CRC for Plant Biosecurity - modelling

Modelling and Simulation with Mathematical and Computational Sciences

VANMEURSA — Mon, 10 Jan 2011 22:56:43 +0000

Modelling and Simulation with Mathematical and Computational Sciences (MODSIM) is uniquely a multi-disciplinary gathering of modellers – from a range of modelling disciplines (statistics, mathematics, computer science, economics, etc) and a diverse range of applications. This provided Dr Low Choy a useful venue for soliciting feedback on her recent research into encoding expert judgements, since expert elicitation is a highly multi-disciplinary and a small but emerging field. Dr Low Choy’s involvement in MODSIM provided useful experience: as session organiser for two strands—on Bayesian statistics and on expert elicitation (including coordinating refereeing of conference papers); presenter of an oral presentation; co-author of two other presentations; and main presenter of a workshop on Expert Elicitation by Design (based on a recent paper, Low-Choy et al, 2009, Ecology).

Dr Low Choy presented a (12 minute) paper on Expert elicitation and its interface with technology: a review with a view to designing Elicitator. Interesting feedback on the talk mostly came from novices to the elicitation field interested in trialling the software. Because of this feedback, Dr Low Choy realised that the workshop planned for the final day, which was originally pitched to elicitation practitioners needed to be modified to target beginners. The workshop provided interesting insights: that researcher interest is growing across many fields (from vulcanology to materials science and planning emergency response), and that software and courses for beginners are lacking.

In 2009, the conference streams were deliberately constructed to encourage multi-disciplinary cross-fertilisation, with no adjacent talks sharing the same modelling discipline. This meant Dr Low Choy was exposed to interesting new approaches. For instance: game theory was used to assess whether small changes in decisions really make a difference in the end-game; to help recover a re-introduced native bird population that was not progressing past fledgling stage, a risk assessment utilised model-based estimates of the probability of survival based on different scientific hypotheses, weighted by expert support for said hypotheses. This has influenced Dr Low Choy’s thinking on decision-theory, a framework for evaluating performance of statistical designs (such as those used for pest surveillance), and on model selection within the Bayesian framework.

Surveillance systems you can count on

VANMEURSA — Mon, 20 Dec 2010 02:32:00 +0000

“You cannot prove an area is free of a particular pest”, says Queensland researcher Mark Stanaway, “you can only estimate the probability. The more information you bring into your surveillance systems, the more accurate your estimate.”

Mr Stanaway and Western Australian researcher Nichole Hammond are both working on CRC for National Plant Biosecurity PhD projects designed to analyse the effectiveness of current surveillance systems and identify areas for improvement.

Mr Stanaway’s project is based on surveillance data and modelling of the spread of spiralling whitefly, a sap-sucking bug from South America first identified in Cairns in 1998. Ms Hammond has focused on an assessment of surveillance systems in WA for the fungal disease Karnal bunt, which affects cereals.

Both are using hierarchical Bayesian models to estimate the uncertainty attached to each level of decision-making in the surveillance process and ranking the importance of each decision, to arrive at a ‘confidence rating’ for the effectiveness of their respective surveillance systems.

For his project, Mr Stanaway has collated more than 11 years of surveillance data to track the spread of the exotic spiralling whitefly (Aleurodicus dispersus) through northern Queensland. His central challenge is to calculate the likelihood that spiralling whitefly is in a particular area, based on the available surveillance strategies and current knowledge about the pest, by quantifying the uncertainty in the surveillance system.

Since arriving more than a decade ago, spiralling whitefly has established itself in Surveillance systems you can count on Statistical modelling provides greater confidence in pest surveillance systems tropical coastal Queensland, from Torres Strait to Gladstone, and has also been found in the Darwin region. Restrictions have been placed on the transport of nursery stock from Queensland because the insect attacks crops and ornamental plants.

Mr Stanaway’s modelling includes all known information about the insect, such as growth and reproductive rates and its spread to date. It also includes uncertainty calculations about the pest’s ecology and the sensitivity of monitoring techniques. It effectively provides a statistical fact check for predictive modelling about how the insect may spread, which can highlight problems in surveillance systems and in the modelling.

