Vegetation classification and mapping: Lava Beds National Monument

DiPaolo, Dominic; Smith, Sean; Odion, Dennis; Ives, Jace; Sarr, Daniel

The National Park Service Vegetation Mapping Inventory Program funded an effort, beginning in 2010, to classify and map the vegetation at Lava Beds National Monument. Southern Oregon University and Klamath Network staff performed the inventory work. The Klamath Network coordinated the effort. To classify the vegetation, we sampled 169 representative vegetation classification plots throughout the 18,936 hectare (46,792 acre) project area using traditional phytosociological methods (i.e., relevé sampling). The relevé plots were located and sampled primarily in 2010 and 2011. We identified a total of 222 vascular plant species in the relevé plots, including several species that were not previously recorded on the monument’s species list. These were added to the flora of the monument by Sean B. Smith published in 2014. We classified the plot data into 24 plant associations using standard vegetation analysis methods. We provide a key to the plant associations and descriptions of each association in Appendices B and D of this report. Mapping of the vegetation polygons was completed via manual digitizing of vegetation types visible on the most recent National Agriculture Imagery Program (NAIP) and ESRI World imagery. A number of plant associations could not be consistently distinguished in the imagery. Based on our analysis, what could be distinguished accurately on the imagery, and a minimum mapping unit of 0.5 ha, we defined 15 vegetation mapping classes that were related to vegetation associations. Up to two of these vegetation classes were mapped as present in each polygon, and the percent cover that each one occupied in a polygon was coded in the database. The percent cover attribute for the primary and secondary (if present) vegetation in each polygon was estimated from photographic and field evidence. In addition, the occurrence of up to two associations and their percent cover were identified as composing the primary and secondary vegetation. We conducted map accuracy assessment in 2013 by visiting 505 stratified random accuracy assessment plots throughout the project area. The field data were compared to the vegetation map to determine accuracy. Our initial map had an overall thematic accuracy that was slightly below the 80% standard set by the National Program. To improve accuracy, we merged two vegetation types resulting in an overall thematic accuracy of 85% for the primary vegetation. Accuracy was difficult to assess for some rare types. We also assessed the accuracy of the primary association in each polygon, which was 82%.

Vegetation classification and mapping: Lava Beds National Monument

Series

Collections