Team meeting afternoon session

13.00-14.30: WP3 Global Phytophthora risks to the UK – Louise Barwell (CEH)

Louise began with a run through of the WP3 team members, objectives and associated milestones. For objective 3.1 (Risk of Introduction) a year 1 milestone has been to compile a global country level database of Phytophthora occurrence. The team has now amassed 13,853 country-level records covering 1601 Phytophthora species x country combinations including year of first record of that species in the country and invasion status. The team are aware of potential sources of bias in country-level occurrences such as level of pest recording activity per country, biosecurity, and the lag phase between pest arrival and first report. Louise has collected metrics on national pest reporting activity and posed the following question to the team; how can national recording and biosecurity efforts best be measured, ie are there descriptors of Plant Health Inspector activity ? Can we weight records against the number of inspections per country ?

Currently Louise is looking at assessing national pest reporting activity through the IPPC, legal instruments such as ECOLEX and level of biosecurity investment (from the FAO). The team have been doing a preliminary analysis developing the first models based on post-2000 occurrence data using predictors such as trade connectivity, recording effort and biosecurity. Live plant imports on their own explained 44.36% of variation in Phytophthora arrivals per country since 2000. When combined with metrics of recording effort per country, the model explained nearly 60% of the variation in Phytophthora arrivals. There followed some discussion on the use of the term ‘arrivals’ vs ‘records’. Some of the post-2000 increases in Phytophthora records per country might not necessarily relate to new arrivals but rather to increased surveillance effort (ie expansion of molecular tools) reporting species that may have been present in that country for some time. David Cooke suggested weighting the new records according to species description date.

The team have also looked at trade pathways and tried to link Phytophthora reports in each country with different commodity imports. The resource trade data allow for live plant imports to be broken down into different commodity types so it should be possible to get a sense of which types of hosts seem to be transporting the most Phytophthoras. The next steps are to collate further trade networks data, to break down total Phytophthora ‘arrivals’ to the species level, and build in country by country climate matching metrics. Once species-level arrivals can be used as the response variable, particular traits can be assessed for whether they make species more or less likely to arrive. The final output will be to predict the arrival risk of different Phytophthora species to the UK and to simulate how this might change under different kinds of trade scenarios. So data will be sought on different projections for future global trade.

For objective 3.2 (Risk of Establishment and Spread), the team have so far collated 9907 georeferenced global Phytophthora records with 1-10 km precision. The data encompass 81 species from 38 countries. In collating these data the team are prioritising regions that are climatically similar to the UK. Louise showed a slide listing the data sources and highlighted the sources of uncertainty (ie taxonomic uncertainty due to differences in use of different ID methods, positional uncertainty, recording effort resulting in false absences etc) and how such uncertainties might be overcome.

As part of objective 3.2 the team are developing niche models for different Phytophthora species predicting global impact and risk of establishment and spread in the UK. As an example, Louise showed the global distribution model for P. cinnamomi published by Burgess et al. 2016 (in Global Change Biology) and presented a list of the potential environmental drivers of Phytophthora distributions ie animal activity, pollutants, hot and cold stress. The pathologists on the Phyto-threats project sent Louise a list of 54 focal Phytophthora species with known or potential impacts on UK tree species for the niche models, including 24 species currently known to occur in the UK – this list can be found on Huddle. These species will be used to validate how well the potential distribution of Phytophthora species in the UK can be predicted with the global niche models. The other 30 species are absent from the UK and 15 of these are thought to also be absent from Europe. For these species risk maps will be created and each one will be ranked for its potential impact in the UK.

Louise also described the global Phytophthora traits database, part of which has now been merged with a similar traits database developed by a team based in Australia and New Zealand. As a result of this merging of databases we now have a comprehensive traits database complete for 172 Phytophthora species which it is thought could be managed and updated by Scion (New Zealand) over the next 10 years as part of their long-term Phytophthora research programme. The two groups aim to publish the traits database in the near future, after which it will be made freely available online. Louise then ran through some of the traits included in the merged database (ie characteristics of sporangia, oospores and other survival structures, thermal tolerances). Not yet included in the merged database are data on species distribution, host species, genome size, disease symptoms and impact metrics. These data will be added later following the publication of some manuscripts currently in preparation.

