Effect of Fertilization Timing on Nitrogen Uptake in Spring Tea of Different Sprouting Phenological Cultivars: A Field Trial with ¹⁵N Tracing

Abstract

Applying a top dressing of nitrogen fertilizer before harvesting spring tea is vital for producing high-quality spring tea. However, the interaction between the sprouting phenological characteristics of various cultivars and the timing of top dressing remains unclear. A field trial was conducted to investigate such interaction. Urea enriched with ¹⁵N was applied to soil of the early-sprouting cultivar Jia-ming-1 (JM1) and the late-sprouting cultivar Tie-guan-yin (TGY) on 29 January (early application, EApp) or 10 March (late application, LApp), respectively. The bud density and yield of young spring shoots were significantly decreased in LApp compared to EApp. Such differences were more remarkable in the early-sprouting cultivar (JM1) than in the late-sprouting cultivar (TGY). The N_dff (N derived from ¹⁵N-enriched urea) in mature leaves and young spring shoots as well as the amount of ¹⁵N in young spring shoots were all higher in EApp than in LApp. N_dff in both mature leaves (R² = 0.99, p < 0.001) and young spring shoots (R² = 0.61–0.89, p < 0.01) could be well predicted by the growing degree days of the duration between the N fertilization and sampling. N_dff and ¹⁵N concentrations in mature leaves were significantly correlated with the content of nitrate and the ratio of ammonium to total inorganic nitrogen. Partial least squares path modeling revealed that thermal condition directly affected soil N supply and soil pH and thereby affected N_dff in mature leaves and young spring shoots. Our findings highlight the importance of early pre-spring topdressing of N fertilizer to improve the yield and N use efficiency of spring tea in both early- and late-sprouting tea cultivars. The work identified a synergistic effect of N uptake by tea plants, N transformation, and soil pH related to the thermo-conditions of early and late N topdressing.

1. Introduction

The tea plant (Camellia sinensis L.) is an important cash crop widely cultivated in subtropical and tropical regions. As a foliar plant, tea tree has a considerably higher nitrogen (N) requirement compared to other crops, due to its frequent harvesting and pruning practices throughout the growing cycle [1,2]. The N uptake in tea trees exhibits seasonal variability [3], with spring growth requiring more nutrients than summer or autumn [1,4]. Tea harvested in spring has high economic value due to the richness in flavor compounds caused largely by ample N nutrition [5,6]. Therefore, topdressing of inorganic N fertilizer before the harvesting of spring tea is an important practice which has been adopted in approximately 80% of tea plantations in China [7]. However, the timing of topdressing in tea production is highly discretionary and varies substantially [8]. Nutrient uptake by plants is directly affected by air and soil temperature through effects on root physiology in combination with soil-mediated controls on nutrient availability [9]. Different timing effects are related to differences in root growth, development stage, and storage and remobilization of nutrients among cultivars [10,11]. Our previous works showed that the uptake of topdressing N in spring tea was the highest when it was applied 50–60 days before the harvesting of young spring shoots and the timing effect was related to the thermal condition between the dates of fertilization and harvest of young shoots [12]. In response to the increasing temperature from winter to early and late spring, the spring budding dates of different tea cultivars vary by up to 40 days [13]. The temporal change in nutrient uptake in spring might also vary according to cultivars of different phenological characteristics in tea [14,15] and also other evergreen or deciduous trees [16,17]. On the other hand, soil N transformation processes, including nitrification, denitrification, mineralization, and immobilization, exhibit strong temperature dependence, as evidenced by previous studies [18,19,20,21]. These thermal sensitivities fundamentally regulate the spatiotemporal distribution of soil N pools, thereby influencing N availability for plant uptake and the temporal effectiveness of fertilization practices. Emerging evidence from cereal crop systems has demonstrated that maize N assimilation patterns are shaped by genotype-specific interactions with N-cycling processes, which are regulated by soil thermal and hydrological conditions during key phenological stages [22,23]. Notably, the thermo-regulated soil–plant N coupling mechanism remains poorly characterized in perennial woody crop systems. For spring tea, a critical knowledge gap persists in optimizing the timing of N fertilization to synchronize with both the plant’s endogenous nutrient demand rhythms and the dynamic patterns of soil N flux. This knowledge deficit is particularly consequential because the early spring growth of tea shoots, which determines both yield and quality, coincides with rapidly changing soil thermal conditions. Elucidating the complex interactions among genotypes, environmental factors, and management practices is essential for developing precision N fertilization strategies in tea cultivation systems.

