The impact of water and nitrogen limitation on maize biomass and resource-use efficiencies for radiation, water and nitrogen

Field Crops Research 168 (2014) 109–118

Contents lists available at ScienceDirect

Field Crops Research journal homepage: www.elsevier.com/locate/fcr

The impact of water and nitrogen limitation on maize biomass and resource-use efficiencies for radiation, water and nitrogen Edmar I. Teixeira a,∗ , Michael George a , Thibault Herreman a,b , Hamish Brown a , Andrew Fletcher c , Emmanuel Chakwizira a , John de Ruiter a , Shane Maley a , Alasdair Noble a a

The New Zealand Institute for Plant & Food Research Limited, Private Bag 4704, Christchurch, New Zealand Institut National Polytechnique de Toulouse (INP ENSAT), Avenue de l’Agrobiopole, Auzeville Tolosane, 31326 Castanet-Tolosan, France c Commonwealth Scientific and Industrial Research Organisation (CSIRO) Ecosystems Sciences, CSIRO Sustainable Agriculture Flagship, Private Bag 5, Wembley, WA 6913, Australia b

a r t i c l e

i n f o

Article history: Received 20 April 2014 Received in revised form 5 August 2014 Accepted 5 August 2014 Keywords: Corn Efficiency Nitrogen Radiation Sustainability Water

a b s t r a c t The impact of limited water and nitrogen (N) supply on maize productivity and on the utilisation efficiency of key production resources (radiation, water and N) was quantified in two field experiments during consecutive seasons in Canterbury, New Zealand. In experiment 1 crops were subjected to five N treatment rates (0–400 kg N/ha) and, in experiment 2, to three N (0 to 250 kg/ha N) and two water regimes (dryland and fully irrigated) using a rain-shelter structure. Limited N and water reduced yield and affected resource-use efficiencies. Total biomass ranged from 8 Mg DM/ha for dryland nil N crops to up to 28 Mg DM/ha for fully irrigated and N fertilised crops. Radiation use efficiency declined with N and water limitation from a maximum of 1.4 g DM/MJ to 0.6 g DM/MJ. Transpiration water use efficiency was higher in water stressed crops than irrigated crops (50–70 kg DM/ha/mm) and increased linearly with N fertilizer rates in proportion to the increase in radiation use efficiency. The crop conductance decreased from 0.19 mm/MJ in irrigated crops to 0.07 mm/MJ in dryland crops with negligible response to N fertilizer rates. Nitrogen use efficiency declined with N input rates from 100 to 150 kg DM/kg N, being inversely related to the efficiency of both water and radiation use. Dryland crops recovered 25% less N from applied fertilizer than irrigated crops. These results highlight that benchmarks of resource efficiency need to consider the level of intensification of the production system and illustrate trade-offs between yield targets and the efficiency of water and N use, that depend on the scale of analysis. To establish a balance between economic returns and environmental impacts, these trade-offs need to be managed depending on the relative values assigned to the use-efficiency of each input resource in relation to crop productivity. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The global demand for food is projected to increase by more than 70% by 2050 (Tilman et al., 2011). In order to minimise the need for further expansion of cultivated lands, this production increase needs to be partially achieved through the intensification of crop productivity, i.e. more produce per unit of cultivated land (Mueller et al., 2012). Much of the enhancement in crop productivity in the last half century has been achieved by increasing rates of agronomic inputs (Borlaug, 2003). However, to optimise use of increasingly limited resources and to avoid adverse environmental impacts,

∗ Corresponding author. Tel.: +64 03 325 9659. E-mail address: [email protected] (E.I. Teixeira). http://dx.doi.org/10.1016/j.fcr.2014.08.002 0378-4290/© 2014 Elsevier B.V. All rights reserved.

future yield increases will need to focus on increasing the efficiencies with which inputs such as fertiliser and irrigation water are utilised in agricultural systems (Tilman et al., 2002). When plant growth is unconstrained by water, nutrients and biotic stresses, total above-ground biomass dry-matter (DM) yield (Ydm , g DM/m2 ) is determined by the product of total solar radiation (R0 , MJ/m2 ), the fraction of this radiation that is intercepted by the crop canopy (fR , 0–1) and the efficiency (εR , g DM/MJ Ri ) by which intercepted radiation is converted into biomass via photosynthesis (Monteith, 1972). However, in many production systems worldwide, yields are limited both by water and nitrogen (N) availability. Under these conditions, both fR and εR can be reduced through stresses on canopy expansion and photosynthesis rates, respectively (Lemaire et al., 2008) and contribute to low Ydm . Intensively managed cropping systems achieve high yields using irrigation and

110

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

fertiliser to maintain water and N supplies at levels adequate to satisfy crop demands. However, water (εW , kg DM/ha/mm) and N (εN kg DM/kg N) use efficiencies tend to decrease as N and water inputs increase beyond crop demand, compromising farm profitability and increasing the risk of negative impacts on ecosystems, such as from eutrophication, nitrous oxide emissions and leaching of nitrate to groundwater (Tilman et al., 2002). To determine the best management practices to optimise both yields and resource use efficiencies it is important to understand how crop production and resource efficiency-coefficients respond to both optimum or limiting water and nutrient supplies. This understanding can support the setting of benchmarks to evaluate and improve future agricultural systems in order to balance the need to increase yields and the efficiency of resource use (Kant et al., 2011; Mulvaney et al., 2009; Sadras and Angus, 2006). Information regarding the effects of water and N on yield and multiple resource use efficiencies (εR , εW and εN ) is usually pieced together from distinct datasets with contrasting scales of analysis (e.g. leaf, plant or crop level), crop species, and experimental conditions (e.g. controlled or field environments). For example, previous analysis reported on relationships between εR and εW in lucerne (Brown et al., 2012) and wheat (Caviglia and Sadras, 2001) in the field and mixed stands of annual weeds in pot trials (Hirose and Bazzaz, 1998). For wheat, a strong positive linear relationship between εW and εR indicated that the rate of transpiration per unit of intercepted radiation (i.e. crop conductance, gc , mm/MJ) was conservative across contrasting N availability (Caviglia and Sadras, 2001). The relationships between εR and εN were also studied in wheat under controlled environment (van den Boogaard et al., 1995); and between εW and εN for maize (Zea mays) in pot trials (Eghball and Maranville, 1991). In this study, we quantify these relationships for field grown maize crops under a wide range of water and N supply. Maize is the most produced grain cereal worldwide (FAOSTAT, 2013) and also a key forage option for the New Zealand pastoral-based livestock industry (de Ruiter et al., 2009). Under unconstrained growth conditions, maize crops have inherently higher potential biomass production than temperate cereals, such as wheat, due to the C4 photosynthetic metabolic pathway that increases the efficiency of radiation, water and nitrogen use (Ghannoum et al., 2011; Sage and Zhu, 2011). There is a well established positive relationship between Ydm and εR (Muchow and Sinclair, 1994) in maize and indication of an inverse relationship between εN and εW from controlled environment trials (Eghball and Maranville, 1991). Nevertheless, there is limited information regarding how N and water supply interact to influence multiple efficiencies in maize crops and how these compare with previous reports for other crops. In this paper, we first report on the relationships among total biomass, εR , εW and εN for field maize crops subjected to limited water and N supply in Canterbury, New Zealand. Then, by analysing how resource capture and conversion efficiency are influenced by crop physiological characteristics (e.g. leaf area index, gc and specific leaf nitrogen) we provide insights on the mechanisms explaining useefficiency patterns in response to water and nitrogen limitation.