Louise ran through some of the questions they are testing using the traits database. These include examining the value of traits data for pathogen groups, the extent to which traits databases been compiled for pathogen groups in the past and whether these traits can be used in analyses of pathogen behaviour. Also, is there a phylogenetic signal in the traits ?, ie to what extent do closely related species share traits, which traits seem to be experiencing the strongest selection pressures, and are there groups of traits that are co-evolving in response to global movement of Phytophthora ?. Another question is do hybrids share traits with parent taxa ?.

Louise then showed a Phytophthora phylogeny based on the ITS region that had been sent to her by Treena Burgess in Australia. Louise has had a first attempt at measuring the phylogenetic signal in some traits, with the outcome being that there is generally a lack of phylogenetic signal in traits. Only sterile and non-caducous species had a non-random relationship, and there was a nice correlation between trait and cumulative branch distance between two species. This analysis used a ‘Brownian motion’ model of traits applied by ‘crawling’ down branches and flipping at random. Louise asked the project team what other appropriate models could be used for trait evolution for Phytophthora ?. The response was that the project team aren’t familiar with the different models but David Cooke suggested focusing on nodes as that is where species split. If Louise could circulate some of the models that she has in mind then others will able to advise. It was also noted by members of the project team that a multigene phylogeny might be better for the analyses (ie Martin et al, 2014), although the ITS-only phylogeny would probably not differ that much. David Cooke commented that it would also be nice to relate genome evolution with the trait database.

The other question being addressed is can species traits predict or explain the global impact of Phytophthora species ?, ie what makes an invasive species ?. Two impact metrics, geographic extent and host range, are being used to test this by modelling individual traits vs groups of traits (‘trait syndromes’); the idea being that particular combinations of traits will help a species get all the way along the invasion pathway (introduction > establishment > spread > impact). Louise has looked at trait syndromes and the extent to which certain traits might be positively or negatively correlated. For example being caducous and causing foliar disease are positively correlated, whereas being sterile and causing root disease are negatively correlated. Heterothallism and invasive potential on roots (in particular) and foliage are key traits that co-vary and are positively correlated. It could be anticipated that heterothallism allows greater sexual recombination and adaptation.

In terms of whether trait syndromes can explain global impacts, initial results suggest that high impact species (ie those with the broadest host ranges) tend to be heterothallic, have one or more survival structures, broad thermal tolerances and fast growth rates. Host range can also be predicted by root (and foliar) disease symptoms. In general, trait syndromes are more ecologically informative about the global impact of Phytophthoras. Louise went on to suggest that it might be possible to develop a trait-based early warning signal for Phytophthora pathogens which are known to have certain traits but for which there are no data on impact, ie to try and predict the behaviour of a novel species at an early stage. In this way potential threats can be ranked in order to inform Pest Risk Assessments.

In terms of next steps, the WP3 team will publish the traits database and phylogenetic analyses, and they aim to submit a paper linking traits and global impacts this November. Louise will continue to collate data on global Phytophthora occurrences, fine tune niche models and begin preliminary niche modelling approaches. They aim to co-develop final model outputs with policymakers.

14.30-15.00: WP4 Predicting risk via analysis of Phytophthora genome evolution – Ewan Mollison

Sarah Green introduced Ewan Mollison who recently started (August 1 st 2017) on the project as a PDRA working with Paul Sharp at the University of Edinburgh. Ewan started his presentation with a project overview and talked about the different mechanisms of evolution in oomycetes ie vertical gene transfer, horizontal gene transfer and hybridisation. Horizontal gene transfer, which is the direct transfer of genes between two species rather than from parent to offspring (known as vertical gene transfer), is of interest because oomycetes are known to have acquired genes from fungi in this manner. Ewan is interested in knowing to what extent this occurs in Phytophthora. Such events can be traced in genomes since ‘recently’ acquired genes from another species may look quite different to neighbouring genes as they tend to retain the characteristics of their donor species. Hybridisation events can also be identified in genomes by increased genome size and chromosome number or by analysis of gene sequences. Ewan presented a slide illustrating hyphal anastomosis and zoospore fusion; two possible mechanisms by which horizontal gene transfer and hybridisation may occur in Phytophthora.

