The spatiotemporal patterns and dynamics of grassland established in the US conservation reserve program (CRP)

The significant cropland expansion in the United States (US) has resulted in many environmental issues, such as soil erosion, biodiversity loss, soil and water pollution, etc. To mitigate adverse effects associated with cropland expansion, the United States Department of Agriculture (USDA) implemented the Conservation Reserve Program (CRP) in 1986, one of the US’s largest and most well-known acreage-reduction programs, to alleviate the environmental cost. However, the lack of time-series maps illustrating the spatial distribution of CRP constrains the evaluation of environmental benefits derived from the program. In this study, through integrating CRP statistics and multiple satellite-based land cover datasets, we developed a knowledge inference-based approach by considering CRP practices, contract period, and soil erosion effect to reconstruct CRP grassland maps. The map products developed here are designed to approximate the actual distribution of CRP grasslands, enabling quantitative assessment of CRP outcomes while reserving privacy and proprietary interests of individual farmers. The developed datasets cover the conterminous US from 1986 to 2020 with a resolution of 1 km by 1 km. Using CRP ground points across the US from the Rapid Carbon Assessment (RaCA) program in 2011, we validated the developed CRP map, achieving an overall accuracy of 90 % within a tolerance of 5 km. The results indicate CRP grasslands expanded from 0.60 Mha in 1986 to 10.19 Mha by 1995, then remained stable at around 10.00 Mha for the next decade. However, the area declined steadily to 5.78 Mha by 2020, driven by rising crop prices and the reduced CRP enrollment cap. Spatially, the areas with high CRP percentages were observed in the Midwest and Northwest, the west of the Southern Great Plains, and the north of the Northern Great Plains. In addition, the developed maps include two kinds of CRP grass practices, “the grasslands converted from the former cropland (CPCP)” and “the existing grasslands that were previously replanted from cropland (CPEG)”. We further found that 9–11 % and 12–24 % of CPCP and CPEG have enrolled in CRP twice during 1985–2020, respectively, which may lead to different potentials of CRP-associated carbon sequestration and greenhouse gas reduction compared with one-time or continuous CRP lands. Overall, the developed time-series maps can serve as a good reference to identify the spatiotemporal changes of CRP grasslands and to inform the modeling studies for assessing the CRP-associated environmental benefits and supporting future policymaking.