M. van Ginkel, I. Ortiz-Monasterio, R. Trethowan, and E. Hernández

Introduction


Nitrogen Use Efficiency in Modern Cultivars
Source: I. Ortiz-Monasterio

Modern CIMMYT semidwarf varieties respond more to nitrogen fertilization than old varieties and, at the same time, do not require more nitrogen than older cultivars at lower levels of fertility. They also outyield earlier semidwarf wheat varieties under both low and high nitrogen levels.

In CIMMYT's Bread Wheat Program, breeding is conducted under intermediate nitrogen levels (about 150 units/ha). Emanating lines have shown improved levels of N uptake efficiency (UPE), N utilization efficiency (UTE), and N use efficiency (NUE).*

High yielding varieties are needed in both high and low fertility situations. UTE is important in both high and low input environments because higher yields are obtained with the same amount of N in the plant. UPE is important in both types of environment but for different reasons. In high input agriculture, high UPE avoids leaching of excess N, which contributes to sustainable agriculture. In low input agriculture, UPE is desirable so that increased levels of N are taken up. This enables farmers to get "more bang for their bucks."

CIMMYT breeds widely adapted germplasm and seeks to improve both UPE and UTE. Luckily genetic diversity has been found for both traits.

* N efficiency terminology:

  • NUE = yield per unit of available N (from the soil and/or fertilizer)

  • NUE can be partitioned into two components and is the product of the two:\

  • UPE = ability to extract N from the soil (N in the plant, per unit of total available N in the soil)

  • UTE = ability to convert absorbed N into grain yield (yield per unit of N in the plant)

 

The Questions

Although CIMMYT's breeding approach has been very successful, we wondered whether there might not be an even better method to breed N use efficient wheats. In considering this possibility, we took into account that in selecting for N efficiency:

The level of N in soil determines the expression of UPE and UTE:
  • UPE is better expressed under low N conditions,
  • UTE is better expressed under high N conditions;

Therefore,

  • UPE is more easily selected for under low N conditions,
  • UTE is better selected for under high N conditions.

Based on the above, we asked ourselves: Would manipulating soil fertility during the selection process enhance selection efficiency for both UPE and UTE? And would alternating selection regimes at different nitrogen levels result in combining both traits and high overall NUE?

If these strategies turned out to be effective, they would result in wheats with higher NUE that produce higher yields and generate better returns for farmers.

 

The Experiment

An experiment was conducted to attempt to answer these questions by comparing five different selection regimes. The parental lines used were Genaro 81 and Ciano 79, two wheat lines with good UPE, and Siete Cerros and Opata 85, two wheat lines with good UTE. They were intercrossed in an incomplete 4 x 4 diallel cross, giving six different populations.

Comparing various selection methods

The populations were selected by the breeder under five selection regimes from F2 to F6:

  • LN: All generations grown under low N levels (zero added, in N-deficient soil)
  • MN: All generations grown under medium N levels (150 kg N ha-1 added)
  • HN: All generations grown under high N levels (300 kg N ha-1 added)
  • ALN: Generations grown alternately under low N and high N, with the first segregating generation (F2) being grown under low N (zero added)
  • AHN: Generations grown alternately under high N and low N, with the first segregating generation (F2) being grown under high N (300 kg N ha-1 added).

Comparing breeder's selection with natural selection

To provide a measure of natural selection (= Nature), unselected bulks were grown in each generation and bulk harvested without any selection by humans.

Yield trials

Balanced sets of uniform F7-derived lines emanating from the above selection regimes were grown in large-scale replicated yield trials.

The yield trials were carried out at three N levels on N-deficient soil: 0 N added, 150 kg N ha-1 added, and 300 kg ha-1 added.

Yield, harvest index, biomass, UPE, UTE, and NUE were compared under the five selection regimes.

 

Results and Discussion

Comparing selection methods

Results for yield, harvest index, biomass, UPE, UTE, and NUE under the five selection regimes were as follows.

