critical nitrogen concentration declines with soil water availability

318 www.crops.org crop science, vol. 54, january–february 2014

RESEARCH

An accurate diagnosis of the N status of crops is required for the optimization of the N management at farm level.

This issue has permanent interest due to the environmental con-sequences of excessive N dressings and the high relative cost of fertilizer N. An efficient N management should avoid the occur-rence of episodes of excess N, aiming to match as best as possible N availability (soil plus fertilizer) with crop N demand. Further, in the case of perennial forage crops, N also influences sward per-sistence (Mackay et al., 2001) and species composition (Schwin-ning and Parsons, 1996).

Crop N demand at any time of crop growth cycle is the result of crop growth rate and its Ncr (Lemaire and Gastal, 2009), Ncr being the minimum N concentration in SB allowing to achieve maximal instantaneous growth rates (Greenwood et al., 1990).

Critical Nitrogen Concentration Declines with Soil Water Availability in Tall Fescue

Pedro M. Errecart,* Mónica G. Agnusdei, Fernando A. Lattanzi, María A. Marino, and Germán D. Berone

ABSTRACTThe diagnosis of the N status of crops is based on the concept of critical N concentration (Ncr), which is the minimum N concentration in shoot biomass (SB) required for maximizing growth. A reference curve of Ncr decrease (ref-Ncr) with SB increase proposed for C3 species (ref-Ncr = 48 SB-0.32) was validated for several crops growing without water deficiency in dif-ferent sites and seasons; however, the validity of ref-Ncr is uncertain when water is limiting. The objective was to assess whether water stress affects Ncr. Five regrowths of a temperate-type tall fescue [Lolium arundinaceum (Schreb.) Dar-bysh.] were followed during autumn, spring, and summer in Balcarce, Argentina. Several N rates were applied and SB accumulation and N concentration were measured in each of four to six sequential SB harvests performed at every regrowth. SB, Ncr, available soil water, refer-ence evapotranspiration (ET0), and real evapo-transpiration (rET) were estimated. Ncr agreed well with ref-Ncr when soil water was nonlimit-ing, but it was consistently lower than ref-Ncr whenever crop rET was reduced (rET/ET0 < 1). Indeed, crop average Ncr during an entire regrowth scaled linearly with the average level of water stress in the period: (Ncr/ref-Ncr)avg = 0.83 (rET/ET0)avg + 0.22 (R2 = 0.90, p < 0.0001). Hence, while ref-Ncr remains appropriate for assessing crop N status under adequate water availability conditions, the N nutrition manage-ment of water stressed crops should be guided by their actual Ncr.

P.M. Errecart, M.G. Agnusdei, and G.D. Berone, Instituto Nacional de Tecnología Agropecuaria (INTA), Estación Experimental Agropecuaria Balcarce, Ruta 226 km 73.5, Balcarce, Argentina; F.A. Lattanzi, Lehrstuhl für Grünlandlehre, Technische Univ. München, D-85350, Freising-Weihenstephan, Germany; M.A. Marino, Facultad de Ciencias Agrarias, Univ. Nacional de Mar del Plata, Ruta 226 km 73.5, Balcarce, Argentina. This publication is a partial requirement for earning a PhD degree at the Univ. Nacional de Mar del Plata by P.M. Errecart. Received 21 Aug. 2013. *Corresponding author ([email protected]).

Abbreviations: D13C, carbon isotope discrimination; DM, dry matter; ET0, reference evapotranspiration; FTSW, fraction transpirable soil water; Ncr, critical N concentration; NNI, N Nutrition Index; Ref-Ncr, critical N concentration of reference; RET, real evapotranspiration; SB, shoot biomass; SBcr, critical SB; SD, standard deviation.

Published in Crop Sci. 54:318–330 (2014). doi: 10.2135/cropsci2013.08.0561 © Crop Science Society of America | 5585 Guilford Rd., Madison, WI 53711 USA

All rights reserved. No part of this periodical may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher. Permission for printing and for reprinting the material contained herein has been obtained by the publisher.

crop science, vol. 54, january–february 2014 www.crops.org 319

Empirical observations of the decline of Ncr with SB increase performed in several species (Lemaire and Salette, 1984; Greenwood et al., 1986, 1991) led to the formulation of reference curves of Ncr dilution (Ref-Ncr) of the type:

Ref-Ncr = a SB-b

where a and b are coefficients physiological group-specific. Such functions were suggested to be applicable at any growth stage and to not vary substantially with major environmen-tal factors (Gastal and Lemaire, 2002). Reference curves of critical N concentration have been validated under different pedoclimatic conditions (Justes et al.,1994; Colnenne et al., 1998; Herrmann and Taube, 2004; Ziadi et al., 2008; Agnus-dei et al., 2010), although in some cases, Ncr dilution curves have shown to be species-specific (Justes et al., 1994; Col-nenne et al., 1998; Marino et al., 2004), to be higher in the seeding year of perennial forage crops (Bélanger and Rich-ards, 2000), to be cultivar-specific (Bélanger et al., 2001), or to decline steadily as perennial forage crops age (Bélanger and Ziadi, 2008). The Ref-Ncr criterion then gave rise to the method of reference for the assessment of crops N nutrition: the N Nutrition Index (NNI), which is computed as:

NNI = crop current SB N concentration/Ref-Ncr

Crop yield is closely related to the NNI (Lemaire and Gastal, 1997; Ziadi et al., 2008; Agnusdei et al., 2010), which confirms the robustness of the concept of Ncr and endorses the NNI as an efficacious tool for the analysis and interpretation of agronomical data (Lemaire et al., 1995; Lemaire and Meynard, 1997; Gonzalez-Dugo et al., 2005).

To be applicable, N nutrition diagnosis techniques should ideally comply with at least two fundamentals: (i) expeditiousness, and (ii) applicability under any cir-cumstance. Concerning the first, Errecart et al. (2012) reported a satisfactory field performance of two NNI proxies of rapid measurement. Regarding the applicabil-ity of the NNI, a yet unclear point is whether the Ncr is indeed constant under any environmental condition. An eventual lower Ncr would lead to underestimations of crop N status and excessive fertilizer N loadings.

Discrepancies between Ref-Ncr and actual Ncr have indeed been observed in several studies. In wheat, short-ened growth season and reduced soil water availability have both been suggested as putative causes of the vari-ability in Ncr among sites ( Justes et al., 1994; Ziadi et al., 2010). In potato, a lowered Ncr was observed under water stress (Bélanger et al., 2001). In forage crops, a lowered Ncr has been observed under nonoptimal growth condi-tions by Agnusdei et al. (2010), who suggested water stress and low temperatures as likely causes of the drop in Ncr.

The occurrence of restrictions to plant growth is the rule rather than the exception in most areas of crop and

forage production. Water stress episodes in particular are highly recurrent, not only in the warm season, but also in spring and autumn. However, we know of no study ana-lyzing the relationship between Ncr and water stress. The aim of the present study was, hence, to assess whether the Ncr of a C3 forage crop is affected by water stress and, if so, to test whether the magnitude of the change in Ncr is related to the intensity of water stress.

