Chapter 1. Introduction

Japanese tea fields receive large amount of nitrogen (N) inputs from chemical fertilizers and organic matters. While N promotes the production and quality of tea leaves, surplus N can cause environmental problems by leaching into waterways or being emitted as a potent greenhouse gas: nitrous oxide (N2O).

The Denitrification and Decomposition (DNDC) model was originally developed for estimating greenhouse gas emissions from U.S. agro-ecosystems. This model incorporates a suit of processes that describe soil hydrology, plant growth and biogeochemical reactions governing transport and transformation of C, N and water in plant-soil systems.

This thesis reports on an attempt to improve DNDC for estimating N2O emissions and nitrate (NO3-N) leaching from Japanese tea fields. I modified DNDC to extend its capacity to simulate agronomic and biogeochemical processes in the tea fields in Shizuoka and Kagoshima prefectures, the two major tea producing areas in Japan. I then estimated the N2O emission and NO3-N leaching from the tea fields across Japan by using the improved model, and assessed the efficacy of an alternative fertilization practice to mitigate N loadings.

Chapter 2. Testing DNDC for its applicability to Japanese tea fields

The comparison of the simulated N2O emission flux with the observations in a field in Aichi prefecture indicated that DNDC is applicable to the simulation of N2O emissions from Japanese tea fields. However, the results indicated the needs for modifications to accurately estimate of the N loading from Japanese tea fields. Modifications are also required to the harvest routine of tea cultivation system and biogeochemical processes in acid soils, which is prevalent in Japanese tea fields due to the very heavy doses of N fertilizers.

Chapter 3. Nitrous oxide emissions from Japanese tea fields as simulated with modified DNDC model

DNDC was modified according to the observations in the tea fields having very acidic soils in Shizuoka prefecture. The modifications included: (1) new parameterizations for tea plant physiology and phenology, (2) partitioning of plant biomass in regard to the harvest routine of tea cultivation system, (3) new parameterization for microbial activities: nitrification and denitrification under low pH, and (4) addition of chemodenitrification occurring in acidic soil conditions. The modified DNDC was better able to capture the magnitudes and temporal patterns of N2O emissions under different fertilizer treatments. Sensitivity tests showed that the alternative fertilization practices could enhance or mitigate N2O emissions at the experimental site.

Chapter 4. Nitrate leaching from Japanese tea fields as simulated with modified DNDC model

DNDC was tested in simulating the observations in a tile-drainage experiment in Kagoshima prefecture. This result indicated that DNDC could simulate discharge flow and N leaching from Japanese tea field. Sensitivity tests showed that NO3-N leaching could be reduced by soil properties and alternative managements at the experimental site.

Chapter 5. Estimating N loading from the tea fields and assessing alternatives for mitigation of N loading across Japan

In this chapter, I estimated the N2O emissions and NO3-N leaching from tea fields across Japan for 10 years from 2001 to 2010 using the modified DNDC (DNDC-Tea). The simulation with DNDC-Tea estimated N2O emissions and NO3-N leaching from Japanese tea fields as 0.64 Gg N2O and 13.01 Gg N, respectively, on average across the 10 years. The process-based model allowed the effects of soil properties, climates and farming practices on the estimates of N2O emission and NO3-N leaching. While DNDC estimates (0.64 Gg N2O) of N2O emission from Japanese tea fields was less than half the estimate (1.3 Gg N2O) based on a constant fraction of N2O emission in total N supply, the former is arguably, more realistic than the latter. The alternative management scenario across Japan was tested with DNDC-Tea, and the results suggested that the efficacy of the alternative management in mitigating the N loading from tea agro-ecosystems depends on the fertilization pattern, local soil properties and climate conditions.

