Ensemble models to assess the risk of exotic plant pathogens in a changing climate
In recent decades, species distribution models (SDMs) have been widely used in many ecological, environmental and climate-change research studies to model invasive species establishment. These models associate recorded locations of species with environmental variables. Nevertheless, the few studies that attempt to model the climate suitability of plant pathogens before their arrival into a new area mainly rely on a single model projection. In this research, eleven species distribution models (in the form of three modelling approaches which include correlative and mechanistic models) were used to project the climate suitability of three target species; kiwifruit bacterial canker (Pseudomonas syringae pv. actinidia) (Psa), dwarf bunt of wheat (Tellitia controversa) and guava rust (Puccinia psidii) for New Zealand and over a global scale. The climate suitability of target species was modelled using CLIMEX as a semi-mechanistic model, MaxEnt as a presence-only correlative model and Multi-Model Framework (which includes nine correlative models). While there were similarities with regard to climate suitability for target species projected by the models over both local and global scales, there were differences in their projection with respect to the degree and extent of suitability, making it hard to select one “best” model. All models were found to have their differences and weaknesses that are largely the result of difference in the theoretical basis and structure of each model. For example, compared with CLIMEX and the Multi-Model Framework, MaxEnt showed lower transferability of projection into new areas. Additionally, as a semi -mechanistic model, the uncertainty of CLIMEX projections was found to be increased by subjectivity in the parameter setting process. To illustrate the impact of parameter variability on the uncertainty of CLIMEX projections, a sensitivity analysis was performed on one of the target species (dwarf bunt) to measure the effect of error in important parameters on model output. The sensitivity analysis showed that for dwarf bunt, CLIMEX outputs were very sensitive to upper temperature threshold and soil moisture parameters, which highlight that sensitivity analysis, should be an integral part of any CLIMEX modelling. For Multi-Model, despite the advantages such as calculating different performance criteria, the importance and contribution of selected variables and their influence on model output is not given. Because of differences in model projection, a method was developed to benefit from the information provided by all the types of models, by combining the results of different model output into an ensemble, or more specifically, a consensus model. A variant of committee averaging was used where model outputs are converted to binary maps (presence- absence) which allow any kind of algorithm and output to be included. The resulting consensus model highlighted the areas where more than half of the models agreed on the climate suitability for target species establishment. Such a model that relies on agreement of model projections indicates with a level of certainty or uncertainty what is likely to happen and consequently can highlight areas, both locally and globally, that have a higher risk of target species establishment. Finally, the effect of climate change on climate suitability of target species was investigated using two scenarios (A1B and A2) for 2030 and 2090. The results showed that, the suitable areas decreased for Psa and dwarf bunt at different levels while guava rust suitability increased. The results of this thesis confirm that models with different theoretical foundation will give dissimilar predictions, and it is difficult to determine conclusively whether one model is superior to others. Among other recommendations, I strongly advise that researchers and risk assessors should not rely on a single-model projection. If time and resources are available, an appropriate ensemble of models should be used to investigate the climate suitability of plant pathogens. Keywords: Plant pathogens, kiwifruit bacterial canker (Psa), dwarf bunt, guava rust, climate suitability, Species distribution models (SDMs), CLIMEX, MaxEnt, Multi-Model Framework, correlative models, semi-mechanistic models, sensitivity analysis, consensus model, ensemble models, climate change, range expansion.... [Show full abstract]
KeywordsPlant pathogens; kiwifruit bacterial canker (Psa); dwarf bunt; guava rust; climate suitability; Species distribution models (SDMs); CLIMEX; MaxEnt; Multi-Model Framework; correlative models; semi-mechanistic models; sensitivity analysis; consensus model; ensemble models; climate change; range expansion
Fields of Research050103 Invasive Species Ecology; 050206 Environmental Monitoring; 070104 Agricultural Spatial Analysis and Modelling; 060704 Plant Pathology
Access RightsRestricted item - embargoed until 01 September 2021
- Bio-Protection Research Centre 
- Doctoral (PhD) Theses 
Showing items related by title, author, creator and subject.
An empirical investigation of the effectiveness and robustness of asset pricing models in Australia : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University Malik, Saba (Lincoln University, 2020)With a plethora of asset pricing models, it is hard to determine: Which model is relatively better than other models in successfully explaining the returns of the capital assets in an equity market? and if any of the ...
A pilot optimisation model for the 1972/73 N.D.C. plan Agricultural Economics Research Unit (Lincoln College. Agricultural Economics Research Unit, 1970-08)This paper represents a further stage in the programme of work initiated at Lincoln some years ago under the general heading of Studies in the Structure of the New Zealand Economy, This has involved the compilation of data ...
Using activity time windows and logical representation to reduce the complexity of biological network models: G1/S checkpoint pathway with DNA damage Khazaaleh, M; Samarasinghe, Sandhya (Elsevier, 2020-05-01)Biological systems are difficult to understand complex systems. Scientists continue to create models to simulate biological systems but these models are complex too; for this reason, new reduction methods to simplify complex ...