Yield

  • Under low N there were no significant differences in yield among the five selection regimes.
  • Under intermediate and high N, lines from AHN (alternating selection starting with no N stress in the F2) significantly outyielded most of the other selection regimes.
  • Lines from ALN (alternating selection with N stress applied in the F2) were significantly lower yielding than all other selection regimes.

Harvest index

  • HI increased in a similar magnitude regardless of the breeder's selection regime applied at any of the N levels.

Biomass

  • Under low N there were no significant differences in biomass among selection regimes.
  • Under high N AHN produced significantly more biomass under intermediate N, and ALN produced the least biomass.

UPE

  • Under intermediate and high N the AHN strategy produced the highest UPE lines, while ALN gave the lowest UPE values.

UTE

  • Under high N HN gave the highest UTE lines.

NUE

  • Under low N there were no significant differences among selection regimes.
  • Under intermediate and high N ALN gave the lowest NUE values.

 

N-Level
kg N/ha		 Breeding
Method		 Yield
t/ha		 HI Biomass
t/ha		 Nt1
kg N/ha		 UPE2
kg N/kg N		 UTE
kg N/kg N		 NUE
kg/kg N		 N-Grain
%		 N-Straw
%
0 Selected 2.99 a 0.364 a 8.70 a 72 a 72.02 a 42.02 a -3 1.77 a 0.34 b
  Bulk 2.86 b 0.345 b 8.67 a 72 a 71.68 a 40.60 b -3 1.76 a 0.36 a
150 Selected 5.56 a 0.376 a 12.99 a 149 a 0.99 a 37.62 b 37 a 2.04 b 0.52 b
  Bulk 4.55 b 0.345 b 11.34 b 137 b 0.91 b 33.34 a 30 b 2.16 a 0.61 a
300 Selected  5.46 a 0.359 a 13.84 a 189 a 0.63 a 30.80 a 18 a 2.38 a  0.76 b
  Bulk 4.35 b 0.321 b 12.11 b 162 b 0.54 b 28.40 b 15 b 2.39 a  0.81 a
All Selected 4.65 a 0.366 a 11.84 a 137 a 24.80 a 34.64 a 28 a 2.06 b 0.54 b
  Bulk 3.92 b 0.337 b 10.71 b 124 a 24.70 a 31.97 b 23 b 2.11 a 0.59 a

Note: Means within each column and N-Level followed by the same letter are not significantly different at P = 0.05
1 Nt = Total N in the crop/ha
2 UPE = Nt/N level ; a value of "1' was used for the "0" N level.
3 Non estimable. 

 

 

Comparing breeder's selection with natural selection

In the above table the "selected" values represent those obtained following selection by the breeder for good agronomic type. The "bulk" values represent the effect of Nature.

When zero N fertilizer was added, the breeder's selection outperformed Nature:

  • Yield by 5%
  • Harvest index by 6%
  • UTE by 4%
  • NUE by 4%

Under conditions of intermediate and high levels of N, the breeder's selection even more clearly outdid Nature:

  • Yield by 22-26%
  • Harvest index by 9-12%
  • Biomass by 14-15%
  • UPE by 8-18%
  • NUE by 23-25%


Wheat plots with zero N fertilizer added in the foreground, and 150 kg N/ha added to plots in the back.

Conclusions

Under conditions of zero N fertilizer added, no differences were found among selection regimes in terms of yield, biomass, NUE or straw N content. Even LN (selection always conducted at low N) failed to result in lines with superior performance when tested in yield trials at low N.

Under conditions of intermediate and high levels of N, selection regime AHN (alternate selection, i.e., high N in the F2, low N in the F3, etc.) gave the highest yields, and ALN (alternate selection, i.e., zero N in the F2, high N in the F3, etc.) gave the lowest. The yield advantage of AHN over ALN was 7%.

Observed differences in yield between selection regimes were driven by biomass, not harvest index. UPE and NUE increased concurrently as yields increased with AHN.

 

© CIMMYT July 2001

 Kronstad Symposium Poster List | Wheat Program | Wheat Research Results