ZHAO Jin, XUE Qing-wu, Kirk E Jessup, HOU Xiao-bo, HAO Bao-zhen, Thomas H Marek, XU Wenwei, Steven R Evett, Susan A O’Shaughnessy, David K Brauer

1 Texas A&M AgriLife Research and Extension Center at Amarillo, TX 79106, USA

2 Department of Soil, Water, and Environmental Sciences, University of Arizona, AZ 85721, USA

3 School of Science and Technology, Xinxiang University, Xinxiang 453003, P.R.China

4 Texas A&M AgriLife Research and Extension Center at Lubbock, TX 79403, USA

5 USDA-ARS Conservation and Production Research Laboratory, TX 79012, USA

1. Introduction

Maize, the largest global cereal crop in production (FAO 2014), is an important cereal crop for food security and will provide at least 30% of the food calories to more than 4.5 billion people in 2050 (CGIAR 2016). Annually, the United States supplies 40% of global annual maize production (FAO 2014;Lobellet al. 2014). In the Texas High Plains, maize remains the most significant irrigated crop (Colaizziet al. 2009), and grain yield was almost double that produced by the other regions in Texas (Farfanet al. 2013). In this region, rainfall during the maize growing season is about 233 mm, which is only 31% of the estimated seasonal water requirement(crop evapotranspiration, ET, 735 mm) for the maximum grain production (Kapanigowdaet al. 2010). However,the Ogallala Aquifer, which provides approximately 90%of the irrigation water in the Texas High Plains, declined in irrigated area from a peak of 2.42 million ha in 1994 to 1.87 million ha in 2000 (Colaizziet al. 2009), and the irrigation water supplies for agricultural production were considerably affected by drought in 2011 and 2012(Ziolkowska 2015).

Greater yield stability, either in grain or total shoot biomass (for silage purpose), through improved drought tolerance is commonly considered an important objective of breeding programs under water-limited conditions.Therefore, drought tolerant (DT) can generally de fined as ability to maintain greater yield or yield stability under drought (Camposet al. 2004; Blum 2009; Cooperet al.2014). Hybrid selection based on performance in multienvironment trials has increased grain yield under drought through increased yield potential and kernel set, rapid silk exertion, and reduced barrenness, though at a lower rate than under optimal conditions (Camposet al. 2004;Boomsma and Vyn 2008). Drought tolerance in maize is likely to entail the selection of plants with a reduced leaf area (especially in the upper part of the plant), short thick stems, small tassels, erect leaves, delayed senescence,smaller root biomass, and a deep root system with little lateral root branching (Ribautet al. 2009). DT genotypes are also expected to have robust spikelet and kernel growth at the cell-division and expansion-growth phases, with good osmotic adjustment to assist in cell retention of water during drought (Ribautet al. 2009). Proper selection of DT hybrids can increase maize yield with greater biomass and harvest index (HI), heavier kernel weight, and higher resource use efficiency (e.g., water and radiation) under water-limited conditions (Boomsma and Vyn 2008; Haoet al. 2015a, b,2016).

Under water limited conditions, plant growth is directly determined by the ability to capture and convert water resources to biomass, which depends on the root system(Hammeret al. 2009). Therefore, the Identification and understanding of rooting characteristics is essential for the development of more DT crops (Eghball and Maranville 1993; Gao and Lynch 2016). A combination of high water use efficiency (WUE) and suf ficient water extraction by a deep root system can help plants to sustain drought stress(Hundet al. 2009). Low crown root numbers improved DT by increasing rooting depth and water acquisition from the subsoil (Gao and Lynch 2016). A more efficient water use and a higher root to shoot ratio (R:S) were also found to be a major reason for a higher stress tolerance with a DT hybrid (Grzesiaket al. 1999).

In the Texas High Plains, because of the declining water table of the Ogallala Aquifer, newly developed DT hybrids will be essential to maintain maize production under limited irrigation (Haoet al. 2015a). Our previous studies in DT hybrids focused on water use, soil water extraction, biomass production, and grain yield. The results showed that DT hybrids had greater yield but used similar amount or less water than a conventional hybrid (Haoet al. 2015a, b).Nevertheless, the rooting characteristics of DT hybrids were still unknown. The objective of this study was to investigate the differences in shoot and root traits, water use, and WUE in DT maize hybrids under deficit irrigation conditions.

