Document relating to the history of the CANBR, by John Connors, 2006
EUCLID
EUCLID Eucalypts of Australia is the culmination of many years of hard work, not only by the authors, but by many others as well. It all began back in 1991 when Judy West, the director of the Australian National Herbarium, asked Ian Brooker, the then curator of eucalypts at the herbarium, to put together a list of distinguishing characters that could be used as a foundation for an interactive key to the eucalypts. Isobel Crawford, with some help from Terena Lally, used Ian’s character list to code up data in the computer software DELTA to produce an interactive key to the eucalypts of Victoria. (This was not the first interactive key to the eucalypts however. George Chippendale and Ludek Wolf from the CSIRO Division of Forestry must be given the credit for this, as they had already produced a pilot version of a key prior to the start of EUCLID). Isobel’s achievement was a formidable task considering that back then the entire CANB eucalypt collection was housed in three separate sets of specimens, along with the eucalypt spirit collection, in a small demountable building behind the Black Mountain Library. One set represented the old CANB collection (on large herbarium sheets), another represented the then recently donated collection from FRI (on small herbarium sheets) and the third was the new incoming specimens, obtained from field collecting or donated from other herbaria. The building was not air conditioned and in the winter was very cold, while in the summer very hot and when it rained it was necessary to manoeuvre the compactus units so that the specimens did not get wet from the leaking roof. It is also worth mentioning here that the information stored on the herbarium specimen labels was databased for the entire collection while it was housed in this demountable. Maggie Nightingale, Eva Bugledich, Colin Miller, Kate Mashford, Marion Garratt, Sam Usback, Libby Viccars and Joan Graham were some who braved the elements to accomplish this significant achievement. (This initial database formed the basis for the early EUCLID maps). In 1995 the entire collection was relocated into the new wing of the CANB herbarium. Coinciding with this was the donation of Denis and Maisie Carr’s collection from the Australian National University, both collections forming what is now arguably the biggest and one of the most important eucalypt collection in the world. Work on EUCLID now began in earnest.
About this time a number of important decisions were made that changed the scope and direction of EUCLID. The usefulness of Interactive keys was starting to become evident. Bernie Hyland and Trevor Whiffin had just published the highly acclaimed Australian Tropical Rain Forest Trees, An interactive Identification System and many other workers were planning their own keys. At this time, DELTA was the only freely available software in Australia suitable for compiling data to produce interactive keys. Intkey, the program within DELTA that generated the interactive key was regarded by some as needing major modifications to keep it up-to-date with the rapidly changing advancements in computer technology. The DELTA team seemed reluctant to make further changes to their program and this led to the development of another computer software package LUCID. It was decided by the EUCLID team to switch to this new program in which we had significant input in the early stages of its development (DELTA has since been improved).
It was obvious to the compilers of EUCLID that the most effective way to supplement character states and species descriptions was with drawings or photographs and that we should try to utilise the existing CANB slide collection to achieve this. It was agreed to experiment with this approach by adding the species of the ACT to the Victorian species. Proving successful, this was soon expanded to include the species from Tasmania, New South Wales and those species from South Australia that were thought to have affinities with the south-eastern eucalypts. Kevin Thiele produced many excellent line drawings, Andrew Slee made use of the CSIRO Scanning Electron Microscope to produce some excellent close-up photographs and we made use of the many excellent Brooker & Kleinig photographs from the slide collection. Field work was undertaken to obtain vital information not available from dried herbarium specimens and also to fill in the gaps for photographs not present in the collection. This then formed the basis of the first edition of EUCLID, Eucalypts of south-eastern Australia, published on CD in 1997 by CSIRO Publishing.
It was also obvious that the efficiency and effectiveness of the key would depend very much on the quality of the data behind it. For this reason we were determined from the outset to code up the data behind EUCLID on sound information obtained from the herbarium collection. This meant that our collection had to be in good shape curatorially before we started coding. Much time-consuming work was carried out curating our collection, updating the database and checking distribution maps. It was also apparent early that a lot of the information needed for the key was not available from the information on the herbarium specimen labels and would have to be obtained in the field. Many long and arduous weeks were spent in the field chasing this data and also taking photographs not represented in our slide collection. The identity of many of the specimens collected on these trips was rigorously tested through the key.