MATERiAlS ANd METHodSExperimental Site, Soils, and ClimateExperiments were performed at the Estacion Experimen-tal Agropecuaria Balcarce (Instituto Nacional de Tecnología Agropecuaria), Balcarce, Buenos Aires, Argentina (37°45¢ S and 58°18¢ W, 130 m asl). The climate is temperate subhumid-humid. Monthly mean temperature ranges from 7.8°C in July to 21.4°C in January. Average annual rainfall and ET0 are 990 and 950 mm, respectively. Despite the high rainfall, water stress episodes are common in the warm season, and also in spring and autumn.

Five regrowths were followed in a 9-yr old sward of a temper-ate type tall fescue [Lolium arundinaceum (Schreb.) Darbysh., for-merly Festuca arundinacea (Schreb.)], ‘El Palenque MAG INTA’. Soil tests were performed at the start of experiments. Four out of five regrowths were carried out on a loamy textured Natraquoll (Soil Survey Staff, 2010). Plant available water holding capacity up to 1 m depth was 56 mm, measured with the Richards membrane pressure method (Dane et al., 2002). The 20 cm depth topsoil had an organic matter content of 38 g kg-1, pH 9 (soil: water 1:2.5), P content of 7 mg kg-1 (Bray I), an electric conductivity of 1.0 dS m-1, and 19% exchangeable sodium. The regrowth followed in early spring 2009 was performed on a loamy textured Argiaquoll located nearby within the same paddock, with 59 mm plant avail-able water holding capacity up to 1 m depth and topsoil organic matter content of 96 g kg-1, pH 7.2, P content 8 mg kg-1, an electric conductivity 0.1 dS m-1, and 11.3% exchangeable sodium.

Water BalanceSoil water balances were performed for each regrowth accord-ing to Della Maggiora et al. (2003), taking into account mea-sured rainfall, irrigation, ET0, and soil plant available water. Soil plant available water was constrained to a 1 m depth because living roots were uncommon in deeper soil (data not shown). Soil water balance computations started in 1 Aug. 2008 assum-ing a soil at field capacity, a soil condition assured by 112 mm of rain over the previous 2 mo. ET0 was calculated after Allen et al. (1998) from data recorded at the experimental site (iMETOS ag weather monitoring station, Pessl Instruments GmbH, Weiz, Austria). ET0 was assumed not to be affected by crop N status (Caviglia and Sadras, 2001; Neves Lopes et al., 2011). Runoff was assumed to be zero. RET was assumed equivalent to ET0 when-ever the fraction of transpirable soil water (FTSW) was above 0.4 (Weisz et al., 1994; Allen et al., 1998; Ray and Sinclair, 1998). For FTSW below 0.4, RET was assumed to decrease linearly, yielding nil RET values at zero FTSW. Then, the daily RET/ET0 ratio was estimated, which was considered as an instanta-neous water stress index, theoretically ranging from 1 (RET = ET0, no water stress) to 0 (nil RET, most severe stress possible).


from the maximum SB registered were defined as N-nonlimited. The average SB of all N-nonlimited treatments was considered as the critical SB (SBcr). Then, a linear function of the form:

N = a + b SB

was fitted to all replicates of N-limited treatments, and Ncr was estimated as the N concentration at SBcr.

The Ref-Ncr (g N kg-1 DM) was calculated according to Lemaire and Salette (1984) as:

Ref-Ncr = 48 SBcr-0.32

for SBcr values above 1.55 Mg ha-1. When SBcr was lower than 1.55 Mg ha-1, Ref-Ncr was assumed constant at 41.7 g N kg-1 DM ( Justes et al., 1994):

Ref-Ncr = 48(1.55)-0.32.

The NNI was estimated as the ratio of crop current N con-centration to Ncr. A time-weighted average NNI (NNIavg) was computed for each treatment with all the NNI values estimated during regrowth, as proposed by Lemaire and Gastal (1997). N nutrition index values above 1.0 were assumed as 1.0 when NNIavg was regressed against treatment relative SB accumulation (the ratio of treatment maximal SB accumulation to the maximal SB accumulation observed in the regrowth), since improvements in N status above the optimal condition would not affect plant growth and would underestimate the detrimental effect on SB accumulation of a period of N deficiency during regrowth.

Experimental design and TreatmentsSwards were cut at 5 cm height at the beginning of each regrowth. Subsequently, a P amendment was surface broadcasted as calcium triple superphosphate at a rate of 20 kg P ha-1 to provide nonlim-iting P availability. Thereupon, treatments (four to five N rates according to the regrowth, Table 1) were applied either as urea or calcium ammonium nitrate. Immediately after fertilization, an irrigation of 30 mm was applied to facilitate fertilizer N incorpo-ration and minimize N losses through volatilization.

Treatments were arranged in a split plot design, replicated in two blocks. N fertilizer levels were randomly applied to the main plots. Main plots (18 m2) were divided into subplots which were randomly assigned to harvest dates. Four to six forage harvests, depending on the regrowth, were performed every 7 to 10 d (Supplemental Table S1).

As crop carbon isotope discrimination (D13C, in ‰) usually decreases under water stress (Farquhar et al., 1989) and correlates closely with plant available soil water in grasslands (Schnyder et al., 2006), the accuracy of the RET/ET0 ratio as an estimator of water stress was assessed by D13C measurements made as follows. SB carbon isotope composition (d13C, in ‰), calculated as:

d13C = (13C/12Csample)/(13C/12CV-PDB standard)– 1

was determined in 0.7 mg of SB collected at the last harvest date of regrowths with an elemental analyzer (NA1500, Carlo Erba Strumentazione, Milan, Italy) interfaced to a continuous-flow isotope ratio mass spectrometer (Deltaplus, Thermo-Finnigan MAT, Bremen, Germany). Samples were measured against a working gas standard previously calibrated against a second-ary isotope standard (IAEA-CH6, accuracy ± 0.06‰ standard deviation [SD]). A laboratory standard (wheat flour) was run after every 10th sample to estimate the precision of the isotope analyses ( ± 0.09‰ SD). The D13C was then estimated as:

D13C = (d13Catm–d13Csample)/(1000+d13Csample)×1000

where d13Catm is the 13C content of atmospheric CO2 (assumed -8.3 ‰).

Sampling and MeasurementsSB and N ConcentrationIn each subplot, a 0.1 m2 (0.2 ´ 0.5m) quadrat was randomly selected. Crop SB inside the quadrat was cut at ground level with battery-powered shears. Senescent material was discarded. Thereafter, samples were lyophilized (Rificor LA-B4, Rificor SH, Buenos Aires, Argentina) and weighed to estimate accu-mulated SB (Mg ha-1). Samples were subsequently ground to pass a 40-mesh screen in a Thomas Wiley Mini-Mill (Thomas Scientific, Swedesboro, NJ, USA) and analyzed for total N con-centration (g N kg-1 dry matter [DM]) according to Nelson and Sommers (1973; Method A, without salicylic acid modi-fication). Total N uptake in shoots (kg N ha-1) was estimated as the product of accumulated SB ´ SB total N concentration.

