Relationships between the Water Uptake and Nutrient Status of Rubber Trees in a Monoculture Rubber Plantation

Abstract

Rubber cultivation is primarily rainfed agriculture, which means that water supplies are not stable in most rubber cultivated areas. Therefore, improving the water use of rubber trees through fertilization management seems to be a breakthrough for enhancing the growth and latex yield of rubber trees and carrying out the intensive management of rubber agriculture. However, the relationships among the nutrient status of rubber trees, their water uptake, and soil resources, including water and nutrients, remain unclear. To address this issue, we measured C, N, P, K, Ca, and Mg concentrations in soil and leaves, stems, and roots in a monoculture rubber plantation and distinguished the water uptake depths based on stable isotope analysis throughout the year. We found that the rubber trees primarily absorbed water from the 5–50 cm depth layer, and soil water and nutrients (usually N, P, K) decreased with depth. In addition, the water uptake depth of rubber trees exhibited positive correlations with the nutrient status of their tissues. The more water the rubber trees absorb from the intermediate soil layer, the more nutrients they contain. Therefore, applying fertilizer to intermediate soil layers, especially those rich in C content, could greatly promote rubber tree growth.

1. Introduction

In Southeast Asia, natural ecosystems are being damaged by the increasing demand for natural rubber (Hevea brasiliensis), which has resulted in rapidly expanding rubber plantations [1]. By 2012, rubber plantations had covered more than 1.5 million hectares in mainland Southeast Asia, including the Yunnan and Hainan provinces in China [2]. The rubber cultivated areas are expected to increase by more than two or three times by 2050. With the dramatic expansion of rubber plantations, serious ecological and environmental problems in rubber cultivated areas have received increasing attention worldwide, such as biodiversity loss, water and soil erosion, and fertilizer pollution [3,4]. All of these problems are related to the improper management of rubber plantations. However, there are several inevitable realistic factors for expanding rubber plantations. First, rubber agriculture is a typical resource-constrained industry due to the irreplaceability of planting regions in the tropics [5]; second, natural rubber is an important strategic material and a scarce resource crucial to national defense [6]; third, natural rubber generates considerable income for rubber farmers and promotes the development of the regional economy [6]. Accordingly, rubber agriculture has already become a highly competitive pillar industry in Southeast Asia, and the planting area continues to grow. Thus, current rubber agriculture appears to be in conflict with economic development and environmental protection [1].

Despite the fact that many initiatives have been proposed to change the adverse conditions of rubber agriculture, such as the establishment of rubber agroforestry systems, their practical promotion seems limited in the current situation. This is mainly because empirical researches and applications on the establishment and management of rubber agroforestry systems are still insufficient, especially for the selection and management of intercropped species [7]. Therefore, it would be more practical to search for effective management strategies to improve the productivity and latex yield of existing rubber plantations rather than simply expanding the rubber cultivated area to obtain a more intensive rubber plantation system [4].

The main water source for rubber plantations is precipitation [8]. As a result, rubber tree water supplies are unstable, especially during the dry season, when serious drought events frequently occur as a result of global climate change. Rubber trees are sensitive to soil water availability, and they can absorb large quantities of soil water due to their strong transpiration capacity and well-developed roots [9]. Therefore, most of the water in the soil supplied by rainfall would be consumed by rubber plantations directly or indirectly in the rubber-cultivated area. For example, water may be consumed via rubber tree transpiration, via soil evaporation in nearly bare soil, or via runoff loss, which varies greatly depending on the topography and soil texture of the planting field [7,10]. Therefore, the soil in which rubber trees grow has a lower moisture content than other soils in natural forests. Furthermore, the water demand of rubber trees is higher during the rainy season because the transpiration rates of rubber trees are higher during this period than during the dry season, since the trees have more leaves in the wet season than in the dry season [11]. Therefore, although rain is abundant during the rainy season, the water supply may not be sufficient due to the reliance of rubber trees on shallow soil water, which is prone to instability even in the rainy season [12].

It is necessary to note that the depth at which rubber trees take up water reflects the distribution of their root hairs, which are the most important organs for the absorption of water and nutrients in most plants. However, the distribution of rubber tree roots is affected not only by soil water availability but also by soil nutrients [13]. Therefore, understanding the relationship between rubber tree water use and the nutrient status of soil and rubber trees will help improve agricultural rubber management, especially fertilization management, and thus help rubber plantations maintain stable latex yield and growth under optimal soil water and nutrient conditions.

