Community Research and Development Information Service - CORDIS

FP7

PRATIQUE Report Summary

Project ID: 212459
Funded under: FP7-KBBE
Country: United Kingdom

Final Report Summary - PRATIQUE (Enhancements of Pest Risk Analysis Techniques)

Executive Summary:
PRATIQUE has addressed the major challenges for pest risk analysis (PRA) in Europe through three principal objectives:
1. to assemble the datasets required to construct effective PRAs valid for the whole of the EU,
2. to conduct multi-disciplinary research that enhances the techniques used in PRA,
3. to ensure that the PRA decision support scheme is fit for purpose, efficient and user-friendly.

Pest risk analysts, phytosanitary experts, invasive alien species specialists, ecologists, economists and risk modellers from 13 leading institutes in the EU, one from Australia and one from New Zealand with subcontractors from institutes in China and Russia have undertaken targeted research to review and improve existing procedures. They have also produced the first structured inventory of PRA datasets for the EU and identified large numbers of pests in eastern Asia that are highly damaging to European trees but have yet to invade Europe.

Improved methods have been developed for:
a) the assessment of economic, environmental and social impacts,
b) summarising risk using effective, harmonised, consistent techniques that take account of uncertainty,
c) mapping endangered areas,
d) pathway risk analysis and systems approaches,
e) guiding actions during emergencies caused by outbreaks of harmful pests.

The new methods for pest risk analysis have been tested with a variety of the major pests and invasive alien species affecting the cultivated and uncultivated habitats of the EU and independently validated by phytosanitary experts. Considerable assistance has also been provided by the project's observers, notably from experts in the EU Plant Health Standing Committee and from the Canadian and USA phytosanitary services.

The deliverables are provided as protocols, decision support schemes and computer programs with examples of best practice that are available to pest risk analysts through modules and direct links to the PRA scheme. The revised PRA decision support scheme based on that prepared by the European and Mediterranean Plant Protection Organization (EPPO) has been computerised providing:
(i) new users with context-sensitive guidance,
(ii) experts with not only a more efficient and user-friendly process but also greatly enhanced access to key datasets and analytical tools,
(iii) policy makers with an improved and robust scientific basis for managing risks,
(iv) stakeholders with a more transparent presentation of the risks.

EPPO has already adopted the decision support scheme for PRA revised by PRATIQUE and is committed to its future maintenance and improvement.
Project Context and Objectives:
The Context for PRATIQUE

The Context for Research on Pest Risk Analysis in the EU

According to the International Plant Protection Convention (IPPC) and the Agreement on the Application of Sanitary and Phytosanitary measures (SPS Agreement) of the World Trade Organization (WTO), countries may require phytosanitary measures only for regulated pests (quarantine pests and regulated non-quarantine pests). A PRA following the IPPC's International Standards for Phytosanitary Measures (ISPMs) is required to determine whether a pest should be regulated and measures must be based on scientific principles and cannot be maintained without sufficient scientific evidence. Phytosanitary legislation in the EU, which is harmonized for all Member States under Council Directive 2000/29/EC and includes the capacity for individual Member States to take emergency measures against the introduction and spread of new harmful organisms, must therefore also be justified by PRAs.

When the call for research in this area was announced in December 2006, it was recognised that the existing PRA methodology, data and tools were insufficiently developed to provide the satisfactory scientific assessments of plant health risks required to justify the protective measures formulated in Council Directive 2000/29/EC and to take the optimal emergency actions following outbreaks. This was hampering effective, sustainable plant health policy and decision making by the European Commission's Standing Committee on Plant Health. The research call was initiated and supported by the EU Council Working Party of Chief Officers of Plant Health Services (COPHS) to meet this important EU policy and science need.

The development of PRA science is vital to the underpinning of European Plant Health. With the dramatic increases in the volumes, commodity types and origins of trade in plant material from third countries, the introduction of new crops, the continued expansion of the EU with new border countries added and the impact of climate change, the threats posed by new plant pests are now greater than ever. In the USA the cost of non-native pests and diseases and their control has been estimated to be $120 billion and invasive alien species may cause over $314 billion per year worldwide in damage and control costs. In a recent study, the number of plant pests and pathogens establishing in Europe has been predicted to increase significantly in the next 10 years based on current trends. Such organisms can be damaging to both agriculture, horticulture, forestry, the environment and natural ecosystems. This has been amply demonstrated by the impacts of a range of different exotic plant pests. Historically, Phytophthora infestans the pathogen causing potato blight and responsible for the Irish potato famine in the 19th century, has shown the potentially enormous impact that introduced pathogens can have. More recently the introduced bacterial pathogens Clavibacter michiganensis subsp. sepedonicus causing potato ring rot and Ralstonia solanacearum causing potato brown rot, are serious on-going concerns for European potato production. Other recent examples include the sudden oak death pathogen, Phytophthora ramorum, which is causing serious problems to the horticultural industry and the environment; the pine wood nematode, Bursaphelenchus xylophilus, which was first reported in Portugal in 1999 and now threatens to spread to the whole of the EU coniferous tree production and the western corn rootworm (Diabrotica virgifera virgifera) which is one of a suite of rootworm pests whose control and contribution to crop losses have been estimated to be $1,000 million per annum in the USA. D. v. virgifera is spreading rapidly in the EU where it has been calculated to cost €147 million per annum. The earlier a pest risk is identified and measures are implemented, the greater the chance of preventing such impacts. Enhancing methods for assessing and managing risks, as developed by PRATIQUE, can help tackle these threats and thus serve to protect European agriculture, horticulture, forestry and plant resources in the natural environment through evidence-based and risk-based policy.

In addition, challenges to EU plant health regulations by non-EU countries seeking new markets are also increasing. Since regulations can only be defended by PRA, this further highlights the importance of PRA as a key component of EU policy. Efficient, effective PRAs ensure that measures prevent the entry of pests that pose unacceptable risks while allowing trade to flow as freely as possible, maintaining EU competitiveness and providing consumers with the greatest choice, quality and value for money. It was therefore vital that expertise from across Europe and other key nations was brought together to develop more efficient PRA techniques and PRATIQUE provided a unique opportunity to do so.

PRATIQUE has also conducted a significant amount of work to enhance monitoring/surveillance strategies, contingency planning and eradication/containment programmes following new outbreaks that currently tend to be undertaken in an ad hoc manner, without learning from past campaigns and using structured protocols.

Pest Risk Analysis: The state of the art at the beginning of the project

Identifying those pests that can enter, establish, spread and cause significant impacts is inherently difficult. There are many species that are of minor importance in their area of origin but become serious pests when invading new areas. The opposite also occurs very frequently. The PRA process can never be perfect but provides, through the assembly and consideration of all the important factors in a logical sequence, the key mechanism for providing sustainable protection for our crops and wild flora. A great deal has been learnt since the early 1990s when PRA procedures were first formulated. The research undertaken by PRATIQUE came at an ideal time to capitalise on these advances by ensuring the data required are more readily available, enhancing the techniques used by multi-disciplinary research and providing a PRA scheme that is straightforward to understand and apply.

Although the IPPC PRA standards set out the structure for PRA, they do not provide a decision support scheme that enables the analyst to work through a logical series of questions for each pest or pathway in order to reach a conclusion. To meet this requirement, PRA schemes continue to be developed and revised by regional, national plant protection organisations (NPPOs) and other institutes. At the beginning of the project, the EPPO PRA scheme was the de facto standard for undertaking detailed PRAs throughout Europe and had even been successfully adapted for non-native species by the UK. EPPO had also published a shortened version of the PRA scheme to support action in emergencies and both the UK and the Netherlands also used short schemes. Although these schemes had a number of limitations (that have been addressed by PRATIQUE), they enabled NPPOs in the EU to produce large numbers of PRAs that follow IPPC standards and justify regulations. Recognising the key role played by PRA in the EC Plant Health Directive, the increasing global scrutiny of phytosanitary legislation and the need to ensure that the decision making process is separated from risk analysis, an independent panel within the European Food Safety Authority (EFSA) had been set up to review PRAs and enhance pest risk assessment procedures. Although PRATIQUE has based its research on the enhancement of the EPPO PRA scheme, it has liaised closely with EFSA and all those involved in plant health within the EU.

The three major challenges for Pest Risk Analysis

At the start of PRATIQUE, despite the importance of PRA for plant health in the EU, the data required to make accurate analyses of the risks throughout the EU were often lacking, PRA processes were insufficiently exploiting important new scientific and technological developments and the PRA procedures were complex, discouraging take-up among all EU member states. The three major challenges that required concerted action to resolve by PRATIQUE were therefore:

(i) The lack of sufficient data required to analyse effectively the risks posed by pests to all member states of the EU,
(ii) The need for multi-disciplinary research to enhance the techniques used in PRA to ensure relevant, effective and efficient risk analysis and action,
(iii) The importance of ensuring the PRA process is fit for purpose and user-friendly.

The Objectives of PRATIQUE

The three major challenges for PRA identified above were directly reflected in PRATIQUE's three principal objectives (and the organisation of the project into six technical work packages (WPs)):

1. To assemble the datasets required to construct PRAs valid for the whole of the EU (WP1)
2. To conduct multi-disciplinary research to enhance the techniques used in PRA for:
* the assessment of economic, environmental and social impacts (WP2)
* producing accurate, consistent PRAs, mapping endangered areas and summarising risk and uncertainty (WP3)
* pathway analysis and systems approaches (WP4)
* guiding actions during pest outbreaks (WP5)
3. To ensure that the PRA scheme is fit for purpose and user-friendly (WP6)

The objectives of each work package are described in more detail below.

Workpackage 1 - Identifying and integrating key national and international datasets
* To construct an inventory of the key national, European and global datasets required for the production of PRAs relevant to the EU
* To assemble the key datasets, attempt to fill gaps, identify opportunities for integration and ensure they are readily accessible to pest risk analysts
* To collect data, particularly on pests damaging native European trees planted in Eastern Asia, and build the datasets that are needed for the development of new tools, the enhancement of existing techniques, test cases and examples of best practice undertaken in WP2 to WP6.

