1. Introduction
In irrigated paddy fields, mineralization of organic nitrogen (N) in soil during a cropping period often limits rice growth and yield. Nutrient omission trials conducted in more than a thousand paddy fields throughout Japan have demonstrated that lowland rice responds to the application of N to a greater extent than to the application of phosphorus or potassium [
1]. To optimize the rate of application of N fertilizer, many researchers have proposed laboratory methods to evaluate soil N availability suitable for routine use, and some of them have been validated by comparing the results with the content of potentially mineralizable organic N estimated by the long-term laboratory incubation [
2]. Several biological and chemical methods such as short-term anaerobic incubation and extraction with hot KCl have proved useful for rapid estimation of soil N availability [
2,
3].
In addition to analysis of the labile fraction of N in soil, analysis of total N is also a widely accepted method for estimating soil N availability [
2,
4]. For example, in 31 surface paddy soils in Japan classified as non-volcanic ash soils, the content of potentially mineralizable organic N evaluated by long-term aerobic incubation was significantly correlated (
r = 0.53,
p < 0.05) with the total N content [
5]. By reviewing previous papers, Sahrawat [
6] also reported that the total N content in paddy soil usually showed a significant positive correlation with the amount of ammonium ion (NH
4+) produced under anaerobic incubation.
The total N content in soil has been measured by two methods: the Kjeldahl method which is a wet oxidation procedure, and the Dumas method which is fundamentally a dry oxidation (combustion) procedure [
7]. These methods were originally developed during the 19th century and the revised versions are still in use. Total N and total carbon (C) contents can be measured simultaneously by the dry combustion method. However, both methods require laboratory facilities, so total N analysis are typically carried out by specialists in well-equipped laboratories.
To reduce the time and cost for measurement, Sharifi et al. [
4] proposed the sodium hydroxide direct distillation method. In this method, the digestion step was fully eliminated, and the amount of N liberated by steam distillation of soil with a strong alkaline reagent for about 10 min was determined. The amount of N obtained from this method had a significant positive correlation with the content of soil total N measured by the dry combustion method. Christianson and Holt [
8] also proposed a rapid digestion procedure with H
2SO
4 and H
2O
2 (the peroxy reagent) without addition of metal catalysts such as Se, Hg, and Cu. The total time for digestion was only 38 min, and the rate of N recovery from six soils ranged from 89% to 98% compared to Kjeldahl digestion. However, these methods are still limited to the laboratory use, as they require hazardous reagents.
To develop a simple method for estimating soil total N usable by non-specialists, we recently proposed a method in which a dilute hydrogen peroxide (H
2O
2) solution is used as the soil extractant [
9]. In this method, surface paddy soil is extracted with a commercially available 3% H
2O
2 solution at 25 °C for 40 h. Then, the electrical conductivity (EC (H
2O
2)) of the soil extract is measured with an EC electrode. Because 3% H
2O
2 commercially available in Japan is called “oxydol” (a Japanese pharmacopoeia term), this method was called the oxydol method in the previous paper [
9].
The concept of this method is based on previous findings on the decomposition of soil organic matter with H
2O
2 and the detection of NH
4+ by EC. Robinson [
10] first introduced H
2O
2 digestion as a pretreatment of soil texture analysis. Considering the volatile nature of both H
2O
2 and CO
2 produced by the reaction of soil organic matter with H
2O
2, another Robinson [
11] proposed a gravimetric method for determining the content of soil organic matter by the loss in weight caused by H
2O
2 digestion. Robinson [
11] also reported that practically all of the soil nitrogen was transformed to NH
4+ during H
2O
2 digestion. This finding was confirmed by Harada and Inoko [
12], who reported that, for the eight soil samples tested, 59–100% of the total N in the original soil was present in the digested solution and that most of the water-soluble N was present as NH
4+. Guan et al. [
13] also reported that, for the eight samples including six volcanic ash soils, more than 90% of organic C and less than 11% of N in the original soil were lost by H
2O
2 treatment. The concentration of NH
4+ in a solution can be measured quantitatively by non-selective EC detection after its separation from other cations by ion chromatography [
14]. If NH
4+ is the dominant cation in a sample solution, the NH
4+ concentration can be roughly estimated from the solution EC without a separation pretreatment.
To validate the oxydol method, we applied it to 83 surface soils collected from paddy fields throughout Japan [
9]. Then, there was a significant positive correlation between EC (H
2O
2) and total N content (
r2 = 0.70,
p < 0.01) when 11 volcanic ash soils were excluded from the analysis. When we applied the method to two sets of paddy soil samples collected at farm scale, the correlation between EC (H
2O
2) and total N became different between the two farms (
r2 = 0.23 and 0.68). The low accuracy of the estimation for volcanic ash soils and some farm-scale samples was mainly due to the low rate of mineralization of soil organic N by the H
2O
2 treatment. This means that the estimation accuracy was decreased by the uneven decomposability of soil N within a sample set. The variations in both soil total N content and its decomposability at farm scale were smaller than those observed at national scale [
9]. At any scale of investigation, the estimation accuracy is expected to increase with an increase of the variation in total N content and also with a decrease of the variation in its decomposability. However, the number of sample sets collected at farm scale was too small to evaluate whether or not the oxydol method can be useful for estimating the farm-scale variation in soil total N.
Moreover, only one type of H
2O
2 sold in Japan was used in previous experiments [
9]. It is uncertain whether other types of H
2O
2 solutions can produce results comparable to those obtained previously. This is because commercial H
2O
2 solutions contain various stabilizers to minimize decomposition of H
2O
2 into water and oxygen under normal storage conditions. For example, the H
2O
2 used in the original method [
9] contained phenacetin as a stabilizer. Other stabilizers typically used in commercial H
2O
2 are colloidal stannate, sodium pyrophosphate, organo-phosphates, and colloidal silica [
15]. These stabilizers are expected to increase the EC of the commercial H
2O
2 solution and consequently the EC (H
2O
2) value; the estimation accuracy of soil total N from EC (H
2O
2) might be significantly affected.
The objective of this study was to further evaluate the oxydol method in terms of (1) the applicability to farm-scale samples and (2) the possibility of using different types of commercially available H2O2 solutions.