Analysis and Interpretation of Factors Which Contribute to Efficiency of Nitrogen Utilizationr

Kamprath, and W. A. Jackson2R. H. Moll, E. J.

ABSTRACTDifferences in N response among corn (Zea mays L.) genotypes

reflect variation in numerous processes involved in N use efficiency.In order to facilitate the study of such variation, we develop anddemonstrate a concept for evaluating the contribution of N uptakeand utilization processes to variation in N use efficiency, Eight hybridswere grown in a replicated field experiment at two levels of N fertilizeron a Dothan loamy sand (Typic Plinthic Paleudult). Differences amongthe hybrids for components of N use efficiency were evaluated frommeasurements of grain yield, N accumulation in the plant at sitking,and N accumulation in the grain and stover at harvest. Significantdifferences were found among hybrids and between N levels for alltraits. Interactions among hybrids and N levels were significant forall traits except grain yield, At low N supply, differences amonghybrids for N use efficiency were due largely to variation in utilizationof accumulated N, but with high N they were due largely to variationin uptake efficiency. Variation in proportion of N translocated tograin was also important at the low N supply. Variation in N ac-cumulated after silking was not important at either level of N supply.Yariation in N remobilization from vegetative tissue to grain wasmoderately important at the low N supply. Hybrids with similar levelsof N use efficiency showed marked differences in component traitswhich contribute to efficiency.

Aitditionat irdr, ,Fr.'N upt"L.J'{ t."n.to."tion, zea mays L.

HE effectiveness with which N is used by corn(Zea mays L.) and other non-legume crop plants

has become increasingly important because of in-creased costs of manufacture and distribution of Nfertilizer. Differences in N utilization among corn gen-otypes have been demonstrated, not only in differ-ential responses to N fertilizer (Smith, 1934; String-f,eld and Salter, 1934), but also in differences inabsorption and in utilization of absorbed N (Beau-champ et al., 197 6; Chevalier and Schrader, 1977 ; Molland Kamprath, 1977; Pollmer et al., 1979; Reed et al.,1980). Thus, the potential for developing superior, N-efficient hybrids appears to exist.

Efficiency in uptake and utilization of N in the pro-duction of grain requires that those processes asso-ciated with absorption, translocation, assimilation,and redistribution of N operate effectively. The rel-ative contribution of these processes to genotypicdifferences in N use efficiency is unknown and mayvary among genetic populations and among environ-ments, including N supply. It therefore is importantto characterize efficiency of N use in terms relatedto variation in the major processes involved. How-ever, a severe constraint in doing so is that directmeasurements are not feasible in large field studiesinvolving both genotypic and fertilizer variables be-cause of high costs in terms of land, labor, and time.As a consequence, most studies to date have beenlimited to estimates of N accumulation in certain plantparts at certain stages of growth. We contend hereinthatjudicious use of such data can afford appreciableinsight about genotypic variation in response to Niupply.

'Paper No. 6842 of the Journal Series of the North CarolinaAgric. Res. Service. Raleigh, NC 27650. Received 30 Mar. 1981.

' Professor of genetics and professors of soil science, North Car-olina State University.

Our objective is to develop and to demonstrate aconcept for using data on N accumulation to evaluatethe relative contribution of various N acquisition anddistribution processes to variation in overall efficiencyof N use.

THEORYWe deflne N use efficiency as grain production per unit

of N available in the soil. Nitrogen use efficiency is GwlNsin which Gw is grain weight and Ns is N supply expressedin the same units (e.9., g/plant). There are two primarycomponents of N use efficiency: (1) the efficiency of ab-sorption (uptake), and (2) the efficiency with which the Nabsorbed is utilized to produce grain. These are expressedas foilows: uptake efficiency : Nr/Ns, and utilization ef-ficiency : GwlNr, where Nr is total N in the plant atmaturity.

It follows that:

Gw/Ns = (Nr/Ns)(GwlNr).

We let Y : log GwlNs, Xl : log Nr/Ns, and X2 : log Gwl rN/, so that Y1 = Xrr * X21 for the ktlr experimental unit.The sum of squares for y : Y - p over all experimentalunits will be:

:oYi: Ilxlu + I1xi, + 2I1x11x21.

The expression can be expanded to include additionalfactors. For example, N uptake during grain filling and trans-location of N to grain may be important variables (Pollmeret al., 1979). Let Na represent N accumulation after silkingand Ng represent N accumulated in grain at harvest. Uti-lization efficiency, Gw/Nr, can be expressed as:

GwlNr = (QwlNg)(NgiNr), andNg/Nr = (Na/NrXNg/Na),

in which

GwlNg : grain produced per unit of grain NNg/Nr : fraction of total N that is

translocated to grainNa/Nr : fraction of total N that is

accumulated after silkingNg/Na : ratio of N translocated to

grain to N accumulated aftersilking.