The growth and development of rubber trees depends largely on the combination and concentration of minerals in the soil. Due to the relative mobility of nutrients, rubber trees often face major challenges in obtaining sufficient nutrients to meet basic cellular processes. For plants, nitrogen (N) and phosphorus (P) are the most important nutrients in protein and genetic material composition [14,15,16], and carbon (C), which comes from assimilated carbon dioxide (CO₂) through leaf photosynthesis, is a substrate and energy source for various physiological and biochemical processes in plants [15]. These three nutrients are the most crucial elements for plant morphogenesis, and their absorption and allocation have been shown to be essential for all organisms [17,18,19]. In addition, potassium (K) is required for maintaining the plant structure and improving the root system, which allows the plant to absorb more water and nutrients, thus improving crop yield [13]; calcium (Ca) helps plants form cell walls and is essential to activate certain enzymes and coordinate certain cell activities [13]; and magnesium (Mg) greatly affects chlorophenol synthesis, photosynthesis, cell division, and DNA repair [13].

Because of the important functions of the above nutrients, fertilization is still a common management practice in current rubber agriculture, which has a positive and significant effect on latex yield, but irrigation is rare, as we discussed previously [20]. The most common rubber tree fertilizer used is compound N-P-K fertilizer [20,21]. Organic fertilizer (e.g., farmyard manure, compost, or water-logged compost; optimum fertilizer rate was more than 25 kg tree⁻¹ year⁻¹ for mature rubber plantations), which contains abundant C and is always mixed with fused calcium–magnesium phosphate or superphosphate calcium, is becoming increasingly important in current mature rubber plantations because of its important role in maintaining soil water and nutrient levels [20,21,22].

To increase the understanding of improved fertilization management for rubber plantation intensification, fertilization should address not only the nutrient status but also the water use of rubber trees. In other words, the analysis of the effects of water and nutrient levels on rubber tree growth should not be separated. The reason for this is that water regulates nutrient availability by controlling the moisture available for microorganisms, leaching, and plant uptake. As a result of their water uptake, plants also indirectly affect nutrient availability [23]. However, combined research on plant nutrient status and water use patterns and soil nutrient status is still insufficient [24], especially in rubber cultivation areas [3,25]. It is still unclear how the water use patterns or root uptake patterns of rubber trees, together with soil nutrient interactions, affect the nutrient status of rubber tree organs (e.g., leaves, stems, and roots).

In general, for better management, we must better understand the connection between the nutrient status and water use of rubber trees and the environmental concentrations of nutrients and water resources. To this end, we measured C, N, P, K, Ca, and Mg concentrations in the soil, leaves, roots, and roots of rubber trees in a rubber plantation, and we distinguished the water uptake depth of rubber trees in the rainy and dry seasons based on the stable hydrogen and oxygen isotope method. Therefore, to elucidate the interactions between rubber trees and soil in terms of water and nutrients, we addressed two main questions: (1) How are water and nutrients distributed throughout the soil profile? (2) How do environmental resources, including soil water and nutrients, and rubber tree water uptake influence the nutrient status of rubber trees?

2. Materials and Methods

2.1. Study Sites

The study site is located in Menglun town (21°55′39″N, 101°15′55″E), Xishuangbanna Prefecture, Yunnan Province, southwestern China. This region is affected by the tropical monsoon, which generates a rainy season (from November to April) and a dry season (from May to October), and the precipitation falling in the rainy season accounts for more than 80% of the annual precipitation [25,26]. To conduct periodic samplings, we selected a monoculture rubber tree plantation and established a standard and permanent quadrat (20 m × 20 m). The rubber tree planting density was approximately 3 m × 6 m. The slope aspect of the quadrat was approximately 102 degrees, and the gradient was 23 degrees. The rubber trees at this site were planted in the 1990s. Since the rubber trees were planted when they were fourth-year seedlings, their age during the sampling year was approximately 32 years. In the establishment stage of the rubber plantation (3–6 years), the optimum fertilizer rate was as follows: more than 15 kg tree⁻¹ year⁻¹ of organic fertilizer (farmyard manure, compost, or water-logged compost), 0.46–0.68 kg tree⁻¹ year⁻¹ of urea, 0.2–0.3 kg tree⁻¹ year⁻¹ of calcium superphosphate, 0.05–0.1 kg tree⁻¹ year⁻¹ of potassium chloride, and 0.1–0.15 kg tree⁻¹ year⁻¹ of magnesium sulfate [22]. In addition, base fertilizer (including organic and inorganic fertilizer) and topsoil were mixed evenly and then backfilled into the rubber tree planting hole (70 cm diameter × 70 cm deep) when establishing plantations [22].