In the present work, a field trial was conducted with the common early-sprouting cultivar ‘Jia-ming 1’ (JM1) and the common late-sprouting cultivar ‘Tie-guan-yin’ (TGY) by applying ¹⁵N-enriched urea at two times, i.e., 29 January (early application) or 10 March (late application). We hypothesized that (1) the N uptake of spring tea would respond differently to the timing of N fertilizer between the early- and late-sprouting cultivars, i.e., the late-sprouting cultivars would be less sensitive than the early cultivars to the topdressing timing; (2) the timing effect might be associated with soil properties and soil N transformation which changed accordingly alongside that of temperature. The N derived from ¹⁵N labeled fertilizer in plants and soil was subsequently measured. The purpose of the study was to identify the main factors influencing the uptake of fertilizer N in spring shoots and to ascertain whether the sprouting phenological characteristic of the tea cultivar affect the optimal timing N topdressing.

2. Materials and Methods

2.1. Description of the Field Site

The field experiment was conducted in a long-term field experiment investigating the interaction of tea cultivar and nitrogen fertilization at Shengzhou Experimental Station, which is located in Shaoxing, China (29°44′43″ N, 120°49′3″ E, 23 m a.s.l) (Figure 1). The plantation was established in 2015 with tea trees of different cultivars planted in a single row (0.33 m between bushes and 1.5 m between rows) at the same year. All cultivars received uniform field management including fertilization, pruning, irrigation and harvesting. Before the study, tea plants were fertilized with annual totals of 150 kg N ha⁻¹, 90 kg P₂O₅ ha⁻¹ and 120 K₂O kg ha⁻¹ of chemical fertilizers. Chemical N fertilizer (urea, 46% N) was applied three times, each in late February (45 kg N ha⁻¹), May (30 kg N ha⁻¹) and October (75 kg N ha⁻¹). Phosphorus as calcium superphosphate and potassium as potassium sulfate were applied together with rapeseed cake manure (1200 kg ha⁻¹) in October. All fertilizers were manually applied to ploughed furrows with a depth of 10–15 cm, along the edge of the tea canopy. The site is characterized by a subtropical humid climate with a mean annual temperature and rainfall of 16.4 °C and 1447 mm, respectively. The soil was classified as a granite-weathered red soil according to Chinese classification (Ultisols in US soil taxonomy), with a silt loam texture in the 0–30 cm soil layer. The soil had an initial pH (H₂O) of 4.40, soil organic carbon of 11.1 g kg⁻¹, total N (TN) of 0.93 g kg⁻¹, total phosphorus of 0.52 g kg⁻¹, total potassium of 8.97 g kg⁻¹, alkali-hydrolyzed N of 173 mg kg⁻¹, available phosphorus of 8.37 mg kg⁻¹, available potassium of 111 mg kg⁻¹, and bulk density of 1.22 g cm⁻³.

2.2. Experimental Design

Two cultivars, ‘Jia-ming 1’ (JM1) and ‘Tie-guan-yin’ (TGY), representing the early- and late-sprouting cultivars, respectively, were selected from ten cultivars [24] for this experiment with ¹⁵N labeling in 2021. A total of 8 bushes with visually uniform sizes were selected in two neighboring rows of tea plants for each cultivar, with each bush containing two plants as a micro-area with a total of 4 replicates. Each treatment was set up with 4 blocks, and each block contained two microplots, which were distributed in the same or different tea rows. Four plastic plates (4 mm thick, 40 cm high) were buried vertically at a depth of 35 cm to isolate a microplot (0.66 m × 1.5 m) containing two bushes (Figure 1). To avoid disturbing the soil in the microplot, narrow deep pits were dug according to the preset boundaries to place the separation board and slowly tap it into the preset depth. Each microplot received ¹⁵N-enriched urea (5.15 atom% ¹⁵N, purchased from Shanghai Research Institute of Chemical Industry Co. Ltd., Shanghai, China) at a dose of 45 kg N ha⁻¹ on one of the two dates: 29 January (early application, hereafter referred to as EApp) or 10 March (late application, hereafter referred to as LApp). Urea was mixed with the surface soil (5 cm deep) removed from the microplot and then evenly spread by hand back to the soil surface in the microplot. The two ¹⁵N-urea application times were randomly arranged to each cultivar. The management measures of tea trees in the microplot, except for the fertilization method which was changed from the local traditional manual shallow trench fertilization to mixing the fertilizer with the surface soil, were consistent with the local traditional field management measures, including fertilizer, fertilization rate, pruning, and harvesting.

2.3. Weather Recording

Air temperature and precipitation were measured in an automatic weather station (SP200, LSI-LASTEM, Milano, Italy) located about 100 m away from the experimental field. Soil temperatures at the surface and 20 cm depth were recorded using a tube probe (JZSX Corp., Beijing, China) installed in the field [24]. Growing degree days (GDDs) equal the value of the average daily temperature minus the base temperature (T₀ = 8 °C).