Most of the slide photographs from our early field work were taken using a Nikon F60 camera body, a Nikkor 105 mm micro lens for close ups and a Nikkor 28-85 mm zoom lens for habit and bark shots. The leaf venation photography was carried out using a technique developed by Ian Brooker. He had a purpose-built frame with a Nikon F301 camera body attached at one end. This camera was equipped with a Nikkor 55 mm micro lens, with a set of two PK13 extension tubes between the camera body and the lens. At the other end of the frame was an adjustable sliding clamp that held the leaf at one tip, with the user sliding the clamp to obtain correct focus. The other end of the leaf was held by hand where final adjustments on the focus were made. The camera and frame would be held up to face the sun (but not directly at the sun) and the photograph taken with the transmitted light revealing the leaf venation and oil gland pattern, with the clear blue sky as the background. Apart from needing three hands to operate it, this was a very effective method to document this important information from fresh leaves while in the field.
We soon amassed a large number of slide photographs that needed to be digitised and burnt onto CD, before they could be utilised in the key. The operating costs for items such as digitising images and running field work were considerable. We were fortunate to obtain a BushCare grant from the Department of Environment and Heritage that helped alleviate the financial burden to CSIRO. This grant also provided some resources for the engagement of a graphic designer for the second edition.
The early stages of EUCLID posed many new challenges for the authors. Working with digital images was one of them. Understanding image resolution, screen resolution, image compression, and different file formats created many headaches for us early on. Fortunately CSIRO Publishing had a talented graphic designer who came to our rescue, Linda Kemp. Linda’s contribution to the first edition of EUCLID was significant and many of the background colours and layout designs chosen by her still survive in this latest version. Working with html pages was another challenge for us. Once again CSIRO Publishing helped out. The EUCLID project officer for the second edition Andrea Jordan introduced us to the world of html and she is responsible for much of the design that is still present in EUCLID today.
Computers have improved so rapidly over the last few years that it is now difficult to imagine how we managed to work with the many large image files and memory hungry programs like Photoshop on our early machines. These were slow 486 machines, with 4 megabytes of RAM and 200 megabytes of hard disc space. We were very fortunate indeed to have Dac Nguyen working with the IT staff for CSIRO Plant Industry. He built for us, what was at the time, the best computer in the Plant Industry fleet, with a fast processor, 8 megabytes of RAM, two hard disc-drives, a 17 inch screen and a CD ROM drive. Dac continued to upgrade our machines to keep abreast of the ever increasing demands for more RAM and more disc space and we are grateful to him for his dedication to our project. Dac, in many ways, typifies the expertise and dedication that exist in many other areas of CSIRO Plant Industry. There are many other individuals (too many to mention) from the Administrative Services, the Visual Resources group, the IT section, the Potting Shed, the Scanning Electron Microscopy group and Site Services who have contributed to the compilation of EUCLID in an indirect way.
EUCLID has progressed a lot since the early days. The data matrix has been continually checked and updated, the taxa list has been expanded to now include all the eucalypts of Australia and the character list has been modified to accommodate features for some of the northern species. As well as improvements in EUCLID, the LUCID software has been progressing steadily as well, with the new LUCID 3 version now available.
Eucalypts are a complex group of plants, with each related group having its own unique set of distinguishing characters. We realised early that if EUCLID was to be successful, then we should endeavour to highlight these distinguishing characters with a comprehensive set of images. We have endeavoured to represent the entire taxa list with a set of images that includes a habit shot, showing if possible the habitat in which the plant is growing; a bark shot showing the texture and colour of the bark, as well as the amount of rough or smooth bark on the trunk; a leaf venation shot showing the tertiary venation and the position of the oil glands (if present); a bud shot showing the number of buds per umbel, the colour and shape of the bud and operculum, and also trying to show if possible the presence or absence of the operculum scar; flower shots showing the colour of the stamens, the presence or absence of staminodes and any clustering of the stamens into groups; a fruit shot showing the colour and shape of the fruit, and the position of the disc and the open valves; a seed shot once again showing the colour and shape of the seed and associated chaff and other features like the hilum position and the presence or absence of a wing or flange; a juvenile plant taken from the field showing leaf position, colour and shape of the leaves and any other important features like leaf or stem hairs; a cultivated seedling, germinated in the glasshouse but grown outside in natural light, showing the same set of features as the field juveniles and a map plotted from the CANB herbarium specimen database, representing real spots on the ground. This standard was very difficult to achieve and we have fallen short in some areas.