Rubber trees during the sampling were approximately 20 m tall, and the average diameter at breast height was 28.6 cm. The rubber tree clone RRIM600 was planted. The soil type is classified as Oxisol (USDA-SCS, 1994) or Acric Ferralsols (World Reference Base), which are derived from Cretaceous yellow sandstone [27,28]. The soil profile was approximately 2 m deep, and intensive weathering and leaching left the soil with high levels of iron and aluminum oxides [28]. The soil pH was approximately 4.8, the bulk density was 1.32 g cm⁻³, the soil compaction was 43.85 kg cm⁻², and the total porosity was 55.02%. Every year, 40 kg of N-P-K fertilizer was applied as compound fertilizer in March and August per hectare. It is worth noting that the common fertilization ditch specifications for rubber trees are 20–30 cm depth, 30–40 cm width, and 50–80 cm length, and fertilization is conducted within the vertical projection of the crown perimeter, with fertilizer typically applied in the root zone, especially in the most intensive site of root absorption [21].

2.2. Sampling

The rubber tree quadrat was divided into sixteen 5 m × 5 m blocks. Soil samples were collected in four blocks at random on each sampling day. Using an earth-boring auger, soil samples from 6 different depths (0–5 cm, 5–15 cm, 15–30 cm, 30–50 cm, 50–75 cm, and 75–105 cm) were collected. The soil samples were mixed as quickly as possible and then divided into three parts [26]. To extract soil water, approximately 2 g of soil sample was sealed in a screw-top glass vial with parafilm and then kept in the refrigerator at −20 °C until extraction. Approximately 100 g of soil was sealed in a plastic Ziplock bag for oven-drying (105 °C), and the last part was used for soil nutrient concentration analysis.

For xylem sampling of rubber trees, the selected rubber trees (4 individuals) were screened randomly. Rubber tree xylem samples were collected through increment boring from their trunks at a height of 1.2 m. Three to four small xylem samples of 3.5 cm in length were collected from each sampled tree. The xylem samples were sealed in 15 mL clear glass tubes, sealed immediately with the screw cap and covered with parafilm, and then kept in the refrigerator at −20 °C until extraction of xylem water.

The shoots of the randomly selected rubber trees (one from each individual at each site) were collected for leaf and stem sampling. Using a 10 m tree pruner, we collected leaves mainly from sunny slopes near the upper canopy edges. After that, we divided the shoots into two parts, leaves and stems, which were placed in different envelopes for drying. Additionally, we collected lateral roots from the soil at a depth of 0–5 cm, and the diameters of the collected roots were approximately 5–10 mm. The collected roots were carefully rinsed in purified water after they were brought back to the laboratory and then dried at 65 °C in an oven to constant weight for at least 48 h.

Soil and plant samples were collected at four different times in the dry season (13 November 2017, 15 January 2018, 5 February 2018, and 15 March 2018) and at three different times in the rainy season (14 May 2018, 12 July 2018, and 25 September 2018).

2.3. Pre-Treatment and Measuring Methods

The water in the soil and xylem samples was extracted with a low vacuum distillation and extraction system at low temperatures (liquid nitrogen, −196 °C) and sealed in a 2 mL autosampler vial. To ensure that the water in the samples was extracted completely, the extraction lasted for at least 90 min. The values of δ²H and δ¹⁸O of the extracted water were determined by a Thermo-Finnigan Delta V Advantage Mass Spectrometer (Thermo Fisher Scientific, Waltham, MA, USA). The isotopic ratios are expressed in per thousand (‰) relative to V-SMOW (i.e., Vienna standard mean ocean water), and the measurement accuracy was better than ±1‰ for δ²H and was better than ±0.1‰ for δ¹⁸O.