2.4. Plant and Soil Sampling

Ten mature leaves on the surface of canopy were collected from every microplot after 0, 6, 12, and 30 days after ¹⁵N application and then oven-dried at 60 °C until reaching a constant weight. Young shoots consisting of one bud with one expanding leaf were harvested by hand in the subsequent spring. The harvesting of young shoots was started on 12 March (42 days and 2 days after EApp and LApp, respectively) and finished on 15 April totally in seven batches for the early-sprouting cultivar JM1 and started on 2 April (63 days and 23 days after EApp and LApp, respectively) and finished on 23 April totally for six batches of the late-sprouting cultivar TGY. The number of young shoots in each batch was used to determine the bud density of the plant. The samples of young shoots were treated in a microwave oven (800 W) for 2 min to inactivate the oxidase enzymes and then dried in an electric oven at 60 °C. Tea plants were pruned on 27 April at the height of 60 cm and the pruning litters were collected, divided into branches and leaves, and then oven-dried at 60 °C to a constant weight. All the sampled mature leaves, harvested young shoots, and pruned material were ground using a ball mill (MM400, Retsch, Duesseldorf, Germany) for the determination of N concentration and ¹⁵N abundance. In addition, representative young shoots of one bud with one leaf were collected at the peak of harvest, frozen in liquid N, stored in a −80 °C refrigerator, and crushed using a ball mill after vacuum freeze-drying to determine the content of quality components.

Soil samples were collected at four sites in each labeled microplot using an auger (Φ = 5 cm) from the 0–10 cm layer on the same day of mature leaves’ sampling and pruning. Four soil cores from the same plot were combined into one composite sample, passed through a 2 mm sieve, and then separated into two portions for further analysis. One portion was stored in a refrigerator at 4 °C for the determination of the contents of NH₄⁺-N and NO₃⁻-N. Another portion was air-dried, passed through a 0.15 mm sieve, and finally oven-dried at 105 °C for the determination of the TN content and ¹⁵N abundance.

2.5. Measurements of ¹⁵N, Soil Properties and Tea Quality

Total N concentrations and the ¹⁵N abundance were determined using an elemental analyzer (EA, Flash 2000 HT, Thermo Fisher Scientific Inc., Waltham, MA, USA) coupled (ConFlo IV, Thermo Fisher Scientific Inc., Waltham, MA, USA) to an isotope ratio mass spectrometer (IRMS, Mat 253 Plus, Thermo Fisher Scientific Inc., Waltham, MA, USA). Inorganic N was extracted from the refrigerator stored fresh soil using 2 mol/L KCl (1:5, w/v) for 30 min. The concentrations of NH₄⁺-N and NO₃⁻-N in the extract were determined using an automatic discontinuous chemical analyzer according to the user’s guide (Smartchem140, AMS Alliance, Frépillon, France).

The properties of the initial soil were determined according to the following methods. Air-dried soil sieved through a 2 mm sieve, and soil pH was determined by a pH meter (Orion 3 star, Thermo Ltd., Waltham, MA, USA) in a 1:2.5 solution (soil: deionized water, w/v); alkali-hydrolyzed N content was determined by diffusion of sodium hydroxide and acid titration; available phosphorus content was extracted with ammonium fluoride-hydrochloric acid and measured using a UV–visible spectrophotometer (TU-1901, Puxi Ltd., Beijing, China); available potassium content was extracted with ammonium acetate and an inductively coupled plasma emission spectrometer (Optima 8000, PerkinElmer, Waltham, MA, USA). The air-dried soil samples were sieved through a 0.15 mm sieve, and an element analyzer (Vario Max, Elementar, Langenselbold, Germany) was used to determine the SOC and TN contents.

The total free amino acid (FAA) and total polyphenol (TP) of young shoots were determined through the ninhydrin colorimetric method and ferrous tartrate colorimetric method using a spectrophotometer (UV-2550, Shimadzu, Kyoto, Japan) [25].