He uses a scoring system that identifies how good the ‘presence’ and ‘absence’ data collected by the surveillance program is – how likely a ‘false negative’ might be and whether the insect is actually present, despite the fact that it hasn’t been found.

His analysis has already identified the most effective host plants to monitor for the whitefly and a method of assessing how effective inspectors are. He says his analysis will be used to estimate whether spiralling whitefly has spread as far south as it is likely to, based on ‘absence’ data from southern-most monitoring points for several years, and the fact that once the pest spreads to a new location numbers build up to the point where it is easily identified in surveillance. This finding may allow restrictions on the movement of nursery plants to be reviewed.

He says the statistical surveillance information is being used in conjunction with a geographic information system (GIS) database and predictive modelling to develop a risk map for the spread of the insect. “The risk map is a much more powerful tool than the surveillance data because it provides a visual representation of changing insect or disease presence over time; it gives a sense of movement or change, which can be otherwise difficult to assess.”

In WA, Nichole Hammond has considered both active and passive surveillance systems for the exotic fungal pathogen Tilletia indica, which causes Karnal bunt of wheat, rated an ‘extreme’ threat to the grains industry if it ever gets through Australia’s quarantine defence.

The Department of Agriculture and Food, WA (DAFWA) has run an active surveillance for the past 10 years, collecting grain samples from the bulk grain handler the CBH Group. Each round of DAFWA surveillance tested approximately 200 grain samples and no Karnal bunt has been detected.

Passive surveillance systems involve the constant scanning of crops by farmers, farm workers and agronomists for anything ‘out of the ordinary’. Ms Hammond says passive surveillance has historically been used to support claims of freedom from plant pests, but its effectiveness has not previously been formally evaluated.

Her research analysed the community detection and reporting processes – the likelihood that growers would detect a possible infection and report it, based on existing surveillance practices for reporting pests and diseases.

She found that the active and passive surveillance systems provided greater than 90 per cent confidence that WA is free of Karnal bunt. This was based on 10 years’ worth of harvest data and the level of infection expected, if the disease was present in WA.

Ms Hammond’s analysis identified how sensitive each stage of the surveillance and identification process was. It highlighted where these could be improved, such as providing more information to growers and agronomists about signs and symptoms of exotic grains pests to improve confidence in on-farm detection.

Photo caption: Nichole Hammond recently submitted her thesis

Photo acknowledgement: GRDC Ground Cover

Article written by: Catherine Norwood

Charting pest rankings: a concert of challenges

K.Scott — Fri, 29 Oct 2010 00:59:46 +0000

The universal principle of 'like attracts like' applies to plant pests as much as to people.

Like rockstars, roadies, fans and promoters congregating at a concert to feast on the music and money, groups of plant pests with shared appetites can also be found gathered at common venues.

CSIRO entomologist and CRC for National Plant Biosecurity researcher Dr Dean Paini calls these pest groupings ‘assemblages’. He uses the concept of selforganising maps to identify pest groupings that affect Australian agriculture and the environment. By identifying those pests currently absent from these groupings in Australia, but present in similar groupings overseas, he determines the pests that are most likely to establish here.

“If 10 species have established in a particular region, and nine of the same species have established in another region with a similar environment, then the likelihood of the tenth species establishing in the second region is very high,” he says, “like the ticket scalpers finally arriving on the concert stadium steps.”

His research has matched assemblages of 844 plant pathogens and insect pests across 459 different geographic zones around the world. Most regions represent either a country or, for larger nations such as Australia, the United States, China and Canada, the various states or provinces that make up that country. Every species was given a risk rating, on a scale of zero to one, according to how likely it was to establish in Australia based on the mapping of pest assemblages.

“All around the world biosecurity departments are looking for this kind of information to help them prioritise risk and allocate resources where they will be most effective,” he says. “This is an objective, analytical tool that can be used to support expert knowledge.”

Dr Paini says identifying whether a pest is likely to establish in Australia is only part of the risk-assessment equation. Other elements include the likelihood of a pest actually arriving and the potential cost to industry and to the Australian environment of an exotic pest incursion.