A lot of our later photographs, particularly those from northern Australia, have been taken using a digital Nikon D110 camera body. Our seed photographs were also taken on a digital camera, using a high quality Prog-Res camera mounted on a dissecting microscope, where a series of photographs were taken for each subject from the top focus position to the bottom focus position and this series of photographs then blended together using the Automontage software to produce an image with no depth of field problems. This highly technical piece of equipment was further complemented by a very simple technique developed by Andrew Slee, where the seeds were suspended well above a blue background on a deep clear glass dish to eliminate reflected light and a series of moveable white lights positioned close to the subject to highlight shadows, thus emphasizing important features of the seed.
As EUCLID grew bigger, so too did some of the problems associated with it and none more so than managing the large number of accumulating files. EUCLID now consists of well over 10,000 images and over 1,000 html pages. It was clear after the publication of the second edition that we needed a better method of managing our html pages and the associated files that link to them. Kevin Thiele produced a basic but effective program that stored all the text and image information associated with our html pages in a database. Running his program, in association with a template and style sheet, produced the early versions of our taxa fact sheets. This initial program of Kevin’s has now developed into a highly sophisticated program within LUCID, called Fact Sheet Fusion. This program is now producing all our html pages for EUCLID. Using this program we can alter the presentation of each and every html page by simply changing the template. Fact Sheet Fusion also produces the links to other pages, including those to other fact sheets and those to the glossary.
Others have benefited from the production of EUCLID. We have managed to see and collect almost all the Australian species of eucalypts in their natural habitat. The field work associated with doing this has sent us to some of the most remote places on the continent and has enabled us to collect many valuable herbarium specimens, not just eucalypts. On a trip to the Great Victoria Desert, co-collector Dean Nicolle was able to collect many valuable eucalypt specimens for his work on the Series Subulatae as well as rediscovering Melaleuca apostiba, up until then lost to science. On a trip to the remote areas of the Northern Territory, co-collector Brendan Lepschi made several valuable collections, prior to that not represented in the CANB herbarium and was fortunate to experience an unforgetable stay in the Top Springs Hotel. On a trip to the Kimberley Region of Western Australia, co-collector Teguh Triono was able to collect material for his study in the Sapotaceae. On a trip to south-east Queensland, co-collector Dave Mallinson made many valuable collections, for some species the first fertile material for the CANB herbarium.
EUCLID is not, and was never intended to be a taxonomic revision of the Eucalypts. Our aim was to produce a working key to all the eucalypts of Australia, with a comprehensive set of images and descriptions for all, complemented with comparable notes for the closely related species. In trying to achieve this, we found it difficult to accept many of the recently published names and have been forced to make difficult decisions regarding the taxonomic status of many new species, relegating some to synonymy. These decisions were never made lightly and were always made after careful examination of herbarium specimens and in some cases field observations, and always with due respect to the authors of the names involved. In most cases we have offered reasons in the notes for our decisions.
Perhaps it is fitting that the final paragraph is dedicated to Siobhan Duffy, the graphic designer for the last two versions of EUCLID. Siobhan’s dedication to this project (and many other projects as well) has always been well above what was expected. Her expertise in graphic design and many other computing software packages, especially LUCID and Fact Sheet Fusion, has contributed a great deal to this final version of EUCLID. The other contributors of EUCLID are extremely grateful for her valuable contribution. We sincerely hope the users of EUCLID enjoy using it as much as we have enjoyed putting it together, and that much is learnt about this fascinating group of plants and with many successful identifications made along the way.
John Connors