2.6. Calculations and Statistics

2.6.1. Calculations

The N derived from the ¹⁵N-enriched urea (N_dff, %) in the plant sample was calculated as follows:

$${\mathrm{N}}_{\mathrm{d}\mathrm{f}\mathrm{f}}\left(\%\right)={\displaystyle \frac{{}^{15}\mathrm{N} \mathrm{a}\mathrm{t}\mathrm{o}\mathrm{m}\% \mathrm{i}\mathrm{n} \mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}-\mathrm{t}\mathrm{h}\mathrm{e} \mathrm{n}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{a}\mathrm{l} {}^{15}\mathrm{N} \mathrm{a}\mathrm{t}\mathrm{o}\mathrm{m}\% \mathrm{i}\mathrm{n} \mathrm{s}\mathrm{a}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{e}}{{}^{15}\mathrm{N} \mathrm{a}\mathrm{t}\mathrm{o}\mathrm{m}\% \mathrm{i}\mathrm{n} \mathrm{u}\mathrm{r}\mathrm{e}\mathrm{a}-\mathrm{t}\mathrm{h}\mathrm{e} \mathrm{n}\mathrm{a}\mathrm{t}\mathrm{u}\mathrm{r}\mathrm{a}\mathrm{l} {}^{15}\mathrm{N} \mathrm{a}\mathrm{t}\mathrm{o}\mathrm{m}\%}}\times 100$$

(1)

where the natural ¹⁵N atom% is 0.3663.

The ¹⁵N content in young shoots or pruned material was calculated by multiplying the biomass, N_dff, and N concentration of the plant organ:

$${}^{15}\mathrm{N} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{e}\mathrm{n}\mathrm{t}\left(\mathrm{m}\mathrm{g} {\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}}^{-1}\right)={\displaystyle \frac{{\mathrm{N}}_{\mathrm{d}\mathrm{f}\mathrm{f}}\left(\%\right)\times \mathrm{N} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{c}\mathrm{e}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}\left(\mathrm{m}\mathrm{g} {\mathrm{g}}^{-1}\right)\times \mathrm{b}\mathrm{i}\mathrm{o}\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}\left(\mathrm{g} {\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{t}}^{-1}\right)}{100}}$$

(2)

2.6.2. Statistical Analysis

T-tests were used to compare cultivars with the same sampling dates and fertilization dates for soil properties, mature leaves, and shoots’ dynamics analyses of N uptake. For the assessment of yield and N uptake efficiency, a post hoc LSD test was performed using a one-way ANOVA with the “stats” package. The effect of fertilization date and cultivar was tested using a multivariate analysis of variance (MANOVA) with the “bruceR” package.

A correlation analysis was conducted to determine the relationship between N_dff in mature leaves and soil properties using the “ggcor” package. Quadratic regression was performed to describe the relationship between N_dff in mature leaves and that in young shoots with GDD using the “stats” package.

A partial least squares path modeling (PLS-PM) analysis was conducted using the R package “plspm” to investigate the causal relationships between latent variables influencing nitrogen uptake in tea plant shoots. The model incorporated four latent variables: (1) thermal condition, indirectly represented by growing degree days above 8 °C (GDDAir8 °C) accumulated during 12-day and 30-day periods post-fertilization; (2) soil N, assessed through total nitrogen (TN, 12 days post-fertilization), nitrogen derived from fertilizer (N_dff, 30 days post-fertilization), ammonium-N (NH₄⁺-N, 30 days post-fertilization), and nitrate-N (NO₃⁻-N, 12 days post-fertilization); (3) soil pH, quantified as the mean pH value measured over 30 days after nitrogen application; and (4) N_dff in mature leaves, evaluated using N_dff of mature leaves at 12 and 30 days post-fertilization. Path relationships among these variables were analyzed to elucidate their direct and indirect effects.

In addition, the relationship among thermal condition, soil N, N_dff in mature leaves, and N_dff in all of the young spring shoots also was investigated using PLS-PM. In the pathway model, the latent variable “thermal condition” was reflected only by the manifest variable of GDD between fertilization and last harvest, which also was calculated using daily average air temperature with a threshold of 8 °C. The latent variable “soil N” was measured by TN (at 12 days after N application), NH₄⁺-N (at 30 days after N application), NO₃⁻-N (at 12 days after N application), and the mean value of total inorganic nitrogen (TIN) between fertilization and pruning. The latent variable “soil pH” was measured by the mean soil pH value between fertilization and pruning. The latent variables “N_dff in mature leaf” and “N_dff in shoot” were measured by N_dff in mature leaves (at 12 and 30 days after N application) and N_dff in young spring shoots. Path coefficients represented the strength and direction of relationships between the variables. The goodness-of-fit of the model was assessed, considering that a value greater than 0.7 indicates a “good” and acceptable fit.

All the analyses in this study were conducted by R software (version 4.0.3).

