The Department of Biogeography and Geomorphology, Australian National University, proposes to develop a print and electronic record of the pollen flora and associated microfossils likely to be recovered from river, swamp and lake and other deposition sites in Australia, Eastern Indonesia, New Guinea, the high western Pacific islands (Solomons, Vanuatu, New Caledonia, Fiji) and the sub-antarctic islands. The major purpose of this record is to make identification of pollen possible for regional users who wish to carry out studies into past vegetation and environmental histories on a variety of timescales. Such studies can contribute to understanding the stability of existing ecosystems, the impact of human activities and the effects of climate change. In the Australian region pollen analysis has produced remarkable results, but it has the potential to be used with much greater precision by a wider group of users if the problem of access to regional pollen identification can be improved. This work will make a major contribution to understanding the ecological effects of climate change and hence can contribute to strategies for managing the present problems of global warming.
The Atlas will also be a major resource for plant systematists interested in the relationships of the living floras of the region, as pollen is a taxonomic feature closely tied to the floral structures used to classify the flowering plants. Palynology is a major geological tool, and the atlas will help stratigraphic palynologists to refine their fossil types in relation to living floras. In converse, more accurate knowledge of palaeofloras will improve understanding of the evolution of Australian plant communities and environments. No pollen atlas exists for the Australian region, and in fact workers make do with atlases from Taiwan, Argentina and Britain. The atlas will also be of interest to other branches of pollen research such as honey, allergy and plant reproductive biology.
Background. Since 1965 the Department of Biogeography and Geomorphology in the Research School of Pacific Studies has been developing a comprehensive collection of the spores and pollen of Australian and Pacific plant species. The collection currently stands at about 16,000 slides covering about 7500 species. Although the total flora of the region probably exceeds 35,000 species, pollen is a conservative element, and for many large families only a sample of representative genera is necessary. Some families (for example the Orchidaceae, the family with the most species in the region), are virtually never represented in sedimentary sites, and hence do not have a high priority for comprehensive collection. Although still incomplete, the department's collection is by far the largest and most comprehensive for the region in the world. It has cost some millions of dollars in equipment, materials and staff time to collect, prepare and catalogue modern pollen. Further effort will be required to develop more complete holdings in groups with potential to contribute to understanding past ecology. Examples are the eucalypts and related genera, the daisies, and mangroves.
The pollen obtained from flowers from identified specimens is suspended in silicon oil in small sample bottles, and is also held as microscope slides. A card file record including a formal description and microphotographs is continually being enlarged, and such coded specimens are being entered in a database at present. The collection and these records are only available to users in Canberra, and it is constantly visited by interstate researchers. Using the material is difficult and time consuming as slides deteriorate and often need to be re-made to allow a researcher to carefully compare unknown pollen grains with likely matches in the collection. Obtaining a slide and examining a few pollen grains under 400 or 1000x magnification usually takes several minutes. Photographs which are kept on reference cards are useful, but often not sufficient. Even more effort is needed to obtain scanning electron micrograph images, although these provide detail which is usually quite diagnostic, and they can greatly help the interpretation of normal light images.
The Pollen Atlas Project plans to make the data available as descriptions and images that will allow identification of unknown grains from all regions in the Australian region. This will take the form of a computer database incorporating scanned micrographs, which will incorporate SEM images where these are available. From this database, which will be continually upgraded and extended, a major pollen atlas will be prepared using a high definition laser printer to handle the images. The advantage of the database is that it will allow the production of updated versions and increased detail on selected groups. The total region is very large and variable so the main atlas may not prove suitable in day to day identifications at specific sites. There may also be too much detail in the database to allow it all to be published in book form. Read/write compact disk technology appears to be the most suitable method for making the data available. There are advantages over a book, if images can be displayed at standard size on a split video screen which is also showing unknown grains via a connection to a microscope camera.
In addition to the complete atlas, the project is intended to make sub-regional or purpose built atlases and databases available to users. These would normally cover natural floristic regions, for example western Tasmania, south eastern montane - alpine Australia or lowland New Guinea. Using locational data, often at generic or family level, we expect to be able to produce pollen guides consisting of 150 - 300 pollen types for any site in our region. These could be tailored for particular purposes, such as food plant identification or aquatic hydroseral analyses. Such guides are going to be interactive, in that obvious gaps in the existing collection will be apparent, and updated versions can be prepared. Currently the collection is strongest for Tasmania, coastal and subalpine Victoria - NSW, northeast Queensland, inland NSW and South Australia and montane New Guinea. Such regional atlases will be easier to use, and should be capable of being installed on common PC facilities. This is essential if a wide base of users in Australia and the Pacific is to be encouraged to make use of pollen analytical techniques.
Atlas Development. The collection is currently catalogued using a database on a PC, and 4000 of the preliminary complete records (without images) are held in a more complex computer database capable of storing a range of image formats and producing publication quality output. Because pollen characteristics are coded in the database, sorting routines can be used to locate groups of pollen with similar characteristics for comparison. We have tested several scanning techniques using photographs of pollen, but the file sizes produced are very large, and photographs are not available for more than part of the collection. We expect that a video capture of the selected grains under the microscope will provide the highest quality image with the least delay or risk of error, and this system is being developed by pTIZAN Computer Services Pty Ltd. The problem of image file storage and retrieval seems most economically solved by videodisk technology, and practical options are being investigated. In the interim the database is being expanded and is stored in a mainframe computer at ANU, but is accessed by personal computers. Interactive user access to this database could be arranged as an alternative to book or videodisk products. This might save on some equipment costs, but could cut out some routine users.
The project will be capable of continuing expansion once the methodology is in place. Future developments will include the incorporation of other fossil groups of interest, including diatoms, algae, dinoflagellates, microfauna and phytoliths. Larger objects commonly found in stratigraphic contexts could also be included, such as seeds less than 5mm, standard SEM micrographs of wood and charcoal or insect carapaces. It would also be advantageous to incorporate the well-archived Tertiary fossil pollen from the region in the database.
The project will require the cooperation of laboratories in the region to provide sample data, to make corrections, and to house non-pollen collections. Significant collections in Sydney, Melbourne and Bandung (Indonesia) exist. Because preparation of the data is already well advanced, the atlas should start to appear in the next three years, if external support for the project becomes available. Although the project has received substantial development support from internal funds, the goals are not directly those of the Research School of Pacific Studies, and thus cannot claim a disproportionate extent of resources. Similar projects in Europe and North America have received grants from geological and taxonomic sources, as well as industry funds.