Therefore: GwlNs = (Nt/Ns)(GwlNS)Nd/Nr)(NSiNa).Other expressions might be devised to evaluate different

factors. As long as the expression is a multiplicative identity,the general expression in terms of logs will be:

Y1 = liX;1,

in which Y1 : log of N use efficiency (GwlNs) for the ktlrexperimental unit and Xi1 - log of the irft multiplicativecomponent.

. By simple algebra, sums of squares of log N use efficiency15:

I,,yi = I*tlrxi * 11;;xpxlr,J = Ir(lryrx;J.Therefore, the contribution of the ifle component to the sumof squares of log N use efficiency is l1y1xi;, which includesnot only the sum of squares of the 1og of the ith componentbut also the sum of products between it and the logs ofother components in the model. This is related to the directand indirect effects of a path analysis (Wright, 1921), whichbecomes obvious if the above expression is substituted into

f

562

I

I

MOLL ET AL.: EFFICIENCY OF N UTILIZATION 563

Appliednitrogen

(Ns) H1-br-c

Nitrogen accumulation

Before AfterTotal silking silking Grain{\o {Nv) (Na) (Ng)

Table 1. Mear: performance of eight hybrids in terms of grainproduction ald \ accumulation at two levels of N fertilizer.

METHODS AND MATERIALSEight experimental hybrids were evaluated in a split plot

experiment with two rates of N fertilizer, (56 and 224 kgN/ha), supplied as ammonium nitrate. The hybrids weresingle crosses between unselected inbred lines () S,* gen-eration) from 'Jarvis Golden Prolific' and 'Indian Chief.'The experiment was conducted with eight replications atClayton, N.C. on a Dothan loamy sand (Typic Plinthic Pa-leudult) in 1978. Each subplot was a single lO-plant row ofa hybrid between experimental inbred lines, with approxi-mately 96 cm between rows and 46 cm between plants. Endsof the plots were planted with a purple marker stock toprovide competition. One competitive plant per plot washarvested at silking and analyzed for total N. Six to eightcompetitive plants per plot were harvested at maturity.Stover was chopped and dried to determine dry weight, anda sample was analyzed for percent N. Shelled grain wasweighed, and a sample analyzed for percent H2O and percentN. Weight measurements were adjusted to a dry weightbasis.

The following data, all in g/plant, serve for analysis:: grain dry weight: total aboveground plant N at maturity: total aboveground plant N at silk: Nt - Nv : aboveground plant N accumulated

after silk: N accumulated in the grain at harvest

Analyses of variance were computed for these five traits.Contribution of the various components to variation

among hybrids in efficiency of N use at each level of Nfertilizer was determined by sums of squares and sums ofproducts of log ratios of hybrid means as outlined in theprevious section.

RESULTS AND DISCUSSION

Analyses of variance showed significant differencesamong hybrids and between N fertilizer rates for allfive primary traits (Table l). Interaction of hybridsand N rate was significant for all traits except grainyield. At the low rate of N fertilizer, comparison ofhybrid means for N accumulation in aboveground tis-sue (Nt) with fertilizer N (Ns) shows that approxi-mately half of the N accumulated in plants at maturitymust have come from mineralization of soil N becauseNt - 2(Ns). The range among genotypes was from43 to 557o.

Logarithms of N use efficiency (Gw/Ns), uptakeefficiency (Nt/Ns) and utilization efficiency (Gw/Nt),which will be denoted oS Y, X1, and X2, respectively,were computed from the means given in Table l. Sumsof squares for Y, and sums of products of Y with Xrand X2 were computed and the contribution of X1 andX2-to Ly2 expressed as (Ixry/IylttOo) and (Lx2y/>y')(100) (Table 2). The relative contribution of thetwo component traits, uptake and utilization, to vari-ation in efficiency among hybrids was considerablydifferent for the two levels of N applied. At low Nsupply the correlation between N use efficiency andN uptake efficiency was small, and variation in thelatter was also relatively small. Therefore, variationin N uptake efficiency contributed very little to vari-ation in N use efficiency among the hybrids. This isin sharp contrast to its substantial contribution to vari-ation under high N.