Workpackage 2 - Enhancing techniques for economic, environmental and social impacts
* To determine the extent to which the analysis of species traits can be used to identify those species that can cause significant impacts in cultivated and uncultivated habitats
* To develop a novel scoring system to assess impacts and determine thresholds for phytosanitary action
* To enhance existing techniques and develop new tools for assessing economic, environmental and social impacts
* To develop a generic integrated model to assess pest spread and impacts

Workpackage 3: Enhancing techniques for standardising and summarising pest risk assessments
* To enhance the consistency and standardisation of pest risk assessments by identifying and applying appropriate criteria
* To develop and test new techniques for quantifying uncertainty in pest risk assessments
* To enhance techniques for mapping endangered areas taking current and future climate, land use and economic impacts into account
* To develop and test new techniques for summarising and communicating overall risk in pest risk assessment

Work Package 4 - Enhancing techniques for pathway analysis and systems approaches
* To review current approaches to pathway analysis in PRA
* To review the current application of systems approaches in PRA
* To develop a pathway risk analysis module for the PRA scheme with a protocol for the application of neural networks and methods for enhancing consistency
* To develop a systems approach module for the PRA scheme

Workpackage 5 - Developing a decision support system for the eradication and containment of pest outbreaks
* To carry out a meta-analysis of the successes and failures of pest eradication and containment programmes worldwide
* To provide guidance for analysing the cost-effectiveness of pest eradication and containment measures
* To develop a decision support scheme to support actions to be taken following pest outbreaks
* To provide recommendations for the application of pest surveillance techniques in detecting pest incursions and managing outbreaks

Workpackage 6 - Project validation and dissemination with the development of a web-based PRA scheme
* To validate the outputs from work packages 1 to 5 using independent experts and a wide range of pests and pathways
* To create a web-enabled EPPO PRA scheme incorporating outputs from work packages 1 to 5
* To consolidate and disseminate project outputs by providing a manual and examples of best practice with the web-enabled PRA scheme

Project Results:
1.3.1 Introduction

The main scientific and technical results and foregrounds are described according to the different work packages and tasks tackled by the project:

WP1: Identifying and integrating key national and international datasets WP2: Enhancing techniques for economic, environmental and social impacts WP3: Enhancing techniques for standardising and summarising pest risk assessments WP4: Enhancing techniques for pathway analysis and systems approaches WP5: Developing a decision support system for the eradication and containment of pest outbreaks WP6: Project validation and dissemination with the development of a web-based

1.3.2 Description of the results by work package and work package task

WP1: Identifying and integrating key national and international datasets
Summary:
Over 800 electronic datasets and information sources have been collected, not counting books, representing the largest effort ever made to collect datasets and information sources needed for carrying out PRAs. To obtain datasets from EU member states, a detailed survey was carried out by sending a questionnaire to EU member states and EPPO PRA experts. Twenty-seven descriptors were chosen to reflect the categories of information needed when performing a PRA to allow the datasets to be linked to specific components of the EPPO PRA scheme. The quality and usefulness of the web-based datasets and information sources were also assessed. Over 500 datasets, that were considered to be of sufficiently good quality and usefulness, have been stored in a database that is directly accessible from the computerised PRA scheme developed by WP6. Significant gaps were identified showing that some European countries provided no or very few datasets and information for some of the 27 descriptors were particularly poorly represented in the database. Schemes for pest risk analysis and related topics, e.g. risk analysis for invasive alien species, animal health, food security and GMOs were collected and collated for use throughout the project. Novel methods have been employed to obtain lists of potential pests before introduction to the EU by surveying arboreta that contain European trees in Siberia and sentinel European woody plants that have been planted in two different climatic zones in China: Beijing (continental-temperate climate) and Hangzhou (warm-humid sub-tropical climate). A large number of invertebrate pests and pathogens, many of which are causing serious damage, have been collected and identified.

WP1 provides: (i) an inventory of relevant EU datasets prepared and structured according to the PRA scheme (D1.2, 1.3), (ii) a gap analysis of datasets required for PRA, particularly for conducting PRAs relevant to all EU member states (D1.4), (iii) a common source for the information required to assess the reliability, efficacy and cost effectiveness of pest management measures to prevent entry and combat outbreaks (D1.2), (iv) a dataset of potential Asian pests of selected woody plants not yet introduced into Europe obtained by the sentinel tree technique in China and observations of arboreta in eastern Russia using methods developed by PRATIQUE (D1.5) and (v) recommendations for a protocol to establish lists of potential plant pests before they are introduced into a new continent (D1.6).
Task 1.1. Information on the pest in its current area of distribution Task 1.2. Information on pathways, including trade, production and economic datasets Task 1.3. Information on the area under consideration for the PRA Task 1.4. Information on pest management
The work on these tasks was carried out in parallel, each of these being coordinated by a task coordinator. The work followed the description of the tasks in the Description of Work. In brief, datasets and information sources were collected, assessed and integrated in the following way: The first step was to identify existing national and international datasets and information sources that are already available. These datasets and information sources were collected by all PRATIQUE partners and integrated in four Excel spreadsheets representing the four key information components required for PRAs. For each dataset and information source, basic information and indicators were collected. To obtain additional national datasets, a European survey was carried out by contacting the national plant protection organisations of all EU member states either by sending a questionnaire by email (20 non-PRATIQUE members) or through participants of an EPPO workshop on PRA held in Cyprus in October 2008. Over 800 electronic datasets and information sources were collected, not counting books. Descriptors were identified for each dataset. In order to enhance the evaluation of the datasets, 27 descriptors were chosen to reflect the categories of information needed when performing a PRA according to the EPPO Decision support scheme for PRA and thus to allow the datasets to be linked to specific sections (or questions when relevant) of the scheme.
The quality and usefulness of the web-based datasets and information sources were assessed. Assessors were asked to evaluate the datasets per category descriptor, as well as for the entire set, by giving an appropriate quality mark. To facilitate accessibility, over 500 datasets, that were considered to be of sufficiently good quality and usefulness, have been stored in a database that is accessible through a simple portal (a data explorer). This is directly linked to the computerised PRA scheme, CAPRA, as described in PRATIQUE deliverable 6.4.
Relevant datasets collected by PRATIQUE were analysed to determine the temporal and spatial trends in interceptions and in the entry and establishment of alien plant pests, pathogens and plants in Europe.
An analysis to identify significant gaps in the data needed to carry out PRAs in Europe has been made using the results from the dataset collection carried out by tasks 1.1 - 1.4 and a survey among the PRATIQUE partners. In this gap analysis, we identified: (i) the European countries for which we had limited or no information, in particular on pathways, trade and other economic datasets; climate; soil; habitat, land use; host plants, etc, (ii) the descriptors used to describe datasets which were particularly poorly represented in the database and (iii) datasets and information sources that were made available during the PRATIQUE project by PRATIQUE partners and collaborators. In particular, we assessed the applicability of the extensive dataset compiled by SEAMLESS, a recently completed EU FP6 project; (iv) datasets and information sources that may not currently exist but would be useful to create or to build into the framework of other projects in order to improve PRA activities in Europe. The detailed results of the four tasks are provided in three project deliverables (the first two have been combined into a single document): D 1.2. Inventory and description of all datasets needed for PRA in the EU that are currently available provided as tables and accessed via hyperlinks in structured project web pages D 1.3. Key datasets in a variety of formats accessible via hyperlinks in structured project web pages with a trend analysis and available for integration into the PRA scheme developed in WP6 D 1.4. Report identifying the gaps in the key data required to undertake PRAs valid for the whole EU During the project, we continued to add new datasets of direct relevance to PRA production in the EU as they were revealed or created. In particular, we also exchanged our datasets with those of PRASSIS, an EFSA-funded project, also designed to deliver an inventory of data sources for pest risk assessment. The PRASSIS database was carefully examined and any dataset that filled a gap in the PRATIQUE database was incorporated.
Task 1.5. Collection of PRA schemes Schemes for pest risk analysis and related topics, e.g. risk analysis for invasive alien species, animal health, food security, GMOs etc., were collected by all PRATIQUE teams and collated by CABI. An Excel spreadsheet listing the different risk analysis schemes and examples of best practices was produced. The schemes and examples were provided as Internet links and/or as PDFs. Hundreds of documents were found, originating from 34 countries or regions worldwide. These have been used by all PRATIQUE partners to perform various tasks, in particular to identify best practices for assessing impact (task 2.1), enhancing consistency (task 3.1), quantifying and capturing uncertainty (task 3.2), mapping endangered areas (task 3.3), summarising risk (task 3.4), and performing pathway analyses (task 4.1). Some of the schemes and examples cannot be provided on a public website since they were provided in confidence. For these, we have provided the document title and reference together with Internet links and contact details for the authors and organisations on the PRATIQUE website. For those that are publically available, we have provided the links to the web pages from which the documents were obtained. Task 1.6. Novel method to obtain lists of potential plant pests before introduction A method was developed and tested to detect new potential pests in their region of origin, before they are introduced into a new continent. Eastern Asia was selected because it is currently the main source of alien woody plant pest introductions into Europe. This method consists of two approaches: i) in Siberia and the Russian Far East, surveys were carried out in several arboreta and botanical gardens to detect arthropods and pathogens attacking European trees and shrubs. Surveys in Russian arboreta were carried out in collaboration with the sub-contractors of the V. N. Sukachev Institute of Forest, Krasnoyarsk and collaborators. (ii) sentinel European trees were planted in two areas in China and monitored at regular intervals for damage caused by indigenous arthropods and pathogens. This activity was carried out in collaboration with the sub-contractors of the Chinese Academy of Sciences and collaborators.

In Central Siberia, five arboreta were surveyed in summer 2008, 2009 and 2010 to identify pest damage on European woody plants. An additional survey was carried out at two arboreta in the Russian Far East in 2010. Severe damage by arthropods has been identified on several plants. Many species have been determined but others are still being identified by specialists. In particular, a very serious dieback of exotic walnut species has been observed in the Far East but the primary cause still remains to be identified and further investigations are planned. Surveys for woody plant fungal pathogens were carried out mainly by Dr. Maria Tomoshevich, a plant pathologist of the Central Siberian botanical garden in Novosibirsk. She started surveying foliar pathogens of woody plants in Siberia in 1997 and has kindly agreed to provide us with her observations on European woody plant species for further analysis.

Data collected from arboreta in Siberia have also been used to test various ecological hypotheses linked to biological invasions. In particular, we have tested whether alien plants are attacked by fewer herbivore species compared to native plants, a prerequisite for the enemy release hypothesis and whether some native pest species may survive better on alien plants because they did not co-evolve with them, following the new association theory. The results show that exotic trees are less attacked by native leaf miners and gall makers than native trees. In contrast, external defoliators attack alien and native trees equally. In the second approach, sentinel European woody plants were placed in two different climatic zones in China: Beijing (continental-temperate climate) and Hangzhou (warm-humid sub-tropical climate). Four species were planted in Beijing and seven in Hangzhou. Survival rates following plantation were much higher in Beijing than in Hangzhou. Damage by arthropods and pathogens was monitored at 2-4 week intervals. At both sites, a high number of insects and pathogens were observed damaging, and sometimes even killing, seedlings. Most insects have been identified, in contrast to pathogens.. Interestingly, many of the damaging insects are polyphagous species that are found more commonly in agricultural fields.A total of 34 insect species showed more than five colonization events during 2008-2010. These species are potentially the most dangerous pests for the European trees tested. Their relative importance was determined by comparing their biology, the damage caused and annual population densities. Six species have been proposedfor PRA and are currently being studied by INRA.

The methods and results of these studies are presented in Deliverable 1.5 (A dataset of potential Asian pests of selected woody plants not yet introduced into Europe). This includes a database of arthropods and pathogens found by arboretum surveys and sentinel trees highlighting the most serious problems requiring PRAs or further investigations. In this deliverable, we discuss the strengths and weaknesses of the methods. In deliverable D.1.6 (A written protocol to establish lists of potential plant pests before they are introduced into a new continent), we suggest recommendations for implementing these methods in future projects. In particular, we recommend the establishment of an early warning system through a global network of arboreta and botanical gardens.
Highlights:
* Tasks 1.1 to 1.4 is considered to represent the largest effort ever made to collect datasets and information sources needed for carrying out PRAs.
* In Task 1.6, we developed and tested new methods to detect new potential pests in their region of origin, before they are introduced into a new continent.

WP2: Enhancing techniques for economic, environmental and social impacts
Summary:
Following a review of six techniques for quantitative economic impact assessment, the use of Partial Budgeting and Partial Equilibrium methods was recommended. The review of techniques for environmental impact assessment indicated that a qualitative scheme is more appropriate than quantitative methods because of the complexity of ecological processes, the wide variety of impact types and the paucity of methods to evaluate the different types of impact. No generic methods for social impact assessments could be identified. The analysis of traits which may serve as indicators for predicting environmental and economic impacts was undertaken with insects (environmental and economic impacts), forestry pathogens (environmental impacts) and plants (environmental impacts) but limitations in the data (quality and quantity) prevented a generic approach from being adopted and robust conclusions from being made. For insects, the mode of reproduction was significantly related to environmental impacts while the reduction of photosynthetic activity was a good predictor for economic impacts. The study on plant traits showed that shrubs and short trees are most likely to have significant environmental impacts and traits predicting invasiveness are similar to traits predicting impacts. The most important traits for fungal pathogens were long distance dispersal and sexual reproduction for invasiveness and optimal growth temperature and infection of perennial organs for ecological impacts. A wide host range for fungi was negatively correlated with invasiveness.