Chapter 6. General discussion

The findings in the preceding chapters are synthesized and discussed against the wider context of climate change mitigation and sustainable agriculture. The findings showed that DNDC-Tea facilitates reliable prediction of N2O emissions and NO3-N leaching at regional and national levels and that the model shall be useful for comprehensive assessment of climatic impacts and sustainability in Japanese tea agro-ecosystems. DNDC, with the new capability of simulating the tea agro-ecosystems shall be a powerful tool for farmers and crop and environmental managers to accurately predict crop production and environmental impacts for a wide range of farming operations and climate conditions in tea agro-ecosystems. This study is an initial step towards this direction.

背景と目的

日本の茶園では,茶葉のタンパク含量と品質を向上させるために,多量の化学肥料窒素と有機物が投入される.その結果生じる過剰の窒素が,地下水の硝酸汚染や大気中への温室効果気体N2Oの放出を引き起こしている.この問題解決の一助として,申請者は生物地球化学モデルDNDCを用いて,日本茶園における窒素の動態をシミュレートすることにより,茶園からの窒素負荷の総量を推定し,窒素負荷低減対策を評価しようとして,本研究を行った.

DNDCの日本茶園への適用可能性

愛知県のある茶園でのN2O放出量観測結果を,DNDCでシミュレートしようと試みた結果,DNDCは茶園へ適用可能であるが,幾つかの修正が必要であることが分かった.特に,茶独特の収穫方式や強い酸性を示す茶園土壌からのN2O発生は,現行のDNDCでは表現できず,修正を必要とした.

修正したDNDCによる日本茶園からのN2O放出シミュレーション

静岡県の極めてpHが低い土壌にある茶園からのN2O放出をシミュレートするよう,DNDCを修正した.主な修正点は,茶の成長と発育・茶葉の収穫を新たにパラメタ化すること,低pH下で生じる硝化・脱窒用に微生物活動を新たにパラメタ化すること,低いpHと高い有機物含量で生じる化学脱窒のためのルーチン追加である.修正後のDNDCは,対象茶園での2水準の施肥窒素量におけるN2O放出を,総量・タイミングともに比較的良くシミュレートできた.モデルの感度分析の結果は,施肥法の変更や気象条件の変化により,N2O放出量が増減することを示した.

修正したDNDCによる日本茶園からの硝酸態窒素リーチングのシミュレーション

鹿児島県の暗渠排水茶園からの硝酸態窒素リーチングを,DNDCでシミュレートした.その結果,DNDCは茶園からの窒素リーチング量を概ね適切に推定できた.感度分析の結果は,施肥方法の変更により硝酸態窒素リーチング量を低減可能であることを示した.

日本茶園全体からの窒素負荷量の推定と低減対策の評価

日本で茶を生産しているすべての県について,県ごとにN2Oの放出と硝酸態窒素のリーチングを,茶用に修正したDNDC (DNDC-Tea) によってシミュレートし,2001年から2010までの10年間について,窒素負荷量を推定した.推定されたN2O放出量は,IPCCの排出係数を用いて推定したN2O放出量を大幅に下回った.植物の窒素吸収量を超えた施肥窒素がN2Oの大量発生を生じることから,施肥窒素量に対するN2O放出量の応答は,本来下に凸の非線形ないし閾値型の線形を示すと考えられる.特に,施肥窒素量が550 kgN ha-1前後に減少した現在では,固定した排出係数を仮定するIPCCの推定は,相当の過大推定となり,DNDC-Teaの推定のほうがより合理的と考えられる.

窒素負荷低減対策の効果をDNDCで評価したところ,化学肥料窒素の一部をコンポストで代替することは,N2O放出量と硝酸態窒素リーチング量を減少させる効果があることが示された.

総合考察

本研究の結果,日本の茶園用に改良したDNDCは,茶園からのN2O放出と硝酸態窒素のリーチングを推定できること,そして窒素負荷軽減対策の評価にも役立つことが示された.今後いっそうの改良と検証によって,さらに正確で適用範囲の広いモデルに育つことが期待される.

以上のように,本論文が,日本茶園からの大気と水への窒素負荷を,生物地球科学モデルを用いたシミュレーションにより推定したことは,学術上,応用上貢献するところが大きく,よって審査委員一同は本論文が博士(農学)の学位論文として価値あるものと認めた.