Search for a command to run...
Fungal pathogens cause major yield losses in many of the world’s crops. It is important to understand which climatic factors drive pathogen occurrence, also in view of climate change. Here, we use a global electronic reporting system for outbreaks of emerging diseases (ProMED) to assess the factors that predict fungal disease outbreaks. Our results based on the ProMED database demonstrate that high temperature is an important driver of pathogen incidence. We further demonstrate that humidity is about twice as important as high temperature in predicting fungal plant disease outbreaks. We highlight here that in order to predict disease incidence, amplicon-sequencing data should be complemented by reports on actual disease outbreaks and the prevalence of specific pathogen species should be confirmed by using species-specific markers. Climate change is expected to lead to adverse impacts on global agricultural production, as it influences plant disease occurrence and severity at spatial and temporal scales (Howden et al., 2007). Furthermore, climate change also affects plant growth through reduced water availability and other abiotic stress factors. Recent studies have shown that climatic variables (e.g. humidity, temperature) are important drivers explaining the global distribution of potentially pathogenic soilborne microorganisms (Větrovský et al., 2019; Delgado-Baquerizo et al., 2020). These studies are fundamental for understanding the global incidence and potential impact of fungal plant pathogens on terrestrial ecosystems. However, it remains poorly understood to what extent the prevalence of soil pathogens and the factors affecting them translate into increased incidence of disease outbreaks. To provide empirical evidence, we analyzed actual disease incidence reported worldwide using the ProMED database (www.promed-mail.org). ProMED is a global electronic reporting system for outbreaks of emerging diseases from the International Society for Infectious Diseases (www.isid.org), which has already been used in the past to address the relative contribution of different factors to emerging plant diseases (Anderson et al., 2004). The ProMED database is based on outbreak reports by authorities, stakeholders and individuals that are submitted to a panel of moderators who review the quality of the data and publish a post containing only reliable data (Yu & Madoff, 2004). Within the ProMED database, we reviewed 264 posts on fungal diseases covering the period 2014–2019 and all continents (see Supporting Information Fig. S1). For every post, we retrieved the archive number, date, disease, location, host, pathogen and driver of emergence (see Dataset S1). Our results demonstrate that climatic variables are important drivers of pathogen incidence and highlight the prevailing role of humidity as a driving factor for actual fungal plant disease outbreaks. For all the posts reporting fungal outbreaks within the period 2014–2019, 36.4% indicated that an increase in relative humidity could have provoked or facilitated the outbreak (Fig. 1). High temperature (17.7%) and low temperature (9.9%) predicted disease in fewer cases. Our results further indicate that pathogens colonizing predominantly aboveground plant organs, including the fungal genus Puccinia and most species within the genus Fusarium, as well as the fungal-like (i.e. oomycete) genus Phytophthora were most frequently reported as causative agents of plant disease (Fig. 2). Fungi that predominantly occupy such niches may be underrepresented if the focus is only on pathogenic fungi inhabiting the soil. Soilborne potential plant pathogens can be relatively abundant in soils from natural ecosystems worldwide. A recent global study revealed that, on average, 14.4% of all internal transcribed spacer (ITS) sequences at a given sampling site are classified as potential pathogenic phylotypes, with Alternaria, Fusarium, Venturia and Phoma being most abundant (Delgado-Baquerizo et al., 2020). According to the information reported in the ProMED database (2014–2019), Fusarium (14.0%), Puccinia (12.9%) and Phytophthora (12.9%) were the genera most frequently detected as the causative agents for plant disease outbreaks (Fig. 2). Of the dominant pathogenic phylotypes reported in the database, Fusarium was the only one being detected widely in soil. The difference between the current study and the study by Delgado-Baquerizo et al. (2020) can also be explained by the ecosystems under study: for example, the ProMED database largely focuses on agricultural systems, while Delgado-Baquerizo et al. (2020) included many natural ecosystems in the analysis. Moreover, the primer pair used in Delgado-Baquerizo et al. (2020) (targeting the ITS gene) did not match to oomycetes such as Phytophthora and as such these sequences were not detected. That makes direct comparisons between both studies to be made with caution. Studies that report on the distribution of pathogens in soils often rely on