| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
UMR INRA-ENSAT ARCHE, Chemin de Borde Rouge, BP27, 31326 Castanet Tolosan, France
* Corresponding author (mduru{at}toulouse.inra.fr)
Received for publication September 15, 2003.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: DD, degree-days (0°C base temperature) • DM, dry matter • LAI, leaf area index • Na, actual sward nitrogen concentration (g N kg–1 DM) • Ni, sward nitrogen index (varying from 40 to 120) • Nus, nitrogen concentration (g N kg–1 DM) of approximately the upper 7 cm of sward
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Dynamic models for assessing N fluxes (Jeuffroy and Recous, 1997), or simplified approaches based on a balance sheet method for calculating N requirements and supply (Machet et al., 1990), can be used for determining N fertilization strategies. However, there are difficulties in modeling some N fluxes, and data for some N components are hard to obtain. Furthermore, these difficulties are greater for grasslands because different management practices can create substantial differences in patterns of N mineralization and immobilization in soils (Jarvis, 1998) due to the processes involved in the supply of N from organic matter (Jarvis et al., 1996).
Field indicators allow the consequences of the different N fluxes to be assessed without measuring each of the components. Field indicators for management of N fertilizer inputs include monitoring of sward N status. Such indicators can be based on the nitrate concentration in the stem base extract (Justes et al., 1997, for wheat) or a combination of the Na (g N kg–1 DM) and sward mass (Mg DM ha–1), Na = a(DM)–b (Lemaire and Salette, 1984). For C3 grasses growing with nonlimiting N, it was shown that the coefficients a and b are stable over sites and years for swards. The critical N concentration (Nc) is the minimum N concentration allowing the maximum herbage growth rate: Nc = 48 (DM)–0.32 (Lemaire and Gastal, 1997). This equation was used to determine a N nutrition index (Ni) as the ratio, at any time during sward growth, between the Na and the critical plant N concentration: Ni = 100Na/48(DM)–0.32. The Ni is well correlated to the sward growth rate of orchardgrass (Duru et al., 1995), tall fescue (Festuca arundinacea Schreb.) (Bélanger et al., 1992; Duru et al., 1995), and timothy (Phleum pratense L.) (Bélanger and Richards, 1997).
However, there are some practical limitations to conveniently assessing Ni, the main one being the need to assess the sward DM mass. A simplified assessment based on the specific leaf N concentration of the uppermost mature leaves of the canopy was proposed previously (Lemaire et al., 1997). The method is based on the fact that the decline in sward N concentration during growth results from the ontogenetic decline in the proportion of metabolic vs. structural tissue as sward mass increases (Lemaire and Gastal, 1997). In a dense canopy, the N concentration of leaves decreases with increasing light extinction through the canopy (Charles-Edwards et al., 1987). For example, Lemaire et al. (1991) showed that the N concentration of the new leaves appearing in the well-illuminated upper canopy of alfalfa (Medicago sativa L.) stand remained relatively constant while the N concentration of older leaves decreased with shading. Consequently, the N concentration of leaves in the uppermost part of the canopy remained relatively constant despite the general decrease in whole-plant N concentration with increasing sward mass (Lemaire et al., 1997). In those studies, the N concentration was expressed on a leaf area basis rather than on a mass basis because with light saturation, leaf CO2 assimilation rate is correlated with N content per unit leaf area (Muchow and Sinclair, 1994), light interception being an area-based phenomenon. This method was used for maize (Zea mays L.) by punching small disks from each of the mature laminae while excluding the midrib, allowing the N crop status to be successfully assessed (Lemaire et al., 1997). However, for narrow grass leaves, separation of the midrib from the other part of the lamina is difficult and time consuming.
The objective of this paper is to report a method for assessing sward N status from the N concentration in the upper sward canopy, measured on leaf mass basis, including the midrib. Our method was compared with a control—a well-established method based on measurements of total sward DM mass and N concentration (Lemaire and Gastal, 1997).
To be suitable for N diagnosis in on-farm surveys, the method should work at any time after a defoliation event to provide simultaneous information on several paddocks. Sampling conditions, such as thickness of the sampled layer and age of the regrowth, should also be established. With that in mind, we compared swards differing in initial cutting dates (early vs. late). Foreseeable problems are that the sampled laminae may vary in composition because of the number of laminae sampled per tiller or the part of the lamina sampled (i.e., the tip or all of it) could be different for each of the layers sampled. Hence, we studied N distribution along the length of the lamina. To assess the required precision of sampling depth for N analysis, we studied the N distribution vertically down the sward profile, cutting the sward into different layers.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
Average daily temperature ranged from about 11°C over the spring growth periods up to 19°C over the summer regrowth periods. Swards at Auzeville were irrigated during summer when necessary to prevent any water stress. Soil is a clayey loam Fluvisol developed on molasse sediments, a tertiary deposit coming from the Pyrenees. Its pH water (0–10 cm) was 8.0 and organic matter content 16 mg kg–1 soil. Phosphorus and K are not limiting for plant growth.
