Submitted to: Society for Range Management
Publication Type: Abstract Only
Publication Acceptance Date: 10/1/2023
Publication Date: 2/26/2023
Citation: Martinez, P., Novak-Echenique, P., Stauffer, N.G., Maynard, J., McCord, S.E. 2023. Evaluating the sensitivity of ecological site identification in national monitoring programs. Society for Range Management. Abstract.

Interpretive Summary: We evaluate the level of error in ecological site identification at monitoring plots of the Assessment, Inventory, and Monitoring (AIM) strategy conducted by the Bureau of Land Management (BLM). We compare observer-identified ecological sites in the AIM dataset, expert-reviewed identification, and the ecological sites predicted using the Land Potential Knowledge System (LandPKS) soil matching algorithm which leverages information from national soil databases along with site-specific characteristics, such as GPS coordinates, slope, and soil observations (e.g., soil texture and rock fragment volume). We provide insights into the level of error in ecological site identification in response to uncertainties in soil and geomorphic descriptions in the field.

Technical Abstract: Ecological site information allows land managers to make informed management decisions that ensure rangelands are used within the bounds of their land potential. To provide information on ecological sites across public lands, the Bureau of Land Management (BLM) launched, in 2011, the Assessment, Inventory, and Monitoring (AIM) terrestrial strategy with the purpose of characterizing key ecosystem processes following standard soil, vegetation, and geomorphological protocols. Although the AIM terrestrial strategy delivered a large dataset with over 50,000 monitoring plots and 120,000 soil observations, it is still unknown the level of error (sensitivity) in ecological site identification in response to uncertainties in soil and geomorphic descriptions in the field. A better understanding of the level of error in the AIM dataset would ensure that data users (e.g., ecologists, rangeland managers, soil scientists, etc.) are aware of potential limitations in the dataset and can inform training of future field data collectors. Here, we compare observer-identified ecological sites in the AIM dataset, expert-reviewed identification, and the ecological sites predicted using the Land Potential Knowledge System (LandPKS) soil matching algorithm which leverages information from national databases (e.g., SSURGO and STATSGO2) along with site-specific characteristics, such as GPS coordinates, slope, and soil observations (e.g., soil texture and rock fragment volume). Our findings provide insights into the uncertainties in ecological site identification and can be used to improve future ecological site identification by field observers in the United States.