This paper was presented at the 6th International Conference on Plant Protection in the Tropics (6th ICPPT), Kuala Lumpur, 11-14th August 2003.

THE AUSTRALIAN PLANT PEST DATA BASE: A MODEL FOR THE COUNTRIES OF SOUTH EAST ASIA?

Lumb, E., Naumann, I.D., Evans, G and Pheloung, P.

Office of the Chief Plant Protection Officer, Department of Agriculture, Fisheries & Forestry – Australia, GPO Box 858, Canberra ACT 2601, AUSTRALIA

Summary: Assessments of the arthropod pest collections and plant disease herbaria in ASEAN countries indicate that none is in a position to provide an adequate description of the health status of their agricultural industries. This situation is partially attributed to the many dispersed collections making access to pest records difficult.

Plant health responsibilities in Australia are split between the Commonwealth and State Governments. Agricultural production is a State matter, whilst the Commonwealth is responsible for trade and quarantine. This dichotomy has created a situation in which the States, Universities and research organisations hold most of Australia’s preserved specimens and records of plant diseases and arthropod pests.

The Commonwealth Department of Agriculture Fisheries and Forestry — Australia (AFFA) is a major user of pest record information contained in the many Australian collections. Obtaining such information from the geographically isolated collections, has proven at best to be time consuming and resource intensive for the Commonwealth. In part to facilitate trade negotiations and justify quarantine restrictions, the Australian Plant Pest Database (APPD) was developed to address the problem of describing the health status of Australia’s agricultural industries.

The APPD is an internet-based, plant health tool linking databases associated with plant pest collections in Australia. The APPD can be used to support decisions made with respect to agricultural trade and emergency pest management.

Through the APPD, pest record information from multiple databases can be dynamically searched. Results are returned quickly to the user, in an amalgamated report containing information considered to be essential components of a pest record as defined by the International Standard for Phytosanitary Measures #8, Determination of Pest Status in an Area (FAO 1998), plus additional mapping options. By mid 2004, it will be possible to search the APPD for over a million insect, mite, nematode, fungi, bacterial and viral records located in 17 separate pest[1] collections. Ongoing databasing activities will steadily increase the proportion of vouchered pest specimens collected in Australia that are accessible through the APPD and ensure that information is up-to-date and quality assured.

The APPD is similar to other biological collection portals and eventually may be amalgamated into the Global Biodiversity Information Facility network.

Keywords: Pest Collections, Distributed Databases, Australian Plant Pest Database, Plant health status

Materials and methods:

The Australian Plant Pest Database (APPD) began with a series of reviews during the mid 1990’s and workshops (1999, 2001), involving State (regional) and Commonwealth (=Federal) Governments, collection curators and plant health scientists. This process defined the aims and objectives of the system and secured broad support, especially from data custodians.

Addressing the following key needs of the distributed database has been critical to the successful implementation and worth of the APPD.

1. Funding

A review of Australian quarantine (Nairn et al. 1996) , recommended the development of a national pest and disease system. This recommendation was supported by the Australian Government and a Commonwealth Budget Initiative provided funding for the development of the APPD. The funding was administered by Plant Health Australia (PHA), used to design of the distributed database software (see 2. Heterogenous systems), and as seed money to initiate or subsidize ongoing databasing activities within pest collections. Substantial in-kind contributions were provided to the APPD network by agencies who were custodians of the collections, in the form of taxonomic expertise, staff and equipment.

2. Heterogenous systems and data

Australian plant pest collection databases are heterogenous in the way information is stored (e.g. locality data may occupy a single, or several fields), and the platforms used (different database packages, such as Microsoft Access, BioLink or Texpress). Gateway programs designed by CSIRO Mathematics and Information Sciences (CMIS), overcomes these local differences, allowing the central “broker” program (Figure 1) to interrogate the differently formatted information, and then to translate it into a homogenous, consolidated report for the user (See Phase 1: Network Architecture Development).

3. Management

A Steering Committee was formed to oversee the development, population and implementation of the APPD. This included the development of guidelines for all APPD stakeholders in the APPD Rules of Operation. The Rules of Operation addresses data sharing, security, record standards, data reliability, agency commitment and access issues. Membership of the Steering Committee consists of curators, collection and plant health agency representatives, with the OCPPO as secretariat.