Specific indicators have been developed to assist with the assessment of economic, environmental and social impacts in the EPPO Decision Support Scheme (DSS) for PRA and both the questions and the guidance have been revised. In addition, rating guidance has been provided and, where possible, examples have been added to enhance consistency. The environmental impact questions have been subdivided into several sub-questions that address specific environmental aspects. A number of tools have also been added to support pest risk analysts, e.g. flow charts for determining yield variation and indirect impacts, and to summarize impact assessments (rule-based matrix models). The matrix model for environmental impacts consists of two versions: one for plant pests and one for pest plants. A quantitative economic impact assessment module based on the partial budgeting approach to assess direct economic impacts and the partial equilibrium approach to assess indirect economic impacts has been developed in GAMS with a user-interface in Excel. This module has been applied to two case studies: the Potato Spindle Tuber Viroid and the pine wood nematode Bursaphelenchus xylophilus. A prototype matrix model for social impact assessment has been developed, especially to estimate the impacts of the pest on the landscape. Finally a suite of spread models programmed in the R language has been provided to simulate the geographical expansion and build up of population density over time. Some of the models are spatially explicit. Flow charts have been developed to show the linkages between the modules themselves and the EPPO DSS for PRA.
WP2 provides: a review of best practice for impact assessments (D2.1), an investigation of the role of species trait analysis in PRA (D2.2), indicators and a revised qualitative scheme for economic, environment and social impacts (D2.3), guidance for assessing the most vulnerable economic, environmental and social receptors (D2.4) models for quantifying economic impacts (D2.5), a suite of models for predicting spread (D2.6) and a generic integrated model that shows how the qualitative PRA scheme and the quantitative modules are linked together (D2.7).

Task 2.1: Review of impact assessment methods: Deliverable 2.1 provides a review of techniques for economic impact assessment, emphasising the techniques that can be applied to quantify economic impacts. It also includes a review of techniques for environmental impact assessment. Because of the lack of quantitative methods, the review focuses on qualitative assessments contained in pest and weed risk assessment schemes. Techniques for assessing social impacts were also explored. Methods for integrating the different types of impacts were also investigated, particularly to determine whether monetizing methods should be applied. Highlights: Partial Budgeting and Partial Equilibrium methods are recommended for quantitative economic impact assessments. A qualitative scheme that captures empirical studies and expert judgment, based on existing schemes and best practices, is proposed for environmental impact assessment. Environmental impacts are much more difficult to predict than economic impacts because of the complexity of ecological processes, the wide variety of impact types and the paucity of methods to evaluate the different types of impact. No generic methods for social impact assessments are available. A simple multi-criteria analysis based approach is suggested for integrating and summarizing impacts. Monetizing methods will not be applied because of the amount of data and work required and their limited role in decision-making. Documented results: Deliverable 2.1 Task 2.2: Analyzing the traits of pests causing significant impacts The report on species traits analysis (D 2.2) focused on the analysis of pest characteristics that may serve as indicators for predicting environmental and economic impacts. Case studies have been undertaken for insects (environmental and economic impacts), forestry pathogens (environmental impacts) and plants (environmental impacts) based on existing databases. Limitations in the data (quality and quantity) have prevented a generic approach from being adopted and robust conclusions from being made. Highlights: For insects, the mode of reproduction is significantly related to environmental impacts. Reduction of photosynthetic activity by insects is a predictor for economic impacts. The study on plant traits showed that shrubs and short trees are most likely to have significant environmental impacts. Plant traits predicting invasiveness are similar to traits predicting impacts. The most discriminating traits for fungal pathogens are long distance dispersal and sexual reproduction for invasiveness and optimal growth temperature and infection of perennial organs for ecological impacts. A wide host range for fungi is negatively correlated with invasiveness. Documented results: Deliverable 2.2 Task 2.3: Guidance for identifying indicators and scoring levels of impact in the EPPO PRA scheme Specific indicators have been developed to assess economic, environmental and social impacts in the EPPO DSS for PRA and guidance has been provided to enhance consistency in rating the magnitude of impacts. Where possible, examples have also been added. The results were provided to D3.2. During the execution of this task, it became apparent that the questions in the impact section of the PRA scheme needed to be revised. The order of questions has been changed to group all economic, environmental and social questions together and the wording of the questions and notes has been adjusted. Indicators have been included in the notes. The environmental impact questions have been subdivided into several sub-questions that address specific environmental aspects. Two separate versions have been developed: one for plant pests and one for pest plants. Several tools have been added to help answer the questions and to summarize impacts, including a decision tree to determine yield variation, a decision tree to determine how indirect economic impacts are distributed among stakeholder groups and rule-based matrix models to summarize the qualitative assessment of impacts. Following the recommendations in the review of impact assessment methods (D2.1), the qualitative environmental impact assessment approach as scheduled in task 2.3 has been incorporated in the computerised PRA scheme (D6.4). The changes to the scheme have been tested by pest risk analysts attending the EPPO Pest Risk Analysis Development Panel and Pest Risk Analysis Workshops (see D6.2).

Highlights: The impact section of the EPPO PRA scheme has been considerably enhanced by providing:
- revisions to the wording of the questions, especially those related to indirect economic and environmental impacts
- indicators for each question
- rating guidance for scoring impacts
- tools for summarizing economic and environmental impacts.

Documented results: Deliverable 2.3
Task 2.4: Development of modules for assessing (a) economic, (b) environmental and (c) social impacts The recommendations from the review conducted in D2.1 were applied. This resulted in a module for quantitative economic impact assessment programmed in GAMS with a user-interface in Excel. This module enables the risk assessor to calculate direct economic impacts by applying the partial budgeting approach and to determine how indirect impacts are distributed in the whole production and trade chain, including consumers, by partial equilibrium modelling. The module makes use of economic data generated by the EU FP6 project SEAMLESS. The quantitative module is appropriate for use in cases when
(a) at least moderate economic impacts are expected,
(b) these impacts cover a large area or affect different industries and
(c) resources (data, time and skills) are available to conduct a quantitative analysis.

Quantitative assessments are particularly likely to be required where the level of impacts is unclear or the analysis is required to justify measures. Two case studies are provided for the Potato spindle tuber viroid (PSTVd) and the pine wood nematode (PWN) (Bursaphelenchus xylophilus).

A protocol for environmental impact assessment has been developed and integrated into CAPRA (see task 2.3). A prototype matrix model has been developed to assess social impacts with the emphasis on landscape effects. The model combines the damage to host plants with the importance of the host plants to the value of the landscape. The landscape value is based on land use (agriculture, nature and living areas), aesthetic and cultural factors. However, due to lack of data, the model could not be validated.

Following a review of the vulnerability concept in relation to PRA, it was found that a separate framework for assessing the vulnerability of the receptor environment need not be undertaken because the bio-ecological and management types of vulnerability are already included in the EPPO DSS for PRA. The broader concept of social vulnerability related to management options is not included in the International Plant Protection Convention (IPPC) standard ISPM11 (FAO, 2004) and so has not been studied by PRATIQUE. However, guidance for assessing the most vulnerable economic, environmental and social receptors has been provided. This is particularly important in question 6.15 of the EPPO DSS for PRA where assessors are asked to identify the endangered area and look at the different receptors, identify which are at highest risk and locate where the pest presence will result in "economically important loss." Since this is related directly to risk mapping, the identification of receptors that are at most vulnerable (at highest risk) has been undertaken in Deliverable 3.3 (the protocol for mapping endangered areas taking climate, climate change, biotic and abiotic factors, land use and economic impacts into account).
Highlights: An economic impact assessment module has been developed and applied to Potato spindle tuber viroid and the pine wood nematode (Bursaphelenchus xylophilus). A protocol for environmental impact assessment has been produced and incorporated in CAPRA. Documented results: Deliverables 2.4, 2.5.

Task 2.5: Development of a generic integrated model The development of a generic integrated model or framework has focused on two sets of tools: (i) a suite of spread models and (ii) flow charts reflecting the relationships (a) between the modules for quantifying components of risk and (b) with the qualitative EPPO DSS for PRA.

The five spread models cover different aspects of the dynamic spatio-temporal process of pest range expansion. They have been programmed in the computer language R to make them as straightforward as possible for risk assessors to run and build familiarity. Examples are provided to help the users. Both the geographical expansion of the pest and the increase in population density are modelled within the area of potential establishment based on host distribution and climate suitability. The models were presented and tested at a meeting of pest risk assessors in Hammamet, Tunisia on 23rd-26th November 2010 and detailed examples are presented in this document. Further testing is required to clarify their role in PRA.

The flow charts show how the modules for climatic mapping, quantitative economic impact assessment, spread and mapping endangered areas are linked to each other and to the EPPO DSS for PRA. An additional flow chart is provided in which the procedures to be followed in the DSS for mapping the endangered area are set out.

Highlights:
A suite of five spread models are provided to enable risk assessors to explore the dynamics of the invasion process.
Flow diagrams show how the quantitative modules are linked to each other and to the EPPO DSS for PRA.

Documented results: Deliverables 2.6, 2.7. WP3: Enhancing techniques for standardising and summarising pest risk assessments
Summary:
The review of best practice identified a range of approaches that could be used directly or following adaptation to enhance consistency in the EPPO DSS for PRA. The enhancement of consistency in the scheme has focussed on providing readily understandable, unambiguous questions for each stage of the scheme, clear rating guidance for each question, pest examples, standardised answers and an easily applicable method to interpret and summarise risk ratings. A wide variety of methods for capturing uncertainty and summarising risk has been explored. Bayesian Belief Networks (BBN) and a knowledge-based approach (KBA) combined with visualization techniques that resemble risk matrices, were initially shown to be the most appropriate for application to the EPPO PRA scheme. Further testing revealed that the PRA Risk Rating and Uncertainty Visualiser, the rule-based Matrix Models (MMs), and the Invasive Risk Impact Simulator (IRIS) were most useful for quantifying and communicating uncertainty and for summarising and communicating overall risk in PRA. A review of risk analysis schemes and standards revealed that there is no existing guide to best practice in mapping endangered areas and this can currently only be inferred by analysis of the different methods used and examples available. Two decision support schemes have been developed: (a) to help users determine whether they have sufficient information to model and map climatic suitability and select from the range of methods available and (b) to map endangered areas or, at least, the areas at highest risk by combining maps of the area of potential establishment and factors related to impacts using spatial multi-criteria analysis software.

WP3 provides: decision rules and pest examples for selecting one of the five possible levels of risk in most questions in the scheme thus enhancing: consistency (D3.1), procedures for capturing and quantifying uncertainty based on advances in risk analysis science (D3.2), procedures for mapping endangered areas based on abiotic, biotic and economic factors (D3.3) and procedures for objectively summarising risk based on an evaluation of all available techniques (D3.4). Deliverables 3.2 and 3.4 have been combined to avoid repetition and because the concepts involved in the two deliverables are very closely related.