Ncr, Ref-Ncr, and NNICrop Ncr (g N kg-1 DM) was estimated at each harvest date of every regrowth. Harvest dates were not used for Ncr estimation if SB did not differ among N treatments (p > 0.10). At each harvest date, N rates whose SB accumulation did not differ (p > 0.10)

Table 1. Description of applied treatments and climatic conditions registered at each regrowth.

Regrowth(fertilization date)

N fertilization rates and source

Mean incoming global radiation

Mean air temperature (RET†/ET0

‡)avg ratio

kg ha-1 MJ m-2 d-1 °c

early spring 2008 (21 Aug. 2008) 0–75–150–225 Urea 14.4 9.4 0.71

Late spring 2008 (23 Oct. 2008) 0–75–150–225 Urea 22.2 15.1 0.51

Autumn 2009 (19 March 2009) 0–75–150–225 Urea 12.1 13.3 0.63

early spring 2009 (19 Aug. 2009) 0–75–150–350–500 cAn§ 13.1 9.3 0.85

Summer 2010 (30 Dec. 2009) 0–75–150–350–500 cAn 21.5 20.7 0.71†Real evapotranspiration.‡Reference evapotranspiration.§cAn, calcium ammonium nitrate.


The effect of water stress on crop Ncr was evaluated in rainfed regrowths. An identical set of regrowths was conducted in paral-lel following the same experimental design but under nonlimiting water availability, provided by drip-irrigation (driplines spaced 0.60 m apart bearing 1 L hr-1 emitters every 0.30 m). Data from these irri-gated regrowths, published in Errecart et al. (2012), was considered here to: (i) verify the validity of the RET/ET0 ratio as an estimator of water stress, (ii) evaluate the effect of soil water availability on crop N uptake, and (iii) confirm the reliability of the method of estima-tion of crop Ncr.

Statistical AnalysisTreatment means comparison (least significant difference test, 10% significance level) and ordinary least squares linear regres-sion analysis were performed with the analysis of variance (ANOVA) and REG procedures of the SAS package (v 9.0, SAS Institute, Cary, NC, USA), respectively. Slopes and intercepts of the linear functions were compared using dummy variables (Littell et al., 2002). Nonlinearity was tested by assessing the significance of an additional quadratic term.

data digitizationPublished data (Lemaire and Denoix, 1987a; Bélanger et al., 1992; Justes et al., 1994; Bélanger and Richards, 2000; Plénet and Lemaire, 2000; and Ziadi et al., 2008) were digitized using the Engauge Digitizing software (http://digitizer.sourceforge.net).

RESulTS

Climatic Conditions and Soil Water AvailabilityExperiments were run under a wide range of climatic conditions. Mean incoming global radiation during the experimental periods ranged from 12 to 22 MJ m-2 d-1, and mean air temperature from 9 to 21°C (Table 1).

Likewise, rainfed regrowths developed under a wide range of soil water availability. The estimated FTSW was in general above 0.20 during most part of regrowths, but reached a minimum of 0.03 in November 2008, when the estimated RET/ET0 ratio also reached its lowest of 0.06 (Fig. 1). The average RET/ET0 ratio for entire regrowths ranged from 0.51 (late spring 2008) to 0.85 (early spring 2009) (Table 1). Short-term changes in the level of water stress within regrowths were also substantial, as the estimated RET/ET0 ratio varied mark-edly even between consecutive harvest dates (Fig. 1).

Figure 1. evolution of the FTSW (solid lines) and the ReT/eT0 ratio (dotted lines) for five tall fescue regrowths. Vertical bars indicate dates of shoot biomass harvest (FTSW, fraction transpirable soil water; ReT, real evapotranspiration; eT0, reference evapotranspiration).

Table 2. Tall fescue carbon isotope discrimination (D13C) measured at the last forage harvest of each regrowth, under rainfed and nonlimiting water availability conditions.

Rainfed Irrigated

————————— ‰ ————————— early spring 2008 20.13 20.56Late spring 2008 18.82 20.62Autumn 2009 20.22 21.76early spring 2009 20.30 20.75Summer 2010 19.60 20.66


Support for the accuracy of the RET/ET0 ratio as an esti-mator of water stress is lent by D13C values. Under water stress D13C was always lower (Table 2), and the estimated average RET/ET0 ratio for each regrowth correlated well with the ratio of D13C of rainfed to irrigated crops (Fig. 2). Thus, the relative magnitude of water stress in rainfed plots was propor-tional to the magnitude of change in D13C, measured as the deviation from D13C values observed in irrigated plots.

Sward Growth and N uptakeN fertilization significantly increased both sward SB accumulation and N uptake (Fig. 3). Under irrigated con-ditions fertilization increased SB up to 3.82 Mg ha-1 and N uptake up to 140 kg ha-1 (both during summer 2010 regrowth), whereas under rainfed conditions these figures

Figure 2. Relationship between the average ReT/eT0 ratio during each of five tall fescue regrowths and the ratio of carbon isotope discrimination in rainfed to irrigated plots at the last harvest date (ReT, real evapotranspiration; eT0, reference evapotranspiration).

Figure 3. n uptake in shoots vs. shoot biomass (SB) accumulation for five tall fescue regrowths grown either under irrigation (open sym-bols and dotted lines) or under rainfed conditions (solid symbols and solid lines). colors represent different n rates (red: 0n; blue: 75n; black: 150n; orange: 225n; purple: 350n; green: 500n).


where 2.58 Mg ha-1 and 125 kg ha-1, respectively (both during early spring 2009). Importantly, N uptake did not differ significantly between rainfed and irrigated condi-tions at equivalent values of SB. This indicates that N availability in rainfed plots was not affected by water stress to the point of restricting N uptake.

Crop Ncr dependence on the level of Water StressIn irrigated plots, Ncr estimates agreed well with the Ref-Ncr. Indeed, eleven Ncr values estimated over a wide range of SB (1.09–6.30 Mg ha-1) averaged 97% of Ref-Ncr

(Supplemental Table S1, Fig. 4). This further verifies the validity of the Ref-Ncr under nonlimiting conditions (see also Errecart et al., 2012) and corroborates the reliability of the Ncr estimation method used in the present study.

In rainfed conditions, Ncr estimates were also obtained for an ample range of SB, both under conditions of non-limiting water availability (i.e., RET/ET0 = 1; 1.50–4.02 Mg ha-1) and under reduced RET (0.90–4.57 Mg ha-1, Supplemental Table S1). Whenever water was nonlimit-ing, estimated Ncr agreed well with Ref-Ncr. But under conditions of reduced RET, Ncr was lower than Ref-Ncr (Fig. 4). Further, estimated Ncr values expressed in relative terms, as the ratio of Ncr to Ref-Ncr, correlated closely with water stress. The Ncr/Ref-Ncr ratio associated sig-nificantly with several estimations of RET/ET0 made at different time intervals before Ncr estimation (Fig. 5). Even though the period of time before the Ncr estimation in which soil water availability better predicted variations in crop Ncr changed to some extent among regrowths, the average level of water stress during the previous 11 d always explained most variation in crop current Ncr and was the best predictor when all regrowths were considered simultaneously (R2 = 0.65, Fig. 6). This indicates that the interaction between crop Ncr and soil water was relatively rapid. Over complete regrowths, the relationship between both variables was also linear and very close. In fact, the relative decrease in crop time-weighted average Ncr was almost entirely accounted for by water stress (Fig. 7).

