Potassium and phosphoric acid fertilization and planting density analyses were conducted across 31 fields in Tokachi District, Hokkaido, in 1990. The three test plots were arranged side-by-side in the same field, each with an area of 15 m² and designed in a fully randomized method, with no repetition. Seven levels of potassium fertilization [range, 0–249 kg-K ha⁻¹ (0–300 kg-K₂O ha⁻¹)], six levels of phosphoric acid fertilization [0–273 kg-P ha⁻¹ (0–625 kg-P₂O₅ ha⁻¹)], and planting density [50,000–90,000 plants ha⁻¹] were examined. To quantify the effect of fertilization and planting density on the yield and quality of sugar beet, these factors were varied, and linear regression coefficients were calculated for each studied field. Next, the correlation between these regression coefficients and soil analysis values was examined. Some significant correlation was found between these variables, and these signs of significant correlations were opposite between potassium and phosphoric acid fertilizations. A negative significant correlation was found between regression coefficients of potassium and phosphoric acid fertilization effects on the root weight and sugar yield.

The effect of potassium and phosphoric acid fertilization can be classified into the following three groups: group A, strong effect of potassium; group B, strong effect of phosphoric acid; and group C, strong effect of both potassium and phosphoric acid. Hierarchical cluster analysis was performed for the regression coefficients of potassium and phosphoric acid fertilization effects on the sugar yield. Therefore, we investigated a method to classify cultivation fields into the abovementioned groups using canonical discriminant analysis for soil parameters and developed a prototype for automatic calculation using Microsoft Excel.

This model is a field discrimination model derived from the field test of sugar beet, but it seems to be helpful in estimating the potassium and phosphate acidic characteristics of the field including other crops.

To measure water discharge, we instrumented 23 river basins ranging from 1.2 to 341 km² in a 1280 km² study area that included the Nasu alluvial fan (497 km²), lying between the Naka and Hoki Rivers, and surrounding areas in Tochigi, northern Honshu, Japan, for a period of 3 years and 4 months (May 2007 to September 2010). In tandem, we developed a model of the water dynamics of this area based on the tank model and used it to calculate water movement in the basins from actual meteorological data. The model estimated that subsurface flow production from the whole study area was 678 mm y⁻¹, corresponding to 59％ of the 1150 mm y⁻¹ difference between precipitation and evapotranspiration. The model calculated that subsurface flow was high in the mountainous northern portion of the study area and river basins located on the upper slopes of the alluvial fan, involving the Sabi, Kuma, and Hoki Rivers. The model also indicated the existence of a subsurface water storage zone in the upper portion of the alluvial fan (at an altitude of 250–350 m a.m.s.l) and that a large proportion of the flow into the basins in the middle–lower portion of the alluvial fan was from subsurface flow, involving the Yusaka, Momura, Kabuchu, Fuka, and Maki Rivers. The model results indicated that subsurface flow is the principal pathway governing water dynamics in the Nasu alluvial fan. The amount of subsurface outflow from the Nasu alluvial fan was 848 mm y⁻¹, considerably less than the inflow of 1630 mm y⁻¹, but measurement of the subsurface water movement from the Takaku colline, outside of the Naka River, is required before these figures can be confirmed. The amount of subsurface water used for paddy field irrigation across all basins was 725 mm y⁻¹, representing 72％ of the difference between precipitation and evapotranspiration and 44％ of subsurface inflow to stream channels, indicating that paddy fields exert a strong influence on the water dynamics of the Nasu alluvial fan.

We measured nutrient (N, P, K) and Cl concentrations in surface water for 3 years and 4 months in 23 instrumented river basins in and surrounding the Nasu alluvial fan. The procedures that calculate nutrient budget were built into the water dynamics model documented in Part I of this paper, and the dynamics of nutrient transport in the Nasu alluvial fan were interpreted. The calculated annual nutrient discharges from the individual basins were linearly related to measured values, with correlation coefficients ranging from 0.84 to 0.98 and regression line slopes ranging from 0.6 to 1.4. The predicted annual changes in nutrient concentrations in surface flow downstream of the whole study area were similar to the measured values. The annual N load from livestock and fish farms in the Nasu alluvial fan was estimated to be 36 kg ha⁻¹ y⁻¹, accounting for 62％ of the total input of 58 kg ha⁻¹ y⁻¹. The annual level of denitrification in paddy fields was estimated at 19.8 kg ha⁻¹ y⁻¹. The largest source of primary Cl input was mineral fertilizer at 125 kg ha⁻¹ y⁻¹, equivalent to 65％ of the total input of 191 kg ha⁻¹ y⁻¹. Nutrient concentrations in surface water tended to be lower in summer as a result of higher precipitation. However, in basins in the middle–lower areas, the N concentration of surface water increased in the summer–fall season, presumably due to increased input from subsurface flows containing high N concentrations. The discharge rates of nutrients from the whole study area were N 51％, P 6％, K 93％, and Cl 91％ of the annual load (resulting from processes such as organic binding, adsorption, and denitrification), and of the total, approximately 60％ was carried by subsurface flow.

Nitrogen mineralization in soil depends on many factors such as time, temperature, and soil fertility. Under soil solarization, a method of disinfecting arable soil using solar heat, the high temperatures achieved may enhance nitrogen mineralization in soil, allowing farmers to use less chemical nitrogen fertilizer. We measured nitrogen mineralization rates in incubation experiments at 30°C, 45°C, and 60°C, and based on the results, we estimated nitrogen mineralization under field conditions. Nitrogen mineralization during incubation was enhanced as soil temperature increased. To evaluate soil fertility, we extracted 28 soils in water at 80°C. The amount of organic carbon extracted was positively correlated with the amount of inorganic nitrogen released during incubation for 21 days at 45°C. We developed a kinetic model to estimate nitrogen mineralization under field conditions in Andosol in Tsukuba in the Ibaraki prefecture. N₀, Q, and k₀ values were determined from incubation experiments in which soil temperatures were measured at depths of 5 and 15 cm. Results showed that measurement of soil temperature at a depth of 5 cm allows to estimate nitrogen mineralization rates under soil solarization.