Weaving a Web Around Siam Weed: Disseminating Scientific Information in the Developing World
Siam Weed is a perennialshrub native to the humid tropics of Central America and theCaribbean. But in recent decades it has become a voraciousagricultural pest in lands far from its native habitat and naturalenemies. Many of these newly "colonized" coountries are among thepoorest in the world, where farmers struggle to raise subsistencecrops and small herds of livestock in the face of this growingperil.
In 1996, a small group ofresearchers from the United States and Australia pondered ways tobetter distribute information on the biological control andmanagement of Siam Weed, scientifically known as Chromolaena odorata.This article details the creation of a web site specifically designedfor audiences in the less technologically developed areas of theworld.
Siam Weed is a woody shrubthat thrives in the warm, humid and low altitude lands in a regionroughly spanning 30o north and south of the equator. In its nativehabitat, it presents little if any threat to cultivated crop landbecause it is easily kept in check by competing plants and effectiveinsect enemies. But late last century, Siam Weed began an outwardmigration, hiding in the cargo holds of ships bound for India andbeyond, where natural environmental restraints did notexist.
By 1905, it began itssweep throughout the Bengal region in India into Burma, Thailand,Bhutan, and Indo-China, and southward in the Ghats and Sri Lanka(then Ceylon). By 1920, the plant was reported in Malaysia. By 1944,cyclone winds and activities associated with World War II had spreadthe weed throughout the Philippines, Sumatra, Java, Taiwan, southernChina, and many of the islands of Micronesia and Melanesia. In the1940s, contaminated seeds from Sri Lanka were imported into Nigeria,giving Siam Weed a foothold on the African continent where, in ashort 20 years, it spread from Cote d'Ivoire to South Africa.
'Killerof the Forest'
Today, the pest is foundfrom Brazil to Northern Australia, from Guam to Nepal, where thefarmers there have named the plant "banmara," or killer of theforest. Genetically aggressive, Siam Weed easily outcompetes nativeplants and overruns cultivated farmland and grazing lands as well asyoung forest plantations such as teak coconut, and rubber. Impactedcrops include such important commodities as coffee, mango, maize,sweet potato, cashew, palm oil, citrus, tea, cassava, sugar cane,rice and cardamom.
Livestock browsing on SiamWeed will sicken and die due to the plant's high nitrateconcentration, nearly six times above toxicity levels. Farmers whoattempt to fight the weed by hand often complain of skin rashes andallergies. One of the tragedies of the Siam Weed invasion is that ithits subsistence farmers in some of the poorest countries of theworld.
Without extensivehand-weeding, Siam Weed will quickly blanket fallow fields andpastures. Even management by fire is ineffective because the flamesleave the plant's root ball intact, allowing it to be one of thefirst plants to regenerate during the next season's rains. Indrought-prone countries, Siam Weed presents a constant fire hazardwhich can sweep flames into virgin parts of the forest. And if thestory isn't grim enough, Siam Weed grows into dense tangled thickets,perfect hiding places for marauding pigs, bandicoots and evenelephants who later disrupt farms and plantations.
AnIntriguing Research Subject
Fortunately, Siam Weed isan intriguing research subject for entomologists, horticulturists,and other scientists in Europe, the United States, and Australia. Forexample, the plant has the interesting capability to temporarilychange the chemical content and color of its leaves when under attackby the enemy Arctiic moth, Pareuchaetus pseudoinsulata. Thecaterpillar stops feeding on the changed leaves, but once it has fledthe plant, Siam Weed reverts back to its original chemicalcomposition and color.
One of the frustrationsfaced by Siam Weed researchers has been deciding upon an effectivemeans to communicate their findings. Researchers are scatteredthroughout the world, including France, Germany, the United States,the United Kingdom and even Norway. French and English are thedominant languages for the researchers, but their target audiencesused literally dozens of languages. Government officials, universitycolleagues and, of course, the affected farmers in the besieged areasare often without the technological or financial resources to availthemselves of the growing body of scientific data on SiamWeed.