(USDA Soil Taxonomy) with an available water holding capacity of around 190 mm/m of depth (Martin et al., 1992, 2004). The depth of the groundwater table for both sites was not identified through core sampling or neutron probe measurements (Section 2.2.2), indicating that saturated soil layers are deeper than the effective root zone for maize, in agreement with regional surveys (Hanson and Abraham, 2009). In the first experiment (E1, 2011–2012) fully irrigated crops were sown on 31st October 2011 at a population of 12.5 plants/m2 and row spacing of 0.76 m. A complete randomised design, with five N application rates (0, 50, 100, 200 and 400 kg/ha), was established with 4 replicates. Pre-trial soil tests showed that available P and K where moderately high (Olsen P 19 mg/kg; available K 160 mg/kg). Even so, base fertilizer of 250 kg triple superphosphate (20.5% P) and 200 kg/ha potassium chloride (KCl; 52% K) was applied on 31 October 2011 to ensure that P or K did not constrain plant growth. Nitrogen treatments were applied in the form of urea (46% N) in split applications with 50% at plant emergence (17 November 2011) and 50% at the 6th expanded leaf stage (29 December 2011). In the second experiment (E2, 2012–2013), the same maize hybrid was established under a mobile rain-shelter structure (Martin et al., 2004) located 350 m west of the first season site. Crops were subjected to fully irrigated and dryland (i.e. not irrigated and protected from rainfall) conditions using the rain-shelter. Briefly, the rain-shelter consists of a greenhouse-like structure (55 m long × 12 m wide) that is positioned in a “parkingmode” ∼50 m from experimental plots. During rainfall events, the rain-shelter automatically moves along metal rails to cover the experimental area, enabling total control of water inputs through a drip irrigation system. Prior to the experiment establishment on 19 October 2012, base fertilizer in the form of 488 kg/ha of triple superphosphate (100 kg P/ha and 68.32 kg Ca/ha) and 100 kg/ha of KCl (50 kg K/ha) was applied on the entire experimental area. Prior to the experiment, the soil was intentionally depleted of N by removing above-ground biomass through sequential harvests of an unfertilised perennial ryegrass pasture (Lolium perenne) followed by an unfertilised oat (Avena sativa) crop. On 23 October 2012, the maize crop was sown at a density of 12 plants/m2 with 0.71 m row spacing. Six experimental treatments combining two water regimes (dryland or fully irrigated) and three nitrogen application rates (0, 75 and 250 kg/ha) were arranged in a randomised complete block design with 4 replicates giving 24 plots (3.6 m × 5.0 m). For fully irrigated crops, water was applied weekly to replenish evapotranspiration measured at the site (Section 2.2.2). The N fertilizer was applied as urea (46% N) in either two or three splits depending on treatment. The first urea application was at sowing (0, 25 and 50 kg N/ha), the second at the sixth expanded leaf stage (0, 50 and 100 kg N/ha) and the third at crop anthesis (0, 0 and 100 kg N/ha) for the respective N treatments. In both experiments, crops were frequently monitored and agrichemicals were applied at commercial rates as required to minimise biotic stress (George et al., 2013 for details). Damage resulting from biotic stresses (insect, pathogen and weed pressure) was negligible in both experiments. 2.2. Measurements

2. Materials and methods 2.1. Field management and experimental design Data from two experiments in which the maize hybrid Pioneer 39G12 was sown in consecutive growth seasons (2011–2012 and 2012–2013) and adjacent sites (spaced by around 350 m) at Lincoln, Canterbury, New Zealand (43◦ 37 S, 172◦ 28 E) were used in this study. The soil is similar in both sites being a deep (>1.7 m) Templeton silt loam (New Zealand Soil Bureau, 1968) or Udic Ustochrept

2.2.1. Site weather and initial soil conditions Daily weather variables (solar radiation, maximum and minimum air temperatures, vapor pressure deficit, Penman potential evapo-transpiration and rainfall) were accessed from measurements at the Broadfields weather station (National Institute of Water and Atmospheric Research – NIWA 17603) which is located ∼200 m from the rain-shelter site (Fig. 1). Total soil moisture through the 1.6 m soil profile (Section 2.2.2) was estimated at ∼240 mm in Experiment 1 in the early stages of crop establishment (8 December 2011) and 330 mm in Experiment

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

111

Fig. 1. Weather conditions experienced by maize crops in the first (2011–2012, Experiment 1, E1) and second (2012–2013, Experiment 2, E2) growth seasons. For E2, the experimental area was protected from rainfall events with a rain-shelter structure. Irrigation in E2 refers to fully irrigated plots (774 mm/season), dryland plots received 101 mm of irrigation between October and November to enable even crop germination and 8.5 mm subsequent as a carrier for N fertiliser. Penman potential evapotranspiration (ETo ) was accessed from the National Institute of Water and Atmospheric Research website (www.niwa.co.nz).

2, with 55 and 86 mm in the top 0.4 m, respectively. Initial total mineral N to a depth of 1.0 m was on average ∼167 ± 10 kg N/ha for Experiment 1 on 15 November 2011. For experiment 2, initial soil mineral N to 1.5 m depth ranged from 24 to 79 kg N/ha, with less than 36 kg/ha in the top 600 mm.

weighed separately and samples were finely ground with a Cyclone Sample Mill (Udy Corporation, Fort Collins, CO, USA) to pass through a 1-mm screen for N analysis using the Dumas high temperature combustion method with a LECO TruSpec C/N analyser.

2.2.2. Soil water monitoring Absolute soil moisture content (SMCA , mm) was estimated from volumetric soil moisture content (SMCV , %) taken at the top soil layer (0–0.20 m) with a Time Domain Reflectometer (TDR) Trase system, Model 6050X1 (Soil Moisture Equipment Corp., Santa Barbara, USA). From 0.20 to 1.60 m soil depth, SMCV was measured using a neutron probe Troxler model 4300 (Research Triangle Park, NC, USA). In the first season (E1), 8 SMCV readings were taken from 8 December 2011 to 26 April 2012. In the second season (E2), 23 readings were taken at approximately 14 days intervals from 23 October 2012 to 4 April 2013 at 0.2 m soil depth intervals down to 1.6 m.