relative abundance data obtained from next-generation sequencing (NGS; Riddell et al., 2019; Ruiz Gómez et al., 2019; Větrovský et al., 2019; Delgado-Baquerizo et al., 2020). Although amplicon sequencing provides important information on the distribution of microbial organisms, relative abundance outputs of sequencing studies do not allow for quantifying the host’s actual exposure (Knight et al., 2018) and the resulting disease severity. Unfortunately, absolute abundances cannot be deduced from compositional data using statistical methods, yet recent studies have shown that absolute abundance of microbial taxa is biologically meaningful (Vandeputte et al., 2017). In addition, the taxonomic resolution of ITS-based NGS data often remains at the genus level, although the pathogenicity of a given taxon can vary widely within the same genus for a range of species, including beneficial, saprotrophic, necrotrophic and pathogenic species (Walder et al., 2017). Therefore, a higher taxonomic resolution such as that provided by exact sequence variants (ESVs) in combination with longer sequenced reads and different target genes could be useful in identifying pathogenicity hotspots (Knight et al., 2018). Further tools, including quantitative species-specific PCR and/or omics approaches, should be used to assess the abundance of fungal pathogens, as has already been done for air- and residue-borne fungi (Dannemiller et al., 2014). According to Delgado-Baquerizo et al. (2020), climatic variables disproportionately contribute to the relative abundance of fungal pathogens in soil compared with elevation, latitude, longitude, vegetation type or soil characteristics. In line with this, a recent meta-analysis of 3084 soil samples revealed that 38.7% of the variability in soil fungal composition could be explained by variations in temperature (Větrovský et al., 2019). Our analysis using the ProMED database confirms that climatic variables are indeed the factors most likely to explain the emergence of disease. However, when analyzing climatic variables in detail, data from the ProMED database indicate that humidity, and not temperature, is the leading factor (Fig. 1). In line with this, other authors have also indicated that mean annual precipitation is the climatic factor best predicting fungal global diversity (Tedersoo et al., 2014). Humidity mostly increases the risk of aboveground infection by organisms such as the highly noxious fungal genus Fusarium as well as the oomycete genus Phytophthora. According to the information retrieved from ProMED, 12.9% of the outbreaks reported in the period 2014–2019 were caused by Phytophthora species (Fig. 2); most of these outbreaks had an increase in humidity identified as a likely driver of emergence. We stress here the importance of including oomycetes (e.g. Phytophthora) in future global surveys, as they include important plant pathogens whose incidence is expected to increase as a result of agricultural intensification and climate change (Corredor-Moreno & Saunders, 2020). Moreover, the relative contribution of fungal vs oomycete plant pathogens to plant disease outbreaks in a climate change context is still poorly understood. Furthermore, beside their effects on plants, fungal pathogens can act as keystone taxa and have a big impact on the functioning and structure of microbial communities (Banerjee et al., 2018). In conclusion, our work demonstrates that climate variables (relative humidity, high and low temperature) are key factors explaining the occurrence of disease outbreaks. Thus, our analysis of the ProMED database on emerging plant diseases together with previous studies on the distribution of potential soil pathogens, highlights the need for agricultural adaptation to climate change (Anderson et al., 2020). Our analysis highlights that studies that predict disease outbreaks using high-throughput sequencing benefit from complementing such molecular data with actual information on disease outbreaks and severity. Finally, a higher resolution to species level as well as the elucidation of indirect effects of climate change on pathogens, such as modified farming practices (e.g. crop rotation, tillage, etc.), merit further research efforts. This work was supported by a grant from the Swiss National Science Foundation (grant no. 310030 188799). The authors declare there are no competing interests FR and SC reviewed and synthesized the entries in the ProMED database. FR wrote the first version of the manuscript and managed the subsequent versions. MGAvdH initiated the work and contributed, together with FW, SFB and SV, to discussion and revision of the manuscript. Dataset S1 Data extracted and synthesized from the ProMED database and used in this letter. [Correction added after online publication 12 April 2021: The Dataset has been revised.] Fig. S1 Percentage of disease outbreaks reported in the ProMED database (period 2014–2019; n = 264) among geographical locations. Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.