In 1988, Dr. R. Muniappanof the University of Guam, one of the leading researchers intobiological control of Siam Weed, helped organize an internationalworkshop in Bangkok on the control and management of the pest. Sincethen three additional workshops have been held -- Indonesia in 1991,Cote d'Ivoire in 1994, and India in 1996. South Africa is thetentative locale for the 1998 workshop.
Efforts are made to placethe conferences near affected farmers. The United Nations' Food andAgriculture Organization and other international agencies haveassisted with travel funds for participants from the DevelopingWorld. But typical of most informal research organizations, the SiamWeed researchers have little but their own grant funds to drawupon.
ReducingCosts and Rivalries
Dr. Muniappan, now in his28th year of research into biological control, also has an interestin computers and the Internet. In mid-1996 he and I hit upon the ideaof placing the many proceedings, newsletters, research reports,photographs and other material relating to Siam Weed on a web site.Frequently, the production of such print materials becomes boggeddown with translations (primarily French to English) and the costs oftypesetting, printing and mailing.
If the materials were madeavailable via the Internet, we could circumvent some of theseobstacles, as well as reduce the small rivalries that tend to arisein such diverse international organizations. But the University ofGuam, a 3,500-student campus serving Guam and the freely associatedstates of Micronesia, is hardly a magnet for Internet activity ortropical entomology. Most U.S. web sites dealing with entomology andthe other environmental sciences, such as those at Colorado State andOklahoma State universities, tend to focus on issues of thetemperate, not tropical, world.
Dr. Muniappan placed aquick phone call to his close colleague, Dr. RachelCruttwell-McFadyen at the Cooperative Research Centre for TropicalPest Management, a joint venture involving three Australiangovernmental agencies and the University of Queensland. Shegenerously offered space on her unit's web server, which is one ofthe busiest Internet sites in the world for information on tropicalpest management. Easy as that, we were in business.
Taking aDifferent Design Approach
The Internet and those whouse it are, by and large, products of the United States. Sitedevelopers tend to push the envelope of technology and bandwidth byadding extras like video and audio files, Java and CGI coding, andpush-pull server technology. Our small web team went in nearly theopposite direction when designing our site. We knew from previousfield experience that most of the computers available in the researchstations and field offices in the Developing World would be PCplatform and older technology, probably 386s and 14.4-baud or slowermodems.
As design leader, I alsoassumed our target audience would be using older versions of Internetbrowsers. Some would probably even be using text-only browsers, as wedo at the University of Guam. As I worked to devise a style sheet forour new web site, it became readily apparent that most of the designstandards for web work, as espoused by such popular web designers asDavid Siegel and Lynda Weinman, would be difficult to apply toeducational or scientific web sites.
By definition, the SiamWeed site is heavy on text. Editing articles was not an option --there would be far too many authors to consult -- so navigationalelements, such as the use of anchors at the top of each article, wereimportant to guide viewers to information. Also, we had limitedflexibility on how much we could "repurpose" the print material toadapt to the dynamics of the new web medium. The proceedings had beenavailable in print format for years and were logically organized bypresenter and topic.
We chose to focus oncontent over interactivity and hope researchers who wished tointeract would use the e-mail addresses we included for workshopparticipants. For a few months we did offer an online registrationform for the Fourth International Workshop in Bangalore, India, butreceived no online responses. (Most workshop participants came fromthe Indian subcontinent and they simply mailed or faxed theirregistration forms to the local organizing committee.)
Text preparation wasfairly straight-forward but highly laborintensive. The proceedingsand newsletters, some of which had been typed manually by workers inIndia, did not exist in electronic text format and so had to bescanned via optical character reading software and then heavilycopy-edited. Graphs and maps were scanned and saved as graphics (GIFformat), but most charts and other pieces of critical data wererebuilt in the HTML (Hyper Text Markup Language) format.