4. Priorities

A streamlined approach towards electronic datacapture was established, so that the benefits of the APPD would be realised in the immediate future. Select pest groups were chosen based on their significance as plant pests and well-defined pest taxonomy. Priority taxa included taxa from fungal, arthropod and nematode groups (Table 1). Some additional (non-priority) taxa were permitted for datacapture, on a case-by-case basis.

Support was given to Australian Biological Resources Study (ABRS, Environment Australia) for the development of authoritative master name lists for the priority taxa. This assisted the dissemination of important pest taxonomy data that may be less accessible through written publications and is freely available through the ABRS website (ABRS 2002).

Essential elements toward successful development of the APPD have been:

  1. Innovative but proven internet software;
  2. Agreement on data standards and priorities for data capture (including master names lists and use of mapping tools);
  3. Continuing guidance from a Steering Committee representing data providers, users and information technology specialists;
  4. Coordination and management by the Office of the Chief Plant Protection Officer and Plant Health Australia;
  5. Ongoing commitment from Plant Health Australia to host the broker and maintain the linkages
  6. Federal funding made available through Plant Health Australia;
  7. Substantial in-kind contributions from data providers; and
  8. Well defined process for secure data access, intellectual property and gateway security.

The development of the APPD was undertaken in two major phases.

Phase 1: Network Architecture Development

Network architecture software was developed by CSIRO Mathematical and Information Sciences (CMIS), Canberra (CMIS 2002). The key components of the architecture include a central “broker” and a collection of individual locally deployed “gateways”, each tailored to the structure of local collection databases including BioLink (using MS SQL Server), Texpress, Oracle and MS Access. Gateways are specific to each local database system and schema.

APPD user queries are distributed over the internet (APPD 2002) by the broker to the gateways. From there they are translated to a form meaningful to the local database and results returned to the broker in a standard, xml format. The broker combines the results into an html-formatted table with a downloadable distribution map, and returns these to the user (figure 1). It is possible to save the results for later inclusion into reporting frameworks.

Deployment of gateway software can be varied to suit the individual needs of the IT network managers. Options include placement of gateway on a computer outside the firewall along with a copy of local database, or deployment within firewall. It is also possible to vary the way the gateway interrogates the local database, for example, to exclude records that are flagged as unverified.

Access to the APPD is currently limited to data providers and government agencies with responsibility for plant health. At present approximately 170 individuals have usernames and passwords, and it is anticipated that access will increase as more data providers join the network, and confidence in the quality of the data grows.

Data providers retain ownership of their own data and responsibility for maintenance of local databases. The APPD Rules of Operation specifies the terms under which data are shared and addresses matters such as intellectual property rights.

Figure 1. How the APPD works. The user specifies the search criteria (e.g. the scientific or common name of the pest, host plant common name or a particular geographic locality), which is sent to the APPD broker software. The broker translates the search criteria and queries the individual, heterogeneous collection databases. Information is returned to the broker and is translated and amalgamated into a single, html report consisting of summary and detailed results and map. This information is returned to the user.

Phase 2: Data Standards and Capture

Physical (i.e. vouchered) specimens exist for arthropod, nematode, fungal and bacterial records held in the APPD. The scope of the APPD pest records are for those species detrimental to plant health, that are found in mainland Australia and Tasmania, but not as quarantine interceptions nor exotic incursions subject to eradication. Data fields of pest records include scientific and common names; collection accession number; habitat; locality; latitude, longitude; host plant association; date of collection; collector; repository, and meet the requirements listed under the FAO’s ISPM 8 (1998).

A standard for a virtual, plant virus collection has been drafted so that well documented virus records, what often don’t include a physical specimen, can be included in the APPD (Gibbs 2002). The development of the standard was co-funded by Rural Industries Research and Development Corporation and APPD funds administered by PHA. Plant virus virtual record vouchers from Australia that do not include physical specimens would require a minimum set of documentation describing the collection event and analysis done to identify the virus.

