Complex systems analysis of invasive species in heterogeneous environments

TitleComplex systems analysis of invasive species in heterogeneous environments
Publication TypeThesis
Year of Publication2016
AuthorsLustig, A.
DegreeDoctor of Philosophy
Date Published2016
UniversityLincoln University
CityLincoln, N.Z.
KeywordsAnnual Report 2015-16

iven that so many ecosystems currently face major habitat change conducive to the rising global spread of invasive species, there is growing awareness of the need to adopt proactive management strategies. Among models most used to predict future changes in distribution of invasive species, few explicitly incorporate characteristic of the population dynamics at the invasion front and the spatial heterogeneity of the environment. In particular, the influence of landscape composition and configuration on population dynamics and ecosystem susceptibility to invasion, remain dependent on broad generalization. The purpose of this study was to investigate how biological characteristics of invasive species interact with the structure of the landscape, to determine establishment and spread success. Critical to this research was the development of a unique spatially-explicit model that allowed for a systemic investigation of the impact of landscape structure on population dynamics of a species. The modelling framework has three components,1) a spatially-explicit, individual-based dispersal simulation framework, 2) a landscape generator allowing independent change in the composition and configuration of landscape components, and, 3) appropriate landscape measures that establish a quantitative relationship between demography, dispersal and the environment. The framework allowed a shift in focus from an individual species, to a more general approach where the pattern of invasion over multiple species and landscape scenarios were used to infer key drivers of invasion. To identify appropriate landscape measures for this research, a multi-scale analysis of widely used landscape metrics was carried out. That analysis highlighted that landscape metrics are sensitive to complex interactions between the intrinsic characteristics of a landscape, and scale-dependent factors, making it difficult to isolate landscape pattern driven effects from the effects of changing spatial scale. As a solution, the self-organising map (SOM) clustering approach is proposed as an efficient way to disentangle the relationships among landscape metrics and spatial scale when accurate characterization of landscape pattern is a key input in spatially explicit ecological models. The investigation of the effect of landscape structure on the establishment and spread of invasive species showed that both population density and rate of spread are affected in significant ways, and sometimes interactively, by landscape based components such as, suitable habitat amount, habitat patch aggregation, core area, edge density and habitat shape complexity. A key result of this research suggests that areas that are vulnerable to invasion can be better predicted by quantifying the elements of the landscape that significantly influence the density and spread of a species. However, the identification of an optimal set of landscape metrics for a species will require case specific study as clearly different species will respond to landscape structure in different ways. This research also demonstrated that dominant processes shaping population density and spread of invasive species can be identified and prioritized, as well as those of secondary importance. Variables representing an Allee effect, the intrinsic rate of increase and propagule pressure were found to be the three most influential on both density and spread over multiple invasive species scenarios. The frequency of long-distance dispersal events, in combination with measures of suitable habitat amount and habitat aggregation, was found to be a better indicators of population dynamics than the intrinsic abilities of a species to disperse in fragmented landscapes. This research has shown that the presence of an Allee effect leads to a balance between the effectiveness of spread and invasion success. Spread is maximized at an intermediate dispersal rate and inhibited at both low and high rates. The configuration and composition of the landscape, by either increasing or mitigating or the dispersal abilities of a species, can lead to a rate of spread under a dispersal threshold for which density and spread is at the highest. This research highlights how complex interactions between propagule pressure, species traits and habitat characteristics can determine patterns of invasion across fragmented landscapes. Successful management of invasive species, particularly for prioritisation and design efficient surveillance and control strategies, will depend on understanding this context dependent effect across habitats. More important, however, the research highlights the need for implementing multi-scenario modelling frameworks to reduce model uncertainty and to identify optimal trade-offs between model precision and complexity. Such development is in its infancy, and further research to correctly and consistently assess, and communicate uncertainty, surrounding spread modelling is needed. Informative assessments and clear communication of uncertainty will allow end-users and practitioners to make more informed decisions about the potential for invasive species establishment.