2. Materials and methods

2.1. Location and experimental design

Two greenhouse studies were conducted in 2014 and 2015 at Texas A&M AgriLife Research Station at Bushland, TX,USA (35°11′N, 102°06′W, 1 170 m a.s.l.). Maize plants were grown in PVC pots, with one plant per pot. Each PVC pot(inner diameter of 15.24 cm) was 70 cm tall with an effective soil depth of 60 cm. Finished pots were lined with a 0.10-mm poly bag to prevent drainage of water in order to accurately calculate ET values. They were then filled with 7.82 kg dry weight of Turface Quick Dry?, a calcined clay material with bulk density of 0.61 g cm–3, and a particle size that is 80% smaller than 30 mesh (0.595 mm) (http://www.turface.com/). This soil weight was determined by a preliminary test of water holding capacity, so that the pot could sustain an average hybrid at the maximum water usage for at least 2 d(in this way it was not necessary to irrigate each day except during pollination). Wooden racks were constructed to hold 11 pots per row at spacing of 21 cm, with 76 cm between rows. This approximates a field plot con figuration with a density of 6.3 plants m–2. Although the roots of each plant were isolated to their own pots, the rack design permitted the upper architecture of the plants to interact in the manner of regional field settings.

Four maize hybrids that differed in their DT characteristics were planted in both years: one conventional hybrid(33D53AM, 115 d), two commercial DT hybrids (P1151AM,111 d; N75H, 115 d), and one experimental maize hybrid(ExpHB, 117 d, developed by Dr. Xu Wenwei, maize breeder, at Texas A&M AgriLife Research in Lubbock).Hybrid 33D53AM was a conventional Pioneer hybrid rated as poorly suited for Drought-Prone soil by DuPont Pioneer in 2012. The P1151AM is Pioneer AquaMax hybrid and N75H Syngenta Agrisure Artesian hybrid. The drought tolerance scores for the two hybrids were excellent based on their rating scales from each seed company. ExpHB was a non-transgenic, experimental hybrid with a very large root system. It can be used as both a grain and a silage crop with high biomass production (Xu Wenwei, personal communications, 2016).

In 2014, the sowing and harvest dates were 26 June and 12 November, respectively. In 2015, the sowing and harvest dates were 8 July and 11 November, respectively.Four seeds were planted per pot, at a 5-cm depth. Plants were thinned to one per plot at the two-leaf stage. Pots were fertilized with Peter’s Professional Water Soluble Fertilizer 20-20-20 (20% N, 20% P2O5, and 20% K2O). Each pot was given a total of 51 g of fertilizer, but divided among three doses; one at planting, one at V6 (six-leaf) stage, and one at V10 (10-leaf) stage.

There were two treatments (water regime and hybrid) in each year, which were arranged as a split-plot design with three replications. Water regime was assigned to the main plots and hybrid assigned as subplots. The two targeted water regimes (i.e., two irrigation treatments) were 100%(I100) and 50% (I50) of ET requirement. For both years, the amount of irrigation for the I100treatment was determined by adding water to keep the soil water level at 100% pot capacity (PC) based on the procedure of Xueet al. (2012).Before planting, the weights of empty PVC pots and the air-dried soil mixture were recorded. Each pot was then saturated with water for 2 d, and water was allowed to drain for approximately 48 h. When drainage ceased, the weight of each pot was recorded again, as the weight at 100% PC (Xueet al. 2012). After emergence, plants were irrigated every 2 d, by adding the amount of water needed to bring pots back to 100% PC in the I100treatment. For the I50treatment, the amount of water added was half of the amount required of the I100treatment. The irrigation treatments were started at about the V2 (two-leaf) stage.In each hybrid, the amount of irrigation (water use) was calculated by weighing three pots when they were watered.The pots for each hybrid were weighed at the same time of the day and the amount of irrigation was the average of three pots. The accumulative water use at three different stages (V9 (nine-leaf), R1 (silking), and R6) was determined by accumulating the water use from emergence to each stage (Xueet al. 2012).

2.2. Measurements

At the V9, R1, and R6 stages, three plants, representing three replications, in each hybrid were harvested from each water regime. Each plant was separated into shoots(aboveground parts including ears) and roots (underground parts). The maize plants were grown to maturity and ears collected for grain yield determination in both years.However, there were large variations in yield data due to poor pollination during silking periods in some hybrids.As such, yield data are not reported in this paper. Shoot samples were then separated into stems and leaves. Leaf area was measured with a leaf area meter (LI-3000A,LI-COR Inc., Lincoln, NE, USA) at the V9 and R1 stages.Plant height of each plant was measured at the V9 and R1 stages. To determine root traits, the growth columns were subdivided into six different segments that were 0–10,10–20, 20–30, 30–40, 40–50, and 50–60 cm from the soil surface. Excavated root crowns (consisting of the number of belowground nodal whorls and the number of roots per whorl) and seminal roots were rinsed followed by manual counting of the root crowns and seminal roots. Root samples from the single depth segments were washed. Afterwards,shoots (including stem and leaves) and roots (including six depths) were oven-dried at 70°C to constant weight. R:S was determined by the ratio of root dry weight and shoot dry weight (Fenget al. 2016). The shoot/root growth rate of each hybrid was calculated using shoot/root dry weight during the V9–R1 and R1–R6 periods using the equations of Radford (1967). WUE was calculated as the ratio of shoot dry weight at maturity and seasonal water use.

