Zeng-Yei Hseu

Andisols formed from young volcanic materials are among the most fertile soils globally, particularly along the Pacific Ring of Fire, where recurrent volcanic activity has produced extensive ash an...

Stable carbon and oxygen isotopes in carbonate nodules were controlled by the type of vegetation (C ₃ or C ₄ ) and soil temperature, respectively. These signatures can reflect the paleoclimatic conditions during the formation of the carbonate nodules. Carbonate nodules were sampled along a soil chronosequence in southwestern Taiwan, representing different ages of shoreline changes associated with phases of marine transgression and regression. The stable carbon and oxygen isotopes in carbonate nodules were analyzed. The ages of the carbonate nodules were determined by radiocarbon ( ¹⁴ C) dating. The climofunctions of mean annual precipitation and temperature were calculated to interpret the climatic conditions. The carbonate nodules were relictic pedofeatures that formed episodically under favorable pedogenic conditions between 4063 and 690 years before present (years BP). The δ ¹³ C values indicated a mixed vegetation of C ₃ and C ₄ plants, with a predominance of C ₄ plants (68.0 to 98%). The mean annual temperature estimated by the climofunction of δ ¹⁸ O value (12.3 to 14 °C) was lower than the present (24.7 °C). The mean annual precipitation (1036 to 1342 mm yr ^–1 ) calculated by the geochemical climofunction was lower than the present (1829 mm yr ^–1 ). The climate was drier and cooler during this period, and this may be due to a weaker East Asian Summer Monsoon, which favored calcification in the soils. The climatic conditions, as interpreted by the stable carbon and oxygen isotopes of carbonate nodules, were consistent with the calcification process.

Rare earth elements (REEs) are emerging contaminants; however, information on their bioaccessibility via soil ingestion and its implications for human health risk remains limited. The physiologically based extraction test (PBET) is commonly used to evaluate the bioaccessibility of potentially toxic elements in soils, but the simplified bioaccessibility extraction test (SBET) offers a streamlined in vitro bioaccessibility assessment. This study examined thirty-two surface soil (0–10 cm) samples from Taiwan for the bioaccessibility of REEs using PBET and SBET. Additionally, the tested soils were characterized by fractional REEs with the BCR sequential extraction. The bioaccessible concentration of REEs extracted by the SBET was 21.1 ± 15.7 mg kg⁻¹ with a mean bioaccessibility of 15.2 ± 9.62%. The SBET-based bioaccessibility positively and significantly correlated with that of the PBET (r = 0.85, p < 0.001). These labile pools were mainly from the acid-exchangeable and reducible fractions of REEs in the soils. The bioaccessible REEs, as measured by SBET and PBET, were predominantly retained by pedogenic iron (Fe) oxides in this study. However, the SBET procedure was time-efficient and low-cost compared with the PBET. Thus, the consistent targeting of the reducible fraction of REEs by both methods confirms the SBET as a reliable and promising alternative to the PBET for assessing bioaccessibility of soil REEs.

Carbonate nodules are pedogenic carbonates commonly found in regions of high evapotranspiration, such as arid and semiarid areas. The stable carbon isotopes in these nodules are influenced by the type of vegetation (C3 or C4 plants), while the stable oxygen isotopes are controlled by soil water and temperature. Carbonate nodules persist in the soil, and their isotopic signatures can reflect the paleoclimatic conditions under which they formed. Carbonate nodules are distributed across the alluvial plains of southwestern Taiwan; however, the present humid conditions may not be favorable for their formation. The objective of this study was to evaluate the climatic conditions and vegetation types that influenced the formation of carbonate nodules using stable carbon and oxygen isotopes. Four pedons with different ages of soil formation were sampled in accordance with marine transgression and regression phases, corresponding to 10,000 years before present (yr BP), 8,000 yr BP, 5,000 yr BP, and 3,000 yr BP. Carbonate nodules were collected from the pedons, and their stable carbon and oxygen isotopes were analyzed using an automated carbonate preparation device. The δ13C values indicated a mixed C3/C4 vegetation, with a predominance of C4 plants (68.0 to 98%). The mean annual temperature (12.3-14.0°C), calculated using the climofunction of temperature and δ18O, was lower than the present (24.7°C). The mean annual precipitation (1036 to 1342 mm yr-1), calculated from the geochemical climofunction, was also lower than the present (1829 mm yr-1). The radiocarbon ages of the carbonate nodules ranged from 4063 yr BP to 690 yr BP, implying that climatic conditions may have been drier and cooler than present during this time frame. This may be due to a weaker East Asian Summer Monsoon, which favored calcification in the soils. These climatic conditions are consistent with the formation environment of carbonate nodules.

Highly weathered tropical soils have low soil organic carbon (SOC), intense Fe-oxide pigmentation, and limited nutrient retention, requiring rapid, cost-efficient SOC assessment. This study evaluated whether fusing NixPro color and KT10 κ measurements improves SOC prediction. A total of 363 soil samples across the A, B, and C horizons of Ultisols and Oxisols in Taiwan were analyzed for SOC using wet oxidation, colorimetric indices, and the κ. Pearson correlations guided variable selection, and SOC models were built using Partial Least Squares Regression (PLSR) and Support Vector Machines (SVM) with 10-fold cross-validation repeated 5 times. SOC–sensor correlations were strongly depth-dependent. In the A horizon, SOC showed no measurable response in color indices, indicating strong Fe-oxide masking. In the B horizon, SOC correlated positively with brightness and reflectance indices (L*, X, Y, Z, G, B; p < 0.001) and negatively with a*, b*, and c* (p < 0.001), reflecting clear pigment dilution at low SOC levels. In the C horizon, correlations were weaker but remained detectable for L*, Y, Z, G, and B (p < 0.05–0.01). Across the profile, chromatic patterns were dominated by the B horizon, which drove positive SOC-h° and negative SOC-a*, b*, and c* relationships (p < 0.001). SOC was also strongly negatively correlated with κ (p < 0.001), while κ remained positively related to dithionite-citrate-bicarbonate extractable and total iron, confirming Fe-oxide control of magnetic response. Predictive modelling showed that SVM with fused NixPro + KT10 inputs achieved the highest accuracy in the B horizons (R2 = 0.70, RMSE = 5.19 g kg-1, RPD = 1.80, LCCC = 0.81) and in the A+B+C horizons (R2 = 0.63, RMSE = 6.22 g kg-1, RPD = 1.61, LCCC = 0.75), demonstrating the effectiveness of multisensor data fusion for SOC estimation in highly weathered tropical soils.

Rare earth elements (REEs) are emerging contaminants, but little is known about their soil bioavailability and physiological effects on crops. This study investigated the effects of cerium (Ce) and yttrium (Y) at 0, 25, 50, and 100 mg kg^− 1 on pak choi (Brassica rapa subsp. chinensis (L.) Hanelt) grown in high clay and low clay soils. Measurements included Ce and Y uptake, biomass, photosynthetic pigments, respiratory activity, and proline concentration. The addition of Ce and Y was associated with elevated Ca, Mg, Fe, and Al in the shoots, supported by the overlapping ion mappings of these elements in the leaves with laser ablation inductively coupled plasma mass spectrometry. The root and shoot biomass of the crop in the high clay soil significantly (p < 0.05) increased from the control to 25 mg Ce kg^− 1 and 25 mg Y kg^− 1, respectively. In both soils, Y was more toxic to photosynthetic pigments than Ce, while respiratory activity was sensitive to both Ce and Y, leading to reduced proline levels. This study demonstrated that the soil bioavailability and physiological responses of the tested crop to Ce and Y were controlled by REEs type and soil texture.