Table S1. Location information and soil physicochemical properties under tea gardens in Tan Cuong commune.

Tan Cuong is a rural commune in Thai Nguyen province. The climate is tropical monsoon characterized by the hot and moist season from April to September and the cool and dry season from October to March. According to the meteorological record during the last decade of Thai Nguyen Meteorology Station, the annual average temperature is 23.6℃ with the maximum of 29.0℃ in July and the minimum of 15.8℃ in January. The average annual precipitation is 1767 mm, 84% of which falls from April to September. The study area was located in the flood plain of Cong River flowing eastward. The terrain was composed of back marsh lowland with interspersing rows of the current natural dike and the previous dikes, followed by river terraces away from the river. The previous dikes were intermittent. The topography was flat in the lowland and rising at the dikes and terraces several to some ten meter height in elevation with gentle slopes of 5 to 15°. Tea gardens were located on the lowlands adjacent to the river and on the slopelands while the other lowlands away from the river were used for paddy cultivation. Although this area is geologically composed of Mesozoic schists and sedimentary rocks including sandstones and mudstones, the riverside lowland is loaded with new deposits supplied through river flooding. As mentioned above, the valuation of tea leaf quality depended on the distance from the river, implying that the influences of terrains and the related soil condition would reflect on tea growth and quality. Therefore, three transect lines (A, B, and C) were established perpendicularly to the river. Four gardens on each line were selected as study sites. The age of the gardens after establishment (garden age) ranged from 4 to 50 years. In terms of terrain, study sites were composed of a new dike adjacent to the river (A1), the previous dikes (A2, A3), and a river terrace (A4) on Transect A. On Transect B and C, they were of lowlands adjacent to the river (B1, C1) and the previous dikes (B2 to B4, C2 to C4). For comparison, four forested patches near to the tea garden sites were also selected on the Transect B and named BF1, BF2, BF3 and BF4, each corresponding to B1 to B4. Tea plants were planted in rows without ridging in all gardens, and along the contour lines in case of slopelands. The tea varieties planted in the study sites were Trung Du and LDP1. The slope of A4 and B3 was terraced, and red-colored subsoils were exposed in A4 with the youngest garden age of 4 years, while the appearance of soils in B3 with 35 years was similar to other sites. Meanwhile, B4 with the age of 4 to 8 years was severely affected by the former earth excavation for construction purposes, resulting in abundant rock outcrops (round-shaped pebbles and cobbles). Field survey and soil sampling were conducted in February at tea garden sites and September at forested sites in 2017. Three quadrates (7m x 7m) were set up within each site, and they were allocated to the upper, middle, and lower parts of the slope in case of slopelands. Within each quadrate, soil samples at the depth of 0-10 cm (surface) and 20-30 cm (subsurface) were collected from two or three points between hedges of tea plants and mixed well to make one composite sample. Soil samples were air-dried and then passed through a 2 mm mesh sieve. Soil pH was measured in water and 1 M KCl in a soil to solution ratio of 1:5 (pH(H2O), pH(KCl)) using the glass electrode method (F-21; Horiba). Electrical conductivity (EC) was measured in a soil to water ratio of 1:5 using an EC meter (CM-14P; TOA). The amount of total C (T-C) and total N (T-N) were analysed with a CN Corder (JM 1000; J-Science Lab). Exchangeable bases (exch. Ca2+, Mg2+, K+, and Na+) were extracted with 1 M CH3COONH4 at pH 7.0. The concentrations of Ca2+, Mg2+ and K+ were determined by atomic adsorption spectrophotometry, whereas that of Na+ was determined with flame photometry (AA-6800; Shimadzu). The NH4+ ions adsorbed on the soil residue were replaced by 10% NaCl and determined by the steam distillation and titration method as cation exchange capacity (CEC). The amount of exchangeable acidity and exchangeable Al3+ (exch. Al3+) were determined using a titration method with 0.01 M NaOH and successively with 0.01 M HCl after the addition of 4% NaF solution, respectively. The amount of exchangeable H+ (exch. H+) was calculated as the difference between exchangeable acidity and exch. Al3+. Available P (avail. P) was determined using the Bray II method with the phosphomolybdenum blue method at the wave length 710 nm (UV-140-02; Shimadzu). Inorganic N (NH4-N, NO2-N and NO3-N) was extracted by 2 M KCl and determined by steam distillation method with MgO and the Devada’s alloy. In spite of possible presence of NO2-N due to the defect in nitrification under strongly acidic soil condition, NO2-N and NO3-N were dealt together as NO3-N in the present study. Particle size distribution was determined with the combination method of pipette and sieving after the removal of organic matter by H2O2 with heating.