Figure 5. Percentage of ncr/Ref-ncr ratio variance explained by the ReT/eT0 ratio estimated either the same day, several days before, or averaged during different periods of time immediately preceding the date of ncr estimation (ncr, critical n concentration; Ref-ncr, critical n concentration of reference; ReT, real evapotranspiration; eT0, reference evapotranspiration).

Figure 4. Relationship between Ref-ncr and estimated ncr un-der rainfed conditions, when soil water balance indicated either non-limiting water availability (solid squares) or water stress (open squares). Solid triangles are ncr datapoints estimated in two tall fescue regrowths grown under irrigation (Ref-ncr, critical n con-centration of reference; ncr, critical n concentration).


N Status under Water deficitNitrogen fertilization significantly improved sward NNI (Table 3). Nonlimiting N status was achieved in all regrowths, except early spring 2008. Since Ncr decreased linearly with increasing water stress (Fig. 6 and 7), assessing sward NNI after the Ref-Ncr underestimated actual N status increasingly more so as water stress intensified. In the most extreme case, RET decreased by 49%, causing a decrease in sward average Ncr of 31% and a corresponding 30% underestimation of sward NNI.

Relationship Between N Status and Forage YieldYields were directly related to crop NNI. When the effect of soil water availability on crop Ncr was accounted for, maximal SB accumulations were achieved with close to nonlimiting N nutrition status (Fig. 8a). When NNIavg was calculated after the Ref-Ncr (instead of the actual Ncr), the relationship became biased, as maximal SB accumulations were achieved at NNIavg substantially lower than 1 (Fig. 8b).

diSCuSSioNNitrogen availability and water stress are the two major limitations to crop production (Sinclair and Rufty, 2012). In assessing whether soil water availability levels limiting crop RET affect crop Ncr, the present work confirms the validity of the Ref-Ncr under nonlimiting water condi-tions, and demonstrates that crop Ncr is systematically lower than the Ref-Ncr under water stress, providing a quantitative analysis of such effect on different seasons on tall fescue, a forage crop.

Notably, the magnitude of the deviation of Ncr from the Ref-Ncr scaled linearly with water stress intensity as measured by the RET/ET0 ratio (Fig. 6 and 7). This means that there is no unique Ncr dilution curve valid for all water stress conditions, and we must instead think of a family of Ncr dilution curves. Ncr continuously responds to the prevailing RET/ET0 conditions, and the time of this adjustment seems to be 11 d, on average (Fig. 6).

Cross-validation of the Relationship Between RET/ET0 and Ncr/Ref-NcrAgnusdei et al. (2010) reported Ncr dilution curves sig-nificantly lower than the Ref-Ncr dilution curve in four

Figure 7. effect of the average soil water availability condition during regrowth on sward time-weighted average ncr/Ref-ncr ratio (ReT, real evapotranspiration; eT0, reference evapotranspiration; ncr, criti-cal n concentration; Ref-ncr, critical n concentration of reference).

Figure 6. Relationship between the average ReT/eT0 ratio estimat-ed during the period of eleven days previous to each harvest date, and the estimated decrease in sward ncr (ReT, real evapotranspi-ration; eT0, reference evapotranspiration; ncr, critical n concentra-tion; Ref-ncr, reference n concentration).

Table 3. Effect of soil water availability on critical N concen-tration (Ncr) and sward N status estimation.

Regrowth N rateRET

decrease†Ncr

decrease§

NNIavg¶ calculated

after the

Ncr Ref-Ncr

kg ha-1 ————— % ————— e arly Spring

20080 19.4‡ 11.3 0.44 0.32

75 0.51 0.37150 0.69 0.51225 0.79 0.59

L ate Spring 2008

0 49.0 31.3 0.55 0.3875 0.82 0.57

150 0.87 0.60225 0.97 0.68

A utumn 2009

0 37.0 30.0 0.66 0.5475 0.75 0.61

150 0.90 0.73225 0.98 0.79

e arly Spring 2009

0 15.0 4.3 0.55 0.5075 0.62 0.56

150 0.86 0.78350 1.09 0.99500 1.20 1.09

Summer 2010

0 29.0 21.0 0.50 0.3975 0.79 0.61

150 0.95 0.74350 1.08 0.84500 1.11 0.87

†Real evapotranspiration decrease: [(1– ReT/Reference evapotranspiration)*100].‡computed up to 21 Oct. 2008 (later harvests were not considered in the analysis because all n treatments differed in SB accumulation, hence there was no certainty that the maximal n rate applied did not restrict crop growth).

§Time-weighted average percentual decrease in critical n concentration = {1–[(ncr/Reference n concentration)avg]*100}.

¶Time-weighted average n nutrition index.


regrowths of C3 forage crops in which growth conditions were not optimal. Soil water balances for each of these regrowths were computed to estimate the daily RET/ET0 ratios and calculate the corresponding Ncr using the relationship presented in Fig. 6. The relationship between observed Ncr values (Agnusdei et al., 2010, their Fig. 6a) and our Ncr estimations is presented in Supplemental Fig. S1. The estimated Ncr values compared very well to observed ones for AR94, which was the only regrowth out of the four characterized by low rainfall, resulting in an average RET/ET0 ratio of 0.75. In contrast, predicted and observed NNIavg values disagreed for the other three regrowths, in which growth conditions were suspected as not optimal due to factors other than water stress, like low temperatures. This independent validation of the relation-ship between Ncr and RET in an annual species suggests that the relationship reported in Fig. 6 may be robust.

Can Crop Ncr be Appropriately Estimated under Water Stress Conditions?One issue regarding the estimation of Ncr that arises when crop growth rates are low is that the statistical approach may be biased. This is because the lower responses of SB to N fertilizer when growth is limited, e.g., by water stress, may be regarded as statistically nonsignificant, and thus, both SBcr and crop Ncr would be underestimated. In our study, however, this effect was not substantial as estimated SBcr values were consistently close to maximal SB accumulations.