The key aims of the project are (in the shorter term, ie the next six months) to sequence and assemble the genomes of three Phytophthora species that appear less damaging but are closely related to species highly damaging to woody hosts, and over the next two years to characterise key genes involved in woody host vs non-woody host infectivity, and to look for signatures of horizontal gene transfer and hybridisation between species.

Ewan presented a brief overview of Phytophthora genomes with the most extensively studied being P. infestans, P. ramorum andP. sojae. Genome sizes are variable, for example the P. austrocedri genome is 140 Mbp and consists of 47% repetitive regions. All species with annotated genome data have from 14,000 to 18,000 genes. Ewan then showed where all the current 25 genome-sequenced Phytophthora species sit by clade using the cox2 gene to calculate the trees. He went on to describe the characteristics and phylogenetic positioning of the three species to be target-sequenced in this project. In terms of looking at woody vs non-woody host infecting species, Ewan will look at ways in which their gene complements differ, whether characteristic sets of ‘woody host infecting’ genes can be identified. He will start with a comparison of P. ramorum (woody) and P. sojae (non-woody) as these are the most extensively studied species. Ewan then finished up with an overview of some background work on genome sequences of P. austrocedri isolates from Britain and Argentina, looking at similarities among isolates and between isolate groups, and whether the genome sequences might tell us anything about the origins of the two outbreaks in each country.

Questions and comments:

Q: How much of the difference between Argentinian and British isolates occurred in the source population before their separate introductions into their new regions ?.

A: We have no idea of the size (ie genetic variability) of the source population and can only infer what has happened since the founder population began in each country. In this case spread has been clonal, so the differences between the two isolate groups have largely occurred in the source population.

Q: It wouldn’t be surprising if these invasions were driven by multiple introductions. Is it possible to look at this ?.

A: Yes this scenario is likely, although the separate introductions were likely to have been of the same clonal strain in each country.

15.30-16.00: WP5 Synthesis and integration – Sarah Green

Sarah Green gave an overview of WP5 activities since May 2017, including Board meetings and the reports/research summaries recently posted on the project website. Sarah noted that a four month no-cost extension (ie to July 31st 2019) which applies to the whole project and its work packages has been agreed by the funders. Jill Thompson (THAPBI co-ordinator) followed up on this point where it was confirmed that the whole group can spend existing funds up to the new deadline (but the project will not be in receipt of any additional funding). Project reporting will be three months after the new deadline. Sarah will initiate changes to the Collaborative Agreement to reflect this.

There was some discussion of where and when to hold the next all project team meeting. It was agreed that CEH in Wallingford next April/May would be appropriate and Debbie will circulate a doodlepoll to confirm a date once Louise has confirmed the plan with Beth. Sarah then reported briefly on the Phyto-threat’s attendance at the National Plant Show back in June 2017, and talked through the seven posters that were designed for the stand display with the aim of raising awareness of the link between Phytophthora outbreaks on trees in the wider environment and spread of these pathogens in trade. It is planned that the project team attend the National Plant Show again in 2018 with some information presented on project results. Sarah also reported briefly on the 125 thIUFRO Anniversary Congress held recently in Germany, where the Phyto-threats project was represented by a talk and poster. Jill Thompson also mentioned an event to be held at Dynamic Earth in Edinburgh next month, it will be mostly targeting school children but the project could have a couple of posters there. The next stakeholder event to consider will be the THABPI dissemination event to be held in London on February 7 th 2018. Watch this space!.