W. Dean Hively
Laila A. Puntel
John Shriver
Alison N. Thieme
Daniel K. Manter
Jennifer M. Moore
Bonnie M. McGill
2025
<p><span>Conservation stakeholders looking to quantify the impact of their investments to increase soil health practice adoption over time often face challenges in interpreting practice adoption data due to discrepancies in language and results among data sources. Similarly, efforts to estimate environmental outcomes of practice adoption, such as water quality and greenhouse gas emissions, can vary depending on different practice adoption input data. To help make sense of different adoption data sources, we compared county-level adoption data for winter cover crops (WCC), no-till (NT), and reduced tillage (RT) in three areas of the United States with contrasting climates and production systems: central Illinois (CIL), southern Illinois (SIL), and western New York (WNY). We analyzed data available during 2015 through 2022 from the Operational Tillage Information System (OpTIS, remote sensing), US Census of Agriculture (AgCensus, a farmer survey), and, specifically in Illinois, the Illinois Soil Conservation Transect Survey (Transect, a roadside survey). The magnitude of differences between the datasets depended on the practice and geographic location. For example, OpTIS and AgCensus tillage data were much more similar in Illinois (average difference of less than 4 percentage points) compared to New York (average differences of 20 percentage points). Similarly, there was less variability and smaller differences between OpTIS and AgCensus WCC data in Illinois compared to WNY. AgCensus tended to report lower WCC adoption for Illinois and greater adoption in WNY compared to OpTIS. All data sources agreed that the rate of change in tillage practices is slow (mainly –1% to 1%) and that adoption of WCC is low (assuming linear growth, it could take nearly a century to reach 50% WCC adoption in CIL). Differences among the datasets were attributed to definitional inconsistencies for RT and NT and how WCC data were acquired. For example, the AgCensus asks if a WCC was planted, whereas OpTIS and Transect evaluate the presence of a standing WCC. Data sources also reflect different time periods (calendar years or crop years) and types of cropland assessed (corn [</span><i>Zea mays</i><span> L.], soybean [</span><i>Glycine max</i><span> {L.} Merr.], or all cropland). We propose two recommendations to improve interpretation and consistency: (1) a working group to harmonize definitions and protocols and develop educational materials for data users, and (2) a research effort that integrates different adoption data types and produces publicly available adoption data at HUC-10 and county scales. Such activities could help improve data access and utility for evidence-based conservation decision-making and enhance the accuracy of environmental models that rely on adoption data as input.</span></p>
application/pdf
10.1080/00224561.2025.2580218
en
Taylor & Francis
What is the (real) rate of soil health practice adoption? Making sense of three data sources
article