JARS v49n2 - Modern Mapping Technology Aids Plant Hunters

Modern Mapping Technology Aids Plant Hunters
Clarice Clark
Puyallup, Washington

Rhododendron literature abounds with references to "The Land of Deep Corrosions" as described by Frank Kingdon Ward in his book The Land of the Blue Poppy . This part of China, where the Yangtze, Mekong and Salween rivers flow through deep parallel gorges, was mapped in the early part of the century by explorers like Kingdon Ward and Joseph Rock. Rock's maps were not made to guide adventurers following in his footsteps, but rather to illustrate accounts of his journeys published in the National Geographic Magazine . Yet modern expeditions still rely on his 1947 maps to find routes on the Mekong-Salween divide.
My husband Jerry Broadus, a licensed land surveyor, and I longed for the chance to contribute to the knowledge of the genus Rhododendron by mapping its distribution using modern technology that would provide greater accuracy. We use miniature computers in our business to electronically collect and store measurements in the field. Our opportunity to map in China came in the spring of 1994 with an offer to accompany a group of knowledgeable scientists, growers and collectors to this area rich in history and varieties of rhododendron.
Mapping benefits both the scientist and the amateur enthusiast. Botanists need accurate and detailed information about the geographical distribution of a species to supplement the morphological characteristics that they study in order to classify a specimen. The Cullen and Chamberlain revision of Rhododendron features pages of maps that are the result of years of intensive data gathering, and these are essential to understanding the system. Yet their typical map covers hundreds of square miles, and even the symbols that "pinpoint" the location of the specimens cover 5 square miles! In most cases they are working with field notes that, at best, give a location to within 1 minute of latitude. For example, George Forrest's type specimen of Rhododendron proteoides was found at 28 degrees 12 minutes north latitude, at an elevation of between 12,000 and 13,000 feet. Translated on the ground in this part of the world, that gives us a search band not quite 2 kilometers (1 mile) across. Since Forrest did not specify which side of the divide he was on, we could have been within a half mile when we took a lunch break above our camp at Xia Zhi Na.
Horticulturists and collectors want all the information on rhododendrons they can get. Growers consult maps, or at least the data obtained from them, to learn about the plant's natural environment. Collectors study maps and notes of previous collectors' efforts to narrow down the range of possible species that they will encounter in the wild. Knowing what plants to expect saves time and directs their efforts towards promising habitats. Any serious expedition to rhododendron country is interested in maps, and a search for up-to-date maps of areas such as Western China is very frustrating. Recent maps of the China-Tibet border are either of a small scale with very little detail, or are unobtainable because of the politically sensitive nature of the area.
Every mapping project has a budget, with limitations on time, equipment and other data collection restraints. This expedition was scheduled for 21 days on foot or horseback, traversing steep terrain noted for torrential rains. Our baggage would be stuffed into bamboo baskets and loaded onto the backs of ponies bushwhacking through a remote area without electricity. Any thoughts of lap-top computers for storing our field data were quickly put aside.
We chose as our data collection tool an inexpensive Global Positioning System (GPS) receiver that was lightweight and waterproof. We were able to borrow a bright yellow marine version called the Trimble Ensign that ran for days on standard AA batteries. These mass-produced devices have become a common part of expeditions, and the data they collect is shared around the world. The Kunming-Gothenburg Botanical Expedition to Yunnan in the fall of 1993 took along a similar device and located extensive thickets of R. proteoides on the Pei Ma Shan plateau. By entering their coordinates into our receiver, we hoped to retrace their steps to a search area about the size of a football field. This is a much more reasonable area to try to cover than the hundreds of acres that Forrest's field notes make us search. The receiver will calculate and display the distance and direction that you need to travel (as the crow flies). Unfortunately, we were denied access to the plateau by local officials, part of the bad luck that can accompany any journey.
Understanding the limitations of these instruments will help you understand the information they give you. GPS receivers use radio waves to communicate with a system of 24 satellites circling the globe in very high orbits. The receiver calculates its location by measuring the exact distance to more than one satellite. Imagine drawing circles with a simple compass. If you do your geometry right, you can produce a very small area where three or more circles overlap. The satellite signals make similar circles on the earth's surface, and the receiver computes its location by finding the center of the zone of overlap.
You need to get the data from four satellites to obtain a three-dimensional fix (latitude, longitude, plus elevation). A single civilian GPS unit will only produce readings with a 95 percent certainty of plus or minus 100 meters (300 feet). This is because the U.S. military degrades the signaling code as a national security measure. Greater accuracy is possible using two units, one of which must be left over a known point, or by using a military grade receiver. Our mapping budget did not allow for the former, and military receivers are unavailable to the general public.
You must have a reasonably clear view of the sky - the receiver cannot work underneath thick tree cover - and have a few minutes to stay in one spot. The speed at which you "lock onto" the satellites will depend on whether you are in a narrow valley or on a high ridge. When you take a receiver to a new location, or start a new receiver for the first time, it has to find a satellite at random and read its almanac - a message that tells your unit how well the satellite system is functioning. This process, which may take 15 minutes or more, is called "initializing," but subsequent reading should go more quickly. You have no real control over the satellite geometry, except to wait or to move to a more open location. We tried to wait for a reading from at least four satellites and wrote down the data immediately in a standard waterproof paper field book. Our unit was capable of storing up to 100 readings, but we were concerned with battery power and knew from our surveying experience it was better to be safe than sorry. We took altimeter readings and photographs of the site of each reading. We needed to name them and determine the direction of drainages and exposures, so we took magnetic compass readings and made sketches of the locale.
It's a good idea to take repeat readings at locations you revisit, like trail intersections, on the way in and again on the way out. Then when you plot your readings on a map, you can catch your blunders. There was ample opportunity to take repeat readings on this trip because late snows lingered in the passes and blocked our efforts to cross over the divide into the Salween drainage. As late as June 4, we had to kick steps in the snow to reach the Zhi Dzom La and could not entice our ponies to accompany us. Our consolation was blooming rhododendrons - the R. beesianum just couldn't wait any longer, and pink flowers glowed in the cold fog.

