Dynamics of water uptake by maize on sloping farmland in a shallow Entisol in Southwest China

Catena 147 (2016) 511–521

Contents lists available at ScienceDirect

Catena journal homepage: www.elsevier.com/locate/catena

Dynamics of water uptake by maize on sloping farmland in a shallow Entisol in Southwest China Pei Zhao a,b, Xiangyu Tang b,⁎, Peng Zhao c, Jialiang Tang b a b c

College of Urban, Rural Planning and Architectural Engineering, Shangluo University, Shangluo 726000, China Key Laboratory of Mountain Surface Processes and Ecological Regulation, Institute of Mountain Hazards and the Environment, Chinese Academy of Sciences, Chengdu 610041, China College of Water Resource&Hydropower, Sichuan University, Chengdu 610065, China

a r t i c l e

i n f o

Article history: Received 10 February 2016 Received in revised form 31 July 2016 Accepted 2 August 2016 Available online 11 August 2016 Keywords: Immobile soil water Maize water use Mobile soil water Shallow soil Stable isotopes

a b s t r a c t The water use patterns of maize grown in shallow soils remain poorly understood. To explore the water uptake dynamics of maize from a loamy Entisol with an average thickness of 40 cm, excavation, isotopic tracer analysis and soil water potential measurements were combined to study the source of water used by maize. The differences between the δD of the water used by maize and the δD of the mobile soil water (MSW, the fraction of soil water with high mobility which can be easily replaced by the infiltrating rainwater) indicated that maize did not take up much MSW. The local meteoric evaporation line, the δD of the bulk soil water (BSW, total soil water including mobile and immobile soil water) and the soil water collected using a lysimeter were used in a model to calculate the isotopic compositions of different fractions of soil water and the proportion of immobile soil water (ImSW, the fraction of soil water with little mobility which was tightly bound to the soil particles). ImSW resulted from several heavy rains that occurred before the sampling. The primary water sources for maize varied temporally and spatially. Maize seedlings at the one-leaf stage used ImSW from 0 to 10 cm soil depth; however, maize plants generally used more BSW from deeper soil layers when the roots reached greater depths. The ratios of MSW to ImSW were not equal between the maize stems and soil, with more ImSW in the maize stems, particularly during the seedling stage. This result invalidates the core concept of most watershed hydrology models and classical hypothesis in the isotopic models of general atmospheric circulation. The difference between the MSW and ImSW in the water cycle of the soil-plant-atmosphere continuum should be considered in the future studies of identifying the plant water sources and modeling hydrological processes. © 2016 Elsevier B.V. All rights reserved.

1. Introduction It is important to clarify how crops use various water sources to understand the soil- plant- atmosphere continuum (SPAC) hydrological cycle in a given agricultural system. The variation and ecological plasticity of the water-uptake depth by plant is an important means of evaluating the vegetation controlling of the hydrological balance in agricultural landscapes (Asbjornsen et al., 2008; Zhang et al., 2011b). This knowledge is also helpful for understanding the responses of crops to changing water conditions due to climate change and human activities. Plant access to available water sources depends on the root distribution along the soil profile. Although excavating plant root systems is

Abbreviations: BSW, Bulk soil water; MSW, Mobile soil water; ImSW, Immobile soil water; IR, Isotopic ratio; SPAC, Soil-plant-atmosphere continuum; LMWL, Local meteoric water line; V-SMOW, Vienna-Standard Mean Ocean Water; GMWL, Globe Meteoric Water Line. ⁎ Corresponding author. E-mail address: [email protected] (X. Tang).

http://dx.doi.org/10.1016/j.catena.2016.08.001 0341-8162/© 2016 Elsevier B.V. All rights reserved.

time intensive and cost prohibitive, this method has been traditionally and widely used to study plant water uptake patterns in various ecosystems (Dahlman and Kucera, 1965; Jackson et al., 1996). However, Dawson and Pate (1996) and Thorburn and Ehleringer (1995) concluded that the amount and spatial distribution of fine and coarse roots do not accurately reflect the primary water sources utilized by the roots in different soil layers. Recent studies involving plant water use indicate that measuring the natural abundances of the hydrogen and oxygen stable isotopes in plant stem water and in potential water sources is a powerful and nondestructive method for determining the sources of water used by different plant species (Ehleringer and Dawson, 1992; Nie et al., 2012). The application of this technique has yielded unexpected results regarding the water use strategies of plants in many situations (e.g. Dawson and Ehleringer, 1991; Brooks et al., 2010). These results indicate that the functions of root systems in soils are not fully understood (Gardner, 1991; Brunel et al., 1995; Midwood et al., 1998). For example, the roles of mobile soil water (MSW, the fraction of soil water with high mobility that can easily be replaced by infiltrating rainwater, e.g., the soil water that can be extracted under 80 kPa of suction) and immobile