2.3. Data statistical analysis

Statistical analysis was conducted using R3.3.0 (R Core Team 2016). Analysis of variance was conducted using mixed effects model function lmer from R package lme4 to evaluate main effects and interactions (Bateset al.2015). The water regime and hybrid were treated as fixed effects, and replication as a random effect. Mean values were compared by LSD function provided in the agricolae package at the 5% probability level (de Mendiburu 2017).Relationships between shoot dry weight and the amount of water use were determined by linear regression.

3. Results

3.1. Shoot traits

Leaf areas varied significantly with year, growth stage,water regime, and hybrids (Table 1). Comparing the two water regimes, water stress in the I50treatment resulted in significantly lower leaf area compared with that in the I100treatment. There were significant differences in leaf area among the four hybrids in both years and water regimes,particularly at the R1 stage. In both water regimes, ExpHB had the highest leaf area, while P1151AM had the lowest.For example at the R1 stage, leaf area of P1151AM was 48–67% lower than that of ExpHB, depending on the year and the water regime. The leaf area in the N75H and 33D53AM hybrids was in the middle. In general, 33D53AMand N75H had 16–31% and 6–48% lower leaf area than ExpHB at R1 stage, respectively. The mean plant height of four hybrids was significantly (P＜0.05) lower in the I50treatment compared with the I100treatment at both the V9 and R1 stages. The hybrid difference in plant height was related to water regime and year. In the I100treatment, no differences (P＞0.05) in plant height were found among hybrids in 2014, but the N75H plants were shorter than 33D53AM and ExpHB hybrids at the R1 stage in 2015. In the I50treatment, no differences (P＞0.05) in plant height among hybrids were found at the V9 stage in 2014, but the ExpHB hybrid was shorter than the other three hybrids in 2015. At the R1 stage, P1151AM was taller in 2014 but shorter than the other three hybrids in 2015 (Table 2).

Table 1 Analysis of variance (P＞F) of shoot and root traits, and water use and water use efficiency as affected by year (YR), water regime (WR), and hybrid (HB) in maize

Shoot dry weights at the V9, R1, and R6 stages were significantly (P＜0.05) affected by the three main effects(year, water regime, and hybrid) except year effect at R6 stage (P=0.2365) (Table 1). Moreover, the two-way interactions on the shoot dry weight were also significant(P＜0.05) except the year×water regime at the V9 stage(P=0.5051) and the year×hybrid (P=0.8929) at the R6 stage(Table 1). The three-way interactions on shoot dry weight were also significant (P=0.0028) at the V9 stage. The differences in shoot dry weights among hybrids were similar to differences in leaf area, but with greater magnitudes.In both water regimes, the ExpHB had the greatest shoot dry weight and the P1151AM hybrid had the lowest. The difference in shoot dry weight between the ExpHB and P1151AM hybrids was extremely large in 2015 in the I100treatment, in which the ExpHB had 2.6 and 1.9 times greater shoot dry weight than the P1151AM at the V9 and R1 stages,respectively. Hybrid differences in shoot dry weight were also smaller at the R6 stage; the P1151AM hybrid had 20–30% less shoot dry weight than the other three hybrids in the I100treatment plots in 2015. Compared with the 33D53AM hybrid (conventional hybrid), the two commercial DT hybrids and ExpHB responded differently to water stress.Hybrids P1151AM and N75H generally had less shoot dry weight reduction under water stress (I50) as compared to the well-watered condition (I100). For example in 2015 at the R1 stage, the P1151AM and N75H hybrids only had slight reductions (2–5%) in shoot dry weight but the 33D53AM hybrid had 16% reduction in the I50treatment. Although the shoot dry weight in the ExpHB hybrid was reduced under water stress, this hybrid still maintained greater shoot dry weights than the other hybrids. Shoot growth rate duringthe V9–R1 stages was significantly (P＜0.05) affected by all three main effects (year, water regime and hybrid). However,the growth rate between the R1 to R6 stages was only affected by water regime (Table 1). Compared with the I100treatment, shoot growth rate was significantly reduced, up to 60% in the I50treatment, depending on the growth stage.The hybrid differences in shoot growth rate were limited to the V9–R1 stages. The ExpHB had a greater shoot growth rate compared with the P1151AM and N75H hybrids in the I100treatment,and a greater growth rate compared with P1151AM hybrid in the I50treatment in 2015 (Table 3).