A second issue is that soil N availability is often impaired under water stress, mainly due to reductions in N mineraliza-tion and transpiration-related N fluxes to the roots (reviewed by Gonzalez-Dugo et al., 2010). If N supply is limited to the extent that it cannot meet crop N demand, Ncr would be estimated under nonpotential N availability conditions, and thus, would be underestimated. The consequence of such N-supply limitation is a lowered N uptake at equivalent

values of SB under water stress than under nonlimiting water availability conditions (Lemaire and Denoix, 1987b; Lemaire et al., 1996). In the environment of the present study, how-ever, as Fig. 3 shows, crop N uptake followed fairly the same pattern under both water availability conditions; that is, there were no significant differences in crop N uptake between rainfed and irrigated conditions at equivalent values of SB. The difference in N uptake linked to soil water availability can thus be entirely attributed to the effect of drought on SB accumulation. Even under water stress conditions reducing SB accumulation up to 2.7 Mg ha-1 in late spring 2008 (Fig. 3), tall fescue was able to absorb N at a rate high enough to maintain its SB N concentration. These results suggest that soil N relative availability was not altered by water stress; that is, sward growth was reduced in a similar proportion as soil N availability to the plant. Hence, N uptake should keep increasing when SB has already reached its plateau. This is indeed demonstrated in Fig. 9, which shows the simultane-ous changes in crop SB and N uptake achieved with increases of the N fertilization rate, for those treatments defined sta-tistically–after their SB accumulation did not differ at p = 0.10– as non N-limited. As Fig. 9 shows, the changes in crop N uptake registered in our work are well in agreement with those calculated from several reports from the literature also estimating crop Ncr (Lemaire and Denoix, 1987a; Bélanger et al., 1992; Justes et al., 1994; Bélanger and Richards, 2000; Plénet and Lemaire, 2000; and Ziadi et al., 2008). Finally, the reliability of the Ncr estimations made under the growth lim-iting conditions prevailing in our study was further corrobo-rated when sward growth showed to be much better related to sward N status assessments performed after the actual Ncr than after the Ref-Ncr (Fig. 8).

Why does Ncr decrease under Water Stress?Previous water availability conditions defined sward rela-tive Ncr (Fig. 6 and 7). Such a relationship between Ncr/Ref-Ncr and RET/ET0 implies a fractional decrease in

Figure 8. Relationship between treatment relative shoot biomass (SB) accumulation and its time-weighted average n nutrition index (nniavg) calculated after a) taking into account the effect of soil water availability on the critical n concentration, or b) the critical n con-centration of reference (Ref-ncr). Maximal SB is the highest SB accumulation achieved among the n rates applied at each regrowth. Data from early spring 2008 regrowth was not included, since non-limiting n nutrition status was not achieved.


crop Ncr with water stress, and this type of decrease would only be possible if drought would decrease just the ‘a’ coef-ficient of the Ncr dilution curve, without affecting the ‘b’ coefficient. Variations in the ‘b’ coefficient imply SB-asso-ciated changes in the Ncr/Ref-Ncr ratio; hence, if the ‘b’ coefficient were to change under water stress, the addition of SBcr as regressor variable should improve the percent-age of the Ncr/Ref-Ncr ratio variance explained by the simple regression against RET/ET0. The R2 of the mul-tiple regression including SBcr was, as supposed, not sig-nificantly higher than that of the simple regression (0.654 vs. 0.646). Moreover, when the dataset was split into two SBcr groupings (above and below 2.5 Mg ha-1, [Fig. 6]) and linear regressions were fitted, neither the slopes (p > 0.20) nor the intercepts (p > 0.15) of the regression of Ncr/Ref-Ncr as a function of RET/ET0 differed between groups. Thus, after Fig. 6 and for our growing conditions, we propose the following Ncr dilution curve:

Ncr = a’ SB-0.32,

where

a’ = 20.6 + 28.3 RET/ET0 ratio.

One possible cause of the lowered Ncr under water stress is increased concentration of water soluble carbohydrates lead-ing to a passive decrease in shoot N concentration. This is a plausible mechanism, as the concentration of water soluble

carbohydrates often increase in droughted plants (Karsten and MacAdam, 2001; Shaimi et al., 2009), although it is unclear whether this change is as linearly related to water stress as the drop in Ncr. The Ncr dilution process has been proposed to result from a compartmentation of plant SB in “structural SB (SBs)” and “metabolic SB (SBm)” fractions, having respectively low (Ns%) and high (Nm%) N concen-trations (Lemaire and Gastal, 1997). Then, Ncr dilution results from the ontogenetic decline of SBm/SB as plants get larger. A third component of SB may thus be needed, SBr (for reserves), that being mainly carbohydrates would have a minimal Nr%. As water stress escalates, SBr should become more important leading to a lowered Ncr.

Another hypothesis concerns differential responses to water stress of allocation of dry mass vs. N. Water stress increases allocation belowground (review Poorter et al., 2012). If this change is greater for N than for dry mass, then shoot N concentration would decrease. Again, the framework of Ncr dilution would need to be extended to include roots.

A third possibility is that the lower growth rates under water stress require less N for metabolic purposes. For instance, photosynthetic rates are lower under water defi-cit conditions, and less N is needed in the photosynthetic apparatus to reach maximal assimilation rates (Ghashghaie and Saugier, 1989; Perniola et al., 1999; Shangguan et al., 2000). A fourth hypothesis involves accelerated leaf senes-cence, and consequently increased N mobilization, under water stress (Gan and Amasino, 1997). This would also lower Nm%. The process of N mobilization is not included in the theory of N dilution.

The first two putative mechanisms are congruent with the observation that the magnitude of the decrease in Ncr is independent of SB, i.e., water stress would affect only the ‘a’ coefficient of the Ncr dilution curve. The latter two are not. As both imply effects on Nm%, they would modify the SBm/SB ontogenetic decline, and thus their magnitude would depend on SB, i.e., would affect the ‘b’ coefficient of the Ncr dilution curve. This was not observed in the present study (Fig. 6 and 7). More research is needed to clarify the causes of the Ncr decrease, particu-larly under field conditions.

interpreting the Effect of a lower Ncr under Water Stress in Terms of Crop N demand and its NNiThe balance between soil N supply and crop N demand defines crop N status (Durand et al., 2010; Gonzalez-Dugo et al., 2010). A lowered Ncr under water stress implies that plant N demand for maximizing growth decreases relative to that under not limited RET. Even when soil N supply typically decreases under water stress (Garwood and Wil-liams, 1967), we did not observe a limitation strong enough to restrain N uptake; in general, N continued accumulating in shoots whereas SB did not (Fig. 9). In fact, water stress

Figure 9. Simultaneous changes in tall fescue shoot biomass (SB) and n uptake achieved with increases in the n fertilization rate, for n treatments defined statistically (based on their SB accumulation not differing at p = 0.10) as non n-limited (X-axis: SBnon n-limited treatment- average SBall non n-limited treatments; Y-axis: n uptakeHigher n rate- n uptakeLower n rate). Literature data was obtained from Bélanger et al. (1992); Bélanger and Richards (2000); Justes et al. (1994); Lemaire and Denoix (1987a); Plénet and Lemaire (2000); and Ziadi et al. (2008).