2.2.4. Radiation interception The fractional transmission of photosynthetically active radiation (PAR) through the canopy was measured weekly using a ceptometer (Decagon Sunfleck Ceptometer) by taking one reading above (R0 , incoming radiation) and five readings below the crop canopy (Rt , transmitted radiation). Readings were always taken above the height of senesced leaves, to account for different rates of basal leaf senescence observed across treatments. Readings were always taken on cloudless days between midday and 1400 h.

2.2.3. Biomass samples In the first season (E1), samples of above-ground biomass (Ydm ) were taken at 5-week intervals beginning from 17 January 2011. A total above-ground biomass sample size of 2 m length by 2 rows area (3.04 m2 ) was taken from each plot. From these samples, a representative 2-plant sub-sample was retained to determine dry matter content and yield components. In the second season (E2), total above-ground biomass samples were also taken five times until the onset of physiological maturity. For all measurements, a buffer of two rows in the border of each plot was left intact throughout the entire experimental period to avoid edge effects from neighbouring plots. During each sampling date in E2, 6–24 plants were destructively harvested in each plot. From these, a sub-sample of 3–6 plants was taken for estimation of dry matter content. All sub-samples were dried in a forced air draft oven at 60 ◦ C for at least 48 h to a constant weight. Each plant component was

2.2.5. Estimation of specific leaf nitrogen (SLN) from chlorophyll metre (SPAD) readings In the second season (E2), use efficiencies were analysed in relation to leaf chlorophyll concentration of the largest leaf (∼12th leaf from the base of the stem) estimated using a chlorophyll metre (SPAD-502, Minolta Camera Co, Japan). Values for each treatment were estimated from the average of 9 SPAD measurements per plot (three measurements per plant on three plants), from 9 sampling dates throughout the life span of the largest leaf (16 January–12 March 2013) before senescence, mostly coinciding with the early grain-filling period. Specific leaf nitrogen (SLN, g N/cm2 of leaf) was then estimated using a calibration developed for an independent sample set relating SLN to a wide range of SPAD readings (R2 = 0.86, n = 17, 4 < SPAD < 48). The SLN for calibration was calculated for individual leaves as the product of specific leaf weight (SLW, g DM/mm2 lamina) and nitrogen content in the biomass (g N/g DM) from combustion analysis (Dumas method). Individual leaf lamina size (m2 /leaf) was determined with a LI-COR LI-3100 leaf area metre (LI-COR Corporation).

112

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

2.3. Calculations

anthesis and early grain-filling (20 February 2012 and 2013 for E1 and E2, respectively) when Ydm was <21.8 Mg/ha.

2.3.1. Daily intercepted solar radiation Daily PAR (Ri , MJ PAR/m2 ) was calculated as the product of linearly interpolated fractional interception (Section 2.2.4) and daily incident PAR. The PAR was assumed as 50% of incident total solar radiation (Sinclair and Muchow, 1999; Szeicz, 1974), measured at the nearby weather station (Section 2.2.1). For comparison with previous literature, results were reported and further calculations performed on the basis of total solar radiation.

2.3.2. Resource use efficiencies In biological systems, resource efficiencies quantify the ratio between input and output and can refer to different levels of organisation (e.g. from plant tissue to the crop to the agricultural region) and to different components of the system (Spedding, 1988). In this study, we concentrate our analysis primarily at the crop scale with a physiological focus (Hay and Porter, 2006). Specifically, efficiencies were calculated always in relation to total above-ground biomass production. The efficiency (ε) of radiation, water and nitrogen utilisation was respectively defined as the quotients between maize above-ground biomass (Ydm ) and (i) accumulated intercepted solar radiation (Ri ), (ii) accumulated transpired water (Wt ) and (iii) N uptake (Nu ) accumulated in above-ground biomass at the final harvest. i.

εR =

Ydm Ri

ii.

εWt =

Ydm Wt

iii.

εN =

Ydm Nu

(1)

For consistency with previous literature, the accumulated crop total above-ground dry matter (Ydm ) production is in g/m2 for εR calculation or kg/ha both for εW and εN . The Ri is the accumulated intercepted solar radiation (MJ/m2 ), Wt is accumulated plant transpiration (mm) and Nu is the accumulated nitrogen in the above-ground biomass of the crop (kg N/ha). In addition, to gain further insights on efficiencies at the cropping system/agronomic scale, water efficiencies were also calculated for accumulated evapo-transpiration (Wet ) to give a corresponding εWet value. Similarly, the fraction of applied fertiliser that was taken up by the crop was estimated as the fertiliser recovery percentage (FRP, %), adapted from O’Neill et al. (2004). FRP =

Nur − Nu0 × 100 r

(2)

where Nur is the nitrogen uptake retained in above-ground biomass at a given N fertiliser application rate of r (kg N/ha) and Nu0 is the N retained in biomass when no nitrogen was applied (i.e. r = 0 kg N/ha). For irrigated crops, resource use efficiencies were also analysed in relation to the nitrogen nutrition index (NNI) estimated for maize crops (Plénet and Lemaire, 1999). The NNI (Eq. (3)) was calculated as the quotient between actual N concentration in the above-ground biomass (N%act ) and the critical N concentration (N%crit ) which is assumed constant at 3.4% for Ydm < 1 Mg/ha and declines with Ydm according to an allometric relationship between 1 and 22 Mg/ha (Eq. (4)). This dilution curve was originally derived for crops unconstrained by water supply and therefore only appropriate to evaluate irrigated crops in our study (Errecart et al., 2014; Sadras and Lemaire, 2014). The NNI was estimated between

NNI =

N%act N%crit

NNI = 3.4(Ydm )−0.37 for Ydm > 1 Mg/ha

(3)

(4)

2.3.3. Water use calculations, crop evaporation and transpiration estimates Total water evapo-transpiration (ET) per growth cycle (Wet , mm) was calculated as the sum of daily changes in stored soil moisture content, water input as rainfall (for E1 only) and irrigation applications for both experiments. In both experiments, volumetric soil moisture content (SMCV , fractional) was measured throughout the entire experimental period (Section 2.2.2). On a daily time step, the value of measured evapo-transpiration (Wet ) was portioned into soil evaporation the (We ) and plant transpiration (Wt ) using the appropriate measure of fractional solar radiation interception following the daily water balance methodology described by Brown et al. (2012). This procedure indicated that water losses due to drainage and run-off were negligible during the experimental period, which is consistent with the amount of rainfall and irrigation in relation to ETo (Fig. 1), the small rate of water application through drip irrigation, the high soil water holding capacity (190 mm/m) and the depth of the soil profile (>1.7 m). Total Wet was calculated as the net balance among irrigation, rainfall (for E1 only) and soil moisture change during each sequential SMCV readings. Values of Wet , Wt and We were normalised by vapour pressure deficit (vpd) in a daily basis and reported for a vpd of 1 kPa to account for any differences in atmospheric demand. This adjustment had only a small influence on results because vpd was, on average, 1.22 and 1.28 kPa in the first and second seasons, respectively.