Table 2 Leaf area and plant height at V9 (nine-leaf) and R1 (silking) stages of four hybrids under two water regimes (I100 and I50,referring to 100 and 50% of evapotranspiration requiements, respectively) in 2014 and 2015

3.2. Root traits

Root dry weight was affected by all three main effects,except at the V9 stage where year did not significantly affect root dry weight. Water regime had a smaller effect on root dry weight as compared to shoot dry weight.Averaged across hybrids, root dry weight was greater in the I50treatment compared with the I100treatment at the V9 stage in both years. Although the differences in root dry weight between the I100and I50treatments were smaller at the R1 and R6 stages, the differences were generally less than 30% (Table 4). In contrast, the difference in shoot dry weight between the two water regimes was as high as 48% at the R6 stage (Table 3). Regardless of the year and water regime, hybrid difference in root dry weight was highly significant at all three stages (V9, R1, and R6)(P＜0.0001). In general, hybrid difference in root dry weight was larger than that in shoot dry weight, particularly in the I100treatment. The difference in root dry weight between ExpHB and P1151AM was particularly large. In 2014 in both water regimes, the ExpHB had 1.6 to 2.7 times greater root dry weight than P1151AM. In 2015, root dry weight differences between these same hybrids ranged from 2.2 to 3.5 times greater in the I100treatment, depending on growth stage. The differences in root dry weight among the ExpHB, 33D53AM,and N75H hybrids were smaller compared with differences between the ExpHB and P1151AM hybrids. The ExpHB had at least 40% greater root dry weight than the 33D53AM and N75H hybrids. Among the three commercial hybrids, the 33D53AM and N75H had greater root dry weight than the P1151AM. Similarly, the ExpHB had a greater growth rate than other hybrids (Table 4).

Comparing the two water regimes, R:S in the I50treatment was significantly higher (P＜0.05) than that in the I100treatment at the V9 and R6 stages, except at the R6 stage in 2014 (Table 5). There were consistent differences in R:S among hybrids. In general, the 33D53AM and ExpHB hybrids had higher R:S than the P1151AM and N75H hybrids(Table 5). Water regime did not affect crown root number in both years at the V9 stage. However, water stress in the I50treatment reduced the number of crown roots at the R1 stage. Among the hybrids, N75H generally had more crown roots than the other three hybrids. Water regime did not affect the number of seminal roots but the differences inseminal roots among hybrids were significant. The ExpHB had the most seminal roots (six). In the other hybrids,33D53AM and P1151AM had more seminal roots compared with N75H (Table 5).

Table 3 Shoot dry weight and growth rate of the four hybrids under two water regimes (I100 and I50, referring to 100 and 50% of evapotranspiration requiements, respectively) in 2014 and 2015

Table 4 Root dry weight and growth rate of four hybrids under two water regimes (I100 and I50, referring to 100 and 50% of evapotranspiration requiements, respectively) in 2014 and 2015

In the I100treatment, root dry weights were significantly(P＜0.05) affected by hybrid in soil layers below 20 cmat three growth stages (list stages here). No significant differences (P＞0.05) in root dry weights in the upper layer(0–10 cm) among the four hybrids were found at all three stages except at the R6 stage in 2015 (Fig. 1). The ExpHB

always had the greatest root dry weights in all soil layers at all three stages, but P1151AM had the smallest weights at the V9 and R1 stages (Fig. 1). Among hybrids, the differences in root dry weight along the soil pro file (10 to 60 cm) were smaller in the I50treatment (Fig. 2). However,the ExpHB still had greater root dry weights compared with the other hybrids in most of the soil layers (Fig. 2). At the V9 and R1 stages, below the soil depth of 30 cm, root dry weights of P1151AM and N75H were smaller than the conventional hybrid (33D53AM). No significant differences in root dry weights were found among the P1151AM, N75H,and 33D53AM hybrids at the R6 stage (Fig. 2).

Table 5 Root:shoot ratio, seminal and crown root number of four hybrids under two water regimes (I100 and I50, referring to 100 and 50% of evapotranspiration requiements, respectively) in 2014 and 2015

Fig. 1 Root dry weight per plant in different soil layers at V9 (nine-leaf), R1 (silking), and R6 (maturity) stages in four hybrids at high water regime (I100, 100% of evapotranspiration requirement) in 2014 and 2015. ＊ and ＊＊, significant differences among hybrids at P＜0.05 and P＜0.01, respectively. Horizontal bars are standard error of mean.

3.3. Water use and WUE

The dynamics of plant water use in the two years and two water regimes are shown in Fig. 3. In general, water use was low (＜6 kg plant–1) from planting to the V9 stage. Water use increased rapidly from the V9 to the R1 stages and continued to increase until the late grain filling stage for all hybrids, water regimes and both years. Comparing the two water regimes, water use was reduced significantly (P＜0.05)in the I50treatment (Table 6). At the early stage (V9), water use was generally not different (P＞0.05) among the hybrids.However, ExpHB still had 27% more water use compared with the 33D53AM and P1151AM hybrids in the I100treatment in 2014 (Table 6 and Fig. 3). The differences in water use among hybrids were greater at the R1 and R6 stages when active growth occurred in maize plants (Fig. 3). In general,P1151AM and N75H used less water than 33D53AM and ExpHB, particularly in the I100treatment (Table 6). In both water regimes, ExpHB had the greatest water use at the R1 and R6 stages in 2014 and at the R1 stage in 2015 (Table 6 and Fig. 3). In most cases, P1151AM and N75H had no difference in water use (Table 6 and Fig. 3).