1994 Mekong Rhododendron 
Expedition map
An overall map of the expedition's route from the Mekong River to the passes on the divide above the Salween River.
Squares labeled 1 through 5 are details showing the species found in that area. The contour lines are 200 m (approx. 650 ft.).
Map courtesy of Jerry Broadus

With one week remaining, we conferred with our Chinese and Tibetan guides over a map, looking for likely places to continue to search for rhododendrons. Our requirements were a flat enough spot for camping, running water, and an exposure that promised less snow. Two days' walk found us in a valley to the south of the Dokar La, not far from the tiny village of Lungdre. Here we were able to reach the rhododendrons, and lots of them, including the elusive R. proteoides . About 300 meters (1,000 ft.) above our camp, we were delighted to find a cliff covered in these desirable dwarf plants just beginning to bloom. By roaming around in an area roughly 2 kilometers (1 miles) on a side, our party managed to catalog 24 species, subspecies and varieties. Our AREA 3 detail map illustrates this campsite, at an elevation of 3,800 m (12,500 ft.).

1994 Mekong Rhododendron Expedition, 
Area #3
Enlargement of detail area #3, showing location and key to abbreviation of species names.
The map represents an area 2 kilometers on a side, or about 1 square miles, and has 40 m (130 ft.) contour lines.
Map courtesy of Jerry Broadus

