Network based control strategies for sustainable management of Lantana camara

TL;DR

This paper develops ecological network models to design control strategies for managing invasive Lantana camara, demonstrating controllability with minimal interventions and highlighting the importance of parameter choices for sustainable ecosystem management.

Contribution

It introduces minimal ecological network models with control inputs for Lantana camara, analyzing controllability and intervention strategies using Lie algebra and LQR algorithms.

Findings

Models are controllable with at least two control inputs.

Intervention effectiveness depends on timing, effort, and penalties.

Nonlinear self-loop effects sustain Lantana camara presence.

Abstract

Control studies in ecological models of networks involving invasive plants can offer tangible management strategies. Framing control policies to tackle problems posed by invasive plants on indigenous plant diversity under the sustainable developmental goals framework is the main motivation of this paper. Based on the reported experimental and theoretical observations, we propose minimal ecological models of direct interactions of \textit{Lantana camara} with other entitites e.g. other indigenous plants (denoted by Control-plant) and soil microbes. The species' abundance variables represented by the nodes in the model networks follow generalized Lotka-Volterra dynamics and input signals are incorporated at appropriate driver nodes to control the \textit{Lantana camara} outgrowth. Analysis with Lie algebra and controllability criteria showed that our model systems are controllable with minimum of two control input signals. Analytical results of our model equations indicated that successive interventions are required to increase the abundance of control-plant, the microbes and suppress \textit{Lantana camara}. We observed from the numerical analysis with LQR algorithm that the number of interventions to be taken to control the \textit{Lantana camara} outgrowth depends on the choice of various important parameters such as intervention time, effort and penalty etc. Further, the increase in nonlinear terms contributed from self-loops in the network model shows the persistence of \textit{L. camara} and how it drives the coexistence of the species. We propose that parameters we have identified in our study needed to be considered and analyzed extensively for designing sustainable policies to control invasive species in an ecosystem.