In 2014, the effects of water regime and hybrid on WUE were significant. Comparing two water regimes, WUE was greater in the I50compared with the I100treatment. Among the hybrids, P1151AM and N75H had greater WUE than the ExpHB and 33D53AM hybrid in both water regimes.Although there were no water regime and hybrid effects on WUE in 2015, there was a trend that N75H in the I100treatment and P1151AM in both water regimes had higher WUE than ExpHB (Table 6). However, the hybrid difference in WUE was much smaller (＜18%) as compared to the hybrid differences in shoot dry weight and water use (＞30%). Data pooled from all hybrids and both water regimes in the two years showed a positive linear relationship between shoot dry weight and water use at three growth stages (Fig. 4).There was a linear relationship between shoot dry weight and water use in both years, with 2014 having a greater slope than 2015. The slope of the linear relationship

Fig. 2 Root dry weight per plant in different soil layers at V9 (nine-leaf), R1 (silking), and R6 (maturity) stages in four hybrids at low water regime (I50, 50% of evapotranspiration requirement) in 2014 and 2015. ＊ and ＊＊, significant differences among hybrids at P＜0.05 and P＜0.01, respectively. Horizontal bars are standard error of mean.

Fig. 3 Accumulated water use of four hybrids under two water regimes (I50 and I100, 50 and 100% of evapotranspiration requirements,respectively) in 2014 and 2015. V9, nine-leaf; R1, silking; R6, maturity.

Table 6 Water use at V9 (nine-leaf), R1 (silking), R6 (maturity) stages and water use efficiency (WUE) of four hybrids under two water regimes (I100 and I50, 100 and 50% of evapotranspiration requirements, respectively) in 2014 and 2015

between shoot dry weight and water use also represented the values of WUE, which were smaller than the WUE calculated by the ratio of final shoot dry weight and water use(Table 6). Nevertheless, the linear regressions also showed that the difference among hybrids was generally small.

4. Discussion

4.1. Drought tolerance

The previous information for drought tolerance characteristics with the four hybrids used in this study was mainly from yield trials under field conditions. The commercial DT hybrids generally had higher yields than the conventional hybrid under drought (Haoet al. 2015a, b, 2016; Zhaoet al. 2018).Although the field data were limited for the experimental hybrid (ExpHB), this hybrid produced more biomass than any of the commercial hybrids used in our previous studies based on our unpublished data. Nevertheless, results of this study are consistent with previous studies based on more detailed measurements regarding shoot and root traits.Based on the responses of shoot dry weight, the hybrids P1151AM, N75H, and ExpHB were more drought tolerant than the conventional hybrid, 33D53AM. Although the two commercial DT hybrids (P1151AM and N75H) generally had the same or lower shoot dry weight than the 33D53AM hybrid in the I100treatment, the former had significantly less shoot dry weight reduction under water stress in the I50treatment. The experimental hybrid (ExpHB) consistently had greater leaf area and shoot dry weight in both water regimes. Evidently,the two commercial DT hybrids and experimental hybrid used different mechanisms to respond drought stress.

The smaller leaf area and lower shoot dry weights in commercial DT hybrids were reported in a field study at same location as this study. Mounceet al. (2016) found the mean shoot dry weight for a conventional hybrid (Pioneer 33Y75)irrigated at 50% replenishment of soil water depletion to field capacity was reduced by 59% compared with irrigation at 100% replenishment of soil water depletion (SWD) to field capacity (FC). However, the reduction in mean shoot dry weight for a DT hybrid (P0876HR) grown in the same field and irrigated at 50% replenishment of soil water depletion to field capacity was only 38% compared with irrigation at the 100% level. These results occurred in the 2014 irrigation season, when seasonal rainfall was average (275 mm).The higher yield for commercial DT hybrids was because of higher harvest index across irrigation treatments (Mounceet al. 2016). In a study in Northwest China with semi-arid environment, Heet al. (2017) showed that drought tolerant soybean cultivars consistently had smaller leaf area and lower stomatal conductance under terminal drought, which significantly contributed to yield under drought stress.