512

P. Zhao et al. / Catena 147 (2016) 511–521

soil water (ImSW, the fraction of soil water with little mobility that is tightly bound to the soil particles) in plant water use strategies remain controversial (Brooks et al., 2010). In addition, a lack of knowledge regarding the root distribution in the soil profile often limits accurate identification of the plant water source when using isotope methods. For example, Bariac et al. (1994) observed that the root density in the soil and the isotopic profile of the soil water were considerably heterogeneous with depth, thus the isotopic data were insufficient for accurately identifying the plant water source. For annual plants, such as maize, roots gradually grow downwards to a maximum depth, which determines the depth of the accessible water sources in the soil profile. The, direct inference method which is according to the intersection of the isotopic values of plant stem water and soil water to identify the plant's main water source was widely used in plant water sources studies (Brunel et al., 1995). However, the isotopic profiles of soil water can be curved in shape, particularly in shallow soils (Zhao et al., 2013a). For example, the occurrence of two or more intersections of plant water and soil water isotopic values may obscure the identifying the primary water source used by the plant. Consequently, using stable isotope analysis combined with an excavation method which determining plant root depth and biomass throughout the soil profile could be useful for accurately identifying the primary water sources of crops during different growing seasons. Maize is one of the most important crops grown in the hilly area of southwest China, which is dominated by purple soils. Purple soils have a small available water capacity of 0.06–0.11 cm3·cm−3 (Wang, 2013), and 73% of the sloping farmland in this region has a soil thickness of 20–60 cm. Soil b 60 cm significantly decreased maize yield in this shallow Entisol (Zhu et al., 2009). Cakir (2004) also found that short-duration water deficits during the rapid vegetative growth period caused 28–32% loss of maize yield in such Entisol. This was explained by that the shallow soil layer limits water storage and root penetration (Zhao et al., 2013b). Furthermore, seasonal droughts that decrease crop yields frequently occur in this area. Understanding the water uptake patterns of maize in such shallow soils is important for agricultural water management and hydrological balance in agricultural landscapes. However, the water uptake dynamics of maize in such shallow agricultural soil systems remain unclear. δD and δ18O analysis with excavation methods were used to assess the proportions of water uptake by maize from different soil horizons on sloping farmland at a hilly area in southwest China with purple soils. The objectives of this study were to (1) compare the isotopic signatures of the rainwater, bulk soil water (BSW), MSW, and maize stem water in the area to infer the relationships among potential water sources and maize stem water; and (2) determine the water source for maize and the probabilities of the contributions from soil water in different depths to the maize water in purple soils using single-source direct inference and an alternative multiple-source massbalance approach. 2. Materials and methods 2.1. Study area This study was conducted at the Yanting Agro-ecological Experimental Purple Soil Station of the Chinese Academy of Sciences, which is located in a hilly area of southwest China dominated by purple soils (31°16′ N and 105°28′ E) (Fig. 1a). This region is characterized by a moderate subtropical monsoon climate, with an annual mean temperature of 17.3 °C and a mean annual precipitation of 826 mm (from 1981 to 2006). From 1981 to 2006, 65.5% of the annual precipitation occurred during the summer. The experimental soil is classified as an Entisol according to Food and Agriculture Organization (FAO) Taxonomy (Gong, 1999; Wang, 2013; Zhao et al., 2013a). The bulk density, porosity, particle size distribution and organic matter content are shown in Table 1 under crop conditions after one month of tillage.

2.2. Sampling and measurements Field experiments were conducted in a sloping farmland plot that measures approximately 50 m in length and 30 m in width at the experimental station and has a shallow soil profile (average thickness is 40 cm). An overhead view of the study plot is shown in Fig. 1c. The distance of the slope toe to the downslope position of the sloping farmland was approximately 20 m. The crest of the slope was on the upslope position of the sloping farmland. Maize (Zea mays L.) was sown on May 23, 2013, at a density of four plants per square meter and was harvested on September 12, 2013. The sampling campaigns were conducted on rainless days or 2–3 days after heavy rainfall. The first internodes of each maize stem were collected at 8 a.m. and stored in glass bottles for plant water sampling. The sampling dates were June 6 (one-leaf stage), July 2 (five-leaf stage), July 27 (elongation stage), August 6 (tasseling stage) and August 29 (grain mature stage) in 2013. Three maize stems were sampled from the upslope, midslope and downslope areas, respectively. Soil samples were obtained near the stem sampling locations (b1 m apart) using a hand-operated auger at depths of 0–5, 5–10, 10–20, 20– 30 and 30–40 cm. At each slope position, suction lysimeters (Soilmoisture Equipment Corp., CA, USA) with a diameter of 48 mm were installed at 2.5, 7.5, 15, 25 and 35 cm. Soil water samples were obtained from the lysimeters at a pressure of −80 kPa after 8 h of equilibrium. The soil water and maize stem samples were collected simultaneously. During the maize growing season, each rainfall event was sampled using a glass funnel (20 cm diameter) that was connected to a high-density polyethylene bottle. A table tennis ball was placed in the funnel to reduce evaporation. In addition, water was collected from the pools in the study catchment to build the local meteoric evaporative line (LMEL). Next, 25 pools were sampled on July 10, 2013. For isotope analysis, the soil, maize stem, and lysimeter soil water samples were placed in airtight glass vials and immediately sealed with airtight caps and parafilm to prevent evaporation. The samples were stored in a refrigerator at 4 °C until analysis. To excavate the maize roots, a pit of one square meter was dug to a depth of 40 cm (bedrock depth) in the field. Maize roots were collected on the same day as the plant stem samples using a depth interval of 10 cm and dried at 75 °C for 48 h until a constant weight was achieved. The soil water potential was measured using tensiometers (Type T4e, UMS-GmbH, Munich, Germany). Four tensiometers were installed at the middle of each layer at the upslope, midslope and downslope positions along the 0–40 cm soil profiles to represent the soil water potentials at depths of 0–15, 15–25, 25–35 and 35–40 cm, respectively. No data were recorded at depths of 25–35 and 35–40 cm on the downslope during the one-leaf stage because of equipment failure. The BSW (bulk soil water, total soil water including mobile and immobile soil water) and maize stem water were extracted using a cryogenic vacuum distillation method (Ehleringer and Osmond, 1989). δD and δ18O analysis were conducted using an L2120-i analyzer (Picarro, USA). The isotope ratios (2H/1H and 18O/16O) are expressed per-mille (‰) for δD and δ18O and are defined relative to the Vienna-Standard Mean Ocean Water (V-SMOW) (Eq. 1). The analytical precision for each sample was 0.2‰ for δ18O and 0.5‰ for δD, which were defined as follows: δD or δ18 Oð‰Þ ¼ ðRSAMPLE =RV‐SMOW −1Þ  103

ð1Þ

where RSAMPLE and RV-SMOW are the D/H or 18O/16O ratios of the sample and the V-SMOW, respectively. 2.3. Analysis method First, the hydrogen isotopes were directly compared between the soil water profile and the maize stem water. The depth at which the soil water and stem water isotopic values were similar indicated the

P. Zhao et al. / Catena 147 (2016) 511–521

513

Fig. 1. The distribution of purple soil in China and the study site at the hilly area of purple soil, SW China (a), the study catchment (b), the top view of study sloping farmland (c).

