Modelling integrated weed management of an invasive shrub in tropical Australia.

Published online
14 Jul 2004
Content type
Journal article
Journal title
Journal of Applied Ecology
DOI
10.1111/j.0021-8901.2004.00909.x

Author(s)
Buckley, Y. M. & Rees, M. & Paynter, Q. & Lonsdale, M.
Contact email(s)
y.buckley@imperial.ac.uk

Publication language
English
Location
Australia

Abstract

Where biocontrol programmes for invasive plants are in place, only one-third are fully successful. Integrated weed management (IWM) emphasizes the use of several complementary control measures. We used models of increasing complexity to determine which parameters affect site occupancy of an invasive shrub, Mimosa pigra, in tropical Australia. Two introduced biocontrol agents have spatial effects on both plant fecundity and the probability of recolonization after senescence. We incorporated biocontrol effects into IWM models with small-scale disturbance, such as grazing and pig-rooting, and large-scale disturbance, such as mechanical control, herbicide and fire. The models were parameterized from experimental and field data. The models indicated that reduction in fecundity is not the most important impact of biocontrol; rather it is defoliation at the edges of stands, allowing grasses to out-compete M. pigra seedlings. We demonstrated that biocontrol alone is only successful at low levels of small-scale disturbance and seedling survival and, even then, current biocontrol agents would take decades to reduce a stand to <5% site occupancy. Our model predicts the most successful IWM strategy to be an application of herbicide in year 1, mechanical control+fire in year 2 and herbicide in year 3, with reduction of small-scale disturbance where possible. The addition of biocontrol improves the success of this strategy. Synthesis and applications. Ascertaining how control measures, including biological methods, will influence persistence of an invasive requires models of the target species' dynamics and its ecosystem. As in previous applications of this model, disturbance is the most important regulator of population size in M. pigra; moderate to high levels of small-scale disturbance promotes M. pigra occupancy. We have shown that IWM can control M. pigra and that biocontrol is an effective part of this strategy. Reductions in fecundity alone are unlikely to control invasive leguminous shrubs. However, biocontrol agents affect the probability of recolonization after senescence and enhance control. Our recommended 3-year treatment programme (herbicide:mechanical control+fire:herbicide, with biocontrol) is justifiable in terms of the biology of the system, making it more likely to be acted upon by risk-averse farmers and land managers.

Key words