A further breakdown in utilization efficiency indi-

._..:Gq'

g/plant

2.47 -2

3756;s

l;

.iE

:a

Standard error

5.{8 2.385.25 2.'.17.1.63 2.425.6.1 3.055.15 2.645,01 2.705.05 2.18J.31 2.34

6.{7 3.31;.6-{ 2.696.70 3.26;.15 3.72121 2.00;.{8 3.705.;5 2.41;.66 4.4r

0.29 0.25

3.10 3.942.54 3.592.2t 2.992.59 3.912.5r 3.182.31 4.072.87 3.882.OO 3.50

3.16 4.354.95 4.869.44 3.923.43 5.072,24 3.053.?8 5.763.34 4.281.26 4.26

0.39 0.27

glplant

-: I

Table 2. Contribution of sarietion in efficiency components tohybrid sum of square for efficiency of N use.

\ Fractionofligz:. -{pplied hybridSS::: rkgha) EaylDy' ryxl S11/Sy

GwNtNvNa

Ne

Trait

N use efficierq;

Uptake efficie:c=.

Utiliztionefficiencl'

0.048 0.102 0.4690.829 0.719 1.153

0.952 0.897 1.0610.171 0.209 0.817

0.449 0.813 0.5520.228 0.402 0.568

0.070 0.143 0.4870.031 0.016 1.944

0.379 0.528 0.7180.197 0.088 2.242

0.503 0.814 0.618

-0.058 -0.147 0.392

1-

\,

56

56224

56224

56221

56224

56224

Grain \ to:a.plant \ r.a).:

N uptake a:te: \:l::sill<ing:a:a

Grain \ \ ':=:a:after silh 'i: )i:

Grain lieic 3-- - ); l- l.:-

ryz: = cr1:-i::: :::t=:.:: ixrween Y and X1: S*i, Sv = standard derivario:-. i:: 1., -': !- :-:*<--r'elr'.

:.xrelation coefficient, and the termsproponion of the sum of squares fori:i: component (X) will be:

= r .. Ir: : (r,.,) S*,/Sy,

in *hrch :,, :! :ie.lar-relation coefficient and S, and S*, arethe s:an;-: :e',:":i..ns for log N use efficiency and the ithcomEllanan: \.-:e -,:.4: a negative correlation between a com-ponent :::,: :nJ \ u:e efficiency will result in a negativeconrnbi:i-.. :. ::-e sum of squares of Y.

The je ,:,r'::-e:rt above involves terms representing ourconceF: ..: ::ii;::rr e parameters of N use efficiency. Someof the.e. :-;: i5 available soil N and total plant N aredifficu^: irr rli:!.ire accurately. In lieu of such measure-menrs. :h.:e:::: \s and Nt could be substituted for Ns andNr to reg,r;r.l fenilizer N and aboveground plant N, re-spectire.-, T:e:.i!rre. \\'e use below roman subscripts torepresenl eri,::rmental measurements. This does not changeIhe algeb:at; ;iielopment presented above, but, ofcourse,must ':e :i:r intr account in the interpretation of theanah:t:.

564

cates that variation in the fraction of N translocatedto grain (Ng/Nt) was more than twice as importantat low N supply than at high N (Table 2). This is dueprimarily to differences in magnitude of correlationbetween N use efficiency and NgiNt. Variation in theproportion of total N that was taken up after silking(Na/Nt) contributed relatively little to variation in ef-ficiency of N use. Even though the variation for thatcomponent was relatively large under high N, it wasnot correlated with N use efficiency.

It has been reported that the amount of N remo-bilized from storage in vegetative tissues is importantin utilization of N (Friedrich and Schrader, 1979; Poll-mer et al., 1979). Accumulation of N in grain (Ng) isequivalent to N accumulated after silking, (Na) plusnet remobilization of N stored in vegetative tissuesprior to silking. If Nr : remobilized N, then (Ng/Na): (Na + Nr)/Na : I + (Nr/Na). Therefore, variationin Ng/Na reflects variation in Nr/Na, and the datashow that Ng/Na was moderately important in con-tributing to variation in N use efficiency at the lowN supply (Table 2). Its contribution at high N supplyis almost entirely due to an unusually large standarddeviation.

All of the hybrids were less efficient in N use at thehigh level of N supply (Table 3). Differences in N useefficiency (Gw/Ns) among hybrids appear to have beendue to a number of factors. Even relatively inefficienthybrids may be above average for one or more of thecomponent factors. Furthermore, hybrids with com-parable high levels of N use efficiency may differmarkedly in the way that level of efficiency is achieved.Hybrids designated by nos. 4, 6, and 7 were relativelyefficient at both levels of N supply. Hybrid no. 4 wasefficient in uptake (Nt/Ns), especially at the low Nsupply. Hybrid no. 7 was average or below in uptakeefficiency, but it was highly efficient in utilizing theN taken up in grain production (Gw/Nt). Hybrid no.7 was also above average in the fraction of total Nthat was translocated to grain (Ng/Nt). Hybrid no. 6appeared to be efficient in uptake when N supply washigh, and was highest in efficiency of translocating Nto grain (Ng/Nt) at both N levels.