514

P. Zhao et al. / Catena 147 (2016) 511–521

Table 1 The bulk density, porosity, particle size distribution and organic matter content of the study shallow soil at the sloping farmland.

Upslope

Midslope

Downslope

Depth (cm)

Bulk density (g/m3)

Porosity (%)

Clay (0–2 μm)

Silt (2–50 μm)

Organic Matter

0–10 10–20 20–30 30–40 0–10 10–20 20–30 30–40 0–10 10–20 20–30 30–40

1.32 1.64 1.63 1.84 1.16 1.62 1.76 1.83 1.44 1.62 1.70 1.77

50.3 38.1 38.5 30.4 56.3 38.8 33.7 31.1 45.7 39.0 35.8 33.1

9.7 ± 0.1 12.0 ± 0.3 10.9 ± 0.2 / 10.2 ± 0.2 13.4 ± 0.2 11.4 ± 0.1 / 11.0 ± 0.2 13.3 ± 0.2 12.3 ± 0.1 /

53.9 ± 0.3 57.1 ± 0.3 58.7 ± 0.7 / 56.3 ± 0.2 55.8 ± 0.5 58.0 ± 0.6 / 57.7 ± 0.5 52.3 ± 0.4 57.3 ± 0.3 /

8.64 ± 1.35 7.37 ± 2.60 5.41 ± 2.44 / 10.78 ± 0.52 9.13 ± 2.07 6.57 ± 2.26 / 12.51 ± 0.35 11.37 ± 1.91 8.03 ± 2.21 /

depth at which the soil water was utilized by crops. For two potential water sources, a linear mixing models based on isotopic mass balance was used to calculate the water source proportion. If three or more water sources existed, the IsoSource model was used to calculate the proportional contributions of the various water sources in maize water. In this model, all possible combinations of each source contribution (0–100%) are examined in a small increment (e.g. 1% in this study). Groups that sum to the measured mixture isotopic ratios within a small tolerance (e.g. 0.2 in this study) are considered to be feasible solution, from which the frequency and range of potential water source contributions can be calculated (Phillips and Gregg, 2003). The layers that did not contain root biomass were excluded as possible sources. The fractional increment and the uncertainty level were set at 1% and 0.2 in this study, respectively. If the isotopic soil water profile did not intersect the maize stem water profile, it was assumed that the maize used ImSW. The soil water collected from the soil cores was considered BSW, and the soil water collected using suction lysimeters was considered MSW (Brooks et al., 2010; Zhao et al., 2013b). We hypothesized that the relationship between δD and δ18O in the ImSW would fit the LMEL, where δD = 6.17δ18O − 5.77 (R2 = 0.96). This equation was obtained by performing a linear regression of the δD and δ18O in the 25 water samples of the pools in this catchment. Then, the isotopic composition and proportion of ImSW in the soil were calculated as follows: δ18 Ob ¼ f 1 δ18 Om þ f 2 δ18 Oim

ð2Þ

δDb ¼ f 1 δDm þ f 2 δDim

ð3Þ

f1 þ f2 ¼ 1

ð4Þ

δDim ¼ 6:17δ18 Oim −5:75

ð5Þ

where the subscripts b, m and im represent the BSW, MSW, and ImSW,

respectively, and f1 and f2 represent the proportions of MSW and ImSW, respectively.

3. Results 3.1. Isotopic compositions of water sources Fig. 2 shows the linear regressions of δD and δ18O for the rainwater, maize stem water, BSW, and MSW from May 25th to August 30th, 2013. The δD values of the precipitation ranged from − 115.26 to 10.56‰, with a mean value of −72.00 ± 34.52 (SD)‰, whereas the δ18O values ranged from − 16.11 to 2.01‰, with a mean value of − 10.75 ± 4.33 (SD)‰. Fig. 3 displays the temporal changes of the amount and the δD of the rainfall during the maize growing season. The δD of rainwater was temporally variable, with a depleted trend from spring to autumn. Heavy rainfall always resulted in the depletion of the isotopic ratios (IRs). The linear fit between the δD and δ18O in the rainwater was as follows: δD = 7.85δ18O + 12.29 (R2 = 0.97). The slope was b 8 (GMWL: δD = 8δ18O + 10) during the growing season and this is attributed to evaporation of falling raindrops in air masses, The higher intercept than 10 shows the direct evidence of post-rainfall isotopic fractionation in this region. Table 2 shows the mean, standard deviation, minimum and maximum values of the various water bodies. The δD of the BSW ranged from − 107.94 to − 17.53‰, with an average of − 75.38 ± 28.22 (SD)‰, whereas the δ18O ranged from −15.18 to −3.40‰, with an average of −11.00 ± 3.60 (SD)‰. The amplitude variation of the isotopic values in the BSW was smaller than the values in the rainwater. The linear equation of δD and δ18O in the BSW was δD = 7.62δ18O + 8.30 (R2 = 0.98). Most of the BSW samples fell to the right side of the local meteoric water line (LMWL), and the slope and intercept of the equation generated from the samples were lower than the slope and intercept of the LMWL, reflecting the effects of evaporation on rainwater.

Fig. 2. Linear regressions of δD on δ18O for the precipitation, bulk soil water, mobile soil water and maize stem water during the growing season of maize.

P. Zhao et al. / Catena 147 (2016) 511–521

Date 20/6/13

20/7/13

20/8/13 20 0 -20 -40 -60 -80 -100 -120 -140

Rainfall (mm)

0 20 40 60 80 100 120

D (‰)

20/5/13

Fig. 3. The amount and Deuterium concentration of the rainfalls during the growing season.