To determine the soil water contents (SWCs), the soil samples were dried at 105 °C for at least 48 h. The samples of leaves, stems, and roots were dried at 65 °C to constant weight for at least 48 h. One week of air drying was performed on soil samples for the purpose of measuring nutrient concentrations. After that, a pulverizer was used to smash all these samples, and then an 80-mesh sieve was used to homogenize them to a fine powder. Using an elemental analyzer (Vario MAX CN; Elementar Analysensysteme GmbH; Hanau, Germany), we measured the total concentrations of C and N in all samples. Using an inductively coupled plasma atomic-emission spectrometer (Thermo Fisher; Waltham, MA, USA), the total P, K, Ca, and Mg concentrations of all samples were determined, the nutrient concentrations of plant leaves, stems, and roots were measured after digestion in HNO₃–HClO₄, and the nutrient concentrations of soil were also determined by this spectrometer after digestion in HClO₄–HF.

2.4. Water-Related Calculation

The gravimetric SWC was determined based on mass in this study. Its value is the mass of water (i.e., the difference in mass between the moist soil and the oven-dried soil dried at 105 °C) per mass of dry soil.

Where M_wet represents the mass of the sampled soil before drying, and M_dry represents the mass of oven-dried soil that has been dried for at least 48 h at 105 °C.

Deuterium excess (d-excess) is calculated using the following formula (Dansgaard, 1964):

d-excess = δ²H − 8δ¹⁸O

(1)

Isotopic mass balance [29] is used to determine the water use patterns of rubber trees, and the basic multiple linear mixing models [30] are:

$${{\delta }^2\mathrm{H}}_{\mathrm{plant}}{=f}_1{{\delta }^2\mathrm{H}}_{\mathrm{S}1}{+f}_2{{\delta }^2\mathrm{H}}_{\mathrm{S}2}{+f}_3{{\delta }^2\mathrm{H}}_{\mathrm{S}3}{+f}_4{{\delta }^2\mathrm{H}}_{\mathrm{S}4}{+f}_5{{\delta }^2\mathrm{H}}_{\mathrm{S}5}{+f}_6{{\delta }^2\mathrm{H}}_{\mathrm{S}6}$$

(2)

$${\delta }^{18}{\mathrm{O}}_{\mathrm{plant}}{=f}_1{\delta }^{18}{\mathrm{O}}_{\mathrm{S}1}{+f}_2{\delta }^{18}{\mathrm{O}}_{\mathrm{S}2}{+f}_3{\delta }^{18}{\mathrm{O}}_{\mathrm{S}3}{+f}_4{\delta }^{18}{\mathrm{O}}_{\mathrm{S}4}{+f}_5{\delta }^{18}{\mathrm{O}}_{\mathrm{S}5}{+f}_6{\delta }^{18}{\mathrm{O}}_{\mathrm{S}6}$$

(3)

$${1=f}_1{+f}_2{+f}_3{+f}_4{+f}_5{+f}_6$$

(4)

In Equations (2) and (3), δ²H_plant (or δ¹⁸O_plant) is the δ²H (or δ¹⁸O) value of plant xylem water; δ²H_S1–δ²H_S6 (or δ¹⁸O_S1–δ¹⁸O_S6) are the δ²H (or δ¹⁸O) values of water sources within the six sampled soil layers. In Equation (4), f₁–f₆ is the proportion of water used by plants from the six studied soil layers.

The contribution of n + 1 sources can commonly be determined by linear mixing models based on n isotope tracers. Whenever there are more than n + 1 sources, the linear mixing models become mathematically underdetermined, with more unknowns and equations than solutions [31]. That is, there was no unique solution provided by Equations (2)–(4) in this study. Hence, MixSIAR [32], a Bayesian mixing model based on statistical distributions, was used to characterize uncertainty in source and mixture isotopic values and estimate the contribution of source values [33] and then to quantitively determine plant water use in this study.

To run MixSIAR, xylem water isotope data were loaded as the mixture data, while soil water isotope data were loaded as the source data in “means and SDs” form, and the sampling dates were set as the fixed effects. When plants uptake water from the environment into the root, isotopic discrimination is too small to be noticed [29,30], so discrimination was set to 0. Furthermore, the Markov chain Monte Carlo (MCMC) run length was set to “long”. For plant species with many individuals, the error structure was set as “residual × process”. For plant species that have only one individual, the error structure was set as “process only”. Other settings were left at their default values. Final solutions were determined by averaging the proportions of plant water consumption from different water sources following MixSIAR.