2.4. Analysis of results Statistical analyses were performed using the software R (R Core Team, 2012). Analysis of variance (ANOVA) was used to partition the observed variation between treatment effects and errors. For analysis with significant results (P-value of the ANOVA was less than 0.05) a Tukey’s HSD (honestly significant difference) post hoc test was performed. Regression analyses and graphics were carried out with a model/loss fitting procedure using Sigmaplot 10.0.1.2 (Systat Software, Inc.). When there was uncertainty in independent variables of linear models, error-in-variables regressions were carried out (Fuller, 1987).

3. Results 3.1. Total above-ground biomass The above-ground dry matter yield (Ydm ) was reduced (P < 0.01) by limited N in both experiments (Fig. 2). There was no interaction (P = 0.58) between water and N treatments for Ydm in E2. At final harvest, total biomass for irrigated crops ranged from 20 Mg DM/ha for the nil N treatment to 25 Mg DM/ha for the fully fertilised crops in E1 (Fig. 2a) and from 20 to 28 Mg DM/ha in E2 (Fig. 2b). In the second season (E2), water stress significantly reduced (P < 0.01) Ydm of dryland crops to <12 Mg DM/ha, around 60% less than for irrigated crops (Fig. 2a–b). For the dryland crops, Ydm ranged from 8 Mg DM/ha for the nil N treatment to 12 Mg DM/ha for the fully fertilised treatment receiving 250 kg N/ha (Fig. 2b).

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

113

Fig. 2. Total above ground biomass of maize crops subjected to contrasting nitrogen fertiliser applications in 2011–12 (Experiment 1, a) and to both nitrogen fertiliser and water regimes (irrigated and dryland) treatments in 2012–13 (Experiment 2, b). Error bars are the standard error of the mean for n = 4.

3.2. Total resource use by the crop Total intercepted solar radiation (Ri ) was unaffected (P > 0.42) by nitrogen treatments in both experiments (Fig. 3a). In contrast, water stress reduced (P < 0.01) Ri from 2033 MJ/m2 in irrigated crops to 1340 MJ/m2 in dryland crops in E2. For irrigated crops, total crop transpiration (Wt ) ranged from 368 in E1 mm to 429 mm in E2 (Fig. 3b). The dryland treatment reduced Wt (P < 0.01) to only 84–97 mm in E2. Although, the Wt was similar among fertilised crops in E1 (P > 0.8) and dryland E2 (P > 0.9), a slight (2–7%) but significant (P < 0.05) reduction in transpiration was found for E1 (9 mm less Wt ) and E2 (36 mm less Wt ) when comparing nil N with fertilised irrigated crops. Nitrogen uptake (Nu ) in harvested biomass increased (P < 0.01) with additional N fertiliser application. The Nu increased from ∼150 kg/ha at nil N rates to ∼275 kg/ha for full N rates in fully irrigated crops in both experiments (Fig. 3c). Water stress reduced Nu (P < 0.01), from an average of 217 kg N/ha in fully irrigated crops to 78 kg N/ha in dryland crops. 3.3. Resource efficiencies in response to water and nitrogen supply

to up to 145 kg DM/kg N (P < 0.01), particularly for the highest N fertiliser application rate. 3.4. Relationships among resource efficiencies The relationships among resource efficiencies were empirically analysed by plotting each individual combination of pairs of coefficients (Fig. 5). There was a significant (P < 0.01) decline of both εR and εWt in response to increasing εN (Fig. 5a and c). A separation between data-points from dryland and irrigated treatments was particularly evident due to differences in εWt (Fig. 5a and b). Within the ranges of use-efficiencies calculated for irrigated crops, the relation between coefficients appeared to be linear (Fig. 5) and all slopes were significant (P < 0.01). For dryland crops in E2, there was also a significant positive relationship between εR and εW (Fig. 5b). In contrast, the response to εN for both εWt (Fig. 5a) and εR (Fig. 5b) were not significant (P > 0.44) under the dryland conditions of E2. Nevertheless, there was a consistent trend of having the lowest εR (0.6 kg DM/MJ) and εWt (100 kg DM/ha mm) estimated at the highest εN values of 145 kg DM/kg N. 3.5. Crop conductance and soil evaporation estimates

Radiation use efficiency (εR ) increased (P < 0.01) with N fertiliser applications up until 250 kg N/ha in the fully irrigated treatments (Fig. 4a). Under irrigation, the εR ranged from 1.0 g DM/MJ at nil N to up to 1.4 g DM/MJ at the full fertiliser rate. Under dryland conditions, limited N supply reduced εR (P < 0.05) from 0.84 g DM/MJ in full N crops to 0.60 g DM/MJ in nil N crops. Water stress reduced (P < 0.01) εR by an average of 37% in dryland crops. The efficiency of water use for transpiration (εWt ) also increased (P < 0.01) with increasing N fertiliser rates up until 250 kg N/ha (Fig. 4b). On average, fully fertilised treatments were 20% more water efficient than nil N treatments. The εWt was greater for dryland crops (94–125 kg DM/ha/mm) than irrigated crops (52–66 kg DM/ha/mm) in E2. The nitrogen utilisation efficiency (εN ) was reduced (P < 0.01) by increasing N fertiliser rates, ranging from around 100 to 130 kg DM/kg N in irrigated crops (Fig. 4c). Similarly, under dryland conditions in E2, nil N crops showed a 14% higher εN than fully fertilised crops. Water stress in E2 increased εN on average by 16%

Crop conductance (gc ) estimates were similar (P > 0.19) among irrigated nitrogen treatments, with a pooled average of 0.19 mm/MJ for irrigated crops in both experiments (Table 1). The drought treatment in E2 caused a significant (P < 0.01) reduction in gc to an average of 0.07 mm/MJ, with no effect of nitrogen fertiliser rates. Nitrogen treatments had a negligible effect in soil evaporation (We ) estimates which were ≤11 mm/year greater (P < 0.01) for nil N treatments under irrigation and were unaffected (P = 0.44) by N treatment in E2 dryland treatments. A twice as large fraction of Wet was lost by soil evaporation (∼41% of Wet ) under water limited conditions than irrigated conditions (Table 1). The maximum leaf area index (LAImax ) was unaffected (P > 0.61) by nitrogen treatments in E1 with an average of 5.3 m2 leaf/m2 soil for the first season (Table 1). Similarly, no nitrogen effect was observed (P > 0.16) on LAImax for dryland treatments in E2 (Table 1). In contrast, for irrigated crops in E2, LAImax was reduced (P < 0.04) from 5.3 in the full N treatment to 4.2 for the nil N treatment. Water