In WP3, four key risk assessment problems have been addressed: the lack of consistency in scoring responses to PRA questions, difficulties in capturing and communicating uncertainty, deficiencies in mapping endangered areas and a need to enhance methods for summarising risk. Task 3.1 Techniques and tools for enhancing consistency in pest risk assessment A review of best practice for achieving and enhancing consistency in PRA has been made by evaluating international, regional and national standards and guidelines for plant health pest risk assessment and examples of their use from countries around the globe. These were examined together with similar documents from related fields such as animal health, nature conservation, genetically modified organism (GMO) and weed risk assessment. The results of this review were used to develop a protocol to provide clear guidance for the different risk ratings. The detailed risk assessment stage of the EPPO DSS for PRA contains questions that require the assessor to select one of five alternative responses, e.g. very unlikely, unlikely, moderately likely, likely, or very likely (in some cases additional options such as N/A or no judgement possible/ ask an economist are given). A few questions are "Yes/No" questions. A correct, consistent response to each question is not only of importance because of each question's contribution to the overall assessment of risk but also because several questions in the scheme are correlated and the answers from one question can be used to inform other parts within the PRA process, e.g. the assessment of spread and the reliability of pest free areas. WP3 provided risk-rating guidance for the establishment section of the EPPO DSS for PRA while WP2 and WP4 contributed guidance for the rating of impacts and entry respectively. A range of approaches was identified that could be used directly or following adaptation to enhance consistency in a revised EPPO DSS for PRA. None of the risk assessment documents evaluated was found to contain a suitable mechanism that ensures or guarantees consistency and no single scheme contained all of the approaches identified to maximise consistency. The primary needs required in the EPPO DSS for PRA to enhance consistency were (i) the provision of examples that clearly illustrate the divisions in each scale, or alternative but equivalent descriptors that allow assessors to distinguish between the divisions and (ii) a mechanism to combine risk elements in a consistent and transparent way. Features that would help inexperienced assessors include: an uncomplicated structure, clear rating guidance, questions posed unambiguously, the provision of standardised answers and an easily applicable method to interpret and summarise risk ratings. Providing links to information and suggesting data sources that would help assessors answer questions would also be helpful. For some questions, it has been possible to assign relatively straightforward risk ratings or guidance, while others required considerable additional investigations. One single approach for all questions was not feasible due to the different characteristics of particular questions. Acknowledging these differences, proposals for rating guidance include: (a) transforming the notes to the questions into sub-questions that can be answered with a yes or a no with guidance on how to summarise the answers, (b) by listing factors, characteristics or conditions adjacent to each rating, (c) by providing examples for different types of pests (insects/mites, fungi, bacteria, plants, nematodes, viruses/viroids) so that the assessor can decide on the appropriate rating by comparing the pest with another pest (usually of the same taxon), and/or (d) by using written probability or impact terms tied to accepted international guidance notes and standards. In addition, part of the EPPO DSS for PRA was restructured, now allowing the assessment to focus on the most relevant factors influencing establishment and to make it easier to determine the suitability of the PRA area for establishment by first identifying the area of potential establishment.
Highlights: Development of a consistent rating guidance, clear formulation of questions in the EPPO DSS for PRA, restructuring of parts of the EPPO DSS for PRA, provision of examples for different types of pests.
Documented results: Deliverable report D3.1. Task 3.2 Techniques for quantifying and communicating uncertainty in pest risk assessment merged with Task 3.4 Methods for summarising and communicating risk in pest risk assessment In the review of methods for quantifying and communicating uncertainty and for summarising and communicating risk, many methods have been explored to determine whether they have the potential for rating and combining risk ratings to produce an overall measure of pest risk. Special attention has been given to approaches used to characterise and propagate uncertainty within scoring frameworks, but, for completeness, several methods, that are not necessarily based on scores but which may provide useful tools for risk assessment purposes, have also been explored. The review was undertaken by investigating methods that have been successfully applied in other areas to rate factors that affect the risk that is being assessed. These included methods that transform some ratings, either by converting a qualitative rating into a numerical ranking or aggregating several component scales into a single overall rating.
Similar to the outcome of the review of consistency in task 3.1, it was found that each method has both strengths and weaknesses and no single approach was found to fulfil all requirements. Most methods are simplistic and do not propagate uncertainty very well whilst others perform better but are more complex to implement. Discussions with PRA experts indicated that they prefer a method that is not only simple to implement and use but that is also based on the EPPO DSS for PRA that uses a risk rating system. These requirements limit the range of potential methods and exclude most, if not all, methods that are better at quantifying and propagating uncertainty. A comparison was made of the different methods to select the most appropriate methods, based on the following criteria: 1) transparency, 2) rigour in dealing with uncertainty, 3) consistency, 4) scientific defensibility and 5) ease of use. While many of the methods failed the first four criteria, Probability Bounds Analysis and Bayesian methods match these quite well, but they tend to be difficult to implement and/or use. Following the results of this comparison, a proposal was therefore made to proceed further with Bayesian Belief Networks (BBN) and a Knowledge-based Approach (KBA). BBNs provide a potential compromise because a) they are based on scores and b) they are capable of propagating uncertainties. Their main disadvantage is that they are hard to develop and implement. Therefore, as an alternative, a KBA, which combines many of the approaches that were assessed was initially selected. The KBA is enhanced by Monte Carlo simulation and based on the EPPO DSS qualitative risk rating system. To facilitate the development of the BBN approach, it was originally proposed to develop a model based on a simplified version of the EPPO DSS for PRA using the four major components of the PRA process: entry, establishment, spread and impact. The advantages and disadvantages of both methods were described. For the two methods, demonstrations and prototypes were developed for clarification and understanding. The BBN approach was not developed further primarily because of the significant time required from PRA experts to build conditional probability tables (CPTs), particularly at question level, and test them for logic and consistency. In addition, because experts may have very different views on how answers should be combined, compiling a unique CPT would need a considerable amount of work, communication and agreement. The KBA was also not developed further, since the different experts involved in the development of the model could not agree on the knowledge components and the further time investment needed to try to solve this. However, elements of the KBA can be found in other methods developed in PRATIQUE and the exercise on clustering questions proved to be very useful when drawing up the ontologies for the matrix models.

Building on experiences made with the review and these two methods, three other methods were developed in collaboration with the EFSA funded project Prima phacie: (a) the risk rating and uncertainty visualiser (a tool that graphically displays risk and uncertainty scores according to what the expert has answered - no calculation is involved), (b) rule based matrix models (MMs) (that visualize the logical sequence and relationship of the different links between questions providing a logical hierarchical structure to the EPPO DSS for PRA and summarizing the answers based on rules, also including uncertainty) and (c) the optional invasive risk impact simulator (IRIS) (a tool developed to interpret the subjective overall scores given by risk assessors for entry, establishment, spread and impact, converting scores of likelihood/impact (including uncertainty) into potential expected annual costs with a time horizon). These three methods not only provide the optimal combination of relevant factors identified in PRATIQUE, but, following testing (see D6.2), were also positively received by pest risk analysts. The "daily use" of the models/approaches when conducting PRAs will help to further improve and test the approaches; this experience can only be gained over the years. The results will be followed up by EPPO.

Highlights: The development of three innovative methods to quantify and communicate uncertainty and to summarise and communicate risk in pest risk assessments, namely: the risk rating and uncertainty visualiser, the rule based matrix models (MMs), and the invasive risk impact simulator (IRIS). The Bayesian network analyses developed in WP3 (D3.2/3.4) are being taken up in a WTO STDF project "Beyond Compliance" (STDF/PG/328) in 2011-2013.
Documented results: Deliverable report D3.2/3.4 Task 3.3 Techniques and tools for mapping endangered areas To identify and assess different approaches to mapping endangered areas and identify best practice, the literature and international standards were reviewed. The review revealed that there is no existing guide to best practice in mapping endangered areas. Before PRATIQUE, best practice could only be inferred by analysis of the different methods used and examples available.

For climatic mapping, the appropriate datasets and the deductive, inductive and combined models were compared and a DSS to help users determine whether they have sufficient information to model and map climatic factors was developed and tested. The DSS is designed to help assessors (a) determine whether it is appropriate to model and map climatic suitability in the PRA area, (b) justify why, if not done, the modelling and mapping of climatic suitability has not been undertaken, (c) evaluate the type, quantity, accuracy, reliability and precision of available pest climate response and distribution data and relate this to the performance of the different modelling and mapping methods and (d) distinguish between the different methods by comparing their general usability, functionality, applicability and relationship to ecological processes. Drosophila suzukii is provided as an example.

The DSS for mapping endangered areas begins by assisting pest risk analysts in deciding whether it is appropriate to try and create a map of endangered areas and then provides guidance on the most suitable methods to follow. Four possible stages can be followed. In stage 1, the key factors that influence the endangered area based on the draft PRA are confirmed, the data are assembled and, if appropriate, maps of the key factors are produced listing any assumptions made. In stage 2, the maps that define the area of potential establishment and the area at highest risk from pest impacts are combined, documenting any assumptions and combination rules utilised. Spatial multi-criteria analysis software (MCAS-S) is recommended for this process and the process is outlined using Diabrotica virgifera virgifera as an example. When possible and appropriate, the two subsequent stages can be followed. Stage 3 can be used to identify the endangered areas (those areas where economically important loss is likely to occur) and stage 4 provides a dynamic picture of the invasion process using spread models.

Highlights: The development of decision support schemes (a) to model and map climatic suitability, selecting from the range of methods available, and (b) to map endangered areas or, at least, the areas at highest risk by combining maps of the area of potential establishment and factors related to impacts using spatial multi-criteria analysis software. Documented results: Deliverable report D3.3. WP4: Enhancing techniques for pathway analysis and systems approaches Summary:

In WP4, PRATIQUE has transformed the approach to PRAs through reviews, analyses and demonstrations that result in a consistent, systematic scheme that integrates the concepts of IPPC standard ISPM11 into a practical decision support system. PRATIQUE WP4 has undertaken reviews of pathway analysis (D4.1) and systems approaches (D4.2) in PRA. These have demonstrated that the EPPO PRA scheme could be readily adapted to explicitly incorporate these approaches with practical modification (D4.5, D4.6, D4.7).

A global review of pathway analysis procedures was undertaken with specific comparisons of pathway PRAs (widely conducted in North America and Australasia) and species PRAs (more commonly conducted in Europe). It is recommended that stronger links be made between the assessment (particularly the entry section but also other sections) and management stages of PRA since they are so closely related. An extensive review of applications of the systems approach to pest risk management, described by ISPM14 was also conducted. Systems approaches were found to be more widely applied than is commonly believed and important opportunities for greater application in Europe were identified.

Suggestions for changes to PRA formats in WP4 outputs have now been adopted in the EPPO PRA scheme, both on paper versions and in the new electronic CAPRA version. These involve greater differentiation of the sections within the scheme: rewording of the Initiation section (D4.6) to cover pathways and systems approaches more clearly, consistent scoring and links between the Entry, Establishment, and Spread sections and the Management section (D4.4) and a demonstration decision support tool for systems approaches based on multiple independent management measures (D4.7).

Neural networks and self-organising maps for PRA were reviewed (D4.3). It was found that there were some applications for these approaches in strategic risk assessments, but that the current level of data was generally inadequate for use in specific commodity, pathway or species PRAs. However, if PRAs have greater consistency of formats and are made more widely available there is considerable scope for learning from previous related analyses and for compilation in databases that could be used to search for relevant correlations.