2.5. Statistical Analyses

The general linear model (GLM) with the fixed effects of “season” and “depth” was used to analyze the differences in soil nutrient concentrations (including soil C, N, K, Ca, and Mg), SWC, soil water isotopic compositions (including soil water δ²H, δ¹⁸O, and d-excess), and plant water use proportions between seasons and among depths. Similarly, the differences in the nutrient concentrations among plant tissues (i.e., leaf, stem, and root) were also analyzed through GLM with “plant organs” as the fixed effect. Tukey’s post-hoc test was used to compare differences between groups when results were significant. The normal distribution of all the data were confirmed (as assessed by normal quantile plots with Lilliefors 95% confidence bounds). Levene’s test was used to determine the homogeneity of variances. Furthermore, Spearman’s rank correlation (p < 0.05) was used to identify the relationships between soil depths and soil nutrient concentrations (C, N, P, K, Ca, and Mg), SWC, and soil water isotopic compositions. If the correlations were high and significant, regression analysis was conducted to reveal the effects of soil depths in different seasons. Through regression analysis, the relationship between soil water δ²H and δ¹⁸O in different soil layers was also revealed, and the fitting functions were indicated as the soil water evaporation lines.

To demonstrate the relationships between plant water uptake and environmental resources, multiple correlation analysis was conducted between the nutrient concentrations in plant organs and soil.

All statistical analyses were performed using R 3.6.3 [34].

3. Results

3.1. Soil Nutrients

Different nutrients in the soil exhibited different concentration variations among the soil layers and between the seasons (Figure 1 and Figure 2). In general, C and N were concentrated mainly in the top and shallow soil layers and decreased significantly with soil depth (Figure 1a,b).

Furthermore, the concentrations of soil C and N were higher during the rainy season than during the dry season, especially for the soil at the 15–50 cm depth (Figure 1a,b). The high similarity in the variations in soil C and N concentrations was mainly due to their significantly positive correlations (r = 0.95, p ≤ 0.01; Figure 2a). The negative impact of soil depth on C and N concentrations was important, regardless of season (Figure 2a–c).

Soil depth also exhibited significant negative correlations (r = −0.7, p ≤ 0.01) with the concentrations of P (Figure 1c and Figure 2a,d). However, the soil P concentrations varied little between the seasons, compared with those of soil C and N (Figure 1c and Figure 2d). The concentrations of soil K showed increasing trends with soil depth (Figure 1d and Figure 2a), especially during the rainy season (Figure 2e). However, soil K seemed higher during the dry season than during the rainy season, typically for soil above 50 cm depth (Figure 1d). In addition, the variations in the concentrations of soil Ca and Mg among the soil depths were not significant (Figure 2a). For soil Ca, the concentrations were higher during the rainy season for soil within the 0–5 cm soil layer, but the concentrations were higher during the dry season for soil within the 15–105 cm soil layer. Differences in the concentrations of soil Mg were not significant between the dry and rainy seasons (Figure 1f).

3.2. Soil Water and Plant Water Use Proportions

The soil water δ²H and δ¹⁸O were highly correlated since the dual-isotope dots were closely distributed around a straight line (Figure 3). In addition, with the increase in the soil depths, the dual-isotope dots were more convergent, the fitting lines were shorter, and their slopes generally increased (Figure 3).

The SWC decreased with soil depth and was significantly higher during the rainy season than during the dry season, especially for soil above 50 cm depth (Figure 2f and Figure 4a). The d-excess of soil water was significantly lower in the dry season than in the rainy season, and it was significantly lower in the upper soil layers than in the other soil layers (Figure 4b).

Through the MixSIAR calculation, it was found that rubber trees mainly took up water from the 5–50 cm depth soil layers (approximately 70.1%; Figure 4c). The water use patterns of rubber trees also varied greatly between seasons. Generally, rubber trees absorbed more water from the 30–50 cm soil layers in the dry season and from the 5–15 cm soil layers in the rainy season. In addition, the stable water-absorbing zone for rubber trees was located in the 15–30 cm soil layer, which exhibited no significant difference between the dry season and rainy season, and the water uptake proportion from this soil layer was 23.9% on average.