Causes of variation in N use efficient in terms ofcomponent factors appear to differ between levels ofN supply and among genotypes. The importance ofvariation in uptake efficiency relative to variation inutilization efficiency was in sharp contrast betweenlevels of N supply. This result may have importantimplications with regard to effects of genetic selectionunder specific levels of N fertility. From the limiteddata presented here we speculate that selection underhigh N supplies might favor genotypes which are ef-ficient in N uptake when N is abundant; but theremay be little or no selection pressure to improve ef-ficiency of utilization of accumulated N. Such geno-types might perform poorly under limited N supplies.

In breeding for improved N use efficiency, it would

AGRoNoMy JoURNAL, vot.74, lray-ruNr, 1982

t20.2 2.04109.4 2.28107.0 2.03r03.0 7.7588.4 2.7290.2 2.2274.8 r.8772.8 2.0895.7 2.05

27.9 0.5831.4 0.7232.3 0.7626.0 0.5727.8 0.7724.6 0.6522.0 0.68).9.7 0.4326.5 0.65

Table 3. N use efficiency and components of N efficiency foreight hybrids at two levels of N fertilizer.

Hybridt Gw/Ns Nt/Ns Gw/Nt Ng/Nt Na/Nt Ng/Na Gw/Ng

4b8q

I

5Men

7468tI35

Mean

Ns = 2.47 g/plant

58.9 0.77 0.5748.0 0_69 0.4652.7 0.81 0.4658.7 0.81 0.4641.6 0.68 0.4840.7 0.72 0.5739.9 0.64 0.4835.0 0.62 0.4946.9 0.72 0.50

Ns = 9.88 g/d4!t48.0 0.7 4 0.5843.4 0.71 0.4842.7 0.77 0.5145.4 0.75 0.2236.0 0.64 0.6537.6 0.67 0.4932.4 0.58 0.5r46.0 0.72 0.5347 _4 0.70 0.49

1.35 76.71.51 69.2t.76 64.9t.75 72.97.42 60.87.27 56.61.35 61.91.27 56.77.46 65.0

1.26 64.51.48 61.11.53 55.43.39 60.30.98 56.61.38 55.91.14 55.41.36 64.01.56 59.2

t Hybrids in rank order by Gw/Ns averaged over both N levels.Gw/Ns = Efficiency of use .

Nt/Ns = Uptake efficiencyGw/Nt = Utilization efficiencyNg/Nt = Grain N/Total N uptake rNa/Nt : N uptake after silking/ Total N uptakeNg/Na = Grain N/N uptake after silkingGw/Ng : Grain yield/Grain N

seem desirable fcrr both uptake efficiency and utili-zation efficiency to be improved simultaneously. Thismay require finding a unique fertility environment ordeveloping a selection index based on data at severalfertility levels in order to ensure equal selection pres-sure on the two components.

LITERATURE CITEDBeauchamp, E. G., L. W. Kannenberg, and R. B. Hunter. 1976.

Nitrogen accumulation and translocation in corn genotypes fol-lowing silking. Agron. J. 68:418-422.

Chevalier, Peggy, and L. E. Schrader.1977. Genotypic differencesin nitrate absorption and partitioning of N among plant parts.Crop Sci. 17:987-901.

Friedrich, J. W., and L. E. Schrader. 1979. N deprivation in maizeduring grain-filling. lI. Remobilization of I5N and 3tS and therelationship between N and S accumulation. Agron. J. 7l:466-472.

Motl, R. H., and E, J. Kamprath. 1977. Effects of population densityupon agronomic traits associated with genetic increases in yieldof Zea mays L. Agron. J. 69:81-84.

Pollmer, W. G., D. Eberhard, D. Klein, and B. S. Dhillon. 1979.Genetic control of nitrogen uptake and translocation in maize.Crop Sci. 19:82-86.

Reed, A. J., F. E. Below, and R. H. Hageman. 1980. Grain proteinaccumulation and the relationship between leaf nitrate reductaseand protease activities during grain development in maize (Zeamays L.). l. Variation between genotypes. Plant Physiol.53:825-828.

Smith, N. 1934. Response of inbred lines and crosses in maize tovariations of nitrogen and phosphorus supplied as nutrients. J.Am. Soc. Agron. 26:785-804.

Stringfield, G. H., and R. M. Salter. 1934. Differential response ofcorn varieties to fertility levels and to seasons. J. Agric. Res.49:991-1000.

Wright, Sewall. 1921. Correlation and causation. J. Agric. Res.20:557-585.

f =-