In this study, the characteristics of drought tolerance in the experimental hybrid (ExpHB) are different than the two commercial DT hybrids as described by previous studies (Ribautet al. 2009; Heet al.2017). In contrast,this hybrid had greater leaf area and shoot dry weight under water stress. The hybrid also had large root system,which can increase the water extraction from soil pro file,particularly from deep pro file. In a rice study, Kobataet al. (1996) showed that a drought tolerant cultivar had greater biomass production and extracted more water from deeper soil pro file due to a deep root system. A hybrid with a large root system could require a good moisture conditions for root development and growth at early stages.Therefore, this type of character may not be suitable for the consistent dry conditions without irrigation. In the environment with consistent drought such as Northwest China, characteristics such as smaller leaf area and smaller root system may be more suitable and more likely lead to greater yield under drought (Fanget al. 2010; Heet al. 2017). In the Texas High Plains, however, maize is mainly grown under irrigated conditions (Xueet al. 2017).In this situation, some characteristics of ExpHB may still be useful to soil water use and biomass production.Nevertheless, more studies are needed to elucidate the responses of hybrid ExpHB to different irrigation quantities and frequencies.

Improved water capture resulted from a greater uptake of water, which is associated with a more active root system(Tollenaar and Wu 1999). In this study, both water regime and hybrid had significant effects on root dry weight and root growth rate. Compared with the I100treatment, water stress (I50) slightly increased root dry weight at the V9 stage,indicating a mild soil water stress promoted root growth.Xueet al. (2003) showed that soil drying at jointing stage promoted root growth in wheat under field conditions.However, water stress significantly reduced the root dry weight and root growth rate in all four hybrids at the R1 and R6 stages (Table 4). Among the hybrids, the ExpHB had the largest root dry weight in all soil layers during the whole growing season. Therefore, this hybrid can extract more water from the soil to maintain shoot growth and development under water stress conditions. Nevertheless,this character may be limited by regular and consistent drought conditions in semi-arid and arid environment(Heet al. 2017). In general, the differences of root dry weight and root growth rate were smaller among the three commercial hybrids (P1151AM, N75H, and 33D53AM).Nevertheless, the DT hybrid P1151AM still had the smallest root dry weight at the V9 and R1 stages. The smaller root system resulted in the lower water use in this hybrid during the early growth period.

It is believed that higher values of the R:S are more advantageous for plants to survive under soil drought stress(Eghball and Maranville 1993; Grzesiaket al. 1999). There were consistent differences in R:S among commercial hybrids; the experimental hybrid (ExpHB) had the highest R:S (Table 5). Crown root is an important determinant of soil resource capture; hybrids with reduced crown root number could have deeper root and improved water acquisition ability from drying soil (Gao and Lynch 2016). In this study,water stress did not affect crown root number at the V9 stage, which is consistent with no reduction in root dry weight in the I50treatment at the V9. However, crown root number was reduced significantly under water stress at the R1 stage. Among the hybrids, P1151AM and ExpHB had more reduction in crown roots than the conventional hybrid(33D53AM) under water stress. The reduced crown roots could promote plants to explore water deeper in the soil pro file (Gao and Lynch 2016). In addition, reduced lateral root branching density will also improve DT in maize by reducing the metabolic costs of soil exploration, permitting greater axial root elongation, greater rooting depth, and thereby greater water acquisition from drying soil (Zhanet al. 2015). Also, less root branching was beneficial to shoot biomass accumulation (Heet al. 2017). Although the water regime did not affect the number of seminal roots,hybrid difference in seminal root was significant and the ExpHB had one to two more seminal roots than the other three hybrids. Seminal roots are primarily responsible for water uptake from deep within the soil pro file (Singhet al.2010). Therefore, the DT hybrids have a stronger ability to uptake water from deeper soil, particularly the experimental hybrid. We previously found that the P1151AM hybrid generally extracted less soil water than 33D53AM but can extract water from deeper within the pro file under drought conditions (Haoet al. 2015b).

4.2. Water use and WUE

Under water limited conditions, plant shoot dry weight is determined by seasonal water use and WUE (Blum 2009). In this study, hybrid differences in water use were significant in both water regimes and for both years. Among the hybrids, 33D53AM and ExpHB consistently had more total water use compared with P1151AM and N75H due to their larger root systems. The differences in WUE among hybrids were only found in 2014; the P1151AM and N75H had higher WUE than the 33D53AM and ExpHB. These results were consistent with field experiments performed in Bushland, Texas (Haoet al. 2015b; Mounceet al. 2016),where water use for the conventional hybrid was greater as compared with the commercial DT hybrid. However,WUE was grater for the commercial DT hybrid in the drier trial year based on field experiments. However, no hybrid differences in WUE were found in 2015. Blum (2009) argued that cultivar difference in WUE under water limited conditions was more related to difference in water use rather than that in biomass production, which are consistent with the results in this study. For example, in the I100treatment at the R6 stage, the difference in water use between the ExpHB and P1151AM was as high as 44% but the difference in shoot dry weight between the two hybrids was generally less than 30%. Although the differences in total water use and WUE between the 33D53AM and ExpHB were not significant, the ExpHB used more water than the 33D53AM hybrid due to the larger leaf area as well as the larger root system at the early stages (V9 and R1). The ExpHB had or tended to have lower WUE than the two commercial DT hybrids but had a greater ability to use soil water and maintain transpiration under water stress. The similar results were also reported in the study of Kobataet al. (1996), who found that a drought tolerant rice cultivar can extract more water from deep soil pro file with deep root system. However, the cultivar had lower WUE.