3. Results

3.1. Temperature and Rainfall During the Experiment

During the experiment, the monthly mean air temperature increased from 5.5 °C in January to 17.1 °C in April (Figure 2A). The change in soil temperature followed a similar trend, albeit with less fluctuation (Figure 2A). The mean daily air temperature within 30 days after EApp was 10.3 °C and increased to 14.3 °C after LApp. The GDDs (Air, T₀ = 8 °C) between the first harvest of young shoots and the day of EApp or LApp were 105 and 12 °C·d for JM1 and were 227 and 134 °C·d for TGY, respectively (Figure 2C). The GDDs between the last harvest of young shoot and the day of EApp or LApp were 328 and 235 °C·d for JM1 and were 415 and 322 °C·d for TGY, respectively (Figure 2C). Precipitation during the experiment mainly occurred in March, with monthly precipitation reaching 206.2 mm (Figure 2B). The rainfall between the first harvest of young shoot and the day of EApp or LApp was 130.6 and 9.8 mm for JM1 and was 279.8 and 159.0 mm for TGY, respectively. The rainfall between the last harvest of young shoot and the day of EApp or LApp was 325–330 and 204–209 mm for the two cultivars, respectively (Figure 2D).

3.2. Soil Properties

The content of NH₄⁺-N in the topsoil increased rapidly to peaks after N application and then decreased to pre-fertilization levels by the end of the spring (Figure 3A). The NH₄⁺-N content in EApp and LApp reached peak levels at 30 and 6 days, respectively, after N application for JM1. NH₄⁺-N content for TGY reached the highest level 12 days after N application in both EApp and LApp. The soil NH₄⁺-N content of JM1 was significantly lower than that of TGY on several sampling dates. The content of NO₃⁻-N increased significantly within 12 days for EApp and 30 days for LApp (Figure 3B). In both application timings, the content of NO₃⁻-N reached the highest levels at the end of harvest. The percentage of NO₃⁻-N/TIN decreased to its lowest levels 6 or 12 days after N application then increased until the last sampling (Figure 3C). The soil NO₃⁻-N content and its share in TIN were not different between the two cultivars on the specified sampling dates. In EApp, soil pH increased after 6 days and started to decrease 30 days after N topdressing (Figure 3D). In LApp, the soil pH increased only slightly after 6 days and thereafter decreased significantly.

3.3. ¹⁵N in Mature Leaves

The N concentration in mature leaves slightly decreased between late January and early April and increased after the last harvest of young shoots (Figure 4A). JM1 contained much higher N than TGY in mature leaves, and such differences were more marked in EApp than in LApp (Figure 4A). Within 30 days of N fertilization, the N_dff in mature leaves only slightly increased in EApp but substantially increased in LApp for both cultivars (Figure 4B). By the end of harvest, mature leaves’ N_dff increased considerably to 2.43–2.70% and 1.48–2.86%, respectively. Mature leaves’ N_dff between the EApp and LApp did not differ in JM1 but was significantly different in TGY. N_dff in mature leaves of JM1 increased with increasing GDD between the day of ¹⁵N application and the sampling dates, reached a maximum at GDD 365.4 °C·d, and no longer changed afterwards (Figure 4C). N_dff in mature leaves of TGY kept increasing with GDD. Nevertheless, their relations could be best described by quadratic regression (JM1: R² = 0.99, p < 0.001, n = 7; TGY: R² = 0.99, p < 0.0001, n = 9) (for JM1 excluding GDD beyond the maximum). Correlation analysis showed that within 30 days of ¹⁵N application, N_dff in mature leaves correlated positively with the NO3−-N content in the topsoil (r = 0.511, p < 0.001) and negatively with soil pH (r = −0.588, p < 0.001).

3.4. Dynamic Change in ¹⁵N in Young Spring Shoots

For the young shoots of tea trees, free amino acids (FAAs) and total polyphenols (TPs) are the main quality-related components, and the ratio of polyphenols to amino acids (TP/FAA) is an important biochemical index of fresh leaf quality. The type of cultivar had significant effects on the FAA and TP and TP/FAA of young spring shoots, but timing of N application only had significant effects on TP content for JM1. TP content in EApp was lower than that in LApp for JM1. ¹⁵N application date did not affect the quality of the TGY cultivars.

Compared to EApp, the density, yield and total N content of young spring shoots of JM1 were significantly decreased in LApp by 32.1%, 33.5% and 33.7%, respectively (

Table 1. Quality, density, yield, N content, N_dff, and ¹⁵N content of young spring shoots of two cultivars receiving early (EApp) and late (LApp) applications of ¹⁵N-enriched urea (mean ± se, n = 4). Data followed by different letters within the same column are significantly (p < 0.05) different. The p-values show the effects of topdressing timing, cultivar, and their interaction.