3.3. Nutrients in Rubber Tree Tissues and Their Relationship with Soil Nutrients and Water and Plant Water Use

Compared to the other rubber tree tissues, the leaves had significantly higher concentrations of C, N, P, and K (Figure 5a–d). However, the Ca concentrations of rubber tree leaves were the lowest, followed by those in stems and roots (Figure 5e). In addition, the Mg concentrations of rubber tree leaves and roots exhibited non-significant differences, and they were significantly higher than those of rubber stems (Figure 5f).

Soil nutrient concentrations were not always correlated with the concentrations of the same nutrient in plant tissues (Figure 6), especially for C, N, and Mg. Only soil P and Ca exhibited significantly high correlations with P and Ca in rubber tree leaves, respectively (Figure 6e), and soil K had significantly high correlations with the root K of rubber trees (Figure 6d).

SWCs exhibited almost no significant correlations with the nutrient concentrations of rubber tree tissues, but the water use patterns of rubber trees exhibited significant positive correlations with the nutrient concentrations of rubber tissues, especially for P in all tissues. However, the nutrients in rubber tree stems exhibited nearly no significant correlations with rubber tree water use patterns.

In addition, the interactions among the soil nutrients, SWCs, and rubber tree water use patterns also exhibited different correlations with the nutrient concentrations in rubber tree tissues. However, a clear phenomenon was that the interactions between SWCs and rubber tree water use patterns have significantly positive correlations with the nutrient concentrations of rubber tree stems. High positive correlations can be found between the P concentrations of all rubber tree tissues and the interactions between their water use patterns and SWCs.

4. Discussion

4.1. Variations in Soil Nutrients

The C and N concentrations of soil are determined mainly by soil organic matter (SOM) because most of the N in soil is bound to soil C in organic form, which is one of the typical characteristics of SOM for most soils [35]. Therefore, the soil N variation was almost the same as that of soil C in the quadrat in this study (Figure 1a,b). This is why soil C and soil N had strong and positive correlations (Figure 2a).

The variations in C and N within the soil profile were mainly caused by the input, decomposition, and transport of SOM in the soil [36]. In general, gravity brings most leaves and other plant parts down to the soil, forming a litter layer above the surface soil where litter initially decomposes nutrients that are released. Since most decomposition occurs near the soil surface, where the input of plant residues is concentrated, soils C and N are generally higher in the topsoil layers. As a result, roots are able to access these nutrients by growing in surface soils. Consequently, root residues are also mainly produced in surface soil, strengthening the nutrient concentration of most decomposition surfaces. The mixing of soil and dissolved organic substances by animals, especially insects and worms, transfers surface C to the deep soil layers during leaching [36]. Therefore, soil C and N exhibited decreasing trends with increasing soil depths (Figure 2b,c). SOM decreased greatly with soil depth.

Seasonal changes in soil nutrients in different soil layers are the result of water movements in soil, which are normally controlled by precipitation. In short, the dynamics and flow of soil water facilitate four basic processes of soil formation: the transfer, transformation, addition and loss of soil components in soil profiles [37]. These processes determine soil chemical, morphological, and physical properties, such as depth-to-depth texture changes. Vertical downward movement or redistribution of soil water is called percolation. Percolation occurs when rainwater moves through soils by flowing downward (percolating) into subsoil layers, where it mixes with air spaces between soil particles and becomes part of the mass flow system [37]. In this process, some nutrients would be held by the SOM, which is typical of leaching of soluble nutrients from the topsoil. This process can be accelerated by high temperatures during periods of intense rainfall (i.e., the rainy season) [38,39]. That is, soil nutrients could also be affected by temperature directly or indirectly. For example, soil temperature in tropical regions always decreases with depth; because of the interaction of soil temperatures and humidity, the abundance of soil microorganisms decreases as the depth decreases, and organic decomposition, therefore, decreases [40,41], thus causing nutrients to decrease further with soil depth [42]. Therefore, nutrients would be enriched overall if more SOM could be stored in the intermediate or deep soil layers. For this reason, organic fertilizer, which could significantly increase the SOM content within the soil, is becoming increasingly important in modern agriculture [43] and would be a promising way to help improve the soil quality in rubber plantations [28].

In addition, rainy seasons have higher rainfall and temperatures than dry seasons, and the general concentrations of C, N, and P in soil were high during the rainy season and decreased with the depth of the soil (Figure 1a–c and Figure 2a–d). However, soil K, Ca, and Mg exhibit little variation with depth. We assume that this is mainly because the concentrations of these nutrients that we measured were the total nutrient concentrations, which account for the content of these elements, resulting from the weathering of soil minerals.

