study it is investigate if a simple model, assuming maximization of transpiration, is able to explain what drives a plant to put its roots either deep or shallow, while competing with its roots against another plants’ roots for water. To do so, an existing ecohydrological model has been modified to function for two competing plant species and simplified to allow for the modeling of woody plants as well. The simplifications caused the root length density, the length of the roots per volume of soil, to no longer influence the water uptake capabilities of the root. Both plants are considered to share the same area with their roots, only differing in vertical distribution.

The main assumption in the modeling of this competition is that the potential evapotranspiration for that shared area of the two plants is divided over both canopies equally. The model and analysis of the model can explain that there are situations a plant prefers to be below another plant and situations where it prefers to be above another plant, but fails to give an example of an evolutionarily stable strategy were the two rooting strategies of both plants would be unequal.

Though further analysis of the model and modifications of the model to allow for the calculation of the root length density look promising.

Address
n.a.
Author
Laurens .J. Kilsdonk
Country
Netherlands
Date
Email

n.a.

Mentor(s)
Mark van Wijk
Type
BSc