114

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

Fig. 3. Total amounts of intercepted solar radiation (Ri , MJ/m2 ), water transpired (WT , mm) and N uptake retained in the biomass (Nu , kg N/ha) at final harvest for maize crops subjected to contrasting nitrogen fertiliser applications in 2011–2012 (black circles) and to both nitrogen fertiliser (open circle) and water regimes (irrigated and dryland) treatments in 2012–2013 (triangles). Error bars are one standard deviation. Lines illustrate empirical response patterns pooled for irrigated treatments (solid) and dryland (dashed) conditions.

stress reduced LAImax (P < 0.01) to an average of 2.9 m2 leaf/m2 soil in E2 (Table 1). 3.6. Crop nitrogen status (NNI and SLN) and percentage of N fertiliser recovery The nitrogen nutrition index (NNI), estimated for irrigated crops in both experiments, ranged from 0.6 to 1.3 (Fig. 6). The εR and εWt increased (P < 0.01) at higher NNI with maximum values being reached at NNI of around 1. The εN , by definition the inverse of N%act in our analysis, decreased with increasing NNI (Fig. 6c). For dryland crops, the nitrogen status was evaluated by comparing chlorophyll metre readings with values for irrigated crops taken across all nitrogen treatments in E2 (Fig. 7). Crops in E2 showed a wide range of SPAD values ranging from 10 to 50 units, which corresponded to specific leaf nitrogen (SLN) estimates from ∼0.5 to 1.3 g N/m2 leaf. The SLN estimates were consistent higher (P < 0.01) for irrigated crops (0.9–1.3 g N/m2 leaf) than for dryland crops (0.5–0.6 g N/m2 leaf). The εR and εWt increased (P < 0.01) with

Fig. 4. Use efficiency of intercepted radiation (εR ), transpired water (εWt ) and nitrogen taken-up (εN ) in response to nitrogen fertiliser application for irrigated and dryland maize crops. Error bars are one standard deviation. Lines illustrate empirical response patterns pooled for irrigated treatments (solid) and dryland (dashed) conditions.

increasing SLN for both irrigated and dryland treatments (Fig. 7a and b). In contrast, the efficiency of N use (εN ) decreased with increasing SLN for all treatments (Fig. 7c). The fraction of applied N fertiliser that was recovered by the crop, the N fertiliser recovery percentage (FRP), declined from 80% at 75 kg N/ha to less than 35% at rates of 400 kg/ha in irrigated crops (Fig. 8). Under drought conditions, on average 25% less N (P < 0.01) was recovered at a given N application rate when compared with irrigated crops. 4. Discussion Nitrogen and water limitation affected biomass yield and the efficiencies of radiation, water and nitrogen use in maize crops. The upper limit of ∼28 Mg DM/ha in this study (Fig. 2) is close to potential yields reported under field (de Ruiter et al., 2009) and modelled (Fletcher et al., 2011) experiments for Canterbury, suggesting that nearly optimal growth rates were reproduced with the full irrigation and full fertiliser treatments. The nearly 4 fold

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

115

Fig. 5. The relationship among the use-efficiencies for radiation interception (εR ), transpired water (εWt ) and nitrogen taken up (εN ) in maize crops subjected to contrasting nitrogen and water availability.

reduction in maize above-ground biomass, obtained through imposed water and N stress, enabled the quantification of resource use efficiencies for a wide range of production situations. 4.1. Resource capture in response to nitrogen and water supply Despite the large difference in biomass yields, the magnitude of N limitation in our study was insufficient to significantly affect crop radiation interception (Ri , Fig. 3a) or total crop transpiration (Wt ) of most crops, with the exception of a slight reduction in Wt (<36 mm/year) for nil N treatments (Fig. 3b). This is partially because non fertilised crops were still able to take 50–150 kg N/ha up from the soil under dryland and irrigated conditions, respectively (Fig. 3c). Nevertheless, this relatively low sensitivity of Ri and Wt to N supply has been also found in other studies with maize

(Barbieri et al., 2012; Massignam et al., 2012). Maize was shown to prioritise radiation interception by maintaining leaf area expansion at the cost of nitrogen concentration per unit leaf area that causes photosynthetic rates to be reduced (Lemaire et al., 2008). This can be illustrated by the small changes to leaf area index (LAImax , Table 1) at low N input rates in comparison with the large decrease in SLN that caused εR also to decline (Fig. 7a). The curvilinear response of Ri to LAI (Monsi and Saeki, 2005) and the acropetal pattern of initial leaf senescence in maize crops, which is faster in stressed crops, also explain the limited sensitivity of Ri to N treatments. This enabled the largest leaves positioned near the ear to remain green for longer, through N translocation from basal leaves, extending the period of higher fractional light interception as senescence progresses (Borrás et al., 2003; Kosgey et al., 2013). Nitrogen limitation had also a negligible impact on the amount of

Table 1 Crop conductance, soil evaporation and maximum leaf area index (LAImax ) of maize crops subjected to contrasting nitrogen fertiliser rates and irrigation. Nitrogen treatment (kg N/ha)

Experiment 1 (2011–2012) 0 50 100 200 400

Water treatment

Irrigated Irrigated Irrigated Irrigated Irrigated

HSD Nitrogen (P value) Water (P value)

0.176a 0.180a 0.181a 0.176a 0.178a 0.014 0.73(NS)

HSD Nitrogen (P value) Experiment 2 (2012–2013) 0 75 250 0 75 250

Crop conductance (mm/MJ)

Dryland Dryland Dryland Irrigated Irrigated Irrigated

0.063b 0.073b 0.071b 0.199a 0.204a 0.208a 0.022 0.19 (NS) <0.01

Means with different letters differ significantly by Tukey HSD (˛ = 0.05).

Soil evaporation mm/year

% of Wet

138a 131ab 129b 130b 129b

27.2 26.0 25.5 25.8 25.5

4.65 <0.01 61c 67c 69bc 102a 87ab 93a 18.9 0.44 (NS) <0.01

LAImax (m2 leaf/m2 soil)

4.98a 5.17a 5.72a 5.36a 5.23a 1.50 0.61 (NS)

42.1 40.7 41.4 20.7 17.2 17.8

2.45d 3.05d 3.31cd 4.16bc 5.03ab 5.31a 1.08 <0.01 <0.01

116

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

Fig. 6. Use efficiency of intercepted radiation (εR ), transpired water (εWt ) and nitrogen taken-up (εN ) in response to nitrogen nutrition index (NNI) for irrigated maize crops between silking and the early stages of the grain filling period. Error bars are one standard deviation. Lines illustrate empirical response patterns pooled for irrigated treatments.

water transpired per unit of radiation intercepted, the crop conductance (gc ), similar to findings for wheat crops (Caviglia and Sadras, 2001). Under irrigation, our estimates showed a ∼30% lower gc for maize crops (0.19 mm/MJ, Table 1) than the 0.25–0.29 mm/MJ found for wheat crops by these authors, which illustrates the more efficient use of water in C4 than C3 cereals at the crop scale. In contrast to the small response to N fertiliser rates, water limitation largely reduced Ri and Wt by 38 and 72%, respectively (Fig. 3a). This response was mediated by a concomitant reduction of leaf area by ∼36% and of gc by up to 60% (Table 1). Water limitation also markedly reduced nitrogen uptake causing more fertiliser to be left in the soil instead of recovered by the crop (Fig. 8). These results reinforce the importance of optimising the degree of water and N co-limitation in cropping systems by matching water or N supply according to the most limiting factor, considering both inputs (e.g. N fertiliser, irrigation) and soil storage during the crop cycle (Sadras, 2004).