Cultivar	Treatment	FAA (%)	TP (%)	TP/FAA	Density (Shoots Plot−1)	Yield (g Plot−1)	N Content (mg Plot−1)	Ndff (%)	15N Content (mg Plot−1)
JM1	EApp	4.92 ± 0.09 a	17.35 ± 0.36 c	3.53 ± 0.13 b	1461 ± 75 a	25.17 ± 1.47 a	1429 ± 114 a	4.11 ± 0.46 bc	57.55 ± 4.15 ab
	LApp	5.10 ± 0.09 a	19.03 ± 0.57 b	3.73 ± 0.10 b	992 ± 189 ab	16.73 ± 3.03 b	948 ± 186 b	3.37 ± 0.23 c	31.38 ± 5.65 b
TGY	EApp	3.49 ± 0.19 b	25.93 ± 0.60 a	7.50 ± 0.47 a	933 ± 89 b	20.57 ± 2.26 ab	1077 ± 110 ab	8.50 ± 0.26 a	91.97 ± 10.99 a
	LApp	3.45 ± 0.09 b	25.88 ± 0.13 a	7.52 ± 0.21 a	674 ± 26 b	16.08 ± 1.77 b	856 ± 94 b	5.50 ± 0.73 b	45.24 ± 3.49 ab
p-Value	Cultivar	<0.001	<0.001	<0.001	0.003	0.258	0.115	<0.001	0.004
	Time	0.580	0.097	0.700	0.007	0.013	0.020	0.002	<0.001
	Cultivar × Time	0.388	0.081	0.738	0.366	0.389	0.340	0.032	0.153

). Similar differences were observed in TGY, but they were not statistically significant.

The N_dff of each harvest’s young shoots exhibited different trends between the two cultivars. TE N_dff of JM1 increased rapidly in the first three in EApp or five in LApp harvests and then slightly increased. N_dff of TGY remained rather stable at high levels in EApp and increased only slightly in LApp (Figure 5A). N_dff values were significantly higher in EApp than in LApp in the first three harvests for JM1 and in the first five harvests for TGY (Figure 5A). The N_dff of shoots could be described by quadratic regression against air GDD (T₀ = 8 °C) between the day of ¹⁵N fertilization and harvest dates (JM1: R² = 0.89, p < 0.0001, n = 14; TGY: R² = 0.61, p < 0.01, n = 12) (Figure 5B). This indicates that GDD effectively predicted N_dff dynamics in young shoots of the JM1 cultivar, and predictive accuracy for TGY remained suboptimal when using GDD alone. However, integrating GDD with rainfall significantly enhanced the model’s robustness (R² = 0.92, p < 0.0001).

¹⁵N content in young shoots of JM1 increased during the first three harvests and showed significant differences between the two application times (Figure 5C). Young shoot ¹⁵N content in JM1 peaked at the fourth harvest (around 30 March) and then decreased. Young shoot ¹⁵N content in TGY peaked at the first or second harvest (around 30 March), decreased from then on, and then increased again until the last harvest (23 April). It was noted that the change in ¹⁵N content between 30 March and 23 April coincided with the change in air temperature (Figure 5A).

Both the timing of N application and cultivar had significant effects on the total ¹⁵N content and average N_dff of young spring shoots (Table 1). The ¹⁵N content and N_dff in EApp were 45.5–50.8% and 18.0–35.3% higher than those in LApp. Shoot N_dff was significantly affected by the interaction between the application time and cultivar. The difference between EApp and LApp was smaller for JM1 than for TGY. N_dff and ¹⁵N contents were higher in TGY than in JM1, particularly for EApp (Table 1).

3.5. N_dff and ¹⁵N Content in Pruned Material

The total biomass, ¹⁵N content, and N_dff of pruned material (including twigs and leaves) were significantly decreased in LApp compared to EApp for TGY, whereas they remained unchanged for JM1 (

Table 2. Dry matter, N content, and ¹⁵N content in pruned material of two cultivars receiving early (EApp) and late (LApp) applications of ¹⁵N-enriched urea. Data are mean ± standard error (SE, n = 4). Data followed by different letters within a row are significantly (p < 0.05) different. The p-values show the effects of topdressing time, cultivar and their interaction.

Cultivar	Treatment	Mass (g Plot−1)	N Content (g Plot−1)	Ndff (%)	15N Content (mg Plot−1)
JM1	EApp	93.9 ± 20.5 b	2.23 ± 0.47 ab	2.78 ± 0.07 a	61.6 ± 12.3 ab
	LApp	65.9 ± 12.4 b	1.63 ± 0.31 b	2.77 ± 0.21 a	43.6 ± 6.9 b
TGY	EApp	167.0 ± 22.6 a	3.23 ± 0.54 a	2.91 ± 0.21 a	96.2 ± 21.3 a
	LApp	111.0 ± 12.5 b	2.21 ± 2.69 ab	1.93 ± 0.26 b	40.7 ± 3.4 b
p-Value	Cultivar	0.006	0.079	0.242	0.100
	Time	0.035	0.072	0.015	0.029
	Cultivar × Time	0.446	0.062	0.172	0.034

). The N_dff of pruned material was significantly higher in JM1 compared to TGY under LApp.