did not induce N deficiency but rather improved crop N status (Table 3), as it reduced crop N demand more than N supply, because water stress decreased both SB accumula-tion and the amount of N required per unit of accumulated SB. This is evident in the short-term dynamics of NNI. For instance, in the summer 2010 regrowth, the NNI of all treatments increased while soil water availability was decreasing, up to 28 Jan. 2010 (Fig. 10). From that date on, such upward trends in crop N status reverted when rainfall events increased soil water availability; somewhat later for the higher N rates, surely owing to a larger soil N supply in those conditions. Hence, during the first part of the regrowth period the soil supplied N in excess of a water stress-reduced N demand. This is the first report in the bib-liography describing such an increase in crop NNI under water stress, in contrast with previous studies reporting either no significant changes (Gonzalez-Dugo et al., 2005) or decreases (Duru et al., 1997) in crop N status under such conditions. If crop NNI would be recomputed after the Ref-Ncr, an analysis of the evolution of crop N status dur-ing the Summer 2010 regrowth would indicate that–albeit at a lower degree– crop NNI would still be increasing dur-ing water stress (data not shown). Hence, the discrepancy between the referred works and our study must be ascribed not only to a different estimation of crop N demand under water stress–that is, estimating crop N status after the Ref-Ncr or after the actual Ncr–but also to different levels of soil N supply between droughted environments. Indeed, water deficit seemed to not alter soil N relative availability in the environment where our study was performed (Fig. 3). Thus, soil N fluxes towards roots, which must have been reduced under water deficit (Durand et al., 2010; Errecart

et al., 2010), still provided N in excess for the decreased N demand exerted by the sward under reduced soil water availability conditions. The extensive root system of the 10 yr old tall fescue sward of our work could account, at least partially, for the high capability of N capture of this pasture, since root length density defines the diffusive N flux, which is the component of soil N fluxes gaining rel-evance under water stress (Durand et al., 2010; Errecart et al., 2010). Indeed, observed values of root length density in the 0 to 10 cm soil horizon (35 cm root cm-3 soil) were much higher than those reported by Gonzalez-Dugo et al. (2005) for a tall fescue sward in the establishment year (6 cm root cm-3 soil). Traits like root dry weight and length density, root hair development and viability, have all been already reported to have major effects on tall fescue per-formance under drought (Huang and Fry, 1998; Huang, 2001; Sun et al., 2013).

Practical implicationsA consequence of a lower Ncr under water stress is that in such situations, NNI assessments based on the Ref-Ncr underestimate crop actual N status. This effect can be sig-nificant; crop NNI was underestimated up to 30% under water stress (Table 3, Fig. 8). Further, assessing tall fescue N status after the Ref-Ncr would have wrongly labeled as N deficient several N-nonlimited treatments (see also Agnusdei et al., 2010). This has important practical impli-cations. For one, the amount of fertilizer N required to reach a given N status changes. For instance, under a con-dition of water availability significantly restricting sward RET and growth like summer 2010 regrowth, achieving an NNI of 0.8 required approximately 100 kg of fertilizer

Figure 10. evolution of the ReT/eT0 ratio (solid rhombi) and the nni for five n treatments during the Summer 2010 regrowth (circles: 0n; triangles: 75n; inverted triangles: 150n; open rhombi: 350n; squares: 500n). Bars show registered rainfall (ReT, real evapotranspiration; eT0, reference evapotranspiration; nni, n nutrition index).


N ha-1. Had sward N assessment been performed after the Ref-Ncr, 220 kg of fertilizer N would have been needed. This analysis has the advantage of hindsight, but it is nec-essary to correct future fertilizer N dressings of crops based on the reported effect of water availability on Ncr. As Fig. 7 shows, knowing the average RET/ET0 ratio of a future sward regrowth would allow fine-tuning fertilizer N applications. Incorporating the concept of Ncr decrease under water stress in crop models should help to achieve such an objective. Several challenges could arise, and one of them will be predicting the average RET/ET0 ratio of future regrowths. For this, historical rainfall and ET0 data, or the output of weather forecasting models could be used. Here, we addressed the challenge of developing predict-ing functions for the Ncr/Ref-Ncr ratio, as we obtained one for tall fescue and further validated it in annual rye-grass. Thus, it seems to hold for the edaphic environment predominating in southeast Buenos Aires. However, the threshold in FTSW at which crops start experiencing stress can differ among species and even among environ-ments (Allen et al., 1998), hence it would be necessary to test whether similar relationships emerge with other crops or under differing growing conditions.

CoNCluSioNSThis work demonstrates that when soil water availability limits crop evapotranspiration, the Ncr is lower than under nonstressed conditions. In tall fescue, Ncr increasingly and linearly diverged from Ref-Ncr as the estimated RET/ET0 ratio decreased. Therefore, the use of the Ref-Ncr curves would underestimate the N status of water stressed crops. An accurate estimation of NNI could be made using Ncr values derived from functions relating the rela-tive decrease in Ncr to the magnitude of water stress. In the present study, the ratio of RET to ET0 showed prom-ising value as an index of water stress that would allow to adjust fertilizer N loadings after historical or forecasted climatic data, to better meet crop N demands.

Supplemental Material AvailableSupplemental Material includes Table S1 (measured SB and N concentration, estimated Ncr and calculated Ref-Ncr for each harvest date of five rainfed and two irrigated tall fescue regrowths) and Fig. S1 (cross-validation of the obtained RET/ET0 vs. Ncr/Ref-Ncr relationship with the independent dataset of Agnusdei et al., 2010).

Supplemental Fig. S1. Relationship between the time-weighted average N nutrition index (NNIavg) reported by Agnusdei et al. (2010) and the NNIavg calculated after com-puting soil water balances, estimating daily soil water avail-ability and calculating the corresponding crop critical N concentration (Ncr), for four regrowths of C3 forage crops: tall wheatgrass 1999 (TW99), oats 1995 (O95), tall fescue El Palenque 1996 (TF EP96), and annual ryegrass 1994 (AR94).

AcknowledgmentsThis study was financially and technically supported by the Instituto Nacional de Tecnología Agropecuaria (INTA) Project PE-AEFP 262921. F.A. Lattanzi received funding from DFG/BMZ (LA2390/1-1). Authors wish to thank three anonymous referees for their comments and also Dr. Gilles Bélanger, Dr. Francois Gastal, and especially Dr. Jean-Louis Durand, and Dr. Gilles Lemaire for the interest shown in this work and the valu-able comments made.

ReferencesAgnusdei, M.G., S.G. Assuero, F.A. Lattanzi, and M.A. Marino.

2010. Critical N concentration can vary with growth condi-tions in forage grasses: Implications for plant N status assess-ment and N deficiency diagnosis. Nutr. Cycling Agroecosyst. 88:215–230. doi:10.1007/s10705-010-9348-6

Allen, R.G., L.S. Pereiro, D. Raes, and M. Smith. 1998. Crop evapo-transpiration: Guidelines for computing crop requirements. Irrigation and Drainage Paper No. 56. FAO, Rome, Italy.

