Collection Mapping & GIS for Landscape Management Brian J. Morgan GIS Manager UC Davis Arboretum One Shields Avenue Davis, CA 95616

Introduction The UC Davis Arboretum is a 100 acre botanical garden located on the University of California, Davis campus in the central valley of California. It is considered a living museum with over 22,000 specimens, and is used by university students, staff, and faculty for teaching and research, as well as visitors for recreation and relaxation. The Arboretum has kept digital maps of its plant collections for close to two decades, and currently use them for a number of internal and visitor based uses. Institutional staff uses the maps, and the data contained in them, for planning and maintenance of the landscape and plant collections, and for curatorial functions such as cataloging the living specimens and sharing it with the scientific community. The original collection maps were created in a geographic information system (GIS) called Geo/Navigator in 1989. The information in this system was collected by students over a five year period, and was only considered relative in its accuracy. This means that features such as plants, irrigation, and amenities were only located relative to other features such as pathways and buildings. This lack of accuracy limited the maps use for site planning and utility management purposes. Several years ago, the software company that produced and supported the software went out of business, leaving the platform unsupported, and further limiting the use of the system. In late 2004 the UC Davis Arboretum received a Museums for America grant from the Institute of Museum and Library Services (IMLS) to migrate their collection maps and plant records to the industry standard software platforms of ESRI ArcGIS and BG-BASE respectively. This grant provided us with close to $150,000 over two years for the equipment and staff salaries necessary to remap their collections in a modern GIS using spatially accurate global positioning system (GPS) technology. This paper presents the methods the UC Davis Arboretum used to create a rich GIS for their collections, the information products that resulted from this work, and discusses the steps we are taking to make GIS projects easier for the botanical garden and zoo community. Methods One of the first things we considered when preparing to map our collections was the survey equipment we would use. Given the scope of our grant budget, we weighed purchase price, maintenance cost, accuracy, data collection time, ease of operation, and environmental conditions before selecting our equipment. These factors led us to the purchase of a Trimble Pathfinder ProXR GPS receiver which yielded sub-meter accuracy after post-processing. After using this equipment for some time, we upgraded to the newly available Trimble ProXH receiver which yielded far superior sub-foot

accuracy. We used these receivers in conjunction with a Trimble Recon ruggedized hand held computer to collect data. Due to the amount of dense tree canopy in our collections, we also purchased a LaserCraft XLRic laser rangefinder for offset mapping. When this piece of equipment is mounted to a tripod with the GPS antenna, it enables you to map the locations of features that are in areas where GPS signals may be inaccurate or may not reach at all. After selecting the appropriate equipment for the task, we needed to decide which software we would use with it. We considered different combinations of software from Trimble and ESRI, and eventually settled on the optimal configuration that met our requirements for features and ease of use. On the hand held data logger, we selected ESRI ArcPad 7.0 and the Trimble GPScorrect 2.1 extension for it to allow for full functionality of the GPS receiver. On our desktop computer we selected Trimble GPS Analyst 2.1 extension for ESRI ArcGIS 9.2 to allow for post-processing of the GPS data we collected. We also selected MapLogic Layout Manager 2.2 extension for ESRI ArcGIS 9.2 for the creation of multi-page map books. Together this combination of hardware and software would allow us to collect accurate location data and use it to create sophisticated collection maps. The final step to take before we began collecting data was to select or design a geographic database, or geodatabase, to meet the needs of our project. We started this step by meeting with curatorial staff from the San Francisco Zoo, Huntington Botanical Garden, Los Angeles Zoo, and San Diego Zoo & Wild Animal Park to find out how they stored their collection map data. This process revealed to us that there was no existing successful strategy or standard for collection mapping with GIS, and we would need to create our own geodatabase design for our collection maps. After several iterations of design, testing, and revising our geodatabase design, we settled on one that met the data storage needs of the Arboretum. This design contained feature classes for amenities, irrigation features, utilities, signs, pathways, water features, planting beds, boundaries, structures, and of course, plants. This geodatabase design gave us the final element we needed to start collecting GPS data. To begin the field mapping portion of the project we hired four students from the UC Davis campus to perform the majority of the data collection tasks. These students worked in pairs to map the amenities, irrigation, plants, signs, and utilities of the 36 thematic collections of the garden. One student on each team was responsible for operating the laser rangefinder to obtain distance, compass bearing, and inclination measurements, the GPS receiver to obtain x, y, and z geographic location coordinates, and the data logger to record attributes for each feature. The other student from each team was responsible for positioning a reflector for the laser for each feature to be mapped, providing attribute information to the equipment operator, and notating the existing collection maps to keep a record of the work. Plants were either mapped as a center point for individual plants or as a polygon drawn around the perimeter of a group of plants. For each center point or individual species inside a mass planting, the accession number, taxonomy, growth habit, and radius were recorded. For each mass planting, the number of species, approximate number of plants, and the accession number for each species contained within was recorded. For all other features only the type of feature was recorded. In addition to this information, the equipment operators name and date were recorded for every feature mapped.

