Influence of chemical properties of soils on methane emission from rice paddies

Abstract

The relationships between methane (CH4) emission from flooded rice paddies and soil chemical properties were investigated using eight different soils in a pot experiment. Since CH4 is produced in paddy soil microbiologically when reducing conditions are sufficiently developed, the amounts of oxidizing agents including free iron (Fe)(III), amorphous Fe(III), easily reducible manganese (Mn), nitrate (NO3 ‐), and sulfate (SO4 2‐), and indexes of reducing agents including total carbon (C), total nitrogen (N), and easily decomposable C, were measured as possible decreasing and increasing factors in CH4 emission. The seasonal variations in CH4 emission rates were similar in pattern among the soils used. However, the amount of emitted CH4 varied largely, with the maximum total CH4 emission (from a brown lowland soil, 1,535 mg pot‐1) being 3.8 times that of the minimum (from a gley soil, 409 mg pot‐1). No correlation was found between the total CH4 emission and any single factor investigated. However, a statistically significant equation was found through multiple regression analysis: r=‐2.24x102 a+2.88b+6.20x102; r 2=0.821; P<0.01; where Y is the amount of emitted CH4 (mg pot‐1), a is the amount of amorphous Fe(III) (mg pot‐1), b is the amount of easily decomposable C (mg pot‐1), and r 2 is a multiple correlation coefficient adjusted for the degree of freedom. The amendment of ferric hydroxide [Fe(OH)3] to a gray lowland soil significantly decreased the CH4 emission from 1,099 to 592 mg pot‐1. This measured amount agreed well with that estimated from the above equation, 554 mg pot‐1.