The δD of the MSW ranged from −105.31 to −11.61‰, with a mean value of − 66.95 ± 28.24 (SD)‰, and the δ18O values ranged from −14.90 to −2.84‰, with a mean value of −10.10 ± 3.64 (SD)‰. In addition, smaller variations of δD and δ18O amplitudes were observed in the MSW than in the rainwater. However, the variation amplitudes of δD and δ18O in the MSW were greater than the values in the BSW. This finding indicated that the MSW was not well mixed with the “old” soil water. The following linear regression of δD and δ18O in the MSW was obtained as δD = 7.72δ18O + 10.78 (R2 = 0.99). The slope and intercept of this equation were similar to the slope and intercept of the LMWL, indicating that the MSW was mainly derived from recent rainfalls and was minimally affected by the evaporation process. The δD of maize stem water ranged from − 104.66 to − 34.38‰, with a mean value of − 77.98 ± 18.10 (SD)‰, and the δ18O values ranged from − 15.66 to − 4.69‰, with a mean value of − 12.16 ± 2.83 (SD)‰. Smaller variations of the amplitudes of the IRs were observed in the maize stem water than the values in the rainwater, BSW and MSW. The maize acquired water from source water with depleted IRs which was heavy rainfalls. The linear equation obtained from fitting δD and δ18O in the maize stem water was δD = 6.18δ18O − 2.81 (R2 = 0.94). The slope and intercept values obtained from the equation are less than the slope and intercept values obtained from the rainwater, BSW and MSW and similar to the slope and intercepts of the LMEL, implying that the potential water sources for the maize had evaporative enrichment characteristics. 3.2. Isotopic ratios of the maize stem water and soil water in the soil profile Fig. 4 shows the root biomass, the δD of the BSW and MSW, soil water potential distribution at different depths and the corresponding maize stem water at the one-leaf stage for the upslope, midslope, and downslope. The root dry mass was 1.0 ± 0. 2 g·m− 3 and roots only grew in the 0–10 cm soil layer. The δD of the BSW in the 0–5 and 30– 40 cm layers was lighter relative to the δD in the other soil layers, showing a convex shape. The δD of the MSW was lighter than that of the BSW, particularly at a depth of 0–20 cm. Only a small difference was observed between the IRs of the BSW and MSW at 20–40 cm depth. Furthermore, the curve plotting the δD of the MSW along soil profile was convex. However, the soil water content decreased as the depth increased, and Table 2 The isotopic ratios of various water bodies at the slope farmland. Water body

Isotopes

Mean (‰)

SD (‰)

Min (‰)

Max (‰)

Precipitation

δD δ18O δD δ18O δD δ18O δD δ18O

−72.00 −10.75 −75.38 −11.00 −66.95 −10.10 −77.98 −12.16

34.52 4.33 28.22 3.60 28.24 3.64 18.10 2.83

−115.26 −16.11 −107.94 −15.18 −105.31 −14.90 −104.66 −15.66

10.56 2.01 −17.53 −3.40 −11.61 −2.84 −34.38 −4.69

BSW MSW Maize stem water

515

the soil water content on the upslope was lower than that of the midslope and downslope. The δD of the maize stem water ranged from −34.4 to −37.0‰ at the three positions and was lighter than the δD values of the BSW and MSW at 0–20 cm. At the upslope and midslope, the δD of the maize stem water was similar to the value of the BSW at a depth of 30– 40 cm. At the downslope, the IR of the maize stem water was within the scope of the BSW at 20–40 cm depth, where the δD ranged from −24.4 to −44.6‰. In addition, these layers had higher water content. The intersection of the IR of the maize stem water in the vertical direction and the soil water IR profile was considered as the main location of plant water uptake (Wang et al., 2010). The inferred main depth of water uptake by maize was 30–40 cm at the upslope and midslope positions and 20–40 cm at the downslope position at this stage. However, the average root depth measured by excavation at this stage was 9.8 ± 2.2 cm. Consequently, maize cannot acquire water from a depth of 20– 40 cm. Comparison of the IRs of the maize stem water, BSW and MSW did not indicate the soil depth of water uptake. A comparison of the IRs in the BSW and MSW indicated that the ImSW had a lighter isotopic composition; thus ImSW could be the fraction of soil water that was used by the maize seedlings. The IRs of the ImSW was calculated using Eqs. (2)–(5) when δDim and δ18Oim were located exactly on the LMEL. The IRs of the ImSW revealed that the δD ranged from −49.9 to −23.3‰, with more depleted IRs in the topsoil. The topsoil had lower water contents than the deep soil, implying that the soil water in the topsoil was mainly immobile. At the three positions, the δD of the maize stem water approached the values of water at 5–10 cm depth, which indicated that the seedlings mainly used the ImSW from this depth. Fig. 5 shows the root mass, the δD values of the maize stem water, BSW and MSW, and the soil water potential in the soil profiles at different positions when the plants reached the five-leaf stage and the roots had reached a depth of 20 cm. The δD values of the BSW and MSW were depleted relative to the one-leaf stage due to the recent rainfall, showing a linear trend along the soil profiles. As shown in the Fig. 5e, the soil water content was high. The soil water distribution in the soil profiles followed a similar pattern to that of the δD values. The difference in the IRs between the BSW and MSW was evident in the soil profiles, especially on the downslope. Using the direct inference approach, the maize mainly acquired BSW from a depth of 5–10 cm in the upslope and midslope areas and from a depth of 10–20 cm in the downslope areas. Apparently, the maize did not acquire the MSW from the soil because no intersection on the isotopic profiles between the δD in the maize stem water and MSW was observed. Fig. 6 shows the root biomass, the δD of the maize stem water, BSW and MSW, and the soil water potential in the soil profiles at the three positions at the elongation stage. The maximum root depth reached 30 cm. At the upslope and midslope locations, the MSW became lighter at a depth of 0–10 cm than the BSW due to replenishment by recent rainfalls, with an average δD of −106.3‰. The difference between the IRs of the BSW and MSW was similar to the difference observed at the five-leaf stage. Deeper soil depth behaves higher soil water content. During this stage, it was directly inferred that the maize mainly acquired water from a depth of 20–30 cm. Fig. 7 shows the root biomass, the δD of the BSW and maize stem water, and the soil water potential at the three positions at the tasseling stage. The maize roots reached a depth of 40 cm, and no MSW samples were obtained because the soil was too dry when samples were collected, as indicated by the low soil water content. The δD distribution patterns in the BSW at the three slope positions also indicated a left bracket shape. Based on the direct inference method and the soil water distribution, the maize mainly absorbed the BSW from a depth of 20–40 cm on the slope. Fig. 8 shows the profiles of root distribution, the δD of the BSW and MSW, the soil water potential and the maize stem water at different positions during the mature grain stage. The root biomass in the 0–10 cm