Fig. 7. The use efficiency of radiation interception (εR ), transpired water (εWt ) and nitrogen taken up (εN ) of maize crops in response to average chlorophyll metre readings (SPAD), and its corresponding estimated specific leaf nitrogen (SLN), pooled during the life span of the largest leaf for Experiment 2 (E2). Lines illustrate empirical response patterns pooled for dryland (dashed) and irrigated (solid) treatments.

4.2. Resource use efficiencies in response to water and nitrogen supply Nitrogen and water limitation affected the resource-use efficiencies in maize crops through different mechanisms. Within each water regime, limited N reduced εR by up to 40% in proportion to a reduction in SLN (Fig. 7a) which agrees with previous reports for maize under N stress (Massignam et al., 2012; Muchow and Davis, 1987). Our analysis used intercepted radiation as the bases for εR calculation. It is possible that by considering absorbed radiation there would be a less severe decline in εR for stressed crops due to an increase in radiation reflectance for canopies under N and water limitation, as previously reported for maize (Schepers et al., 1996). Maximum values of 1.4 g DM/MJ for εR and 70 kg DM/ha/mm for εWt in irrigated crops (Fig. 4), equivalent to 55 kg DM/ha/mm in an evapo-transpiration basis (εWet ), are within the range reported for unconstrained maize and sweet corn grown in New Zealand (Fletcher et al., 2008; Neal et al., 2007). The εR in our study is lower

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

Fig. 8. Fertiliser recovery percentage of maize crops subjected to contrasting nitrogen fertiliser application rates and water regimes. Line illustrates linear trends for irrigated crops. Error bars are one standard deviation.

than maximum values of 1.7 g DM/MJ found in warmer climates for maize (Muchow and Sinclair, 1994) because C4 photosynthesis rates are often limited by low temperatures in the cool temperate climate of Canterbury (Wilson et al., 1995). In contrast, N limitation increased εN in both experiments through a dilution of N concentrations in plant biomass. Plant N concentration estimated as 1/εN ranged from 0.7% for dryland crops at low N, which is similar to the 0.8% minimum previously reported for maize crops (Lemaire and Gastal, 1997), to >1.0% in irrigated crops. Although the lowest N concentrations equate to higher εN , this increase in εN at plant scale can be misleading because it comes with a cost of a reduced efficiency of fertiliser and soil N uptake at the agricultural system scale (Fig. 8 and Fig. 3c, respectively). For irrigated crops, the nitrogen nutrition index (NNI) was used to provide a more physiologically meaningful quantification of crop N status by accounting for N dilution effects (Sadras and Lemaire, 2014). The NNI ranged from 60% below optimum for the 0 kg N/ha crops (NNI = 0.6) to a 30% excess N consumption for the 400 kg N/ha crops (NNI = 1.3). The εR and εWt were optimised at NNI close to 1.0 for irrigated crops in both experiments. For dryland crops, the crop nitrogen status was quantified through SLN estimates from SPAD readings (Fig. 7). The SLN was consistently lower in dryland than irrigated crops for the same N fertiliser application suggesting that drought conditions exacerbated N stress. This is supported by the lower N fertiliser recovery observed under dryland conditions (Fig. 8). Nitrogen limitation also consistently reduced εWt (Fig. 4b), in agreement with previous observations for temperate cereals such as wheat and barley crops (Cabrera-Bosquet et al., 2007; Cooper et al., 1983). A positive response of photosynthetic efficiency to N supply, translated into higher εR (Fig. 4a), was the main mechanism explaining increased water use efficiency (εWt ) at high N fertiliser rates in our study. This is illustrated by the strong linear relationship between εR and εWt , within each water regime (Fig. 5b). The slope of this relationship is analogous to the inverse of crop conductance in Table 1. This stability of gc for conditions in which εR and εWt change proportionally has been previously proposed for lucerne crops (Brown et al., 2012). The small sensitivity of gc to N treatments indicates that changes to εWt in response to N supply were mostly driven by non-stomatal limitations, similar to findings for wheat (Caviglia and Sadras, 2001). This suggests that the limited response of stomatal conductance to N supply found at leaf level for other species such as sunflower and sorghum (Jacob et al., 1990) may scale up to the canopy level also in maize

117

crops. Although the trade-off found between εWt and εN (Fig. 5a) agrees with previous reports, its magnitude cannot be generalised due to the influence of site-specific environmental factors (e.g. incident radiation, temperature and vapour pressure deficit) on both crop growth and evapo-transpiration (Sadras and Rodriguez, 2010). Under drought conditions, εWt increased to values of up to 120 kg DM/ha/mm (Fig. 4), equivalent to ∼70 kg DM/ha/mm of water for total evapo-transpiration (Wet ), around 35% more efficient than for irrigated crops. This was largely caused by a decline in crop conductance (gc ) because transpiration (Wt , Fig. 3b) was relatively more reduced than canopy cover (i.e. LAImax , Table 1) in dryland crops. The difference in εWt caused the clear separation of data-points between dryland and irrigated crops in Fig. 5a–c, illustrating another scaling up of leaf processes to the crop level. Specifically, that the additional resistance to gaseous diffusion caused by stomata closure under drought stress has a greater proportional effect on transpiration than CO2 uptake and therefore increases εWt (Hay and Porter, 2006). However, the higher εWt for dryland crops comes with a price of reduced biomass accumulation due to limited CO2 uptake, as illustrated by Ydm < 12 Mg DM/ha in water stressed crops in comparison with >25 Mg DM/ha for irrigated crops (Fig. 2). The impacts of contrasting water and N supply on the carbon footprint of maize crops were not explored in our study. This may be an important aspect for future research as shown by preliminary life cycle assessments in New Zealand (MAF, 2011) and the US (Grassini and Cassman, 2012). These indicate that nitrogen fertiliser (use and manufacture) and energy for irrigation are by far the largest contributors to CO2 emissions per unit area in intensive maize production systems. However, yield improvements at high water and N input levels, when accompanied by increases in resource-use efficiency (Herrmann et al., 2014), may reduce emissions per unit product and minimise the need to exploit additional land to achieve targets of production (Grassini and Cassman, 2012). Further research on the interplay among productivity, resource-use efficiencies and emissions is required to identify possible paths for sustainable intensification of maize production.