Task 4.1 Reviews of current practice in pathway analysis and systems approaches Subtask 4.1.1 Review of pathway analysis in PRA A review of PRA processes has been conducted, with specific comparisons of pathway PRAs (widely conducted in North America and Australasia) and species PRAs (more commonly conducted in Europe). The PRA framework is established in two IPPC standards (ISPM 2 and ISPM 11), however the detail of how PRAs are carried out can vary considerably from country to country because of the flexible approach within the concept standards. Pathway PRAs are normally generated by trade requests, whereas species PRAs arise due to concern over the interceptions of known quarantine pests. As such, pathway PRAs often deal with more uncertainty. The issue of how uncertainty is covered was an issue highlighted in the review and also related to WP3 inputs. Risk assessment and risk management are generally treated as independent stages within PRAs and it was suggested in the review that these stages be more closely linked, either directly, or through iteration between the two stages. It was considered important to treat uncertainty in similar ways within the assessment and management stages and, by looking at potential management early in the risk analysis, it may be possible to limit the range of uncertain scenarios that need to be covered. Most pathway PRAs are prioritised by the sequence of trade requests that generate them, whereas species PRAs are prioritised because of concerns over the pest. In many cases with pathway PRAs, little or no trade actually occurs, despite the request and the effort to approve the trade. Priorities could be set based on reasonable expectations of trade volume, which would be in line with the perceived link between trade volume and pest risk. Highlights: Recommendations for stronger links between assessment and management stages are incorporated in the EPPO DSS for PRA and in a demonstration DSS for Systems Approach (D4.7); it is suggested that pathway PRAs should be prioritised on the basis of the expected volume of trade. Documented results: Deliverable report D4.1 Subtask 4.1.2 Review of systems approaches in PRA A review of applications of the systems approach to pest risk management has been conducted. The systems approach has been described by the IPPC in ISPM 14 in 2002 as a means of applying PRA to pest pathways in which an integrated management approach employing a range of control measures is used. Systems approaches were found to be more widely applied than is commonly believed and have been in practice for many years prior to the international standard. However, a mere combination of approaches does not constitute a genuine integrated systems approach. Several key issues were identified: (i) the equivalence of measures is even more complex when groups of risk mitigation measures are employed than when there is reliance on a single method, (ii) the independence or dependence of measures is important in estimating their combined effectiveness and the risk of failure and (iii) managing compliance can be more difficult with systems approaches where there are options for actions at different points in the production and transport chain. The systems approach has particular application in established bilateral trading regimes where there are strong opportunities for cooperative compliance management.
Highlights: Practical examples of systems approach are widespread and there are important opportunities for greater application in Europe Documented results: Deliverable report D4.2 Task 4.2 Innovative techniques for pathway analysis and systems approaches Subtask 4.2.1 Neural networks and self-organising maps for PRA Much of the work of PRATIQUE WP1 has been to collect datasets related to pest distribution and risk. The purpose of this work package task has been to derive maximum benefit from this information by "mining" it to derive linkages between apparently disparate or scattered pieces of information and to provide insight both for the PRA process itself but also for defining effective and efficient systems approaches for reducing pest introduction and spread. Efforts have been made to identify options for geographical links, such as maps, and broader linkages that could be demonstrated in neural networks. The first steps have been to identify associations between different elements and the risk of pest introduction and spread, that is pathway analysis, and then to investigate how a systems approach to mitigation might be designed. For pathway analysis one of the key dimensions is the spatial component and it is therefore essential that the data can be presented for processing with some sort of spatial reference. One substantial dataset has been examined in relation to risk of entry in Australasia, and this has been tested in reverse for risk in Europe. Most commodities or pest species do not have adequate sets for comprehensive use of neural networks. Potato pests have been examined as one of the best compilations. A more consistent approach to PRAs internationally and more widespread electronic availability of PRAs will enable more effective applications of neural network concepts in future to make use of previous information. The EPPO CAPRA electronic PRA format is an important step in making PRA outputs more useful for cross-analysis.
Documented results: Deliverable report D4.3.
Subtask 4.2.2. Enhancing consistency in the analysis of pathways and risk management options A systems map, or ontology, of the entry, establishment and spread components of the EPPO PRA scheme has been produced. On the basis of the associated clusters of nodes in these maps, suggestions were made to reorder and in some cases regroup some questions within the EPPO DSS for PRA, which has been adopted in revisions of the EPPO DSS for PRA, including the CAPRA version. Rewording has also been suggested to make questions within clusters more appropriate to the context.
Consistency in likelihood, impact and confidence scoring is important. Consistent scales and terminology in scoring schemes makes information more portable and useful between PRA areas. Though it is recognised that the probability of pest entry is specific to each location, the scale on which it is assessed can be made consistent across different pest entry locations. Scales have been suggested for incorporation in the EPPO DSS for PRA that give five defined values for magnitude (on a log scale) in several dimensions: monetary, environmental, social and health. A set of five likelihood score values is suggested, based on the climate change likelihood scores adopted by the Intergovernmental Panel on Climate Change (IPCC). A specific time horizon is suggested for event probabilities, which can then be transformed into annual likelihoods for consistent calculation of risks (likelihood and magnitude on the same time scale). These likelihood scales give wider ranges for middle scores and progressively smaller ranges for the extremes of probability. This is based on experience of risk assessors avoiding extremes. A four-point uncertainty (or confidence) scale is also suggested, based on IPCC use. Highlights: Modifications to the EPPO DSS for PRA template for greater consistency and clarity have been adopted by EPPO, inclusion of ratings and scoring systems and visual presentations of risk components in the CAPRA electronic PRA template. Documented results: Deliverable report D4.4.
Subtask 4.2.3 Specifications for systems components, performance and description of risk likelihood in PRA A report describes the conceptual development of a proposed framework and protocol for enhanced application of Systems Approach within the EPPO PRA scheme. A test version of the framework was presented in M4.8, and considered by an EPPO Panel. It uses graphic illustrations to support clearer interpretation of complex issues related to the components of the risk calculated in the PRA and the estimated impacts of proposed risk management measures, whether alone or in combination. The new framework also responds to the need for consistency in the risk management component (see D 4.4) and enhanced use of pathway analysis (see D4.1). The proposed structure of the framework allows for the addition of a probabilistic component.

Highlights: Test version of a systems approach concept related to the EPPO PRA template; this has been further developed into a demonstration DSS for Systems Approach (D4.7).
Documented results: Deliverable report D4.5. Task 4.3 PRA Modules for pathway risk analysis and systems approaches Subtask 4.3.1 Pathway risk analysis module with a protocol for the application of neural networks The systems approach review highlighted how current management questions in the EPPO PRA scheme could be reorganised in different ontologies (D4.5, D4.6) to evaluate independent control points along a supply chain. A common framework for pathway, species and systems approach analyses is presented.

Systems approach concepts developed within PRATIQUE WP4 (D4.5, D4.6, D4.7) together with Bayesian Network analyses developed in WP3 (D3.2/3.4) are being taken up in a WTO STDF project "Beyond Compliance" (STDF/PG/328) in 2011-2013, in which five national plant protection organisations (NPPOs) in SE Asia (Malaysia, Thailand, Philippines, Vietnam, and Indonesia) working with three PRATIQUE participants (Imperial, Queensland University of Technology (part of the CRCNPB) and CABI) will develop methodologies for regional applications of systems approaches.

Neural networks were found to have potential applications in strategic risk assessments, but the current level of accurate data available was generally inadequate for use in specific commodity, pathway or species PRAs, so neural network concepts were not explicitly applied in the PRA approaches.

Highlights: Modifications to the EPPO PRA template for greater consistency and clarity and more explicit initiation for species, pathways and systems approaches. Documented results: Deliverable report D4.6.
Subtask 4.3.2 Demonstration module for systems approaches in PRA The demonstration decision support tool for systems approach (D4.7) shows how information on a range of control measures and risk components can be collected from existing questions within the EPPO PRA scheme and evaluated, using performance and uncertainty ratings, against a broad set of evaluation criteria in a spreadsheet format. This demonstration decision support tool has been designed to have substantial similarity with the demonstration decision support tool for eradication (D5.3) so that both can be incorporated within the EPPO CAPRA electronic PRA format in due course.

Highlights: Demonstration DSS for Systems Approach produced in spreadsheet format, consistent with the style of the demonstration DSS for Outbreak/Eradication management (D5.3); Demonstration DSS (D4.7) used in WTO STDF "Beyond Compliance" Project (STDF/PG/328) in 2011-2013 on application of Systems Approach in SE Asia. Documented results: Deliverable report D4.7 and DSS D4.7 spreadsheet.
WP5: Developing a decision support system for the eradication and containment of pest outbreaks Summary:
A review of eradication and containment campaigns has been undertaken based on data for 173 campaigns against 94 different species (41 animals, 26 pathogens, and 27 weeds). Eighty four of these campaigns were conducted in the EU. A subset of campaigns (explicit eradications) was analysed in order to identify factors common to successful campaigns. It confirmed that eradication campaigns are more likely to succeed in closed (or bounded) systems and that readiness to act swiftly positively affects the outcome of a campaign. A survey of 41 different published or unpublished studies analysed the costs and benefits of eradication and containment programmes for invasive plants, pests and pathogens using a specially designed template containing 48 questions. A technical analysis of the database was carried out and summarised in the form of a report, including tabulations of the different types of costs and benefits, together with methods and data sources used for their quantification. A review of the different methods used to analyse the costs and benefits of eradication and containment programmes for invasive plants, pests and pathogens led to the construction of a decision support scheme for analysing the costs and benefits of phytosanitary measures during outbreaks. A protocol to identify appropriate actions to take following outbreaks of harmful organisms and when preparing contingency plans has been constructed and evaluated by experts in eradication and containment. It is now available as a module within CAPRA. Pest surveillance techniques have been investigated in three ways: (i) through a review of traditional and modern techniques for detecting new pest incursions at the species level, (ii) an evaluation of the effectiveness of the monitoring programmes for early detection and (iii) the identification of the best monitoring techniques for surveying ongoing outbreaks. Trapping of exotic insect pest species in French and Italian harbours and airports has been undertaken for three years, showing that the multi-lure approach associated with trap designs specific to three important groups of target organisms (wood boring beetles, moths and fruit flies) was the most suitable for use at ports of entry and the most successful in detecting alien species under both indoor and outdoor conditions.

WP5 provides: reviews of best practice in eradication and containment (D5.1) and detection methods (D5.4), a decision support scheme to guide eradication and containment actions during outbreaks (D5.3), guidance on cost benefit analysis techniques for outbreak management (D5.2) and an investigation of novel systems for remote data capture, surveillance and communication to detect new pests and define outbreaks (D5.5).