516

P. Zhao et al. / Catena 147 (2016) 511–521

Fig. 4. The root biomass at the one leaf stage (a), and the δD profiles of bulk soil water, mobile soil water, immobile soil water and corresponding maize stem water at upslope (b), midslope (c), downslope (d) and soil water potential data (e).

soil layer (444.5 ± 59.0 g·m−3) was nearly two hundred times heavier than the root biomass at a depth of 30–40 cm (2.3 ± 0.6 g·m−3). The root biomass consistently accumulated in the 0–10 cm soil layer. The observed isotopic depletion in the BSW and MSW relative to the previous sampling campaigns was attributed to the recent rainfalls. The difference in the δD between the BSW and MSW was small, and the water content of the topsoil was lower than the deeper soil layers'. According to the direct inference and soil water distribution results, maize mainly acquired BSW from 20 to 40 cm depth. 3.3. Proportions of soil water in the maize water The proportional contributions of the soil water to the maize water were calculated using a linear mixed model. At the one-leaf stage, maize acquired 21.5 and 78.5% of the ImSW from depths of 0–5 and 5–10 cm on the upslope, respectively. On the midslope, maize acquired 7.8 and 92.2% of the ImSW from depths of 0–5 and 5–10 cm, respectively. In addition, the maize on the downslope acquired 28.2% and 71.8% of the ImSW from depths of 0–5 and 5–10 cm, respectively. At the five-leaf stage, the proportions of BSW at depths of 0–5, 5–10 and 10–20 cm that contributed to the maize water could not be calculated based on the concentrations of the two isotopes due to the highly linear relationship between the δD and δ18O (Zhang et al., 2011a). The proportion of the water source that contributed to the maize water was calculated using the δD by IsoSource. On the upslope, maize acquired 26–59%, 0–74% and 0–42% of the BSW from depths of 0–5, 5–10 and 10–20 cm, with mean proportions of 42.8 ± 7.1 (SD)%, 36.5 ± 21.8 (SD)% and 20.7 ± 12.6 (SD)%, respectively. On the midslope,

maize acquired 51–68%, 0–49% and 0–32% of the BSW from the three depths. In addition, maize acquired 0–18%, 0–37% and 63–85% of the BSW from the three depths on the downslope. At this stage, the BSW at a depth of 0–10 cm served as the main available water source for maize plants grown on the upslope and midslope. However, the maize grown on the downslope acquired larger proportion of water from the deeper soil (10–20 cm). At the elongation stage, maize used 0–42%, 0–55%, 0–45% and 0–64% water derived from depths of 0–5, 5–10, 10–20 and 20–30 cm on the upslope and with means of 20.0 ± 10.8 (SD)%, 25.8 ± 14.7 (SD)%, 23.3 ± 11.1 (SD)% and 30.9 ± 15.7 (SD)%, respectively. On the midslope, maize acquired 0–42%, 0–69%, 31–99% and 0–10% of the BSW from the four depths. On the downslope, maize acquired 0–15%, 0–26%, 74–87% and 0–8% of the BSW from the four depths. On the upslope, maize acquired approximately equal amounts of water from each soil layer, and on the midslope and downslope, maize mainly acquired soil water from 10 to 20 cm depth. At the tasseling stage when the roots reached a depth of 40 cm, the maize could potentially use water from the entire soil profile. During this stage, the maize acquired 0–11%, 0–47%, 0–44% and 53–57% of the BSW from depths of 0–10, 10–20, 20–30 and 30–40 cm on the upslope, with means of 3.6 ± 2.4 (SD)%, 23.2 ± 12.2 (SD)%, 18.4 ± 11.8 (SD)% and 54.8 ± 10.8 (SD)%, respectively. On the midslope, maize acquired 0–12%, 18–39%, 0–63% and 13–68% of its water from the four depths. On the downslope, the maize could acquire 0–7%, 0–1%, 43–51% and 48–53% of its water from the four depths. During this stage, the maize on the upslope and midslope acquired much water from 30 to 40 cm depth than other depths. Water at other layers was also important.

P. Zhao et al. / Catena 147 (2016) 511–521

517

Fig. 5. The root biomass at the five-leaf stage (a), the δD profiles of bulk soil water, mobile soil water and corresponding maize stem water during at upslope (b), midslope (c), downslope (d) and soil water potential data (e).

However, the maize mainly obtained BSW from 20 to 40 cm depth on the downslope. At the mature grain stage, the maize used 0–22%, 0–42%, 0–12% and 53–80% of the BSW from soil depths of 0–10, 10–20, 20–30 and 30– 40 cm on the upslope, with means of 6.6 ± 4.7 (SD)%, 22.4 ± 8.6 (SD)%, 3.2 ± 2.4 (SD)% and 67.8 ± 5.6 (SD)%, respectively. On the midslope, the maize acquired 0–2%, 0–33%, 52–82% and 2–27% of its water from the four depths. On the downslope, maize acquired 32– 52%, 0–23%, 0–46% and 18–49% of its water from the four depths. Maize plants grown on the upslope and midslope mainly used deep soil water. However, soil water sources at depths of 0–10 cm and 30– 40 cm were both important for the maize grown on the downslope.

4. Discussion 4.1. Bulk, mobile and immobile soil water in the shallow Entisols In most cases, there was large difference among the isotopic values along the shallow Entisol profile. The different isotope values of soil water along the soil profile are suitable for studying potential water sources for plants using the isotope approach (Burgess et al., 2000). However, the IRs of the BSW on August 6th showed minor differences along the soil profiles due to the integrated effect of the hydrological processes, such as rainfall infiltration, soil water movement, evaporation, etc. (Liu et al., 2015). Consequently, the small discernable differences in the isotopic profiles of the BSW have made it difficult to identify the potential water sources for maize. Root distribution and soil water content data were more helpful to identify the water source for the plants in such cases.