5. Conclusion

We investigated the shoot and root traits, water use, and WUE of four hybrids (one conventional and three DT hybrids) under two water regimes. There were significant differences in most of the shoot and root traits, water use and WUE among the four hybrids in both water regimes.Under water stress, the three DT hybrids showed higher drought tolerance than conventional hybrid. There are two different mechanisms to respond to water stress among the DT hybrids. Compared with the conventional hybrid(33D53AM), the two commercial DT hybrids had smaller leaf area, shoot dry weight and root system. As a result,the two hybrids used less water but had higher WUE. In contrast, the experimental hybrid (ExpHB) had greater leaf area, shoot dry weight, and root dry weights compared with the commercial DT hybrids. This hybrid was able to use more water under water stress due to the larger root system. However, how the hybrid performs under more regular and consistent drought conditions in semi-arid and arid environment still requires more studies. For the studies under irrigated conditions, different irrigation management strategies may be required to optimize yields, which need more investigations under field conditions because of the different response mechanisms among hybrids.

Acknowledgements

We are grateful to Texas A&M AgriLife Research staff,Chance Reynolds, Brad Parker, Cole Pope, and Bella Porras for their help in the studies. This research is supported in part by the UDSA-Ogallala Aquifer Program and Texas A&M AgriLife Research Cropping System Program, USA and the USDA National Institute of Food and Agriculture Hatch Project, USA (TEX09438).

Bates D, Maechler M, Bolker B, Walker S. 2015. Fitting linear mixed-effects models using lme4.Journal of Statistical Software, 67, 1–48.

Blum A. 2009. Effective use of water (EUW) and not water-use efficiency (WUE) is the target of crop yield improvement under drought stress.Field Crops Research, 112, 119–123.

Boomsma C R, Vyn T J. 2008. Maize drought tolerance:Potential improvements through arbuscular mycorrhizal symbiosis?Field Crops Research, 108, 14–31.

Campos H, Cooper M, Habben J E, Edmeades G O, Schussler J R. 2004. Improving drought tolerance in maize: A view from industry.Field Crops Research, 90, 19–34.

CGIAR (Consultative Group on International Agricultural Research). 2016. CGIAR strategy and results framework 2016–2030: Overview. [2016-11-01]. http://library.cgiar.org/bitstream/handle/10947/4069/CGIAR%20SRF%20 Overview%20WEB.pdf?sequence=10

Colaizzi P D, Gowda P H, Marek T H, Porter D O. 2009. Irrigation in the texas high plains: A brief history and potential reductions in demand.Irrigation and Drainage, 58, 257–274.

Cooper M, Gho C, Leafgren R, Tang T, Messina C. 2014.Breeding drought-tolerant maize hybrids for the US cornbelt: discovery to product.Journal of Experimental Botany,65, 6191–6204.

Eghball B, Maranville J W. 1993. Root development and nitrogen in flux of corn genotypes grown under combined drought and N stress.Agronomy Journal, 85, 147–152.

Fang Y, Xu B C, Turner N C, Li F M. 2010. Grain yield, dry matter accumulation and remobilization, and root respiration in winter wheat as affected by seeding rate and root pruning.European Journal of Agronomy, 33, 257–266.

FAO (Food and Agriculture Organization of the United Nations).2014. FAOSTAT. [2016-06-16]. http://faostat.fao.org/site/567/default.aspx#ancor

Farfan I D B, Murray S C, Labar S, Pietsch D. 2013. A multi-environment trial analysis shows slight grain yield improvement in Texas commercial maize.Field Crops Research, 149, 167–176.

Feng G, Zhang Y, Chen Y, Li Q, Chen F, Gao Q, Mi G. 2016.Effects of nitrogen application on root length and grain yield of rain-fed maize under different soil types.Agronomy Journal, 108, 1654–1665.

Gao Y, Lynch J P. 2016. Reduced crown root number improves water acquisition under water de ficit stress in maize (Zea maysL.).Journal of Experimental Botany, 67, 4545–4557.

Grzesiak S, Hura T, Grzesiak M T, Pie fikowski S. 1999. The impact of limited soil moisture and waterlogging stress conditions on morphological and anatomical root traits in maize (Zea maysL.) hybrids of different drought tolerance.Acta Physiologiae Plantarum, 21, 305–315.

