Vegetation Mapping in the American Southwest - Stages of a Vegetation Mapping Project

In order to ensure consistency among vegetation maps from numerous parks, the NPS and USGS have established a standard mapping methodology. The steps are described here.

Plan, Gather Data, and Coordinate Tasks

To begin the mapping process, park managers bring together all of the parties who will be working on the mapping project and outline the tasks to be accomplished. If a park has special management issues or concerns that a vegetation map might address, those concerns are integrated into the planning process. Because the vegetation mapping process relies on effective coordination between different agencies and organizations, a thorough plan that accounts for all tasks and assigns responsibility for each one is essential. Also at this stage, park managers determine the boundary of the area to be mapped, and they begin gathering the imagery that will provide the basis for the final map.

Survey Park to Understand and Sample the Vegetation

Prior to beginning the fieldwork, project members gather a preliminary list of vegetation associations and alliances from the NVCS based on previous studies and local knowledge. This list is used to set targets for data collection, with each association marked for 3-5 plots. Botanists also use this list in the field when they survey the entire site for representative stands of vegetation.

Once a stand of vegetation is located that appears to be representative of the plant communities in the area, a study plot is laid out to capture its characteristics. Pertinent information includes: elevation, slope, aspect, landform, topographic position, soil texture and drainage, hydrologic (flooding) regime, and evidence of disturbance or wildlife use. Vegetation structure and species composition are sampled using plots that vary in size, depending on the dominant physiognomy of the vegetation.

Within each plot, botanists visually divide the vegetation into strata, and the height and canopy cover of vegetation is estimated for each stratum. Physiognomic class, leaf phenology, and type of dominant stratum are recorded. The species of each stratum are then listed and percent canopy cover estimated using a cover scale (<1%, 1-5%, >5-15%, etc.). Additional species within the vegetation unit that occur outside of sampled plots are listed separately.

Researchers use a Global Positioning System (GPS) unit to record the plot location. Universal Transverse Mercator (UTM) X-Y coordinates and elevation are recorded both manually and stored as waypoints in the GPS unit. Finally, four representative photos in each cardinal direction are taken of the plot to maintain a visual record and for potential monitoring purposes.

Classify the Vegetation in the NVC using the Field Data

After gathering field data from plots distributed throughout the park, researchers input the data into a national database. The data are checked for consistency, and obvious errors are corrected. In some cases, statistical analyses, including descriptive statistics, ordination, cluster analysis, and other analyses, are used to determine appropriate groupings of plots and vegetation type designations.

After analyses are complete, the database is used to match each plot with a corresponding vegetation association. Any plots that do not have a clear association are analyzed further and may be used to identify new associations. All data are used to further refine and improve the vegetation classifications used by the NVCS. Ultimately, a final set of associations is identified in each park. A consistent, ecologically based vegetation classification forms the foundation for the development of an information–rich vegetation map (Muldavin et al. 2010).

Acquire Current Imagery and Translate Vegetation Types into Recognizable Map Units

Crispleaf Wild Buckwheat Mancos Shale Sparse Vegetation, Mesa Verde NP.
Crispleaf Wild Buckwheat Mancos Shale Sparse Vegetation, Mesa Verde NP (ground photo [top], aerial photo [center], and photosignature [bottom].

from Thomas et al. 2009

If necessary, NPS acquires new imagery or aerial photography that is current and has high resolution and accurate color. This imagery is entered into an ArcGIS, a Geographic Information System (GIS) program, and merged to form a continuous image of the park and surrounding area. Using standard photointerpretation (9x9 inch photos with stereoscope) or computer analysis (sometimes a combination of the two), the imagery is analyzed and distinctive patterns and textures are identified. The imagery is segmented according to these distinct patterns, creating a map that is a mosaic of interlocking sections.

Before they are matched with vegetation associations, each pattern type is “ground-truthed” by botanists working in the field. During this process, photointerpreters often collect “observation point” data, abbreviated samples of vegetation and environmental characteristics, to assist in the determination of vegetation types and their distribution across the landscape in map units.

Finally, the map unit associated with each vegetation type is used to code the appropriate sections of the map. Sometimes, when it is not possible to distinguish certain plant associations on the imagery, more than one plant association will comprise a map unit. The resulting map clearly displays the location and distribution of each map unit present in the park (see images at right).

Assess the Accuracy of the Final Map Product

Once the map is completed and finalized, an accuracy assessment (AA) is conducted. This is a four-step process consisting of a sample design, sample site selection, data collection, and data analysis. The design of the AA process selects random points placed in each map unit based on respective map unit frequency and area. Once the target locations are selected, field technicians are provided with draft field maps, overview maps, map unit definitions, the key to the associations, and digital GPS files containing the location of the target AA sites.

Field technicians travel to the AA target sites and, using the key, determine the vegetation association. At each target, they record the primary and secondary associations that occur within a roughly 50-meter radius. They also record dominant stand stratum (herbaceous, shrub, woodland, forest), environmental data, and percent canopy cover of the major species, as well as other nearby vegetation types outside the 50-meter radius. Data from this survey are cross-checked with the vegetation map to determine the accuracy of the map. The NPS has a minimum standard of 80% accuracy for all vegetation mapping projects.

 

Prepared by Ben Gillock, Patty Valentine-Darby, and Anne Cully, 2009.

Part of a series of articles titled Vegetation Mapping in the American Southwest.

Last updated: July 1, 2015