3.6. Partial Least Squares Path Modeling

PLS-PM was constructed to assess the direct and indirect effects of the thermal conditions, soil N condition, and soil pH on N_dff in mature leaves (Figure 6A). These relationships represent the short-term effects of N topdressing. The results showed that soil pH had a direct negative effect on N_dff (path coefficient (pc) = −0.96, p < 0.001) (Figure 6B), whereas thermal condition was found to have indirect positive effects (indirect pc = 0.84) by controlling the soil’s N condition (pc = −0.86, p < 0.001) and pH (pc = −0.83, p < 0.001). Excluding the indirect effects, the path model explained 81% of the total variance in N_dff in mature leaves (goodness-of-fit: 0.82).

We constructed another PLS-PM model to assess the direct, indirect, and interactive effects of thermal conditions, soil N conditions, soil pH, and N_dff in mature leaves on N_dff in young spring shoots (Figure 6C). These relationships represent the long-term effects of N topdressing. The results showed that thermal conditions had direct positive effects on soil N condition (pc = 0.81, p < 0.001) and soil pH (pc = 0.74, p < 0.001). N_dff in young spring shoots was greatly and directly influenced by soil pH (pc = 0.90, p < 0.01) and N_dff in mature leaves (pc = 0.85, p < 0.01) (Figure 6D). Furthermore, N_dff in young spring shoots was indirectly influenced by the thermal condition (indirect pc = 0.69) and soil N condition (indirect pc = 0.45) (Figure 6D). Excluding the indirect effects, the path model explained 77% of the total variance in shoots’ N_dff (goodness-of-fit: 0.76, Figure 6C).

4. Discussion

4.1. Effect of N Topdressing Timing on N in Young Spring Shoots of Early- and Late-Sprouting Cultivars

The timing of fertilizer application is a crucial factor influencing the N uptake and overall yield of spring tea [12,26]. In this study, we employed a microplot ¹⁵N-tracing field trial to evaluate the interaction between topdressing timing before spring tea harvesting and the sprouting characteristics of two tea cultivars with notably different sprouting dates (about 20 days apart). Our results indicated that the timing of topdressing markedly impacted the bud density, yield, N content, N_dff and ¹⁵N content of young spring shoots, as well as total biomass, N_dff, and ¹⁵N content of pruned material (Table 1 and Table 2). Late application of fertilizer consistently resulted in lower ¹⁵N absorption and N_dff compared to earlier applications, regardless of the cultivar’s sprouting characteristics. The finding confirms the importance of early N application in promoting N uptake and accumulation in spring teas [12,26]. The present work further confirmed previous findings that the uptake and allocation of pre-harvesting N topdressing in young shoots (Figure 5B) and also in mature leaves (Figure 4C) can be well described by growth degree days (GDDs) between the ¹⁵N application and the dates of harvesting or sampling, except for young shoots of TGY (R² = 0.61). The N_dff of young shoots for TGY was significantly correlated with rainfall, and it could be better predicted by rainfall combined with GDDs. This is because soil water plays a decisive role in the N cycle through microbial activities (e.g., soil nitrification and mineralization), while also influencing plant physiological and metabolic processes (e.g., respiration, photosynthesis, nutrient uptake), hence affecting the N uptake and assimilation of plants and even their preference for different forms of N. Generally, increased moisture levels can enhance nutrient availability, root growth, and nutrient uptake, thereby promoting plant N use efficiency. However, excessive soil moisture inhibits N use efficiency, and studies have shown that excessive rainfall during the monsoon season leads to decreased tea yield [27].

Consistent with previous studies demonstrating cultivar-dependent N uptake linked to phenological traits [14,15], our findings reveal a phenology-mediated interaction between topdressing timing and N_dff in young spring shoots. Notably, delayed fertilization (LApp) induced a more pronounced reduction in N_dff in the late-sprouting cultivar TGY (35% decrease, Figure 5A) compared to the early-sprouting JM1 (18% decrease). This highlights that late-sprouting cultivars may be more sensitive to temporal variations in N fertilization, underscoring the importance of aligning topdressing schedules with cultivar-specific phenology to optimize nutrient utilization.

The even larger effect of LApp on N_dff was also observed in mature leaves and pruned material (Figure 3B and Table 2). These findings were contrary to our hypothesis that late-sprouting cultivars would be less sensitive than early ones to topdressing timing. The significant decline in N_dff of TGY with late application underscores the critical role of early pre-spring topdressing for both early- and late-sprouting cultivars in order to improve N acquisition in spring shoots.

