Objective  Spatial and temporal dynamics of and the driving forces behind the agricultural carbon emissions in Fujian were studied.

  Method  The agricultural production data from the Statistical Yearbook of Fujian Province (1990-2016) were compiled to calculate the agriculture carbon emissions, the Geographic Information System (ArcGIS) applied to analyze the spatial variations, and the logarithmic average di-model (LMDI model) employed to decipher the factors that governed the outcomes.

  Result  Overall, the total emission in the province was on a downward trend from 1990 to 2016 with at an annual decline rate of 0.35% from 598.32×10⁴t to 546.69×10⁴t. The emissions differed significantly among 9 prefecture-level municipalities. Nanping had the greatest amount that reached 103.64×10⁴t, and Xiamen the least at 5.29×10⁴t. Materials and land for agricultural operations were the major pollution sources, accounting for 43.85% of the total emissions. They were followed by the CH₄ emissions from paddy fields which contributed 43.04%. The factors that drove the emission reduction included carbon intensity, employment structure and local population, each contributed 35.00×10⁴t, 8.86×10⁴t and 75.72×10⁴t, respectively, to the total. In contrast, the income from agricultural activities was a negative driving force that added 40.45×10⁴t to the total emissions.

  Conclusion  In recent years, the agricultural carbon emissions in Fujian have been reduced overall, and Fujian can further take measures to effectively reduce agricultural carbon emission and to promote low-carbon agriculture development.