BSW is separated into MSW and ImSW according to the water motility. MSW normally originates from the most recent rainwater (Zhao et al., 2013a). In this study, most MSW was isotopically heavier than the BSW throughout the soil profile. That is, the IRs of the ImSW were isotopically depleted relative to the values of BSW and MSW in most soil layers. Normally, heavy rainfall behaves depleted IRs. Thus, the water source for the ImSW was replenished by one or several heavy rainfall events. Brooks et al. (2010) suggested that ImSW was derived from the first rain after a long period of drought because the drought resulted in the evaporation of most of the soil water, especially the water in the micropores. The water from heavy rainfall after a drought would occupy the micropores and would not be easily replaced during the rainy season. However, the study site did not experience a very long-term drought during the sampling period. The ImSW was likely derived from a mixture of several recent heavy rains. The average δD of the ImSW at the one-leaf stage was − 35.15‰, which was similar to the values from the four large rainfall events (N 14 mm) that occurred before the experiment, with a volume-weighted δD of −32.97‰. Furthermore, the IRs of the ImSW cannot be calculated using the LMEL at any of the growing stages, except for the one-leaf stage. This result indicated that the relationship of the δD and δ18O in the ImSW changed over time and was strongly affected by subsequent heavy rainfalls. Furthermore, rainfall infiltration, evaporation, and the exchange of water with the mobile fraction at the boundaries between the pores and peds could affect the isotopic composition of ImSW source. The regressions of δD and δ18O in the BSW and MSW were similar to the LMWL, which indicated that evaporation had a relatively small effect on these water sources. The larger slope and intercept of the δDδ18O equation of the MSW compared with that of the BSW also implied that the ImSW equation must have a smaller slope and intercept than

518

P. Zhao et al. / Catena 147 (2016) 511–521

Fig. 6. The root biomass at the elongation stage (a), the δD profiles of bulk soil water, mobile soil water and corresponding maize stem water at upslope (b), midslope (c), downslope (d) and soil water potential data (e).

the BSW equation and be subjected to strong evaporation. Evaporation mainly influenced the soil water content at 0–10 cm depth (Zhao et al., 2013a). The topsoil did not contribute to water storage because it contained a large amount of macropores, which was indicated by the low soil bulk density of the topsoil (Wang et al., 2015). The MSW was easily replaced by infiltrated rainwater and flowed out from the soil and penetrated the deep soil and mudstone layers (Zhao et al., 2013a). Therefore, the MSW was subject to evaporation over a relatively short period. The ImSW remained in the soil after rapid drainage of the MSW; therefore, the ImSW was the main water body exposed to evaporative fractionation in BSW. 4.2. Maize water use in the shallow Entisol The water use pattern of maize in the shallow Entisol was revealed using evacuation, isotopic methods and soil water potential data. The maize root distribution was restricted to depths of b 40 cm. Liu et al. (2009) stated that the biomass of maize roots is mainly concentrated at 0–40 cm depth. A review of global root distributions also indicated that maize crops have some of the shallowest roots among other plants, with root lengths shorter than 25 cm (Jackson et al., 1996). Although the maximum root depth gradually increases during the growing season, most maize roots are distributed in the upper 10 cm of soil. Maize seedlings acquired the immobile fraction of soil water at 0– 10 cm depth. At this depth, the soil was well structured with abundant macropores due to immediate plowing before sowing. Rainwater quickly drained from the topsoil, and no water could be exacted from the topsoil with suction lysimeters after a period of drought. This phenomenon was also supported by the low soil water content. The MSW in the topsoil did not remain in the soil long due to drainage. Because the soil was tilled, the fine roots of seedlings could become associated with soil

clumps or penetrate into micropores. Large amounts of the mobile water stored in the deep soil pores cannot be reached easily by seedling roots. Under these circumstances, the ImSW may become the main water source for maize seedlings with a limited rooting depth because the seedling roots are separated from the MSW zone (Fig. 9a). ZegadaLizarazu and Iijima (2004) also found that the uptake of deep soil water by maize did not significantly increase under drought circumstances. This result shows that the drought-resistance of maize let it can grow by acquiring the ImSW in the soil. Generally, maize is capable of exploiting deep-water sources as the roots penetrate downwards to meet the increasing water requirements of plants. The deeper soil usually has a greater ability to supply water to maize than shallower soil. At later growth stages, the contributions of the BSW, including the MSW and ImSW in the deep soil layers, to the maize stem water generally increases over time. Larger root systems increase the soil porosity, which is beneficial for water infiltration and storage. Maize can acquire both MSW and ImSW from soil pores. Deeper and larger roots of the later maize can also reach the macropores that store a larger proportion of the MSW and micropores that mainly store the ImSW (Fig. 9b). The observed changes in the water use pattern over time on the upslope and midslope were similar to what was observed by Zhang et al. (2011a), who found that maize acquired significant amount of soil water at a depth of 0–20 cm during the seedling and elongation stages but tended to absorb much deep soil water when the roots penetrated deeper layers. Asbjornsen et al. (2007) also found that maize was likely obtaining more water from the upper soil horizon (0–20 cm). However, Wang et al. (2010) found that the main depth at which maize acquired water increased from the elongation stage to the tasseling stage but decreased at the grain mature stage. The observed maize water-use pattern on the downslope in the present study was consistent with the

P. Zhao et al. / Catena 147 (2016) 511–521

519

Fig. 7. The root biomass at the tasseling stage (a), the δD profiles of bulk soil water and corresponding maize stem water at upslope (b), midslope (c), downslope (d) and soil water potential data (e).