Bélanger, G., F. Gastal, and G. Lemaire. 1992. Growth analysis of a tall fescue sward fertilized with different rates of nitrogen. Crop Sci. 32:1371–1376. doi:10.2135/cropsci1992.0011183X003200060013x

Bélanger, G., and J.E. Richards. 2000. Dynamics of biomass and N accumulation of alfalfa under three N fertilization rates. Plant Soil 219:177–185. doi:10.1023/A:1004749828745

Bélanger, G., J.R. Walsh, J.E. Richards, P.H. Milburn, and N. Ziadi. 2001. Critical nitrogen curve and nitrogen nutri-tion index for potato in eastern Canada. Am. J. Potato Res. 78:355–364. doi:10.1007/BF02884344

Bélanger, G., and N. Ziadi. 2008. Phosphorus and nitrogen rela-tionships during spring growth of an aging timothy sward. Agron. J. 100:1757–1762. doi:10.2134/agronj2008.0132

Caviglia, O.P., and V.O. Sadras. 2001. Effect of nitrogen supply on crop conductance, water- and radiation-use efficiency of wheat. Field Crops Res. 69:259–266. doi:10.1016/S0378-4290(00)00149-0

Colnenne, C., J.M. Meynard, R. Reau, E. Justes, and A. Mer-rien. 1998. Determination of a critical nitrogen dilution curve for winter oilseed rape. Ann. Bot. (Lond.) 81:311–317. doi:10.1006/anbo.1997.0557

Dane, J.H., J.W. Hopmans, N. Romano, J. Nimmo, and K.A. Winfield. 2002. Soil water retention and storage-laboratory methods. In: J.H. Dane and G.C. Topp, editors, Methods of soil analysis, Part 4: Physical methods. SSSA Book Ser. 5. ASA, SSSA, Madison, WI.

Della Maggiora, A.I., A. Irigoyen, J.M. Gardiol, O. Caviglia, and L. Echarte. 2003. Evaluación de un modelo de balance de agua en el suelo para el cultivo de maíz. Rev. Arg. Agrome-teor. 2:167–176.

Durand, J.L., V. Gonzalez-Dugo, and F. Gastal. 2010. How much do water deficits alter the nitrogen nutrition status of forage crops? Nutr. Cycl. Agroecosyst. 88:231–243. doi:10.1007/s10705-009-9330-3

Duru, M., G. Lemaire, and P. Cruz. 1997. The nitrogen require-ment of major agricultural crops: Grasslands. In: G. Lemaire, editor, Diagnosis of the nitrogen status in crops. Springer-Verlag, Heidelberg, Germany. p. 59–72.

Errecart, P.M., M.G. Agnusdei, F.A. Lattanzi, and M.A. Marino. 2012. Leaf nitrogen concentration and chlorophyll meter


readings as predictors of tall fescue nitrogen nutrition status. Field Crops Res. 129:46–58. doi:10.1016/j.fcr.2012.01.008

Errecart, P.M., M.A. Marino, M.G. Agnusdei, and J.L. Durand. 2010. Soil nitrogen fluxes to the roots and their relation with sward nitrogen nutrition. In: C.F. Machado, editor, Interna-tional workshop: An overview of research in pastoral-based systems in the southern part of South America. Editorial Uni-versidad Nacional del Centro de la Provincia de Buenos Aires, Tandil, Argentina. p. 108–110.

Farquhar, G.D., J.R. Ehleringer, and K.T. Hubick. 1989. Carbon isotope discrimination and photosynthesis. Annu. Rev. Plant Physiol. Plant Mol. Biol. 40:503–537. doi:10.1146/annurev.pp.40.060189.002443

Gan, S., and R.M. Amasino. 1997. Making sense of senescence. Molecular genetic regulation and manipulation of leaf senes-cence. Plant Physiol. 113:313–319.

Garwood, E.A., and T.E. Williams. 1967. Growth, water use and nutrient uptake from the subsoil by grass swards. J. Agric. Sci. 69:125–130. doi:10.1017/S002185960001652X

Gastal, F., and G. Lemaire. 2002. N uptake and distribution in crops: An agronomical and ecophysiological perspective. J. Exp. Bot. 53:789–799. doi:10.1093/jexbot/53.370.789

Ghashghaie, J., and B. Saugier. 1989. Effects of nitrogen deficiency on leaf photosynthetic response of tall fescue to water deficit. Plant Cell Environ. 12:261–271. doi:10.1111/j.1365-3040.1989.tb01940.x

Gonzalez-Dugo, V., J.L. Durand, F. Gastal, and C. Picon-Cochard. 2005. Short-term response of the nitrogen nutrition status of tall fescue and Italian ryegrass swards under water deficit. Aust. J. Agric. Res. 56:1269–1276. doi:10.1071/AR05064

Gonzalez-Dugo, V., J.L. Durand, and F. Gastal. 2010. Water defi-cit and nitrogen nutrition of crops. A review. Agron. Sustain. Dev. 30:529–544. doi:10.1051/agro/2009059

Greenwood, D.J., F. Gastal, G. Lemaire, A. Draycott, P. Millard, and J.J. Neeteson. 1991. Growth rate and %N of field grown crops: Theory and experiments. Ann. Bot. (Lond.) 67:181–190.

Greenwood, D.J., G. Lemaire, G. Gosse, P. Cruz, A. Draycott, and J.J. Neeteson. 1990. Decline in percentage N of C3 and C4 crops with increasing plant mass. Ann. Bot. (Lond.) 66:425–436.

Greenwood, D.J., J.J. Neeteson, and A. Draycott. 1986. Quantita-tive relationships for the dependence of growth rate of arable crops on their nitrogen content, dry weight and aerial envi-ronment. Plant Soil 91:281–301. doi:10.1007/BF02198111

Herrmann, A., and F. Taube. 2004. The range of the critical nitro-gen dilution curve for maize (Zea mays L.) can be extended until silage maturity. Agron. J. 96:1131–1138. doi:10.2134/agronj2004.1131

Huang, B. 2001. Nutrient accumulation and associated root char-acteristics in response to drought stress in tall fescue cultivars. HortScience 36:148–152.

Huang, B., and J.D. Fry. 1998. Root anatomical, physiological, and morphological responses to drought stress for tall fescue cultivars. Crop Sci. 38:1017–1022. doi:10.2135/cropsci1998.0011183X003800040022x

Justes, E., B. Mary, J.M. Meynard, J.M. Machet, and L. Thelier-Huche. 1994. Determination of a critical nitrogen dilution curve for winter wheat crops. Ann. Bot. (Lond.) 74:397–407. doi:10.1006/anbo.1994.1133

Karsten, H.D., and J.W. MacAdam. 2001. Effect of drought on growth, carbohydrates, and soil water use by perennial rye-grass, tall fescue, and white clover. Crop Sci. 41:156–166. doi:10.2135/cropsci2001.411156x

Lemaire, G., X. Charrier, and Y. Hébert. 1996. Nitrogen uptake capacities of maize and sorghum crops in different nitrogen and water supply conditions. Agron. 16:231–246. doi:10.1051/agro:19960403

Lemaire, G., and A. Denoix. 1987a. Croissance estivale en matière sèche de peuplements de fétuque élevée (Festuca arundina-cea Schreb.) et de dactyle (Dactylis glomerata L.) dans l’Ouest de la France. I. Etude en conditions de nutrition azoteé et d’alimentation hydrique non limitantes. Agron. 7:373–380. doi:10.1051/agro:19870602

Lemaire, G., and A. Denoix. 1987b. Croissance estivale en matière sèche de peuplements de fétuque élevée (Festuca arundinacea Schreb.) et de dactyle (Dactylis glomerata L.) dans l’Ouest de la France. II. Interaction entre les niveaux d’alimentation hydrique et de nutrition azotée. Agron. 7:381–389. doi:10.1051/agro:19870603

Lemaire, G., and F. Gastal. 1997. N uptake and distribution in plant canopies. In: G. Lemaire, editor, Diagnosis of the nitrogen sta-tus in crops. Springer-Verlag, Heidelberg, Germany. p. 3–43.

