Xu’s research program has been highly successful in generating epidemiological knowledge on diseases of fruit crops, and using the obtained knowledge to develop models to predict disease development. Several of these models have been implemented in the UK and elsewhere as forecasting tools in commercial fruit production. These include models for apple powdery mildew, apple and pear scab, tree canker and fruit rot of apple, and Botrytis fruit rot of strawberry, raspberry and blackcurrant. Use of these models has significantly reduced fungicide input (up to 45%) while maintaining effective control. These models comprise a key component of crop production strategies developed at NIAB EMR to produce fruit with minimum pesticide residues. Recently, Xu has been leading the effort to link the models in a computer server to networked weather stations in order to automate the delivery of disease forecasts to mobile devices throughout the UK.
Xu is a global leader in the statistical analysis and interpretation of spatio -temporal patterns of disease incidence and severity data. He and his colleagues developed stochastic simulation models and used these tools to study the relationship of common spatio-temporal statistics with underlying biological and physical factors. A series of papers clearly demonstrated that all spatial statistics are largely affected by initial epidemic conditions and sampling methods and, to a much lesser extent, by the spatial dispersal characteristics of a given pathogen. Thus, it is crucial to understand initial epidemic conditions and sampling methods in comparative studies; otherwise, conclusions are likely to be misleading or erroneous. Furthermore, Xu and a collaborator identified inherent shortcomings in the popular SADIE methodology for spatial analysis in ecology and epidemiology, and developed a new analytical approach to correct these shortcomings.
Effective use of biocontrol agents (BCAs) is a potentially important component of sustainable agriculture, but extensive research and development has generated a limited number of effective BCAs. Moreover, disease control efficacy is often variable even for the successful BCAs (mostly in greenhouse cultivation). To address this issue, Xu and his collaborators developed a generic modelling framework to understand the dynamics of a foliar pathogen and BCA populations under homogeneous and heterogeneous conditions. A series of significant modelling publications described biocontrol efficacy in relation to biocontrol mechanisms, mixed use of BCAs, and homogeneous and spatio-temporally heterogeneous conditions. Important results showed that: the rate of a BCA colonizing healthy and diseased plant tissues is critically important, but often neglected, and a BCA combining two mechanisms is more effective than combined use of two BCAs, each with a single biocontrol mechanism. Furthermore, the combined use of two BCAs usually results in competitive interactions among BCAs rather than a synergistic interaction that has been common claim in the literature. Their findings indicate that synergy between two BCAs is possible only under spatially heterogeneous conditions and when two BCAs differed significantly in their ecological requirements. These concepts and modelling approach has been adopted by several international research groups.
Xu has made many significant contributions to the epidemiology of Fusarium Head Blight (FHB). Unlike in USA where one pathogen species dominates, FHB in Europe is caused by several pathogens, and hence understanding the consequence of intra- and interspecific interactions is critical in assessing both the disease and mycotoxin risks. Through a collaborative research program led by Xu and involving several European partners, the team demonstrated positive correlations in the presence of pairs of FHB species in a given field. However, the association was indirect, due to co-favorability of microclimatic conditions, rather than due a manifestation of direct synergistic interactions. Furthermore, single- and mixed-species co-inoculations showed that species interact competitively. Even for the dominant species, its biomass was much less in the co-inoculation than when inoculated alone. However, there was a considerable increase in the mycotoxin productivity per unit of fungal biomass in the co-inoculations.
Xu has been recognized several times for major accomplishments in research. He is an Honorary Professor at the University of Kent (UK) and a fellow of Royal Society of Biology (London). In 2015, he was awarded the prestigious JONES-BATEMAN CUP by the Royal Horticultural Society (UK) in recognition of his highly significant contributions on the epidemiology and management of diseases on fruit crops. This cup is awarded every third year to a single person who has made significant contributions to horticulture. Xu is a highly prolific researcher who is regarded as one of the leading plant epidemiologists worldwide. He has written 146 journal articles, 38 technical publications, ten book chapters, and nine computer programs for research and disease management. In addition, he has co-edited two books. Xu has frequently been an invited speaker at international conferences and at universities and institutes in many countries. He has participated in international projects with institutions worldwide and co-led two multi-country European projects. Xu has been equally active in service to professional societies. He has served as an associate or decision-making editor for two journals, and serves on committees of national and international scientific organizations. Currently, he is serving as a senior editor of the APS journal Phytobiomes. Since 2007, he has been acting as an international expert on the invitation of the Chinese government to provide strategic R&D advice for diseases of cereal and fruit crops. Without a formal teaching appointment, he routinely teaches short courses at Chinese universities and advises graduate students.