探討臺灣代表性土壤利用世界土壤參比分類系統之問題及其與美國土壤分類名稱之對應

Abstract

臺灣土壤調查過去皆以美國土壤分類系統 (Soil Taxonomy, ST) 當作土壤分類系統，其中Entisols (新成土)、Inceptisols (弱育土)、Alfisols (淋餘土) 及Ultisols (極育土) 占前四大主要土壤分佈類別。世界土壤參比分類系統 (World Reference Base for Soil Resources, WRB) 為十年來由聯合國糧農組織土壤分類系統(FAO-UNESCO system) 修改的國際土壤分類系統，近五年已於歐亞非三洲盛行，往往被當作全球化下各國土壤資訊轉換的橋樑，可補足ST土壤分類系統的相關缺憾。本論文將選出臺灣面積大於1,000公頃或具有特殊性的土壤作為此研究之臺灣代表性土壤，將這些土壤利用WRB分類系統分類後討論套用WRB時產生的問題，並利用WRB (2006 版本) 及Gray et al. (2011) 得到ST及WRB系統的標準對應表，而對應度 (similarity) 即表示為可正確對應的土壤數除上總土壤數。 本研究選出的172個代表性土系之平均對應度為64%，討論造成對應度低的原因，首先是水成土壤，由於WRB系統十分重視水的作用而較ST新增了許多水成土壤的高階土壤分類單位及診斷特性，在臺灣常被討論的水成土壤，包含地下水造成的Gleysols (灰白土)、Plinthosols (鐵網紋土)、表面浸水一段時間的Stagnosols (淹水土)，診斷特徵如Gleyic colour pattern (灰白色彩型態)、Plinthic horizon (鐵網紋層) 及Stagnic colour pattern (淹水色彩型態)。WRB新增特徵除了造成兩大系統下分類出的水成土壤結果不一致外，實際上在土壤分類時，常造成同一特徵由於同時符合診斷特徵而重複判斷。另外，人為灌水產生的淹水土由於也具有Stagnic colour pattern而可能遭誤解被分類為Stagnosols。 Cambisols (變育土) 對應度低是因為WRB系統明顯比ST系統能有效描述人為耕作的因子，因此有許多ST系統新成土在WRB系統下可符合人為耕作層而進一步分類成變育土，例如Anthraquic horizon (水田耕作層)、Hydragric horizon (人為浸水層) 及Irragric horizon (灌溉沉積層)。其他與對應度無關，但在臺灣土壤往往會發生的特殊情況有：有大比例的臺灣水田土壤由於無法符合Anthrquic horizon對土色的要求，而錯失正確分類的機會。部分風化時間悠久的臺灣紅土即使土壤剖面中黏粒含量分佈均勻，但CEC仍不符合 < 16 cmolc/kg clay的標準而無法分類為Ferralic horizon (鐵鋁層)。具可觀數量的臺灣新沖積土往往因為土層厚度太薄而無法符合WRB系統的Fluvic material (沖積診斷特性)，而無法被分類為沖積土，僅能分類為Regosols (腐石土)。Leptosols (薄層土) 在WRB系統中沒有明顯定義為當地母岩化育而成，容易被沖積或崩積土壤混淆而誤分類為此參考土類。WRB系統的Arenosols (砂質土) 僅提及沙漠或海灘的砂土，沒有將臺灣常見沖積及崩積砂質土包含在內，容易造成讀者對於砂質土分類的誤解。

620 rural soil series were developed in the history of soil survey projects from 1962 to 1979 in Taiwan. These soil series have been classified by the USDA Soil Taxonomy system (ST) and the major Soil Order were Entisols, Inceptisols, Alfisols and Ultisols. In last decade, World Reference Base for Soil Resources (WRB) system was revised from the FAO-UNESCO soil classification system and was significantly used by many countries in Europe, Asia and Africa regions. WRB system can be regarded as the system to compensate the weakness of USDA Soil Taxonomy system. A correlation table were developed to assess the correlation of soil taxa of this study based on the standard correlation tables by using the published paper of Gray et al. (2011). In this study, we selected 172 representative soil series from Taiwan to discuss the problems for Taiwan soil series to be classified by the WRB system. The study showed the mean similarity of correlation was 64% which was produced mainly by the Hydric soils including Plinthosols, Gleysols and Stagnosols in the WRB system. The relative lower similarity of correlation table was produced by various soil formation factors and soil characteristics. WRB system has created more Hydric diagnostic properties and horizons than ST system, including plinthic horizon, gleyic colour pattern and stagnic colour pattern. The other main errors of high category are also produced by the hydraulic soil characteristics, including Gleysols, Plinthosols, Planosols, and Stagnosols. Although these differences can be identified by the pedogenetic processes of the magor soil groups, it make the soil surveyors easily confuse to identify and describe the soil properties and soil formation. Also, A lot of Flooding soils irrigated with water also have stagnic colour pattern, and it can be misunderstood and to be classified into Stagnosols. WRB system created anthropogenic factors to affectly describe the soil formation process by human cultivation which are listed as criteria of Anthrosols, Cambisols and others, and the Entisols in ST with the cultivation evidence can be classified as Cambisols. There are some unique and important problems happened by using WRB systems to classify Taiwan soils. Many Paddy soils of Taiwan can be classified as uncorrected soil group which lack of the criteria of anthraquic horizon. Some red soils with long-term weathering soils even have uniform distribution of clay particle and without argic horizon, but it still can not meet the criteria of CEC value ( < 16 cmolc/kg clay) and can not be classified as Ferralic horizon. Taiwan alluvial soils are too thin to meet the evidences of stratification of fluvic material, and it was classified as Regosols, not to be Fluvisols. The formation of Leptosols can not be identified to be developed in-situ weathering of sediments or rocks, therefore the alluvial and colluvic soils can be misunderstood and to be classified as Leptosols. Arenosols are defined as the soils formed from sandy soils appearing in the beach and desert, but it was not included the sandy soils with fluvic soils and colluvial soils in Taiwan.