5. Conclusions Our results illustrate the interdependencies among productivity and the efficiencies of radiation, water and N use in maize crops under resource constrained conditions. Nitrogen limitation reduced maize yields mainly through a decline in radiation use efficiency. This translated into a proportional reduction in water use efficiency but more efficient use of N taken up. Water limitation reduced yields both through a reduction in resource capture (radiation interception and N uptake) and conversion efficiency (radiation use efficiency) but translated into higher efficiency of water use. As a consequence, radiation and water use efficiencies were positively related through a consistent value of crop conductance across a wide range of N supply but variable depending on water supply. These results highlight that targets and benchmarks of resource-use efficiencies should consider the level of intensification of the agricultural system, e.g. the soil water and N available or supplied through irrigation and fertilizer inputs. To produce yields close to the potential set by radiation incomes, there is an inherent cost of operating at sub-optimal efficiencies of water and nitrogen use that depends on the scale of analysis (e.g. at the crop or cropping system scale). To establish a balance between economic returns and environmental impacts, these trade-offs need to be managed depending on the relative values assigned to the use-efficiency of each input resource in relation to crop productivity.

118

E.I. Teixeira et al. / Field Crops Research 168 (2014) 109–118

Acknowledgements The authors thank Dr Craig Anderson and Dr Denis Curtin who provided constructive suggestions on an earlier version of this manuscript. This research was funded by The New Zealand Institute for Plant & Food Research Limited through the Land Use Change and Intensification Programme (LUCI) and the Pastoral 21 Environment Programme. We thank Genetic Technologies Limited for providing the maize seed for the trials and Ms. Alex Noble for the assistance with data collection.

References Barbieri, P., Echarte, L., Della Maggiora, A., Sadras, V.O., Echeverria, H., Andrade, F.H., 2012. Maize evapotranspiration and water-use efficiency in response to row spacing. Agron. J. 104, 939–944. Borlaug, N.E., 2003. The green revolution: its origins and contributions to world agriculture. J. Bioresour. Sci. 4, 11–22. Borrás, L., Maddonni, G.A., Otegui, M.E., 2003. Leaf senescence in maize hybrids: plant population, row spacing and kernel set effects. Field Crop. Res. 82, 13–26. Brown, H.E., Jamieson, P.D., Moot, D.J., 2012. Predicting the transpiration of lucerne. Eur. J. Agron. 43, 9–17. Cabrera-Bosquet, L., Molero, G., Bort, J., Nogués, S., Araus, J.L., 2007. The combined effect of constant water deficit and nitrogen supply on WUE, NUE and (13C in durum wheat potted plants. Ann. Appl. Biol. 151, 277–289. Caviglia, O.P., Sadras, V.O., 2001. Effect of nitrogen supply on crop conductance, water- and radiation-use efficiency of wheat. Field Crop. Res. 69, 259–266. Cooper, P.J.M., Keatinge, J.D.H., Hughes, G., 1983. Crop evapotranspiration – a technique for calculation of its components by field measurements. Field Crop. Res. 7, 299–312. de Ruiter, J., Fletcher, A., Maley, S., Sim, R., 2009. Aiming for 45 t/ha per annum: yield of supplementary feed crops grown in sequences designed for maximum productivity. Proc. N. Z. Grassl. Assoc. 71, 107–116. Eghball, B., Maranville, J.W., 1991. Interactive effects of water and nitrogen stresses on nitrogen utilization efficiency, leaf water status and yield of corn genotypes. Commun. Soil Sci. Plant Anal. 22, 1367–1382. Errecart, P.M., Agnusdei, M.G., Lattanzi, F.A., Marino, M.A., Berone, G.D., 2014. Critical nitrogen concentration declines with soil water availability in tall fescue. Crop Sci. 54, 318–330. FAOSTAT, 2013. Production Statistics. Food and Agriculture Organization of the United Nations, http://faostat.fao.org/site/567/default.aspx#ancor (accessed 07.04.09). Fletcher, A.L., Moot, D.J., Stone, P.J., 2008. Radiation use efficiency and leaf photosynthesis of sweet corn in response to phosphorus in a cool temperate environment. Eur. J. Agron. 29, 88–93. Fletcher, A.L., Brown, H.E., Johnstone, P.R., de Ruiter, J.M., Zyskowski, R.F., 2011. Making sense of yield trade-offs in a crop sequence: a New Zealand case study. Field Crop. Res. 124, 149–156. Fuller, W.A., 1987. Measurement Error Models, Wiley Series in Probability and Mathematical Statistics. Probability and Mathematical Statistics. John Wiley & Sons, Inc., New York, ISBN 0-471-86187-1. George, M.J., Teixeira, E.I., Herreman, T.F., Brown, H.E., 2013. Understanding nitrogen and water stress mechanisms on maize crops. Proc. Agron. Soc. N. Z. 43. Ghannoum, O., Evans, J., Caemmerer, S., 2011. Nitrogen and water use efficiency of C4 plants. In: Raghavendra, A.S., Sage, R.F. (Eds.), C4 Photosynthesis and Related CO2 Concentrating Mechanisms, 32. Springer, The Netherlands, pp. 129–146 (Chapter 8). Grassini, P., Cassman, K.G., 2012. High-yield maize with large net energy yield and small global warming intensity. Proc. Natl. Acad. Sci. 109, 1074–1079. Hanson, C., Abraham, P., 2009. Depth and spatial variation in groundwater chemistry – Central Canterbury Plains. Technical Report R09/39 ISBN 9781-86937-969-8. Investigations and Monitoring Group, http://www.ecan. govt.nz/publications/Reports/groundwater-report-depth-spatial-variationgroundwater-chemistry-central-canterbury-plains-000209.pdf Hay, R.K.M., Porter, J.R., 2006. The Physiology of Crop Yield. Blackwell Publishing, Oxford, UK, pp. 328. Herrmann, A., Claus, S., Loges, R., Kluß, C., Taube, F., 2014. Can arable forage production be intensified sustainably? A case study from northern Germany. Crop Pasture Sci. 65, 538–549. Hirose, T., Bazzaz, F.A., 1998. Trade-off between light- and nitrogen-use efficiency in canopy photosynthesis. Ann. Bot. 82, 195–202. Jacob, J., Udayakumar, M., Prasad, T.G., 1990. Mesophyll conductance was inhibited more than stomatal conductance in nitrogen deficient plants. J. Plant Physiol. Biochem. 17, 55–61. Kant, S., Bi, Y.-M., Rothstein, S.J., 2011. Understanding plant response to nitrogen limitation for the improvement of crop nitrogen use efficiency. J. Exp. Bot. 62, 1499–1509. Kosgey, J.R., Moot, D.J., Fletcher, A.L., McKenzie, B.A., 2013. Dry matter accumulation and post-silking N economy of ‘stay-green’ maize (Zea mays L.) hybrids. Eur. J. Agron. 51, 43–52.