Once all of the features of a collection were mapped, the data was ready for the post-processing phase of the project, where features were imported into the geodatabase, differentially corrected to improve accuracy, filtered to remove errors, and verified for positional and attribute accuracy. During data collection, Trimble GPScorrect records satellite signal data for each position taken, and stores it for use in differential correction. Once in the office Trimble GPS Analyst uses this information, and the information recoded by other permanently fixed GPS receivers in the region to improve the accuracy of the positions recorded. This process is called differential correction. After correction, the program calculates the estimated accuracy of each position, and filters out the ones that are over a predetermined threshold. For this project, we set that threshold to two feet, meaning that any feature position that was not estimated to be within two feet of its true location was discarded. When this process was completed, the remaining features could be compared to the collection maps from the field, and the geometry and attributes could be verified. At the end of verification, any features that were discarded could be identified, and remapped as necessary. Once all of the features were collected, and had passed verification, the data was ready to be turned into a collection map. Using MapLogic Layout Manager software, templates were designed in ESRI ArcGIS for an overview map, detail map, and index. The overview map showed the entire collection with the boundaries and numbers of the detail map pages, an inset location map that showed the location of the collection in the entire Arboretum, scale bars, scale, north arrow, collection name, and page number. The detail maps contained a zoomed in area of the collection at a scale of one inch equals twenty feet, a twenty foot grid, along with scale bars, scale, north arrow, collection name, and page number. Landscape plan symbols were shown on these maps for every type of feature data was collected for. Plants were symbolized by their growth habits, and scaled using their measured radius. The index showed a list of the scientific names and accession numbers for each plant specimen in the collection, along with the page number and grid location it could be found on. All three of these templates were used together to create a landscaped oriented, letter size, map book for each collection, and were rendered in grayscale for ease of reproduction. Completed map books were then verified for accuracy by checking the locations of features in the field. Results After over two years of field mapping, post-processing, map creation, and verification, field mapping and post-processing was complete for 31 out of 36 developed collections, and map creation and verification was complete for 10 out of 36 collections. Collection map books are currently available to staff, students, researchers, and visitors in print format via a binder at the UC Davis Arboretum headquarters. The map books will soon be available on the Arboretum website in PDF format for easy downloading, viewing, and printing. The UC Davis Arboretum is committed to completing the mapping of their collections, and continuing their work with GIS. The Arboretum recently received a grant from the Slosson Endowment Fund for close to $50,000 to map the remaining collections, and make the maps available via their website. The work on this grant is scheduled to be completed by July 2008. The Arboretum is currently testing a prototype web application that allows for interactive use of the collection maps GIS. This web application was created using ESRI ArcGIS Server 9.2, and allows for

user account limited access to a variety of tools. Visitors can interact with the maps by zooming, panning, inspecting features, measuring, and querying. Staff can perform the same functions as visitors, but can also edit features and their attributes, create new features, and delete existing features. This application allows visitors and staff to use the most common functionality of the GIS without having to install or learn the software. It also saves a substantial amount of mapping staff time by reducing the amount of information and map requests they respond to. All collection map data is also available to staff on their handheld tablet computers, PDAs, or smart phones that run Windows Mobile through ESRI ArcPad. A simplified version of the collection maps can be accessed through ArcPad which allows users to view, query, and edit data. This functionality will be extended to standard cell phones with the release and integration of ESRI ArcGIS Server 9.3. The standard cell phone version will allow for the delivery of map based content to visitors in the form of tour and plant identification guides. Discussion The expanding use and access of GIS at the UC Davis Arboretum has motivated thoughts on using GIS as a method of information management for the entire organization. This way of using GIS, called enterprise GIS, is one of the central principles behind the Arboretum’s latest project to develop a GIS standard for landscape management. This project is currently funded through an IMLS Conservation Program Support grant for close to $230,000 over eighteen months. The project is a partnership between botanical gardens, zoos, and ESRI to develop an ArcGIS data model for landscape management that will provide organizations with a free template, or starting point, for implementing GIS projects. This template will be designed to meet the needs of organizations of different sizes and funding levels by being modular in its design, or allowing for light, medium, and full-featured implementations, and focusing on the features that provide the greatest benefit. The data model will also be designed for compatibility with distributed database protocols such as DiGIR and TAPIR. Compatibility with these protocols will assure that a garden or zoo can make their collection information available on the internet through biodiversity portals like GBIF. Work on this project began in September 2007 with a meeting called a technology seminar that was held on the UC Davis campus. This meeting involved participants from UC Davis, the UC Davis Arboretum, San Francisco State University, San Francisco Zoo, San Diego Zoo, and The Presidio Trust, and focused on determining how an enterprise GIS system could meet the needs of these users. This meeting was followed by a survey of American Public Garden Association (APGA) and Association of Zoological Horticulture (AZH) membership that focused on assessing the GIS needs and priorities of botanical gardens and zoos. Information from this survey will be used to determine the features that will be included in the data model. The results of the technology seminar and survey will then be presented at the first official design team meeting at the Missouri Botanical Garden in St. Louis, Missouri on November 10, 2007, and will be used to decide which features will be designed and their associated timeline. This meeting will be followed by several iterations of design, team meetings, and testing until the final model release date in June, 2008. The released model and supporting documentation will then be made available for free download from ESRI’s data model website.

ESRI has shown their commitment to supporting the GIS efforts of botanical gardens and zoos by announcing their startup grant program for botanical gardens and zoological parks. This program will provide these organizations with free software and training for a period of two years. ESRI (2007). ESRI has also agreed to publish two book titles on the subject to further aid these organizations. The first title will be targeted toward directors and administrators, and will focus on the return on investment that GIS can provide. The second title will be targeted toward GIS users, and will provide technical information on the data model. We feel that the combination of the forthcoming data model, the software and training provided by ESRI, and the two forthcoming book titles will provide botanical gardens and zoos of all sizes and funding levels the tools they need to launch a successful GIS project.

References ESRI, 2007. ESRI Startup Grant Program for Botanical Gardens and Zoological Parks, http://www.esri.com/grants/docs/bg_zp_grant_application_guidelines.pdf.