Lemaire, G., and F. Gastal. 2009. Quantifying crop responses to nitrogen deficiency and avenues to improve nitrogen use efficiency. In: V.O. Sadras and D.F. Calderini, editors, Crop physiology: Applications for genetic improvement and agron-omy. Elsevier, Burlington, Massachusetts. p. 171–211.

Lemaire, G., F. Gastal, and D. Plénet. 1995. Dynamics of N uptake and N distribution in plant canopies. Use of crop N status index in crop modelling. In: G. Lemaire and I.G. Burns, edi-tors, Diagnostic procedures for crop N management (Les Colloques Series N° 82). INRA Editions, Paris. p. 15–29.

Lemaire, G., and J.M. Meynard. 1997. Use of the nitrogen nutri-tion index for the analysis of agronomical data. In: G. Lemaire, editor, Diagnosis of the nitrogen status in crops. Springer-Verlag, Heidelberg, Germany. p. 44–55.

Lemaire, G., and J. Salette. 1984. Relation entre dynamique de croissance et dynamique de prélèvement d’azote pour un peuplement de graminées fourragères. I.- Etude de l’effet du milieu. Agron. 4:423–430. doi:10.1051/agro:19840503

Littell, R.C., W.W. Stroup, and R.J. Freund. 2002. SAS for linear models. 4th ed. SAS Inst., Cary, NC.

Mackay, A.D., A. Gillingham, C. Smith, P. Budding, P. Philips, W. Clarke-Hill, et al. 2001. Effect of soil physical condition, and phosphorus and nitrogen availability on pasture persis-tence. Grassl. Res. Pract. Ser. 15:85–92.

Marino, M.A., A. Mazzanti, S.G. Assuero, F. Gastal, H.E. Ech-everría, and F. Andrade. 2004. Nitrogen dilution curves and nitrogen use efficiency during winter-spring growth of annual ryegrass. Agron. J. 96:601–607. doi:10.2134/agronj2004.0601

Nelson, D.W., and L.E. Sommers. 1973. Determination of total nitrogen in plant material. Agron. J. 65:109–112. doi:10.2134/agronj1973.00021962006500010033x

Neves Lopes, M., C. Feitosa de Lacerda, M.J. Duarte Cân-dido, R.C. Fernandes Franco Pompeu, R.G. da Silva, J.W. Batista Lopes, et al. 2011. Gas exchange in massai grass under five nitrogen fertilization levels during establishment and regrowth. R. Bras. Zootec. 40:1862–1869. doi:10.1590/S1516-35982011000900004

Perniola, M., G. Posca, V. Caqndido, and E. Tarantino. 1999. Gas exchange in greenhouse grown pepper under nitrogen and water stress. Cahiers Options Méditerranéennes 31:263–273.

Plénet, D., and G. Lemaire. 2000. Relationships between dynam-ics of nitrogen uptake and dry matter accumulation in maize crops. Determination of critical N concentration. Plant Soil


216:65–82. doi:10.1023/A:1004783431055Poorter, H., K.J. Niklas, P.B. Reich, J. Oleksyn, P. Poot, and

L. Mommer. 2012. Biomass allocation to leaves, stems and roots: Meta-analyses of interspecific variation and environ-mental control. New Phytol. 193:30–50. doi:10.1111/j.1469-8137.2011.03952.x

Ray, J.D., and T.R. Sinclair. 1998. The effect of pot size on growth and transpiration of maize and soybean during water deficit stress. J. Exp. Bot. 49:1381–1386.

Schnyder, H., M. Schwertl, K. Auerswald, and R. Schäufele. 2006. Hair of grazing cattle provides an integrated measure of the effects of site conditions and interannual weather variability on d13C of temperate humid grassland. Glob. Change Biol. 12:1315–1329. doi:10.1111/j.1365-2486.2006.01169.x

Schwinning, S., and A.J. Parsons. 1996. Analysis of the coexis-tence mechanisms for grasses and legumes in grazing systems. J. Ecol. 84:799–813. doi:10.2307/2960553

Shaimi, N., R. Kallida, F. Volaire, N. Saidi, and C. Al Faiz. 2009. Summer dormancy and drought survival of Moroccan eco-types of orchardgrass. Crop Sci. 49:1416–1424. doi:10.2135/cropsci2008.09.0545

Shangguan, Z., M. Shao, and J. Dyckmans. 2000. Effects of nitro-gen nutrition and water deficit on net photosynthetic rate and chlorophyll fluorescence in winter wheat. J. Plant Physiol. 156:46–51. doi:10.1016/S0176-1617(00)80271-0

Sinclair, T.R., and T.W. Rufty. 2012. Nitrogen and water resources commonly limit crop yield increases, not necessar-ily plant genetics. Global Food Secur. 1:94–98. doi:10.1016/j.gfs.2012.07.001

Soil Survey Staff. 2010. Keys to soil taxonomy. 11th ed. USDA-Natural Resources Conservation Service, Washington, DC.

Sun, J., W. Meyer, J. Cross, and B. Huang. 2013. Growth and physiological traits of canopy and root systems associated with drought resistance in tall fescue. Crop Sci. 53:575–584. doi:10.2135/cropsci2012.05.0292

Weisz, R., J. Kaminski, and Z. Smilowitz. 1994. Water deficit effects on potato leaf growth and transpiration: Utilizing frac-tion extractable soil water for comparison with other crops. Am. Potato J. 71:829–840. doi:10.1007/BF02849378

Ziadi, N., G. Bélanger, A. Claessens, L. Lefebvre, A.N. Cam-bouris, N. Tremblay, et al. 2010. Determination of a critical nitrogen dilution curve for spring wheat. Agron. J. 102:241–250. doi:10.2134/agronj2009.0266

Ziadi, N., M. Brassard, G. Bélanger, A.N. Cambouris, N. Trem-blay, M.C. Nolin, et al. 2008. Critical nitrogen curve and nitrogen nutrition index for corn in eastern Canada. Agron. J. 100:271–276. doi:10.2134/agrojnl2007.0059

critical nitrogen concentration declines with soil water availability

Documents