Not everything that makes its way to Australia will find favourable conditions and not every exotic pest will have a significant impact on industry, even if it does establish itself here.

Among the highest-ranked pests in Dr Paini’s mapping analysis were the yellow stem borer and the purple stem borer, both pests of rice crops, with risk indexes (out of one) of 0.7924 and 0.7722 respectively.

However, the comparatively small size of the Australian rice industry suggests that even should these pests establish here, the impact on the domestic economy would be relatively small.

The same cannot be said of the oriental fruit fly (Bactrocera dorsalis). Although it ranked 31st in Dr Paini’s research, with a risk index of 0.5008, it has the capacity to attack a wide range of fruit and horticultural crops. Complementary CRC risk-related research led by CSIRO economist Dr David Cook has identified potential costs of almost $652 million to the apple and pear industries alone, over 30 years, associated with an incursion of this pest.

Dr Cook’s research team has developed an enhanced risk-analysis model to assess the potential cost impacts of incursions and to help industry groups prioritise risks and areas for biosecurity investment.

Apple and Pear Australia Ltd (APAL) was among a number of horticultural groups that took part in Dr Cook’s research.

APAL general manager Tony Russell says the industry has had a long interest in biosecurity issues, crystallised by the potential of New Zealand apple imports bringing fire blight into Australia. Dr Cook modelled the impact for a range of pests nominated by APAL and fire blight was the pest with the highest impact, with an outbreak estimated to cost $846 million over 30 years.

“The modelling has provided us with an economic assessment, which helps to start the process of prioritising pests and making decisions about investment required,” Mr Russell says.

“The next phase will be the ability to model the spread of an incursion under different circumstances. To do that we realise we need better information about where our commercial orchards are located – so that we can better plan and respond to an incursion.”

Other horticultural groups that took part in the project, as members of Horticulture Australia Ltd, included the potato and vegetable industries, along with the Rural Industries Research and Development Corporation.

Dr Cook says an important part of the modelling project was actually gathering detailed information about the pests involved, which was collated into threat data sheets. Based at the Department of Agriculture and Food, Western Australia, Dr Abu-Baker Siddique spent three years collating detailed information about the 69 nominated exotic pests affecting 13 different horticultural industries involved in the project.

The data sheets include valuable specialist knowledge gathered from around the world, such as information on the different control and treatment regimes. They have provided essential information to the modellers developing the risk management tool.

The modelling produces statistical information about the likely economic impact of the nominated pests over a 30-year period. It takes into account control costs, loss of production and lost markets. When combined with a deliberative multicriteria evaluation (DMCE) framework, it includes an allowance for environmental and social costs of pests. For instance, where a pest has the potential to affect native flora and fauna, the implications of this damage can be compared with the agricultural damage to get a much clearer picture of the threat posed to Australia.

The tools developed by Dr Cook’s team allow industry and government participants to determine the trade-offs they may be willing to make in terms of allocating resources and potential impacts of an incursion. They also provide information that can be used to inform biosecurity policy decisions and establish cost-sharing arrangements for control measures.

“We found, when we began using the tools with stakeholders, that they were extremely interested in the assumptions and calculations used in putting the modelling together,” Dr Cook says. “But the model we had developed was extremely complex and it wasn’t easy to ‘crack it open’ to demonstrate the variables and how they influenced the outputs of the model.”

That’s one of the reasons for his next project – ‘Communicating uncertainty in biosecurity adaptation’. Dr Cook says it’s about developing a simpler form of the model that can be more easily understood and manipulated by non-modellers.

“I call it the ‘war games’ version. We’re planning to develop it as PC-based software that, with some training, industry groups will be able to use to try out different incursion and impact scenarios to help with their decision-making. It will have a map-based interface, which should be easier to use than the statistical outputs of the original model.”

Dr Paini is also moving on to a new risk-analysis project to identify the potential pathways for incursions into Australia – the ‘likelihood of arrival’ part of the risk equation.