findings of Wang et al. (2010). At the mature grain stage, the aerial roots assisted the maize plants to reacquire the surface soil water. Furthermore, Zhang et al. (2011a) found that maize used deeper soil water (e.g., soil water at a depth of 80 cm and groundwater) compared to the depths observed in this study. This discrepancy likely occurred because the water-use pattern of maize was controlled by the water demand at specific growth stages, the maximum reachable root depth and the amounts of available water at different soil depths, which depend on the canopy cover-atmospheric demand relationship, soil evaporation, soil profile restrictions, etc. In addition, maize grown on the upslope tends to acquire much water from the whole soil profile than other slope locations. Upslope always behave low water content than other slope. That may imply that maize may use all available water sources to overcome low water status. Determining the IRs in the BSW and MSW in this Entisol allowed us to gain a deeper insight into the maize water use pattern. The similarity of the δD-δ18O equations between the LMEL and maize stem water demonstrated that maize mainly acquired soil water with obvious evaporation effect. However, evaporation has only a minor effect on the MSW. The difference in the δD between the MSW and maize stem water implied that this potential soil water source, which could be sampled by suction lysimeters, may not play a major role in the maize water use strategy. A comparison of the linear regressions of the δD and δ18O in the maize stem water and the BSW indicated that maize does not contain the same ratio of MSW and ImSW as the soil. The ratio of MSW and ImSW also varied in different soil layers with different water contents. According to the water source of maize at the one-leaf stage and the δD-δ18O diagrams of the maize water and potential

water sources, maize acquires a larger proportion of the ImSW than the proportion of ImSW observed in the soil and uses little of the available MSW. This phenomenon was more obvious at a depth of 0–10 cm, where the root biomass was the largest and the low water content due to MSW was often quickly exhausted. Usually, MSW is the mobile fraction of soil water that the soil holds at field capacity and can be drained from the soil at relatively low suction pressures (between 0 and −80 kPa) over a relatively short time period (i.e., a week). Consequently, the ImSW stored in soil micropores is more stable and available than MSW for plant water use especially in shallow soils or dry days. Most watershed hydrology models support the idea that roots take up water from the same pool that is moving to the stream (Brooks et al., 2010). However, Zhao et al. (2013a) found that the MSW was the main fraction of soil water that participated in the generation of subsurface flow and moved to the stream. That is, the mobile water drained from the shallow soil profile seeping through the bedrock and moving into the underflow. In this shallow soil, maize is proved to mainly use ImSW fraction. Consequently, the water sources for plant water use and flow generation departed from the same pool at least at such Entisol regions. These findings may challenge the core concept of most watershed hydrology models. In addition, the isotopic models of general atmospheric circulation have classically indicated that the isotopic signature of transpiration water belongs to the meteoric water line. This assessment assumes that the water used by plants is not subject to evaporation and is derived from deep soil layers (Boujamlaoui et al., 2005). However, this research found that maize mainly used the water with strong evaporation effect which invalidates the classical hypothesis in the agricultural landscapes of sub-tropical regions.

520

P. Zhao et al. / Catena 147 (2016) 511–521

Fig. 8. The root biomass at the full ripe stage (a), the δD profiles of bulk soil water, mobile soil water and corresponding maize stem water at upslope (b), midslope (c), downslope (d) and soil water potential data (e).

Furthermore, the results of this research were also helpful for building a new irrigation standard (e.g. soil water reach the upper limit of ImSW content) for maize on the sloping farmland in this region. The differential abilities of plants for using MSW and ImSW can be used to explain the phenomena observed in other studies. For example, Midwood et al. (1998) found that the δD and δ18O values of stem water from woody plants in groves were lower than those observed for soil water at every sampled depth increment (in their Fig. 3a and b). The

authors explained that all of the plant species in the groves were obtaining significant amounts of soil water from depths below those sampled (i.e., N 1.5 m), despite the fact that several plant species in the groves had few or no roots below 80 cm. According to the observed lower IRs of the ImSW compared with the BSW, it is also very likely that some woody plants take up ImSW, which has lower IRs than BSW, which may partially explain the findings of Midwood et al. (1998).

Fig. 9. The conceptual model for the difference of maize absorbing mobile and immobile soil water in soils at seedling (a) and other growing stage (b).

P. Zhao et al. / Catena 147 (2016) 511–521

In this study, the soil water was divided into MSW and ImSW using two sampling methods and subsequent calculations. A method for precisely separating the MSW and ImSW based on the accurate soil pressure head and further quantification of the ImSW are desirable in the future. This information is useful for investigating water movement mechanisms in the SPAC on agricultural landscapes and building a more accurate hydrological model. 5. Conclusions Most of the maize roots were distributed at 0–10 cm depth. Water from this layer was important for maize seedlings and maize in the grain mature stage on the downslope. Generally, root growth allows maize to acquire larger proportion of water from the deep soil in later growth stages as its water storage increases, even when the root biomass in the deep soil is small. The results indicated that environmental and physiological conditions both affect maize water sources. The linear regression of δD vs. δ18O in the maize stem water was similar to the LMEL, indicating that the water source for maize was subject to strong evaporation. The maize seedlings mainly acquired ImSW from the top 10 cm of the soil. Thus, it was inferred that the ratio of ImSW to MSW in the maize stem was not exactly equal to that in the soil, with more ImSW occurring in the maize stem. The disparity between these two fractions of soil water should be considered when exploring the mechanisms of water cycling in the SPAC and in ecohydrological processes at various scales. Acknowledgments This study was supported by the National Basic Research Program of the Ministry of Science and Technology of China (No. 2012CB417101), the National Natural Science Foundation of China (41471188, 41101202 and 41371241), and the Hundred Talents Program of the Chinese Academy of Sciences and Sichuan Province. We thank the anonymous reviewers and editors for their important assistance with this manuscript. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.catena.2016.08.001. References Asbjornsen, H., Mora, G., Helmers, M.J., 2007. Variation in water uptake dynamics among contrasting agricultural and native plant communities in the Midwestern U.S. Agric. Ecosyst. Environ. 121, 343–356. Asbjornsen, H., Shepherd, G., Helmers, M., Mora, G., 2008. Seasonal patterns in depth of water uptake under contrasting annual and perennial systems in the Corn Belt region of the Midwestern US. Plant Soil 308, 69–92. Bariac, T., Gonzalezdunia, J., Katerji, N., Bethenod, O., Bertolini, J.M., Mariotti, A., 1994. Spatial variation of the isotopic composition of water (O-18, H-2) in the soil-plant-atmosphere system. 2. Assessment under field conditions. Chem. Geol. 115, 317–333. Boujamlaoui, Z., Bariac, T., Biron, P., Canale, L., Richard, P., 2005. Depth of extraction roots and water isotopic signature uptake by plant roots. Compt. Rendus Geosci. 337, 589–598. Brooks, J.R., Barnard, H.R., Coulombe, R., McDonnell, J.J., 2010. Ecohydrologic separation of water between trees and streams in a Mediterranean climate. Nat. Geosci. http://dx. doi.org/10.1038/NGEO722.