4.2. Variations in Soil Water Content and Isotopic Compositions

The dynamics of water in the soil are governed by many factors that change vertically as depth increases, vertically as terrain changes, and temporally as climate changes [44]. In this study, SWCs during the rainy season were higher than those during the dry season (Figure 2f and Figure 4a). It is unquestionable that frequent precipitation during the rainy season led to such a phenomenon (Huang et al., 2010).

The original factor affecting the vertical variations in water within the soil is the amount of water that enters the soil, which is typically dependent upon the amount of precipitation but also on the terrain, soil texture, SOM, and other factors [39]. Another important reason for the decrease in water content with increasing depth is soil evaporation, which is less efficient at deeper depths. In this study, SWCs decreased with increasing soil depth, and such variation trends were significant in the rainy season but not in the dry season (Figure 2f). Perhaps water isotopes (i.e., δ²H and δ¹⁸O) could explain this phenomenon, since measurements of water isotope compositions provide insights into the study of hydrological cycles at multiple scales [45,46].

In general, the greater the seasonality in evaporative fractionation, the more discrete the dual-isotope dots will be in the soil water isotope plots. Additionally, water parcels evaporate along an evaporation line, where the slope depends on the relative evaporation rates of the various water isotopes [47]. Intense evaporation will result in a shallower slope for this evaporation line. In other words, in dry conditions, the evaporation line has a flat slope. It is likely that climate seasonality will have a greater impact on the isotopic composition of evaporating soil waters and evaporative losses, both of which will decrease soil evaporation slopes [47]. Therefore, the decreasing slopes of the soil evaporation lines with soil depth (Figure 3) indicate that soil evaporation became weaker with increasing soil depth. It is clear that the upper soil cover weakens evaporation. The other indicator of soil evaporation is d-excess. Evaporation commonly results in a lower d-excess value [46,48]. Therefore, the higher d-excess of soil water during the rainy season (Figure 4b) indicated that soil evaporation was less intense and rainwater inputs were more abundant during the rainy season than during the dry season. This is typically due to the seasonality of the climate, and we also found that the soil evaporation decreased with increasing soil depth. In addition, roots and organic matter also decreased with increasing soil depth, resulting in a decrease in soil porosity, which slows soil water infiltration [49,50]. Therefore, with increasing soil depth, the seasonal variation in the SWC decreases (Figure 4a).

Due to the change in soil water, the proportion of water use by rubber trees also changed dramatically. The 5–50 cm depth soil layer is the main absorbing zone of rubber trees, and the 15–30 cm depth soil layer provides a stable water source (Figure 4c), and in the rainy season, the trees used more water from the shallower soil layer, while in the dry season, they used more water from the deeper soil layer. This phenomenon demonstrates that the root distributions of rubber trees are plastic and have a great effect on the seasonal availability of water. However, the distributions of the roots were a compromise between various environmental resources and conditions. This means that water was not the sole determinant of these changes, and nutrients within the soil also played an important role.

4.3. Variations in Plant Nutrient Status

Nutrients and water are absorbed by terrestrial plants primarily through the roots, while photosynthesis (the creation of energy and organic compounds) mainly occurs in the leaves. Therefore, nutrients, especially N, P, and K, are typically concentrated in leaves because of the importance of these elements in metabolism; as a result of the large percentage of dead cells in xylem, wood, such as stems, has low concentrations of N and P, and the same applies to the root system except for fine roots and other young tissues [15]. Therefore, the concentrations of N, P, and K were higher in the leaves of rubber trees (Figure 5a–d). As a result of its association with cell walls, Ca makes up a greater proportion of plant nutrients in wood than in any other part of the plant. Therefore, the Ca concentrations were lower in the leaves of the rubber trees (Figure 5e). In addition, as an important component of chlorophyll, most of the Mg in plants is found in leaves containing chlorophyll followed by roots, since they are in charge of resource uptake. Therefore, Mg concentrations were lower in the stems than in the leaves and roots of the rubber trees (Figure 5f).