He is specifically looking at indirect shipping connections that could bring high-priority pests to Australia. The project, ‘Six degrees of preparation’, is founded on network theory, well known in public health circles for studying the transfer of disease, such as the pandemics of SARS in 2002 and H1N1 influenza in 2009.

“You have to be able to identify what may be only a single link between two otherwise unconnected communities,” Dr Paini says. “For instance, goods coming from the United Kingdom could be transferred to containers from Asian countries that have previously carried grain infested with a pest that’s not commonly found in the UK, and that is not found in Australia. Those goods are then delivered to Australia and before long a bug crawls out of the woodwork and we have an exotic incursion on our hands.”

Article written by: Catherine Norwood

Modelling pest risks

K.Scott — Thu, 16 Sep 2010 06:23:33 +0000

The IVth International Pest Risk Modelling Workshop was held in Port Douglas from 23-25 August. The theme for this year’s workshop was ‘pest risk in a changing world’ and was attended by 29 delegates from eight countries. The workshop is an annual gathering of researchers and pest risk analysts who are interested in advancing methods for improving our ability to inform risk assessments for biosecurity.

In his keynote address, Professor Bob Sutherst from the University of Queensland provided a lively retrospective on pest risk modelling approaches. This year’s meeting follows on the heels of a recent publication in Bioscience of the groups’ roadmap for improvement of pest risk mapping. The group have now agreed to embark on the development of a practitioner’s guide, which would include a collation of techniques and approaches, as well as a gentle policymaker’s guide to the dark art of pest risk mapping. This guide will be a multi-authored affair, edited by CSIRO’s Dr Darren Kriticos.

Sponsorship from the CRC for National Plant Biosecurity, Horticulture Australia and Hearne Scientific Software allowed the meeting to sponsor the attendance of Ni Wenlong from the Chinese Agricultural University. Wenlong is undertaking postgraduate studies on assessing the pest risk due to fruit flies under current and future climates, and is using CRC products developed from our Climate Change project. At the meeting he presented some of his work on the Peach fruit fly. In the immediate term, his research could help Australia understand one of the pest threats posed to peach growers.

The CRC donated a leather-bound compendium to the best presentation and best poster at the workshop. The awards were nominated by the attendees and the best presentation was awarded to Dr Ross Weiss of Agriculture and Agrifood Canada, and the best poster was given to Dr Libby Pinkard. Hearne Scientific Software donated a copy of CLIMEX for a lucky draw, which was won by Dr Sue Worner of Lincoln University.

The next meeting is being planned for Fort Collins around August 2011, where we expect to be able to celebrate the completion of the Practitioner’s Guide.

Written by Dr Darren Kriticos

Spreading the collaboration

K.Scott — Thu, 16 Sep 2010 05:30:26 +0000

Australia faces a significant challenge in order to minimise the entry, establishment and spread of harmful plant pests and diseases which could affect agriculture production, market access and our environment. A plant pest or disease outbreak could have serious economic and environmental impact. To protect Australia from a potentially devastating pests and diseases a tool to predict how they were established, how far they spread and the speed at which they spread is required.

In order to create this tool, CRC researchers from University of Western Australia (UWA) and the Department of Agriculture and Food, Western Australia (DAFWA) recently met to discuss the development of simulation models.

The economic and ecological model node meeting was organised by DAFWA entomologist, Dr Maria Majer and Research Assistant Professor Rohan Sadler (UWA). Dr Majer is working on the CRC project Simulation Surveillance Platform which is developing a computer simulation model that will predict the establishment and spread of invasive plant pests and diseases. Maria is also working with Research Fellow, Dr García (UWA) who brought his expertise to the meeting. Professor Sadler is working on a project looking at the progress in modelling of the benefit cost analysis of market specification and technology options. He also supervises CRC PhD student, Ms Hoda Abougamos who is researching spatial elements of the economics of pest control in grain logistic chains.

In addition to these participants, Research Fellow Dr James Bennett (UWA) and his project leader, Assistant Professor Dr Michael Renton also attended. They are working on a system to rapidly characterise new invading organisms. This model will provide timely information about the pest’s biology and spread potential, as well as guide the appropriate response in the event of an incursion. Dr Renton also keeps busy as a supervisor for CRC PhD students Mr David Savage and Dr Mingren Shi.