521

Brunel, J.P., Walker, G.R., Kennettsmith, A.K., 1995. Field validation of the isotopic procedures for determining sources of water used by plants in a semiarid environment. J. Hydrol. 167, 351–368. Burgess, S.S.O., Adams, M.A., Turner, N.C., Ward, B., 2000. Characterization of hydrogen isotope profiles in an agroforestry system: implications for tracing water sources of trees. Agric. Water Manag. 45, 229–241. Dahlman, R.C., Kucera, C.L., 1965. Root productivity and turnover in native prairie. Ecology 46, 84–89. Cakir, R., 2004. Effect of water stress at different development stages on vegetative and reproductive growth of corn. Field Crop Res. 89, 1–16. Dawson, T.E., Ehleringer, J.R., 1991. Streamside trees that do not use steam water. Nature 350, 335–336. Dawson, T.E., Pate, J.S., 1996. Seasonal water uptake and movement in root systems of Australian phraeatophytic plants of dimorphic root morphology: a stable isotope investigation. Oecologia 107, 13–20. Ehleringer, J., Dawson, T., 1992. Water uptake by plants: perspectives from stable isotope composition. Plant Cell Environ. 15, 1073–1082. Ehleringer, J.R., Osmond, C.B., 1989. Stable isotopes. In: Pearcy, R.W., Ehleringer, J., Mooney, H.A., Rundel, P.W. (Eds.), Plant Physiological Ecology, Field Methods and Instrumentation. Chapman and Hall, New York, pp. 281–300. Gardner, W.R., 1991. Modelling water uptake by roots. Irrig. Sci. 12, 109–114. Gong, Z.T., 1999. Chinese Soil Taxonomy. Science Press, Beijing in Chinese. Jackson, R.B., Canadell, J., Ehleringer, J.R., Mooney, H.A., Sala, O.E., Schulze, E.D., 1996. A global analysis of root distributions for terrestrial biomes. Oecologia 108, 389–411. Liu, J.M., An, S.Q., Liao, R.W., Ren, S.X., Liang, H., 2009. Temporal variation and spatial distribution of the root system of corn in a soil profile. Chinese J. Eco-Agric. 17, 517–521 in Chinese with English abstract. Liu, Y., Liu, F., Xu, Z., Zhang, J., Wang, L., An, S., 2015. Variations of soil water isotopes and effective contribution times of precipitation and throughfall to alpine soil water, in Wolong Nature Reserve, China. Catena 126, 201–208. Midwood, A.J., Boutton, T.W., Archer, S.R., Watts, S.E., 1998. Water use by woody plants on contrasting soils in a savanna parkland: assessment with δ2H and δ18O. Plant Soil 205, 13–24. Nie, Y., Chen, H., Wang, K., Yang, J., 2012. Water source utilization by woody plants growing on dolomite outcrops and nearby soils during dry seasons in karst region of Southwest China. J. Hydrol. 420-421, 264–274. Phillips, D.L., Gregg, J.W., 2003. Source partitioning using stable isotopes: coping with too many sources. Oecologia 136, 261–269. Thorburn, P.J., Ehleringer, J.R., 1995. Root water uptake of field-growing plants indicated by measurements of natural-abundance deuterium. Plant Soil 177, 225–233. Wang, H., 2013. Soil hydraulic parameters of forest land and sloping farmland in the hilly area of central Sichuan Basin. Master Dissertation. Northwest A&F University, Yangling in Chinese with English abstract. Wang, P., Song, X., Han, D., Zhang, Y., Liu, X., 2010. A study of root water uptake of crops indicated by hydrogen and oxygen stable isotopes: a case in Shanxi Province. China. Agric. Water Manag. 97, 475–482. Wang, H.L., Tang, X.Y., Zhang, W., McKenzie, B.M., 2015. Within-year changes in hydraulic properties of a shallow entisol in farmland and forestland. Vadose Zone J. 14(7) http://dx.doi.org/10.2136/vzj2014.11.0163. Zegada-Lizarazu, W., Iijima, M., 2004. Hydrogen stable isotope analysis of water acquisition ability of deep roots and hydraulic lift in sixteen food crop species. Plant Prod. Sci. 7, 427–434. Zhang, C., Zhang, J., Zhao, B., Zhu, A., Zhang, H., Huang, P., Li, X., 2011a. Coupling a two-tip linear mixing model with a delta D-delta O-18 plot to determine water sources consumed by maize during different growth stages. Field Crop Res. 123, 196–205. Zhang, W., An, S., Xu, Z., Cui, J., Xu, Q., 2011b. The impact of vegetation and soil on runoff regulation in headwater streams on the east Qinghai–Tibet Plateau, China. Catena 87 (2), 182–189. Zhao, P., Tang, X., Zhao, P., Wang, C., Tang, J., 2013a. Identifying the water source for subsurface flow with deuterium and oxygen-18 isotopes of soil water collected from tension lysimeters and cores. J. Hydrol. 503, 1–10. Zhao, P., Tang, X., Zhao, P., Wang, C., Tang, J., 2013b. Tracing water flow from sloping farmland to streams using oxygen-18 isotope to study a small agricultural catchment in Southwest China. Soil Tillage Res. 134, 180–194. Zhu, B., Kuang, F., Gao, M., Wang, T., Wang, X., Tang, J., 2009. Effects of soil thickness on productivity of sloping cropland of purple soil. J. Mount Sci. 27, 735–739 in Chinese with English abstract.