Photosynthesis is influenced by the environmental factors that regulate the leaf surface and the seasonality of plant growth. Plant growth is initially supported by stored reserves of C and nutrients acquired in previous years. Early in the growing season, each part of the plant competes for limited carbohydrate and nutrient supplies, resulting in a seasonal progression of production of different plant parts. As an example, leaves are the first to be produced, followed by roots and finally wood or stems [15]. In the dry season (November–February), rubber trees shed their leaves, followed by a flushing of leaves in the rainy season (February–April) [11,51]. The concentrations of N, P, and K in rubber tree tissues, therefore, decrease when leaves shed (November–February) and increase when leaves flush (February–April). In addition, the amount of nutrients available to the plant at any one time depends on how much light, water, and nutrients are available to the plant [15]. If there is not enough water, then the plant will not grow as well as it could have done despite the nutrient amount being sufficient. This can affect both photosynthesis (the process that makes sugars from carbon dioxide) and respiration (the process that uses sugars). Therefore, it is important to understand the relationship between root absorption, plant tissue nutrients, soil water content, and soil nutrient concentrations for rubber tree growth.

4.4. Relationship among Environmental Resources, Plant Root Uptake, and Nutrient Status

Based on stable hydrogen and oxygen isotope compositions (i.e., δ²H and δ¹⁸O) and the results of MixSIAR, rubber trees mainly take up water from the intermediate soil layers (5–50 cm deep soil layer), and the water uptake proportions of rubber trees were positively correlated with the nutrient concentrations of their organs, especially for the leaves (Figure 6). The interaction between SWC and plant water use was highly correlated with the nutrient concentrations in plant organs. This means that the more water is absorbed by rubber trees from the intermediate soil layers (5–50 cm deep soil layers) and the higher SWC in these soil layers, the more nutrients that rubber trees can absorb and transport to leaves.

In general, soil moisture affects plant growth by affecting root distribution and transpiration rate [26]. More specifically, the root distribution determines where plants take up nutrients within the soil profile and determines where they are most vulnerable to drought stress [15]. In general, shallow-rooted plants are more susceptible to drying than deep-rooted plants due to the lack of surface area exposed to air flow during periods when there is little rainfall or irrigation [52]. However, soils, typically in the deep soil layers, that have been compacted over time lose some pore space between particles, making it difficult for rainwater to enter into them, thereby reducing the available water storage capacity. For this reason, rubber trees take up most of their water from the intermediate soil layers (5–50 cm deep soil layers) and greatly depend on water from the 15–30 cm deep soil layers. In addition, by altering the rate of nutrient radial transport through the apoplasm, the transpiration rate may directly influence nutrient uptake [53], and root cells are directly and indirectly affected by the effects of transpiration rate on nutrition supply [54,55]. Reducing transpiration rates might help reduce the nutrient depletion in the rhizosphere caused by plant uptake [56,57]. Therefore, plant water uptake greatly affects the uptake of nutrients.

It is important to note that while water movement in the plant, especially in the xylem, can significantly affect nutrient transport, water uptake is not the same as nutrient ion uptake [13]. There is no direct correlation between high water uptake and high nutrient ion uptake. Mineral transportation between mineral uptake and production sites and mineral consumption sites is essential for higher plants, and the transportation process depends heavily on water. Long-distance transport is primarily accomplished by xylem and phloem, and organic compounds, inorganic ions, and water are the three primary materials transported [15]. Water and minerals are transported from the roots to the leaves through the xylem. In summary, the uptake of rubber trees indicates the depth of its roots, and the depth of its roots represents a compromise between water and nutrient availability. As a result, fertilization management influences the water use of rubber trees and helps them improve their efficiency without irrigation.

5. Conclusions

The majority of nutrients were concentrated in the topsoil layer and decreased as depth increased in the monocultural rubber plantation, and most nutrients were abundant during the rainy season compared to the dry season. In the variation process of soil nutrients, soil water plays an important role in the modification of soil nutrient content. In addition, soil water variations among the seasons make rubber trees adjust their main water-absorbing zone within the soil, and the nutrient status of their organs also appeared to be affected by the water uptake of rubber trees. There is a direct relationship between the amount of water rubber trees absorb from the intermediate soil layers and the amount of nutrients concentrated in their tissues. Therefore, we suggest that more fertilizer should be applied to the intermediate soil layers in mature rubber plantations, especially those rich in C content. This study, therefore, provided an important update on the understanding of the coupling of water uptake and the nutrient status of rubber trees, and these findings would be helpful for the intensive management of current rubber agriculture, especially for fertilization in rainfed regions.