4.2. Role of Soil N Transformation in N Uptake of Tea Trees in Response to Topdressing Timing

In the present work, soil inorganic N contents and pH in the topsoil were analyzed to test the hypothesis that the effect of timing is associated with soil N transformation, which changes in accordance with rising temperature in spring. The content of NH₄⁺-N increased sharply following urea topdressing, which was accompanied by increases in soil pH within 6 days as a result of urea hydrolysis (Figure 3) [28]. The NO₃⁻-N content increased, while its share of TIN decreased concomitantly with increasing NH₄⁺-N, indicating the occurrence of nitrification [29]. Consequently, soil pH decreased 30 or 6 days in EApp and LApp, respectively, due largely to nitrification [28]. We did not measure the rate of nitrification in the present work, but the increases in NO₃⁻-N content and its share in TIN as well as the decrease in soil pH, which were more rapid and intensive in LApp than in EApp, likely indicate stronger nitrification following the former treatment. The lower nitrification in EApp might be explained by weaker microbial activity [30,31] as a result of lower soil temperature [32,33]. On the other hand, the decrease in NH₄⁺-N and increase in NO₃⁻-N share in TIN might be also explained to a certain extent by the preferential absorption of NH₄⁺-N over NO₃⁻-N [34,35], which was intensified under elevated temperatures in LApp. Our study identified strong correlations between N_dff in mature leaves, soil pH, and N availability, likely suggesting a direct linkage between soil N transformation processes and tea plant N uptake. The PLS-PM modelling showed that the soil pH value was one of the most important factors affecting the response of N uptake by tea plants to N topdressing timing. In the present work, we observed opposite effects of soil pH on the N_dff of mature leaves and N_dff of young shoots (Figure 6). The N_dff of mature leaves in the short term decreased with increasing soil pH, whereas the N_dff of young shoots increased with increasing soil pH in the long term after N topdressing. The PLS-PM modelling predicted synergy between N uptake by tea plants, N transformation, and soil pH in line with another recent work [36]. This synergy is closely related to the thermo-conditions of early and late N topdressing.

4.3. The Quality and Economic Value of Spring Tea in Response to Topdressing Timing

The quality of young spring shoots in tea plants is primarily determined by the concentrations of free amino acids, tea polyphenols, and caffeine, as well as TP/FAA ratio [37,38]. These key quality components critically influence the economic value of tea. This study showed that the type of cultivar significantly influenced FAA, TP, and the TP/FAA ratio in young spring shoots (Table 1), suggesting that genetic differences between tea cultivars play a major role in determining key quality components. In addition, N fertilization practices in tea plantations, including N application rate [39] and N fertilizer form (e.g., ammonium vs. nitrate) [35,40,41], also significantly modulate tea quality. In the present study, the effects of N application timing on the quality of young shoots were analyzed (Table 1), and we found that N application timing significantly affected TP content in shoots of JM1, with early application (EApp) resulting in lower TP compared to late application (LApp), suggesting that the bitterness of green tea can be reduced [42] and the quality of the tea thereby improved. Given that higher spring tea quality typically commands a higher market price and tea yield significantly increased in EApp compared with LApp, EApp could substantially enhance the economic value of spring tea production in cultivar JM1. However, ¹⁵N application date had no significant effect on tea quality parameters for the TGY cultivar, indicating that this cultivar’s biochemical composition is less responsive to N fertilization timing. Meanwhile, the yield only increased slightly; thus, considering the economic value of spring tea, the timing of N application has a relatively small impact on TGY. However, under EApp, the N_dff of young shoots significantly increased, indicating that the contribution of fertilizer N to the current season’s shoots was enhanced.

5. Conclusions

The results of this study indicate that the timing of N fertilizer before the spring tea harvest leads to differences in accumulated heat during fertilization and harvesting, affects soil N supply and N concentration in mature leaves, and indirectly impacts fertilizer N uptake by young shoots of the tea plant. Furthermore, topdressing N applied in late January (EApp) enhanced the utilization of fertilizer N in spring shoots across all tea cultivars compared to application in early March (LApp), irrespective of their sprouting times, under the conditions of this experiment. Specifically, the increases in biomass and total N accumulation in young spring shoots were greater in the early-sprouting cultivar than in the late-sprouting cultivar, while N derived from fertilizer and total uptake of fertilizer N in young spring shoots showed an opposite effect. Our findings highlight the significance of early N fertilizer application before spring tea harvesting for sustaining high yields, N accumulation, and N fertilizer efficiency in tea plants. This study provides valuable insights for optimizing the timing of N fertilizer application in order to enhance the uptake and utilization of fertilizer N, ultimately improving the yield of spring tea.