Measurements for Nitrogen and Cutting Treatments
To compute Ni at different times of a growing period, three to five measurements were taken at 1- to 2-wk intervals, each time on one subplot per block (0.5 by 0.5 m) chosen at random in each Auzeville treatment (defoliation and N) or Aveyron sward. For spring growths, the latest samplings were made before heading stage. Sward height before cutting, sward mass, and plant N concentration by layer were measured. Fifteen measurements of sward height were made per subplot using a sward stick (Duru and Bossuet, 1992). The sward was then cut, layer by layer, using a set three-sided squareframe (a square open on one side) to control the cutting height of the motorized hand shears (Wolf, Wissembourg, Germany). The first frame was laid flat on the ground, and the others were spaced at 7-cm intervals above it. The upper layer could be 0- to 7-cm thick. In practice, it was difficult to yield an uppermost layer thicker than 4 cm. Consequently, the thickness of the yielded upper layer varied from 4 to 11 cm. Each sample (layer and replication) was dried at 80°C for 48 h and then weighed and ground through an 0.8-mm screen for measurement of total N concentration (Kjeldahl, 1883).
Specific Measurements for Summer Regrowth in 1997
Three hundred to 900 tillers per treatment (to get enough material for analysis) were selected at random, and the youngest fully expanded laminae (ligule visible) were cut off and saved. This was done on day of year (DOY) 167, 183, and 197. The laminae were separated into segments of 15-cm length, starting from the tips. Length of the basal segment depended on the total leaf length and was measured on 40 laminae.
Furthermore, 80 tillers per treatment were individually marked with plastic rings within the sward in each plot. Numbers of partially or completely green laminae per tiller were recorded once a week between 168 and 196 DOY.
Calculations and Statistical Analysis
The Ni was equal to Ni = 100Na/48(DM)–0.32 when DM accumulation was >1 Mg ha–1 and as Ni = 100Na/48 when DM < 1 Mg ha–1 (Lemaire and Gastal, 1997). Values below 30 indicate very severe N deficiency; the minimum theoretical value is about 20 (Lemaire and Gastal, 1997).
Analysis of variance was performed using SYSTAT software to assess separately for each treatment whether Ni changed over the regrowth period, comparing data for the different sampling dates, which are independent because they correspond to different subplots. Analysis of variance was also performed to assess the effects of N supply, cutting regimes and their possible interaction on the N concentration in lamina segments during summer regrowth in 1997 (two-way analysis of variance). Regrowths and seasons were not compared statistically. Means were compared for the N concentration in the upper sward layer and in the tip of the youngest fully expanded lamina. Linear regression analysis was made to express Nus as a function of Ni and Ni as a function of Nus.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
Nitrogen Concentration in the Different Sward Layers
The N concentration in sward layers is shown (Fig. 2) for three treatments at different times from the initial cut and the N sampling date. For one of the three swards studied in 1994 in Aveyron (square symbols in Fig. 1), Ni was constant throughout growth, and the first sampling date occurred at least 50 d after N application. For the H1 treatment of summer regrowth in 1997 at Auzeville, there were no significant differences in Ni during the regrowth period, but there was a short delay (14 d) between cutting date and the first sampling date. In contrast, for the H1 treatment of Auzeville spring growth in 1997, Ni decreased significantly over the sampling period when the first sampling date was more than 43 d after the N application (Fig. 1). For these three treatments, the N concentration of a given layer decreased from one sampling date to the next (Fig. 2). When a new layer was sampled, its N concentration was greater than the one below. When there were at least three layers, mean comparison tests show that the decline in N concentration between two successive layers was least for the upper layers and greatest for the layers closest to the ground (P < 0.05).
|
When expressed against sward depth (height = 0 at the top of the canopy), there was a decline in N concentration with sward depth, regardless of sampling date (Fig. 2). The N profiles at the different sampling dates were rather similar for the sward studied in Aveyron but variable for the two H1 treatments at Auzeville. The profile of N concentration was similar to those found for other crops although in the literature it is expressed as specific N content and cumulative LAI {for soybean [Glycine max (L.) Merr.]: Shiraiwa and Sinclair, 1993} or relative irradiance at a given depth within the canopy (for alfalfa: Lemaire et al., 1991). These profiles shown for crops other than grasses were consistent with the N distribution at individual leaf level. The N decline at sward level resulted both from the addition of new leaf segments with a lower N concentration (Table 2), and from aging of leaf segments (Lemaire and Gastal, 1997).