Task 5.1 Eradication and containment The review of eradication and containment campaigns has been based on the collection of detailed data for as many plant pest eradications worldwide as possible (data for 173 eradication campaigns against 94 different species were assembled). The review includes organisms defined as plant pests, i.e. insects and nematodes (41 species), bacteria, fungi, viruses or viroids (26 species of plant pathogens) and plants as pests (27 species). The information collected about campaigns to eradicate and contain quarantine pests in the EU and worldwide has been compared and a subset of campaigns (explicit eradications) has been analysed in order to identify factors common to successful campaigns. Contributors answered 70 questions for each campaign and the outcomes were classified in four ways (in summary): 1) the campaign failed; 2) the organism could not be contained; 3) the campaign was successful in the sense that the organism could be contained within a certain area; and 4) the eradication was completely successful. The following hypotheses were addressed: 1) Pest populations with a limited spatial occurrence are more easily eradicated; 2) Island and indoor (closed system) outbreaks are more easily eradicated; 3) Early detection and a quick response will increase the chances of success; 4) Campaigns against well-known organisms are more likely to succeed; and 5) Good cooperation between authorities, the involvement of stakeholders and the availability of a PRA or even a contingency plan could be equally important for the success of the campaign. Linear mixed effect models (LMMs) were applied to test these factors for eradication success, and a more extended analysis was performed using classification and regression trees (CART). Highlights: The statistical analysis of the eradication database confirmed some, but not all, of the general assumptions (i.e. hypotheses) about relevant factors for eradication success. For example, it confirmed: 1) that eradication campaigns are more likely to succeed in closed (or bounded) systems and 2) that readiness to act swiftly and positively affects the outcome of a campaign. The CART analysis included more factors than the LMMs analysis and highlighted the importance of the size of the infested area, including whether it is easily accessible. Accessibility (related to both geographical remoteness or powers of entry) is therefore a key factor in determining successful outcomes. Taxonomy is also relevant as bacteria and plants are easier to eradicate than animals, fungi, and viruses. All of these factors should be taken into account when designing contingency plans.
Documented results: Deliverable 5.1
Task 5.2 Cost benefit analysis A survey was carried out of the different methods used to analyse the costs and benefits of eradication and containment programmes for invasive plants, pests and pathogens. A total of 41 different published or unpublished studies relating to the economics or costs of a diverse range of pest control campaigns were processed using a specially designed template containing 48 questions. The studies were categorized according to their objectives, which were mainly (i) economic impact assessments of a new pest, or a change of policy towards a pest, (ii) cost benefit analyses (CBA) proper, with one or more different scenarios or measures and (iii) studies aimed at determining the optimal control strategy. The timings (ex post or ex ante) and objectives (e.g. eradication, containment or suppression) of the studies varied considerably and in some cases scenarios were not explicitly defined. A technical analysis of the database was carried out and summarised in the form of a report, including tabulations of the different types of costs and benefits, together with methods and data sources used for their quantification. Conclusions and lessons learnt from this analysis of economic methods were also captured. The construction of a decision support system for application in the analysis of the costs and benefits of phytosanitary measures has been completed. The results of the economic literature review (and case studies) provide input and guidance to this task in addition to the work on impact assessment. A protocol on CBA was produced consisting of the following 9 steps:
(i) Define the problem in terms of: the purpose, perspective, scope, scale and time
(ii) Define the baseline scenario
(iii) Select the control scenarios
(iv) Predict the effects of each control scenario
(v) Estimate the costs and benefits of each control scenario
(vi) Consider the time aspect of costs and benefits
(vii) Calculate decision criteria
(viii) Perform a sensitivity analysis
(ix) Report on the CBA

Highlights: In the cost benefit analysis studies, the scenarios were usually eradication, or containment (including delay of spread), or both, and the objective was to determine which particular strategy was the most cost-effective. The benefits were usually taken to be the losses avoided, the nature of which varied according to the particular pest. Losses caused by or incurred as a result of the presence of a pest were distinguished from the costs of (additional) controls. Types of costs and benefits were structured and interpreted differently. For example, some studies considered the effects on tourism and health as economic losses, whilst others considered loss of income (in a population) as a social loss. All studies, however, considered the direct impacts of the pest, such as reduced yield or quality, higher production costs and effects on trade or exports, as economic losses. Environmental or social losses were less frequently captured, although they were often recognized when relevant, but were usually found to be too difficult to quantify or generally considered to be of minor importance. The various outputs from this study (the literature review, case studies, lessons learnt and CBA protocol) form the basis of two publications in progress. Task 5.3 Decision support scheme A decision support scheme (DSS) to identify appropriate actions to take following outbreaks and when preparing contingency plans has been constructed and evaluated by experts in eradication and containment. The DSS was conceptualized and developed in both Microsoft Word and Excel formats. This was then developed as an electronic scheme within the CAPRA application. There are three formats of the scheme:

* CAPRA - the main intended format for the scheme (http://capra.eppo.org/index.php);
* Microsoft Word - this includes guidance and effectively forms a stand-alone manual. However, it has the advantage that it can be completed electronically or even printed and filled in by hand if necessary;
* Microsoft Excel - this version was employed to develop the components of the scheme, although it has now been superseded by the CAPRA format. It provides a useful, simple version of the DSS with drop-down menus. If the Excel version is employed, the Word version should be used as a manual. The Excel version has been added as an annex to this deliverable in case access to CAPRA is unavailable.

Stakeholders involved in the development of this product suggested that the Word and Excel versions should be made available as alternatives for utilising the DSS, if these are preferred or if there are technical reasons that preclude the use of CAPRA. Highlights: The major components of DSS were identified as the factors associated with the biological traits of the organism, with the assessment of costs and with the operational constraints:
* Several biological traits are considered to be critical and thus need to be specified for the area where the outbreak is detected. Among these are the information and any uncertainties about the life cycle, the mobility of the organism, the preferred hosts and the population density. There are also uncertainties regarding the reliability of testing, detection and trapping methods. Designing an effective trapping strategy is essential for successful eradication, correctly demarcating the regulated areas and establishing an optimal grid density.
* The critical factors for successful eradication need to be assessed in terms of financial resources and cost/benefits. The following factors were discussed in particular: the advantages and disadvantages of financial compensation, whether a quick and effective response at an early stage is hampered by limited financial resources, the difficulties that often occur when trying to assess benefits (e.g. trees in urban areas) and assessing the point when the objective of eradication has to be changed either to containment or suppression.
* Contingency plans may help to address operational problems before they arise. They save time by avoiding lengthy deliberations and help to prepare a quick and effective response in an outbreak situation. In situations when it is not possible to control a particular pest, it is even more critical to be able to detect and eradicate it at an early stage. The following operational factors are recognized as being important for successful eradication: conflicts of interest, tracing forward, the need for delegation of authority to companies which implement eradication measures, access to private areas, logistical planning of machinery and manpower and the availability of effective plant protection products

General aspects which are emphasized as helping the implementation of an eradication campaign are: (a) communication with stakeholders, the public, scientists, interest groups and NGOs and (b) willingness of growers/farmers to implement official measures.
Task 5.4 Pest surveillance techniques The objectives of this task were: 1) to review both traditional and modern techniques for detecting new pest incursions at the species level, 2) to evaluate the effectiveness of the monitoring programmes for early detection through commercial trade pathways and 3) to identify the best monitoring techniques for surveying ongoing outbreaks. Their reliability, efficiency, cost-effectiveness and relevance to pest outbreaks in the EU have been identified, wherever possible. The results of the review have been used to develop recommendations for early pest detection programmes and outbreak surveys in the EU, identifying potential constraints.
Highlights: Information on inspection and trapping systems for alien pests worldwide has been collected, starting with the species included in the EU Plant Health Directive 2000/29 and EPPO lists of pests recommended for regulation as quarantine pests. For each species, we compiled an extended questionnaire about the pest (e.g. life-history traits, spread capabilities, feeding behaviour and habitat), trap characteristics (e.g. attractant, dispenser), location-specific and event-specific variables. Questionnaires were tailored to insects and pathogens. The information retrieved from the questionnaires was compiled into a database in which we have collected all the data about the detection of alien insect pests and pathogens. The mostly commonly used methods to detect plant pests and diseases are provided for groups of organisms with high economic relevance, such as Coleoptera (bark beetles, flathead borers, leaf beetles, longhorn beetles and weevils), Diptera (cone flies, seed flies and fruit flies) , Homoptera (aphids, leafhoppers, psyllids, and whiteflies), Lepidoptera (moths), Thysanoptera (thrips), Bacteria (potato brown rot), Fungi (pitch canker disease and brown rot disease). Future perspectives in detection methods are discussed, with particular reference to the considerable increase in the volume, commodity type and origin of trade in plant material from third countries, the introduction of new crops, the continuing expansion of the EU with the addition of new countries to its borders and the impact of climate change affecting the boundaries to the distribution of pests and their vectors. A list of detection methods and semiochemicals available for most important pests included in Council Directive 2000/29 and annexes and in EPPO lists is provided, as well as a critical analysis of the surveillance programmes deployed for the detection of fruit flies worldwide.

The trapping of exotic insect pests included studies on trap design, lure combination, generic lures, trap density and automatic detection for a few of the most important target insects such as wood beetles (Coleoptera), fruit flies (Diptera), and moths (Lepidoptera) associated with green plant tissues. Traps were deployed at ports of entry but also in natural conditions (e.g. forests) to validate the techniques with native species taxonomically related to the target exotics. Thirty-one sampling sites were used in France and Italy and several hundred species including four quarantine species and thousands of individuals were trapped.

We showed that the multi-lure approach associated with trap designs specific to three important groups of target organisms (wood boring beetles, moths and fruit flies) was the most suitable for use at ports of entry and the most successful in detecting alien species under both indoor and outdoor conditions. The same methods can be used to delimit the outbreak areas and to monitor the spread of the organisms. Overall the multi-lure traps were found to be statistically more efficient than single lure traps. In addition, because one multi-lure trap is used instead of many single-lures traps, the cost of the materials and of the man-power is strongly reduced - by approximately as many times as the number of lures minus one, i.e. if previously 5 traps were needed for 5 lures, a fourfold saving would be made by placing them all in a multi-lure trap. We have also contributed to the development of an automatic trap, by defining the most relevant traits of the target insects to be recognized by the system, although a prototype did not become available during the project. The work carried out on this topic has led to the development of a generic method of automatic detection that is now under development in the EU FP7 project Q-DETECT. WP6: Project validation and dissemination with the development of a web-based PRA scheme
Summary:
A provisional list of pests that could be used for testing and validating the PRA scheme was assembled. The list included 2 bacteria, 6 fungi and fungi-like organisms, 14 insects, 2 nematodes, 4 plants and 2 viruses or viroids. In response to an urgent request from the Commission, the pine wood nematode, Bursaphelenchus xylophilus, was also added to the list. In order to test the deliverables, species have been chosen from this list but other pests have also been included in particular those for which an EPPO Expert Working Group had been organized. The testing has been undertaken by EPPO Panels, EPPO Expert Working Groups and also by individuals from project partner institutes and pest risk analysts from other countries who participated in the Workshop for testing PRATIQUE deliverables (Hammamet, Tunisia, 2010-11). In addition, a meeting with the Standing Committee on Plant Health (with a request to complete a written questionnaire) was organized on 1st October 2010 to collect feed-back from risk managers on their expectations on key issues in PRA (for details see D6.3).

A computer program called CAPRA (Computer Assisted PRA) has been developed. The computer program is web-downloadable and standalone allowing the user to work on a PRA even without an Internet connection. All sections of the EPPO DSS for PRA are included, hyperlinks to the definitions of terms of the Glossary of phytosanitary terms (ISPM No.5) are available as well as a link to a dataset explorer that includes the datasets identified by WP1.
The EPPO DSS for PRA, which is used throughout Europe, directly follows ISPM11 and provides pest risk analysts with a comprehensive series of questions that explore all the factors that must be considered in a PRA. This scheme formed the basis for PRATIQUE's investigations and for the dissemination of its results. The validated outputs of PRATIQUE: datasets (WP1) and the guidance and enhancements to the PRA scheme (WPs 2-4) have been directly incorporated into the EPPO PRA scheme or provided as stand-alone modules and simple links. These are now available through the EPPO DSS template of CAPRA (see below). The modules which have not been fully evaluated by the EPPO PRA Development Panel are included in a PRATIQUE version of CAPRA. The Decision support scheme (DSS) for generating contingency plans and prioritizing action during outbreaks developed in WP5 has also been incorporated in CAPRA.