The presence of DAFWA experts in entomology such as Dr Bill Woods, Dr John Botha, Mr Marc Widmer, Ms Rose Fogliane and plant pathologist, Mr Bill MacLeod and modelling expert Dr Moin Salam ensured for an informative meeting and robust discussions.

The interaction of the groups was very positive and was an excellent chance to network with others in the field. The group was very pleased and honoured by the presence of CRC Principal Scientist Dr James Ridsdill-Smith Adjunct Professor University of Western Australia, and Honorary Fellow CSIRO Entomology.

The meeting also provided an opportunity for PhD student Nichole Hammond (DAFWA) to present her research findings to fellow CRC researchers. Nichole is evaluating plant pathogen surveillance and surveillance tools, and in particular using data to demonstrate Western Australia is free from Tilletia indica, Karnal bunt.

Photo caption: Back row, Michael Renton, David Savage, Mingren Shi, James Bennett, Dr García and James Ridsdill-Smith. Front row, Rohan Sadler, Veronique Florec, Maria Majer, Moin Salam and Nichole Hammond.

Image courtesy of Maria Majer

Article written by Maria Majer

Modelling the potential distribution of the Asian citrus psyllid in Australia under increasing temperatures

K.Scott — Thu, 28 Jan 2010 23:27:45 +0000

Hierarchical Bayesian Models for Emergency Plant Pest Management

VANMEURSA — Thu, 21 Jan 2010 04:11:08 +0000

Phosphine resistance modelling: the right genetics is crucial

VANMEURSA — Wed, 13 Jan 2010 00:58:43 +0000

Rapid response needed to rein in biosecurity breaks

K.Scott — Thu, 03 Dec 2009 01:58:07 +0000

Global Biosecurity Media Release: 1 December 2009

Scientists are investigating how Australians can respond more quickly and strategically to biosecurity outbreaks.

Presenting at the upcoming 2010 Global Biosecurity Conference, Dr James Bennett from the University of Western Australia said in the past, surveillance programs for pests and diseases of crops had been developed in an ad-hoc manner, on a case-by-case basis.

He said a rapid response was crucial to minimising the costs and impacts of outbreaks.

“The key element for rapid response is an effective surveillance strategy.”

“When a new harmful plant pest or disease is discovered, a rapid and accurate response is needed because as the organism spreads, the cost of management escalates and the likelihood of eradication quickly declines,” Dr Bennett said.

“Once an outbreak has been detected an assessment of the organism’s potential to spread in the environment where it has been found needs to be carried out immediately.”

Dr Bennett said historically Australia had been reactive rather than proactive in developing response programmes.

“When an incursion occurs, there is often little information about how far the invading organism will spread in the new landscape.”

“In some cases it may be possible to use historical data from a different site with similar climatic and environmental conditions to parameterise a simulation model; however, this is rarely the case.”

“Where there is little available data about the potential size of an incursion, expert knowledge is invaluable.”

The project being carried out by Dr Bennett and his colleagues will result in a list of questions to ask experts about an invading species, which will enable all available spread-related information to be quickly gathered.

“The gathered information will then be used to accurately forecast the spread and to make better decisions about how to best respond to the incursion.”

Understanding the risks of potential biosecurity issues will form the basis of the Threats and Impacts stream of the 2010 Global Biosecurity Conference to be held in Brisbane from 28 February – 3 March 2010.

The conference is a partnership event between the CRC for National Plant Biosecurity, Australian Biosecurity CRC for Emerging Infectious Disease and the Invasive Animals CRC.

The 2010 Global Biosecurity Conference is sponsored by: the Grains Research and Development Corporation (GRDC); the Australian Centre of Excellence for Risk Analysis (ACERA); and the Australian

Government’s Australian Quarantine and Inspection Service (AQIS).

Registrations for the conference are now open. Visit www.globalbiosecurity2010.com

CRC60034: Bayesian Surveillance Systems - PhD

K.Scott — Tue, 12 Aug 2008 06:03:53 +0000