Lemaire, G., Gastal, F., 1997. N uptake and distribution in plant canopies. In: Diagnosis of the Nutritional Status in Crops. Springer, Berlin/Heidelberg/New York, pp. 3–34. Lemaire, G., van Oosterom, E., Jeuffroy, M.-H., Gastal, F., Massignam, A., 2008. Crop species present different qualitative types of response to N deficiency during their vegetative growth. Field Crop. Res. 105, 253–265. MAF, 2011. Ministry of Agriculture and Forestry. Carbon Footprint of New Zealand Arable Production – Wheat, Maize Silage, Maize Grain and Ryegrass Seed. MAF Technical Paper No: 2011/97, August, www.fedfarm.org.nz/ Files/2011-MPIGrainCarbon.pdf. Prepared for Foundation for Arable Research and Ministry of Agriculture and Forestry by AgriLINK New Zealand Ltd. Martin, R.J., Jamieson, P.D., Wilson, D.R., Francis, G.S., 1992. Effects of soil moisture deficits on yield and quality of ‘Russet Burbank’ potatoes. N. Z. J. Crop Hortic. Sci. 20, 1–9. Martin, R., Wilson, D.R., Jamieson, P.D., Stone, P.J., Reid, J.B., Gillespie, R.N., 2004. A simple versatile model of crop yield response to water deficit New directions for a diverse planet. In: Fischer, R.A. (Ed.), Proceedings of the 4th International Crop Science Congress. Brisbane, Australia, 26 September–1 October. Massignam, A.M., Chapman, S.C., Hammer, G.L., Fukai, S., 2012. Effects of nitrogen supply on canopy development of maize and sunflower. Crop Pasture Sci. 62, 1045–1055. Monsi, M., Saeki, T., 2005. On the factor light in plant communities and its importance for matter production. Ann. Bot. 95, 549–567. Monteith, J.L., 1972. Solar radiation and productivity in tropical ecosystems. J. Appl. Ecol. 9, 747–766. Muchow, R.C., Davis, R., 1987. Effect of nitrogen supply on the comparative productivity of maize and sorghum in a semi-arid tropical environment II. Radiation interception and biomass accumulation. Field Crop. Res. 18, 17–30. Muchow, R.C., Sinclair, T.R., 1994. Nitrogen response of leaf photosynthesis and canopy radiation use efficiency in field-grown maize and sorghum. Crop Sci. 34, 721–727. Mueller, N.D., Gerber, J.S., Johnston, M., Ray, D.K., Ramankutty, N., Foley, J.A., 2012. Closing yield gaps through nutrient and water management. Nature 490, 254–257. Mulvaney, R.L., Khan, S.A., Ellsworth, T.R., 2009. Synthetic nitrogen fertilizers deplete soil nitrogen: a global dilemma for sustainable cereal production. J. Environ. Qual. 6, 2295–2314. Neal, J.S., Greenwood, K.L., de Ruiter, J.M., Martin, R.J., 2007. Forages and water use efficiency. In: Dairy Science: Proceedings of the 2007 Australasian Dairy Science Symposium, University of Melbourne 18–20 September, pp. 240–255. New Zealand Soil Bureau, 1968. General survey of soils of the South Island. Soil Bureau Bulletin 27. New Zealand Department of Scientific and Industrial Research, Wellington. O’Neill, P.M., Shanahan, J.F., Schepers, J.S., Caldwell, B., 2004. Agronomic Responses of Corn Hybrids from Different Eras to Deficit and Adequate Levels of Water and Nitrogen Joint contribution of USDA-ARS and Agric. Res. Div. of the Univ. of Nebraska. Published as Journal Ser. no. 14623. Mention of commercial products and organizations in this article is solely to provide specific information. It does not constitute endorsement by USDA-ARS over other products and organizations not mentioned. The USDA-ARS is an equal opportunity/affirmative action employer and all agency services are available without discrimination. Agron. J. 96, 1660–1667. Plénet, D., Lemaire, G., 1999. Relationships between dynamics of nitrogen uptake and dry matter accumulation in maize crops. Determination of critical N concentration. Plant Soil 216, 65–82. R Core Team, 2012. R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0, http://www.R-project.org/ Sadras, V.O., 2004. Yield and water-use efficiency of water- and nitrogen-stressed wheat crops increase with degree of co-limitation. Eur. J. Agron. 21, 455–464. Sadras, V.O., Angus, J.F., 2006. Benchmarking water-use efficiency of rainfed wheat in dry environments. Aust. J. Agric. Res. 57, 847–856. Sadras, V.O., Lemaire, G., 2014. Quantifying crop nitrogen status for comparisons of agronomic practices and genotypes. Field Crop. Res. 164, 54–64. Sadras, V.O., Rodriguez, D., 2010. Modelling the nitrogen-driven trade-off between nitrogen utilisation efficiency and water use efficiency of wheat in eastern Australia. Field Crop. Res. 118, 297–305. Sage, R.F., Zhu, X.-G., 2011. Exploiting the engine of C4 photosynthesis. J. Exp. Bot. 62, 2989–3000. Schepers, J.S., Blackmer, T.M., Wilhelm, W.W., Resende, M., 1996. Transmittance and reflectance measurements of corn leaves from plants with different nitrogen and water supply. J. Plant Physiol. 148, 523–529. Sinclair, T.R., Muchow, R.C., 1999. Radiation use efficiency. Adv. Agron. 65, 215–265. Spedding, C.R.W., 1988. Biological efficiency in agriculture. In: An Introduction to Agricultural Systems. Springer, The Netherlands, pp. 52–72. Szeicz, G., 1974. Solar radiation for plant growth. J. Appl. Ecol. 11, 617–636. Tilman, D., Cassman, K.G., Matson, P.A., Naylor, R., Polasky, S., 2002. Agricultural sustainability and intensive production practices. Nature 418, 671–677. Tilman, D., Balzer, C., Hill, J., Befort, B.L., 2011. Global food demand and the sustainable intensification of agriculture. Proc. Natl. Acad. Sci. 108, 20260–20264. van den Boogaard, R., Kostadinova, S., Veneklaas, E., Lambers, H., 1995. Association of water use efficiency and nitrogen use efficiency with photosynthetic characteristics of two wheat cultivars. J. Exp. Bot. 46 (Special issue), 1429–1438. Wilson, D.R., Muchow, R.C., Murgatroyd, C.J., 1995. Model analysis of temperature and solar radiation limitations to maize potential productivity in a cool climate. Field Crop. Res. 43, 1–18.