The new computerised PRA scheme (CAPRA) provides pest risk analysts in the EU and worldwide with a validated, user-friendly computerised PRA scheme with detailed guidance, modules containing procedures to be adopted in difficult sections of the scheme, a manual and examples of best practice. Pest risk managers, phytosanitary regulators and policy makers will benefit because PRAs will now more clearly highlight the key factors to take into account when developing phytosanitary measures, ensuring that the choices made are consistently based on sound science, reflect uncertainties and represent the most cost-effective options while following IPPC principles of minimal impact, transparency and equivalence. In addition, stakeholders will be able to understand more readily the justification for any measures that are proposed.

Task 6.1 Validation of the outputs from work packages 1 to 5 Production of a list of pests to develop and test deliverables A provisional list of pests used for testing the scheme in WPs 2-5 and validating the scheme in WP6 was assembled during the first six months of the project and covers a range of pests according to the following criteria: (i) pest taxon (insects, nematodes, viruses, fungi, bacteria, invasive alien plants etc), (ii) type of entry pathway (man-made or natural; intentional or unintentional) and (iii) receptor (crop, amenity, habitat, ecosystem, etc). The list includes 2 bacteria, 6 fungi and fungi-like organisms, 14 insects, 2 nematodes, 4 plants and 2 viruses or viroids. Additional species were also used during the testing (e.g. those for which an EPPO Expert Working Group has been organized).

Validation Confirmation of validation procedures It was agreed that validation would be undertaken by different groups (e.g. EPPO Panels and Expert Working Groups) but also by different partner institutes. Feedback procedures were developed to ensure that they are as specific as possible so that the results from the testing process can be reported back directly to the teams that have prepared the deliverable so that it can be adjusted as appropriate. For each deliverable, a validation form was created in collaboration with the lead author of the deliverable. This included what is expected from the validation process (comments on ease of use, whether the deliverable is user friendly, whether it delivers what is expected etc). Whenever possible, lead authors were also invited to attend the relevant parts of the various Panel meetings and Workshops where they were able to present and discuss their results. One Workshop was organised in 2010 where the different deliverables prepared up to that date were presented and tested. Two other workshops were organized during the project where the computer programme CAPRA was presented and tested (Limassol, Cyprus 2008; Hammamet, Tunisia, 2010).
Retrospective validation
The DoW states that: "The procedures developed in WPs 2-5 will be tested with the information available for pests over 20 years ago to determine whether the actual outcomes (in terms of pest entry, establishment, impacts and management) can be successfully predicted." This issue was discussed in depth and it was concluded that there is likely to be greater difficulty in obtaining historic data, especially if the datasets required did not exist in the past and/or the information available requires analysis by experts, before it can be used to parameterise the PRA tools in the new scheme. Consequently a true retrospective validation of tools that have been parameterised using currently available information is not possible since they would have to be reparameterised using historical information. In addition, comments received from statisticians indicate that to produce a statistically significant retrospective validation, at least 50 to 200 PRAs would have to be carried out. Clearly such an unrealistically large number of PRAs could not be undertaken by PRATIQUE to provide a statistically significant validation and, given the doubts regarding the extent to which the parameterization of tools using "old" data was feasible, the focus was put on testing the deliverable based on current data.
Task 6.2 Construction of a web-based PRA scheme
CAPRA was developed by the EPPO Secretariat. The first beta version of this program was produced in August 2008. It only included the categorization and the risk assessment part of the EPPO Decision-Support Scheme (EPPO DSS). This first beta version was tested during the Expert Working Group for performing PRA (EWG) on Xanthomonas axonopodis pv. allii. After this meeting, a second version of the program was prepared and sent to the members of the EPPO Panel on PRA development together with a simple evaluation form. The program was also tested by 70 persons from 18 different countries with 4 case studies during the EPPO training workshop on PRA in Limassol, Cyprus (2008-11-11/14). CAPRA was then used to prepare PRA documents both before and during several EWGs. This provided several opportunities to test the functionalities of the computer programme and its user friendliness. It should be noted that as these groups had to prepare PRAs following the current version of the DSS and ultimately make recommendations regarding the risks and measures proposed, it was not possible to use these EWGs to test the rewording and restructuring of the questions (this was mainly done by the Panels on PRA Development, the Panel on Phytosanitary Measures and the Workshop organized in November 2010). However, one tool developed by WP3, the visualizer, was demonstrated during the EWGs organized in July and September 2010 and was welcomed.

Testing of the computer program both by those preparing PRAs and during other meetings enabled continuous improvements to be made. It also allowed the testing of the reports generation facility (as reports of PRAs have to be generated for risk managers). The improvements made are presented in D6.3.

More on the program itself The computer program is a web downloadable standalone version that has to be installed on individual computers. Compared to a completely web-based version, it allows the user to work on a PRA even when he/she has no Internet connection. This was the main reason for making this choice. When connected to the Internet, the programme will automatically search to see if an updated version of CAPRA is available and propose that the new version is downloaded and installed.
CAPRA is written in Delphi (Pascal object) and CAPRA-net is written in an open source programming language (PHP) and a powerful open source database (PostgreSQL). In the current version of the program the Microsoft Access format has been abandoned as it is closed and dependent on Microsoft.

Main functions of the programme (see section 1.5 for screenshots of CAPRA)

All sections of the EPPO DSS for PRA are included and hyperlinks to the definitions of terms of the Glossary of phytosanitary terms (ISPM No.5) are available The different sections of the EPPO DSS can be performed independently and risk assessors are no longer restricted to a fixed sequence. Screens include all questions from the different sub-sections of the EPPO DSS.
Details about running the programme are given in Annex 2 of Deliverable 6.4.
A "Team mode" option (allowing more than one expert to contribute to a single PRA) was available in the first CAPRA version and will be available again in a few months; the EPPO Secretariat is working on this feature. The principles are presented below. A PRA report can be generated in various formats (Word, XLS or XML). In the case of a software bug, a report is automatically generated and sent to the EPPO Secretariat. With the dataset explorer, the datasets collected by WP1 are integrated into CAPRA. A user's manual has been prepared as well as a technical document describing the program.
The DSS for generating contingency plans and prioritizing action during outbreaks prepared by WP5 has also been integrated into CAPRA following the same principles and incorporating similar tools (e.g. the visualiser).
Task 6.3 Consolidation and dissemination of project outputs with a manual and examples of best practice
As explained earlier, the EPPO DSS for PRA formed the basis for PRATIQUE's investigations. The validated outputs of PRATIQUE, in particular the enhancements to the PRA scheme provided by WPs 2-4, have been directly incorporated into the EPPO PRA scheme. The modifications have been approved by the Working Party on Phytosanitary Regulations in June 2011 and, consequently, the EPPO DSS for PRA revised by PRATIQUE is now recommended for PRA in the EPPO region. Detailed guidance elaborated during the project is available within CAPRA. Examples of PRA outputs that illustrate how the scheme can be utilised are available in the CAPRA network and further examples will be added in the future. A deliverable (D6.5) has been prepared; it includes the manual (i.e. the EPPO DSS for PRA version 2011), an explanation of how the different deliverables have been integrated into the CAPRA network as well as some examples.

Potential Impact:
1.4.1 The Potential Impact of PRATIQUE
Impacts on Pest Risk Analysis in Europe
The revised EPPO decision support scheme (DSS) for PRA developed by PRATIQUE provides (a) considerable enhancements to its functionality and user-friendliness, (b) links to datasets relevant for all EU member states and for the assessment of all aspects of PRA and (c) modules to assist with the modelling and mapping of risk. The new scheme should therefore:
* enhance the production of PRAs that are: consistently accurate assessments of risks and uncertainties, compiled as efficiently as possible and fit for purpose,
* ensure that PRAs more accurately reflect the risk posed to all the member states of the EU,
* make the PRA production process easier, more logical and user-friendly both for inexperienced and for expert PRA practitioners, minimising the inappropriate use of models and maps,
* encourage a greater number of PRAs to be produced,
* effectively reduce the time needed for PRA training,
* provide clearer PRAs so that the risks and uncertainties can be more readily understood by regulators and stakeholders in a format that is straightforward to read,
* help to harmonise methods for PRA production in Europe,
* provide a springboard for further research to enhance PRA procedures. For example, large numbers of consistently produced PRAs can be used to test the scheme, identify improvements and enable other tools, e.g. neural networks, to be applied to identify how pest risk analysts compare and contrast different elements of risk. In addition, the new tools, e.g. for modelling spread and mapping endangered areas, provide a basis for further investigation with different species in different habitats,
* be sustainable because PRA is a fundamental ongoing activity for EPPO, the new DSS for PRA developed by PRATIQUE has already been adopted by EPPO (at the meeting of the Working Party on Phytosanitary Regulations in June 2011) and, even after the project has ended, it will:
o host all the project deliverables, maintaining their availability as appropriate through the computerised DSS for PRA (CAPRA),
o ensure that further appropriate developments are undertaken so that the DSS for PRA remains fit for purpose.
Impacts on Pest Surveillance, Contingency Planning and the Eradication and Containment of Outbreaks in Europe
In addition to the revised DSS for PRA and the additional modelling and mapping modules that can also be used in outbreak situations, e.g. to model spread and defend areas at greatest risk, a variety of tools and datasets has been developed to enhance the capacity of risk managers to:
* learn from past eradication and containment campaigns by providing a database of 173 campaigns,
* design and operate monitoring and surveillance strategies using novel techniques, such as multilure traps,
* take rapid emergency action, selecting the most appropriate management options, efficiently targeting resources following outbreaks by applying a generic DSS for selecting management options based on the likelihood and magnitude of the risk, the current outbreak situation and various attributes of the measures available, e.g. efficacy, cost and acceptability,
* undertake a cost benefit analysis (CBA) of different management options by providing a CBA protocol,
* generate effective contingency plans,
* review existing eradication and containment campaigns, deciding when they need to be modified and are no longer appropriate.
Impacts on Plant Health Policy in Europe

Risk managers, phytosanitary regulators and policy makers will benefit from many of PRATIQUE's outputs, primarily because PRAs:
* will be much easier to read,
* will highlight more clearly the key factors to take into account when developing phytosanitary measures, ensuring that the choices made are based on sound science, reflect uncertainties and represent the most cost-effective options while following IPPC principles of minimal impact, transparency and equivalence.

Since the example pests selected to illustrate the PRA techniques enhanced by PRATIQUE include species of major ongoing concern in the EU (e.g. pine wood nematode, potato stem tuber viroid and the western corn rootworm, Diabrotica virgifera virgifera), these examples can also be used to contribute to plant health policy against these harmful organisms.

PRATIQUE's outputs should:
* ensure that limited EU member state plant health resources are targeted to the most important pests,
* increase the number of high quality PRAs that can be used directly to strengthen the regulations in the EC Plant Health Directive,
* reinforce the EC's ability to provide the technical justification needed to support challenges to EC regulations under WTO and IPPC procedures,
* minimise the trade barriers that phytosanitary legislation can create, allowing trade to flow as freely as possible and markets to flourish,
* significantly reduce the risk of challenges from third countries,
* assist the DG SANCO of the European Commission in implementing the changes currently being proposed to the Plant Health regime. The Food Chain Evaluation Consortium, that was contracted to evaluate the current regime and help plan the revision, requested and was given early sight of PRATIQUE's deliverables,
* assist the Commission's DG ENVIRONMENT in developing a directive on invasive alien species and procedures for assessing risk. Several of PRATIQUE's outputs, particularly the computerised DSS for PRA (CAPRA) and techniques for visualising and summarising risks and uncertainties have already, with little change, been adopted and successfully converted for use in the assessment of the risk posed by non-native species in Great Britain. These techniques are now being evaluated for use as a basis for invasive alien species risk assessments in the EU under the new directive that will be published in 2012,
* highlight the dangers posed to EU plant health by pests in Eastern Asia (based on surveys of native European trees growing in eastern Russia and planted in China).
Impacts on producers, consumers and biodiversity in Europe

Assuming they are taken up by the plant health services in the EU, PRATIQUE's outputs should:
* provide significantly increased protection to EU crops, plant products, wild flora, habitats and ecosystems,
* maintain the competitiveness of key agricultural, horticultural and forestry sectors in EU and export markets,
* enhance EU biosecurity,
* enable stakeholders to understand more readily the justification for the measures being proposed,
* give consumers increased quality in the imported plants and plant products they purchase.
Global Impacts
Although PRATIQUE's work has been targeted primarily at enhancing PRA techniques in Europe, the revised EPPO DSS for PRA and the additional tools that have been developed are not tied in any way to European conditions and can be used to conduct PRAs and manage outbreaks anywhere in the world. The key limiting factor, as always, lies in the data available for conducting PRAs. Although PRATIQUE has made a special effort to collect all the relevant datasets for EU member states, very large numbers of the datasets collected are at a global scale or of global relevance.