Hammer G L, Dong Z, McLean G, Doherty A, Messina C,Schussler J, Zinselmeier C, Paszkiewicz S, Cooper M. 2009.Can changes in canopy and/or root system architecture explain historical maize yield trends in the US Corn Belt?Crop Science, 49, 299–312.

Hao B, Xue Q, Marek T H, Jessup K E, Becker J, Hou X, Xu W,Bynum E D, Bean B W, Colaizzi P D, Howell T A. 2015a.Water use and grain yield in drought-tolerant corn in the texas high plains.Agronomy Journal, 107, 1922–1930.

Hao B, Xue Q, Marek T H, Jessup K E, Hou X, Xu W, Bynum E D, Bean B W. 2015b. Soil water extraction, water use,and grain yield by drought-tolerant maize on the Texas High Plains.Agricultural Water Management, 155, 11–21.

Hao B, Xue Q, Marek T H, Jessup K E, Hou X, Xu W, Bynum E D, Bean B W. 2016. Radiation use efficiency, biomass production, and grain yield in two maize hybrids differing in drought tolerance.Journal of Agronomy and Crop Science,202, 269–280.

He J, Du Y, Wang T, Turner N C, Yang R, Jin Y, Yue X Y,Zhang C, Cui T, Fang X, Li F M. 2017. Conserved water use improves the yield performance of soybean (Glycine max(L.) Merr.) under drought.Agricultural Water Management,179, 236–245.

Hund A, Ruta N, Liedgens M. 2009. Rooting depth and water use efficiency of tropical maize inbred lines, differing in drought tolerance.Plant and Soil, 318, 311–325.

Kapanigowda M, Stewart B A, Howell T A, Kadasrivenkata H, Baumhardt R L. 2010. Growing maize in clumps as a strategy for marginal climatic conditions.Field Crops Research, 118, 115–125.

Kobata T, Okuno T, Yamamoto T. 1996. Contributions of capacity for soil water extraction and water use efficiency to maintenance of dry matter production in rice subjected to drought.Japanese Journal of Crop Science, 65, 652–662.

Lobell D B, Roberts M J, Schlenker W, Braun N, Little B B,Rejesus R M, Hammer G L. 2014. Greater sensitivity to drought accompanies maize yield increase in the U.S.Midwest.Science, 344, 516–519.

de Mendiburu F. 2017. Agricolae: Statistical procedures for agricultural research. R package version 1.2–7. [2016-11-01]. https://CRAN.R-project.org/package=agricolae

Mounce R B, O’Shaughnessy S A, Blaser B C, Colaizzi P D,Evett S R. 2016. Crop response of drought-tolerant and conventional maize hybrids in a semiarid environment.Irrigation Science, 34, 231–244.

R Core Team. 2016. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. [2016-11-01]. http://www.R-project.org/

Radford P J. 1967. Growth analysis formulae - Their use and abuse.Crop Science, 7, 171–175.

Ribaut J M, Betran J, Monneveux P, Setter T. 2009.Drought Tolerance in Maize,Handbook of Maize:Its Biology.Springer, New York. pp. 311–344.

Singh V, Oosterom E J V, Jordan D R, Messina C D, Cooper M, Hammer G L. 2010. Morphological and architectural development of root systems in sorghum and maize.Plant and Soil, 333, 287–299.

Tollenaar M, Wu J. 1999. Yield improvement in temperate maize is attributable to greater stress tolerance.Crop Science,39, 1597–1640.

Trachsel S, Messmer R, Stamp P, Hund A. 2009. Mapping of QTLs for lateral and axile root growth of tropical maize.Theoretical and Applied Genetics, 119, 1413–1424.

Xue Q, Marek T, Xu W, Bell J. 2017. Irrigated corn production and management in the Texas High Plains.Journal of Contemporary Water Research and Education, 162, 31–41.

Xue Q, Stewart B A, Lazar M D, Piccinni G, Salisbury C D. 2012.Genotypic variation of osmotic adjustment, water-use and transpiration efficiency among closely related wheat lines.Journal of Crop Improvement, 26, 258–281.

Xue Q, Zhu Z, Musick J T, Stewart B A, Dusek D A. 2003.Root growth and water uptake in winter wheat under de ficit irrigation.Plant and Soil, 257, 151–161.

Zhan A, Schneider H, Lynch J P. 2015. Reduced lateral root branching density improves drought tolerance in maize.Plant Physiology, 168, 1603–1615.

Zhao J, Xue Q, Jessup K E, Hao B, Hou X, Marek T H, Xu W,Evett S R, O’Shaughnessy S, Brauer D K. 2018. Yield and water use of drought-tolerant maize hybrids in a semiarid environment.Field Crops Research, 216, 1–9.

Ziolkowska J R. 2015. Shadow price of water for irrigation - A case of the High Plains.Agricultural Water Management,153, 20–31.