The taxonomic range of the APPD covers pest arthropods, nematodes, fungi, bacteria and viruses (Table 1). For ongoing data capture, high priority arthropod and nematode groups (containing major pest species) were identified and funding assistance given to data providers to facilitate digitising of label data for these groups. Funding was also provided to enhance or validate pre-existing digital records. While the focus of the project is on the capture of pest collections in Australia, records of taxa other than those in priority groups and non pest taxa are also accessible via the APPD.

Table 1: Priority pest taxa for Australian Plant Pest Database

Nematoda / Acari / Isoptera /
Orthoptera
/ Thysanoptera
Longidoridae
Trichodoridae
Rhabditida
Tylenchid families / Tetranychidae
Eriophyidae
Tenuipalpidae
Penthaleidae / Mastotermitidae
Kalotermitidae
Termopsidae
Rhinotermitidae
Termitidae / Acrididae
Gryllidae
Gryllotalpidae
Tettigoniidae / Thripidae
Aeolothripidae
Coleoptera / Lepidoptera / Hemiptera / Diptera / Ungrouped taxa
Cerambycidae
Curculionidae (in part)
Scarabaeidae
Chrysomelidae
Bostrichidae / Noctuidae
Pyralidae
Gelechiidae
Tortricidae / Coccoidea
Aleyrodidae
Aphididae
Psyllidae
Cicadellidae
HETEROPTERA
Pentatomidae (in part)
Coreidae (in part)
Miridae (in part)
Lygaeidae (in part) / Tephritidae / Bacteria
Viruses
Fungi

Environment Australia’s (EA) Australian Biological Information Facility (ABIF) website (ABRS 2002) has been adopted as the reference site for valid names. Joint projects have been developed with EA to address critical gaps in the ABIF website (ABRS 2002), to ensure that preferred synonyms within the database are current for arthropod and nematode groups.

Timetable:

By completion of Phase 2 (Datacapture phase), it is anticipated that roughly 1,000,000 records, including those residing on the Australian Plant Disease Database, will be available.

By completion of Phase 2, 14 databases will be online (representing 17 collections) from the following agencies: Victorian department of Natural Resources & Environment, Queensland Department of Primary Industries, New South Wales Agriculture, State Forests of New South Wales, Department of Primary Industries, Water & Environment, Forestry Tasmania, South Australian Research & Development Institute/ Adelaide University, Department of Agriculture Western Australia, Bureau of Sugar Experiment Stations, the University of Queensland and the Australian Plant Disease Database. Other invertebrate and pathogen collections will be included into the APPD at a later date, or under the Online Zoological Collection of Australian Museums (OZCAM) initiative.

Results:

Distributed database composition

Databasing given provisional APPD funds will result in the projected priority taxa record composition of the APPD by mid 2004 (figure 2).

As at June 2003, over 750 000 records were made available through the APPD.

Figure 2. Estimated proportions of priority taxa in APPD by 2004 (excluding Australian Plant Disease Database records).

User feedback

Initial user survey results have been encouraging, with users finding the database highly useful to their field of expertise, easy to operate and performed the desired queries. Users were able to connect readily to the database, and queries were returned within a short period of time (Figure 3).

The APPD has been used for the following:

·  confirming plant pest and disease detections;

·  establishing species host ranges;

·  determining pest and disease quarantine status;

·  quarantine, Import Risk Analysis and pest management;

·  generation of pest distribution maps; and

·  determining pest and disease ranges.

Comments from users (State and Commonwealth Governments and research organisations) have allowed APPD IT infrastructure developers (CSIRO Mathematics and Information Services) to fine-tune the APPD website and software functions to users needs.

Figure 3. Respondent results from initial APPD useability survey

Examples of APPD benefits

Case study 1: Support of trade decisions

AFFA makes extensive use of the APPD to underpin its negotiations with other countries when Australian industries are seeking access to new markets. Guignardia citricarpa, is a debilitating bacterial disease of citrus also cosmetically damaging to fruit. For many of Australia’s trading partners it is a quarantine pest. By mapping locality data of the disease against commercial citrus areas of Australia (Figure 4), it is revealed that the disease has not been recorded in citrus exporting areas of Australia, and thus supports Australian overseas market access bids based on area freedom.