Following presentations of PRATIQUE at the Enhancement of North American Plant Health Risk Assessment workshop hosted by Canadian Food Inspection Agency in 2010, strong views were expressed that PRATIQUE provides an excellent model for revising PRA procedures and conducting PRA-related research that should be emulated in North America.
There are several other reasons why PRATIQUE's work has global impact, equally appropriate for both developed and developing countries:
* The EPPO DSS for PRA revised by PRATIQUE follows the IPPC's international standards for phytosanitary measures (ISPMs) on PRA that are recognised by the World Trade Organisation so any PRA created using this DSS cannot be challenged for lack of conformity with international standards,
* The many years of prior development and the extensive testing of the revisions to the EPPO DSS for PRA by those from many countries with a wide variety of expertise in PRA provides reassurance that the wording is as clear and unambiguous as possible, understandable equally to those conducting their first or their fiftieth PRA and to those for whom English is or is not their mother tongue,
* Although some of the modules require sophisticated modelling expertise, they begin with caveats that ensure they should only be attempted in relevant situations and by those with the appropriate resources,
* All of PRATIQUE's deliverables are freely downloadable and so are accessible for use by anyone with a computer and internet connection.
It is particularly important to recognise the assistance to PRATIQUE given by countries outside the EU: the project partners from Australia and New Zealand, the sub-contractors from China and Russia and the project observers from Canada, USA and Norway. Both the Canadians and Americans, not only gave us the benefit of their thinking on the best way to undertake PRA but also acknowledged their gratitude to PRATIQUE for helping provide ideas and an excellent forum for evaluating the new techniques they are planning to include in their revised PRA schemes.
In addition to contributing to the development of PRA schemes in other countries, the PRATIQUE outputs can also play a role in the development and application of international standards:
* A number of difficulties with ISPM11 were identified during PRATIQUE, e.g. the position and role of spread in the PRA process, and these will be available to IPPC Secretariat Standards Committee when this standard is revised.
* PRATIQUE's development of systems approach concepts and Bayesian Network analyses are being taken up in a WTO Standards and Trade Development Facility (STDF) project "Beyond Compliance" (STDF/PG/328) in 2011-2013, in which five NPPOs in SE Asia (Malaysia, Thailand, Philippines, Vietnam, and Indonesia) working with three PRATIQUE participants (Imperial, Queensland University of Technology (part of the CRCNPB) and CABI) will develop methodologies for regional applications of systems approaches.
Research Impacts in the EU
PRATIQUE has already provided information, PRA procedures and collaborative research input to a number of EU projects and initiatives, principally:
* A survey of PRA research priorities made by canvassing attendees at the PRATIQUE Final Conference in May 2011.
* The EFSA Plant Health Panel Working Group on Environmental Impacts: providing early drafts of the PRATIQUE environmental risk assessment scheme
* The EFSA funded research project PRASSIS: reciprocal sharing of datasets
* The EFSA funded research project Prima phacie: primarily collaboration on methods for summarising risk and uncertainty
* The FP7 funded project SHARCO: provision of PRA methodology
* The FP7 funded projects ISEFOR and REPHRAME: discussions on priority research for analysing the risk, enhancing surveillance strategies and designing optimal eradication/containment strategies against pine wood nematode
* The FP7 funded project Q-Detect: provision of information on surveillance methodologies
* The FP6 funded project SEAMLESS: exploration and utilisation of the databases for quantitative economic impact modelling and endangered area mapping
* The ERA-NET funded project: EUPHRESCO: provision of PRA methodology to projects funded under its umbrella
Global Research Impacts
* PRATIQUE has contributed significantly to international activities aimed at defining and promoting best practice and exploring new techniques in pest risk modelling and mapping primarily through presentations and discussions at the annual International Pest Risk Modelling and Mapping Workshops. A key objective is to prepare a best practice manual for which PRATIQUE's deliverable on mapping endangered areas, including a DSS for climatic mapping, already provides a recognised platform on which to build.
* PRATIQUE's research on Asian pests of native European trees in eastern Asia by surveying arboreta and planting trees has contributed to discussions on a global network of sentinel plantings with surveys of arboreta and botanic gardens

1.4.2 Principal Dissemination Activities

Dissemination of project results

The results of the project are freely accessible via the Internet from the PRATIQUE website and the CAPRA Network website provided by EPPO. As appropriate, the project deliverables are integrated within or linked to CAPRA, the computerised DSS for PRA developed by PRATIQUE.

A project publicity leaflet was produced in four languages (English, French, German and Italian) at the start of the project outlining its aims and objectives and a paper was published in the scientific literature (the EPPO Bulletin) setting out PRATIQUE's detailed plans and objectives (Baker RHA, Battisti A, Bremmer J, Kenis M, Mumford J, Petter F, Schrader G, Bacher S, De Barro P, Hulme PE, Karadjova O, Lansink AO, Pruvost O, Pysek P, Roques A, Baranchikov Y & Sun JH (2009) PRATIQUE: a research project to enhance pest risk analysis techniques in the European Union. EPPO/OEPP Bulletin 39, 87-93).

The project website has enabled wide dissemination of the project and its findings to relevant agencies, researchers, stakeholders and the general public. The website provides access to all the project's technical deliverables.

The final PRATIQUE conference was held on May 24th-25th 2011. It was attended by 67 participants from 17 countries. In addition to a full day of presentations on the project's findings, opinions on the research priorities for PRA were canvassed through a questionnaire. Half a day was devoted to a discussion on these research priorities and the future of PRA in Europe with presentations from the European Commission, EFSA, National EU Plant Protection Organisations, EPPO, the International Plant Protection Convention Secretariat and representatives of the Canadian and American plant heath services.

Several papers have already been published in peer-reviewed international scientific journals and numerous additional papers are in preparation.

A very large number of presentations on the project and its results have been made, notably to:
* representatives of National EU Plant Protection Organisations
* the EC Plant Health Standing Committee, 1st October 2010
* the European Food Safety Authority Plant Health Panel
* the EPPO Working Party on Phytosanitary Regulations and Panels on PRA Development, Phytosanitary Measures and Invasive Alien Species
* EPPO Training Workshop on PRA in Cyprus, 2008
* The Australian plant health services and risk analysis research community, 2009
* EPPO Training Workshop on PRA for French speaking experts, Tunisia, 2009
* the Technical Consultation among Regional Organizations in Plant Protection, 2010
* The PRATIQUE EPPO Workshop, Tunisia, 2010

Presentations have been made at numerous relevant scientific meetings with publications in conference proceedings. The principal international meetings are noted here (a more detailed list is provided in section 2A:
* EFSA Colloquium on Pest Risk Assessment, 2007
* International Plant Protection Conference, 2007
* European Crop Biosecurity Workshop, 2007
* International Congress of Plant Pathology, 2008
* International Pest Risk Modelling and Mapping Workshops, 2007-11
* International Congress of Entomology, 2008
* Canadian Forest Pest Management Forum Gatineau, 2008
* European Conference on Biological Invasions - Neobiota, 2008 & 2010
* EPPO Workshop on Eradication, Containment and Contingency planning, 2009
* Ecology and Management of Alien Plant Invasions, 2009
* Ecological Society of America, 2009
* IUFRO International Forest Biosecurity Conference, 2009
* International Congress on Biological Invasions, 2009
* Global Biosecurity Conference, 2010
* IUFRO World Conference, 2010
* International Mycological Congress, 2010
* Society for Risk Analysis Annual Meeting, 2010
* Enhancement of North American Plant Health Risk Assessment Workshop, 2010
* CFIA Symposium Getting Ready for Tomorrow: Integrating Foresight and Emerging Trends into Timely Science Advice for Plant Risk Assessments, 2010
* International Symposium of Integrative Zoology, 2010
* EPPO Workshop on Eradication, Containment and Contingency Planning, 2010
* European Association of Agricultural Economists, 2011
* U.S. Department of Agriculture Interagency Forum on Invasive Species, 2011
* European Science Foundation Exploratory Workshop on "Risk assessment analysis: methods and applications for evaluating biological invasions", 2011
* EFSA Colloquium on Emerging Risks, 2011

1.4.3 Exploitation of results

The revised and computerised DSS for PRA, the datasets, , the tools to summarise and communicate risk and uncertainty, the modules for pest risk modelling and mapping and the information and tools provided for taking action following pest outbreaks have been specifically designed for immediate exploitation by pest risk analysts, pest risk managers and policy makers in national plant protection organisations and other bodies, e.g. EFSA and regional plant protection organisations. The fitness for purpose of the project deliverables has been assured by: (a) the presence of experienced PRA practitioners in PRATIQUE, (b) presentation, discussion and testing of project deliverables in panels, workshops and expert working groups and (c) continuous interaction with pest risk analysts, pest risk managers and policy makers in partner institutes, on the EC Plant Health Standing Committee and in plant health services of the EU and worldwide.

EPPO has already adopted the DSS for PRA developed by PRATIQUE (at the meeting of the Working Party on Phytosanitary Regulations in June 2011) and so will apply these methods when conducting the approximately 5 PRAs it conducts per year. EFSA's remit, that excludes assessments of monetary losses and cost benefit analyses means that it cannot utilise the new scheme in its entirety. However, the DSS for PRA revised by PRATIQUE is currently being evaluated by the EFSA funded project Prima phacie to help guide the EFSA Plant Health Panel on future developments of the EFSA guidance on a harmonised framework for pest risk assessment and the identification and evaluation of pest risk management options. Member states are increasingly adopting shortened PRA formats for operational reasons, turning only to a detailed analysis when the risk is so significant that a full PRA is warranted, particularly to influence regulation at the EC Plant Health Standing Committee. In the EU, these detailed PRAs, conducted by member states acting individually or in partnership and by research projects, have been undertaken using previous versions of the EPPO PRA scheme. It is greatly to be hoped that future PRAs will now follow the EPPO PRA scheme revised by PRATIQUE.

The conduct and development of PRA together with the training of pest risk analysts is a core activity of Partner 4 (EPPO) which it fulfils as part of its mandate as a Regional Plant Protection Organisation for the International Plant Protection Convention, administered by FAO (the Food and Agriculture Organisation). As such, EPPO is committed to taking all possible efforts to maintain and enhance the functionality of the DSS for PRA after the project including repairing broken web-links and maintaining contacts and coordination with data owners outside and inside the project.

Related information

Contact

Christina Steveni, (PLH Group Co-ordinator)
Tel.: +44 1 904462226
Fax: +44 1 904462250
E-mail
Record Number: 196163 / Last updated on: 2017-03-24
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top