Returning home with all these wonderful facts and figures, we set about the job of making a map that would be accurate and useful. We had to translate the lines on our base map into vectors, by using computer software called COREL DRAW! This program in turn made a file that we could import into AutoCAD, another drawing program that could merge files with our surveying software. Then we could add the latitude and longitude coordinates from our notes and begin to rectify them with the contours. This procedure revealed that four of our GPS readings were out of tolerance and subsequently not used. Analysis showed these inaccurate fixes to be the result of some constraint such as geography or lack of time which prevented us from acquiring measurements from more than three satellites.
Then came a lot of decisions about mapping conventions and protocol. All maps involve choices and compromises. We had to select from myriad options for symbols, line weights and the size of the print. Our trail is the black line, dashed in the enlargement. Showing the location of the species involves abbreviation and placement so that other information is not obscured. Repetition of the abbreviation indicates a particularly abundant or noteworthy group of plants. We are reluctant to use other researcher's data and only show plants and trails logged on this trip. By leaving areas blank, we are not saying that there are no rhododendrons or trails. There are no doubt many, especially on and over the Dokar La, that are not reflected on our maps.
Even the spelling of place names turned into far more work than previously imagined. The Tibetan names have often been changed by the Chinese and then transliterated into the Pinyin system of Western alphabet. There are three or four spellings of the same mountain in as many references. For example, the name for the entire mountain range has been spelled Ka-gwr-pu (Royal Botanic Garden), K'a Gur Pu (Kingdon Ward), and Kang Karpo (modern usage). We generally adopted Joseph Rock's spelling, because it seemed consistent with our interpreter's valiant attempts to translate for us on the spot.
The last hurdle was how to reproduce the map. Our own computer-driven plotter could only produce colored lines on white paper. Our efforts seemed flat and lifeless. Most map users would expect hypsometric tints, a color scale commonly found on wall maps, using green for low elevations to brown or yellow to white for highest elevations. We did not adopt the brown tones as we felt they do not reflect the abundance of plant life found at high elevations at this latitude. With a great deal of help and access to an electrostatic plotter, we were able to create a background color and a more pleasing paper product. Paper would be the most practical for sending to other members of the expedition, some of whom may return to the area. We were also interested in a digital format for several reasons. Other researchers could directly import the data and combine it with new information to produce maps and reports. Also, each generation of reproduction on paper degrades the image, meaning the lines get fuzzy and the print gets blurred. Magazine publishers recognize this and request articles and maps on a disk which they can adapt to their favorite program and manipulate to produce a final product - all without touching any paper. Local efforts to translate our map into this format resulted in crashed hard disks and locked up terminals. It is such a huge file that it requires immense amounts of memory, and we are forced to fall back on paper as our only means of reproduction. Here was another endeavor foiled by the three goddesses of all projects: time, energy and software!
We hope our efforts inspire other rhododendron expeditions to add a GPS receiver to their data collection tools, whether it be to remote China or populations of native rhododendron in your state or province. A copy of our route map and detailed maps of all five study areas are in the Pierce Library at the Rhododendron Species Botanical Garden in Federal Way, Wash.

Acknowledgments
Jerry and I would like to thank Robert MacOnie, Jr., for countless hours of computer aided design work and his contributions to the final production of the map. We are also indebted to Steve Reding of Geoline and Ron McDuffie of Trimble Navigation for the loan of the Ensign receiver.
On behalf of the entire expedition group, we would like to express our appreciation to the Chinese Academy of Sciences Sichuan Scientific Service Corporation of Expeditions for organizing the trip.

References and Recommended Reading
Dahl, Bonnie. The User's Guide to GPS. Evanston, IL: Richardsons' Marine Publishing; 1993.
Monmonier, Mark. How to Lie with Maps. Chicago and London: The University of Chicago Press; 1991.
Chamberlain, D.F. Revision of Rhododendron II Subgenus Hymenanthes. Notes from the Royal Botanic Garden Edinburgh, Vol. 39, No. 2; 1982.
Cullen, J. Revision of Rhododendron I Subgenus Rhododendron, Sections Rhododendron & Pogonanthum. Notes from the Royal Botanic Garden Edinburgh. Vol. 39, No. 1; 1980.

The maps accompanying the article and the photographs on pages 86-87 first appeared in the December 1994-January 1995 issue of P.O.B. magazine, illustrating an article by Jerry Broadus for professional surveyors.

Clarice Clark, a member of the Tacoma Chapter, has also written about her observations of vireyas in the wild on the island of Borneo in the winter 1992 issue of the Journal.