Relationship between Sward Nitrogen Index and the Nitrogen Concentration of the Upper Sward
For a given sampling method, the height and mass of the upper sward could vary with the sampling date and treatment. On average for a treatment and a growing season, the lamina mass in the upper sward layer varied from 10 to 60 g m–2 (not shown). As a percentage of the standing herbage mass, this was 4 to 22%, and this percentage decreased as the number of layers increased with age.
When the N concentration of the upper sward (Nus) was plotted against the Ni, a single relationship was found, regardless of the N treatment, sampling date, or growing season (r2 = 0.87, SE = 3.9). However, the relationship was improved if, as suggested in the previous section, data were divided into sets including samples collected before or after 500 cumulative DD following the previous cutting or N application (see equations in Fig. 3). At Auzeville, the parameters of both equations were not significantly different for spring and summer regrowths (P > 0.05). Samples taken before 500 DD had a higher N concentration, not only for the summer regrowth in 1997 as seen previously (Fig. 2), but also for the spring regrowth in 1997 (three measurements) and the summer regrowth in 1996 (four measurements) (Fig. 3).
|
To predict Ni from Nus, the following equations were established: Ni = 2.94Nus – 20.59 (>500DD) and Ni = 2.28Nus – 12.07 (<500DD).
The N concentration in the upper layer could be a satisfactory indicator of the N sward status if samples are taken after three leaves have appeared following the latest defoliation or date of N application. Before that, the upper sward layer is composed of one or two laminae that have a higher N concentration. The amount of DM formed in the upper layer varied according to sampling date and treatment but did not decrease the accuracy of the Nus measurement because the decline in N concentration down the sward profile was not linear. As observed previously for several crops (Lemaire et al., 1991; Drouet and Bonhomme, 1999), the N concentration decreased slowly near the top of the canopy and quickly at the base of the sward.
While the N concentration in the upper sward has been expressed on an area basis to relate to C gain (Anten et al., 1995), we found a satisfactory correlation between the N concentration on a leaf mass basis and the sward Ni, as observed too by Lemaire and Gastal (1997). As the specific leaf mass is usually significantly greater when there is a N deficiency (Duru et al., 1995), expressing the N concentration on a mass rather than an area basis increases the difference in Nus between N treatments at the sward level as well as at the leaf level. We conclude that expressing the N concentration on a mass basis did not alter the ranking of the different treatments or sampling dates for the Nus.
The critical N concentration for maximum herbage growth found previously (48 g kg–1) when the sward mass was less than 1 Mg DM ha–1 lies between the two values that we found (41.7 and 49.2). It could mean that for low sward mass, there is an effect of the leaf number as we have described above. Before 500 DD, the number of leaves per tiller had not yet leveled out, but in both cases, the critical N concentration was lower than that found for the lamina tip (Table 2) because the maximum N concentration was reached only when the lamina was fully expanded (Maurice et al., 1997). This relationship suggests that a N diagnosis could also be made from the N concentration of the youngest mature fully expanded leaf at any sampling time, as shown for summer regrowth in 1997.
Two control N concentrations in the upper sward layer were found for N sward status assessment depending on the time elapsed since the last defoliation. For an orchardgrass sward, this threshold corresponds to 500 DD. Before this stage, the control N concentration was higher due to the inclusion of fewer laminae from older leaves.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This method is particularly suitable as a decision aid for farmers and advisors as shown previously for the method based on the critical N concentration for crops (Meynard et al., 1997) and grasslands (Duru et al., 2000). On the one hand, it could be used to decide whether to apply N fertilizer when nutrition is below a given threshold. In this case, the diagnosis should be made early so that the Nus value found before 500 DD can be used as the control value to decide whether to increase the sward N status. On the other hand, it could be used for assessing fertilizer strategy retrospectively. Determining insufficient or excessive N nutrition at a key season, especially in spring, may enable the situation to be rectified for the following year according to the rules adopted for nutrient application planning.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Ziadi, G. Belanger, F. Gastal, A. Claessens, G. Lemaire, and N. Tremblay Leaf Nitrogen Concentration as an Indicator of Corn Nitrogen Status Agron. J., July 7, 2009; 101(4): 947 - 957. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
, and 
while those with low (0–80 kg ha–1) N added are indicated by 
, 
, and 
. Circles and squares are for light (L) and heavy (H) defoliation treatments, respectively, at Auzeville; squares, circles, and triangles are different swards in Aveyron; vertical bars represent the standard error; and symbols joined by a broken line mean that measurements were done before 500 degree-days.