DownloadTimes are Critical
The site is also quitestatic, updated perhaps twice a year as compared to U.S. commercialstandards of once or twice a month. Download times for individualpages on our site would also be critical. Some countries pay forInternet service by the minute. For example, Guam's neighbor to thewest, the Republic of Palau, pays nearly $3 per minute to access theInternet. Others access the Internet for only a few hours a day viasatellite linkage. Even populous countries such as Malaysia offer asfew as three Internet Service Providers, compared to several thousandin the United States.
Our maximum goal was thatall pages should load in 45 seconds or less on a 14.4-baud modem. Wekept to a minimum the use of invisible tables, a common trickdesigners use to build in visual organization or columns on a webpage. Many older graphic-interface browsers and text-only browsersfail to display tables well. Subsequent site updates have usedtables, however, because they provide such an efficient andconsistent means to place navigational elements on a webpage.
We were fortunate to haveexcellent quality photographs of Siam Weed and some of the morepromising insect "control agents" now under study. The photos werecarefully digitized using a 216-color "browser-safe" color palette tohelp standardize how the images would appear across computer platformand browser. Several of these "browser-safe" color palette areavailable online as freeware. (Check Lynda Weinman's site atwww.lynda.comfor starters.) They can be implemented by converting a chosendigitized photograph to "index" color mode, then choosing the newcolor palette as the color guide instead of your computer's defaultsystem palette. In some cases, we took an image that was originallyfour megabytes in size on a straight scan and reduced it to 30K or40K, an acceptable download size.
I did take one designindulgence. Headlines for the main sectional pages were created in anillustration program using a crisp typeface, which viewed well onscreen; these were saved as GIF graphics to be placed on theirrespective page. Viewers using a text-only browser would not be ableto view these headlines or any other graphics, however, so filling inaccurate information in the ALT image tag on all graphics was amust.
We also followed a strictnaming convention for the pages, both to assist search enginesbrowsing the site and to ease the cross-platform shuffle since thepages were being created on Macintosh computers but mounted andviewed on various computer platforms.
Pre-TestingLeads to Improvements
The first version of thesite was mounted in late 1996. We did extensive pre-testing to ensurethe site looked acceptable, although not necessarily beautiful, on awide range of machines, browsers and modems. The second site update,completed in August 1997, included a better navigational scheme sincethe site now contained more than 230 pages of full-text articles fromthree proceedings and nearly four years worth of newsletters andother materials.
We decided to nottranslate articles due to time and fiscal constraints, so theexisting site is bilingual, but not cross-translated, in French andEnglish. Also the traditional standard of italicizing the scientificnames of plants and insects was omitted since italicized type, ingeneral, displays poorly on computer monitors.
In the 18+ months the sitehas been operational, it has not exactly been a ripping traffic hubon the Internet. But comments do roll in somewhat steadily fromresearchers around the world who are pleased to have found such acomprehensive and immediate source of information. One note inparticular seemed to capture the irony of our project. It was from aresearcher in India who accessed the site via a computer at hisuniversity, but his "thank you" note to us was typed on a manualtypewriter on newsprint.
Humble as it is, the siteis a wonderful example of international teamwork. It easily met itstangible goals of distributing scientific information on Siam Weedinexpensively and to a diverse international audience. We hope itserves as a useful weapon in curtailing the spread of Siam Weed.
The site was created onApple Macintosh computers -- LC 575 and Performa Power PC 6116CD.
Linda Austin is anExtension Agent/Media Specialist assigned to the College ofAgriculture and Life Sciences at the University of Guam. Formallytrained in journalism, graphic design and international relations,she now specializes in designing web sites for educational audiencesand teaching educators how to build better web sites. Email:firstname.lastname@example.org
Products and companiesmentioned:
Adobe Photoshop, PageMill,Illustrator; Adobe Systems, Inc., Mountain View, CA, (800) 833-6687,www.adobe.com
OmniPage Direct; CaereSoftware, Los Gatos, CA, (800) 535-SCAN, www.caere.com
This article originally appeared in the 01/01/1998 issue of THE Journal.