J. agric. Sci., Camb.

(1980), 94, 675-689

With 5 text-figures
Printed in Great Britain

675

Genetic improvements in winter wheat yields since 1900 and

associated physiological changes
BY R. B. AUSTIN, J. BINGHAM, R. D. BLACKWELL, L. T. EVANS*, M. A. FORD,
C. L. MORGAN AND M. TAYLOR
Plant Breeding Institute, Trumpington, Cambridge
(Revised MS. received 6 December 1979)
SUMMARY
Experiments were carried out to assess the increase in yield potential of winter
wheat in the U.K. due to variety improvement since the early years of this century.
The effects of other genetic changes were minimized by applying fungicide to control
eyespot and foliar diseases, and by using nets to prevent lodging. The experiments were
carried out in 1978 at Cambridge. One, on soil of high fertility in Camp Field, received
104 kg N/ha and the other, on soil of lower fertility in Paternoster Field, received
38 kg N/ha. Twelve genotypes were tested. Eight were varieties which formed a
chronological series beginning with Little Joss, introduced in 1908. The remaining
genotypes were recently developed selections from the Plant Breeding Institute and
a line bred by the French breeders, Benoist.
The average yield of the 12 varieties and lines tested was 3-96 t/ha in Paternoster
Field and 6-40 t/ha in Camp Field. In both fields the two highest yielding entries,
Hobbit and the advanced breeding line 989/10, outyielded Little Joss by close to 40 %.
Benoist 10483 was the only entry for which the percentage yield advantage depended
on high soil fertility.
The newer, high yielding, varieties were shorter and reached anthesis earlier than
the older varieties. They had lower stem weights/m2 than the older varieties but similar
maximum leaf area indices and leaf weights/ma. Within each experiment the total
dry-matter production of the varieties was similar, the increase in grain yield due to
variety improvement being associated mainly with greater harvest index (ratio of
grain yield to grain + straw yield).
It is argued that by a continuation of the trend towards reduced stem length, with
no change in above-ground biomass, breeders may be able to increase harvest index,
from the present value of about 50% to about 60%, achieving a genetic gain in yield
of some 25%. As the limit to harvest index is approached, genetic gain in yield will
depend on detecting and exploiting genetic variation in biomass production.
improvement in the timeliness of cultivation operaturns, and an increase in the use of herbicides and
The yield of wheat in the United Kingdom has crop protection chemicals, all of which have been
increased by an average of almost 70 kg/ha/year claimed to increase yield. On the other hand the
over the last 30 years (Austin, 1978) or by about increase in the intensity of cereal cultivation and
80%. During these years a succession of new its extension to less suitable soils and climatic
varieties has been introduced, each new one regions will have tended to reduce national
outyielding its predecessor. These new varieties, average yields.
having shorter and stiffer straw, have permitted
Elliott (1962) and Silvey (1979) have summarized
the use of increased amounts of nitrogen fertilizer, data from variety trials carried out by the National
rates of application having more than doubled Institute of Agricultural Botany showing the perover the period. In addition there has been an centages by which successively introduced new
Present address: Division of Plant Industry, varieties of winter wheat have outyielded the ones
CSIRO, P.O. Box 1600, Canberra City, A.C.T. 2601, they replaced. From their figures, Maris Huntsman
Australia.
would be expected to yield 47 % more than Hold0021-8596/80/2828-3700 $01.00 1980 Cambridge University Press

676

R. B. AUSTIN AND OTHERS

out to measure the component of the increase in
yield due to variety improvement since the introduction of Little Joss in 1908 and associated physiological changes. Lodging was prevented and
chemicals were applied to prevent severe infections
of disease, both of which would have complicated
the interpretation of the results and probably
exaggerated differences in yield. To ascertain
whether the yield advantages of the new varieties
depended on high soil fertility, one experiment was
carried out on a low fertility site and the other on
a high fertility site.

fast, one of the most widely grown varieties in

1947. In these trials lodging and disease are not
controlled, and varieties being outclassed are
likely to be more susceptible than those which
replace them. Therefore the yield improvements
calculated from variety trial data include a component due to decreased susceptibility to lodging
and disease. Additionally, as the new varieties
have been selected on soils of progressively increasing fertility, their increased yield may depend on
high soil fertility.
No direct comparisons of the yields of a series
of old and modern varieties of winter wheat appear
to have been made in England since the introduction of effective cereal fungicides. In Holland, Van
Dobben (1962) compared old Dutch varieties with
some modern ones then in use and noted that the
modern ones yielded more grain, less straw but a
similar amount of total above-ground dry matter.
This trend was confirmed by results of Watson,
Thome & French (1963) who compared the old
variety Squareheads Master with Cappelle-Desprez.
More recently, similar results have been reported
by Jain & Kulshrestha (1976), Syme (1970) and
others, and these studies have included varieties
having major dwarfing genes. In none of these
recent studies was it ascertained whether the
changes in height and dry matter distribution were
the only differences consistently associated with
increased yield.
The two experiments reported here were carried

MATERIALS AND METHODS

Both experiments were carried out at Trumping ton in the growing season 1977-8. The low fertility
site, in Paternoster Field, was on a sandy loam
overlying a gravel terrace. It had been under grass
since 1970; during this time it had not received
nitrogenous fertilizer and the grass had been cut
but not removed. The high fertility site was in
Camp Field where the soil was a clay loam overlying a chalky marl. It had been cropped with
cereals and field beans and had received fertilizer
in accordance with modern farm practice.
Each experiment comprised four randomized
blocks of 12 plots of varieties and advanced breeding lines (Table 1). Each plot consisted of three
adjacent subplots, referred to as A, B and C, B
being the centre one. Each subplot, 4-2 m long,

Table 1. Yields, heightsand harvest indices of varieties ir,t, two fields

Yield of dry grain
, moisture)
(t/ha)

Harvest

Height to
base of ear
(cm)

index
(%) i

1
2
3
4
5
6
7
8
9
10
11
12

Variety*
Little Joss
Holdfastf
Cappelle-Desprez
Maris Widgeonf
Maris Huntsman
HobbitJ
MardlerJ
370/500J
730/3637$
989/10J
Armada
Benoist 10483
S.E.

Mean

Camp

Paternoster

Field

Field

Camp
Field

]Paternoster
Field

Camp
Field

5-22
4-96
5-86
5-68
6-54
7-30
6-21
6-34
7-07
7-57
6-86
7-23

112
98
87
96
86
64
57
61
60
64
76
67

142
126
110
127
106
80
77
78
78
84
97
87

34
34
42
39
46
50
47
49
50
49
45
45

36

011

013

0-9

1-3

0-5

0-9

3-96

640

77

99

44

44

Year of
Paternoster
Field
introduction
330
1908
2-91
1935
3-74
1953
3-57
1964
4-04
1972
4-63
1977
4'07
1978

4-25

4-35

4-59
1978
4-24
3-80

36
42

39
46
48
49
48
48

51
43
48

* The numbers 370/500, etc., refer to advanced breeding lines from the winter wheat breeding programme of
the Plant Breeding Institute. The numbers 1-12 are used to identify the varieties in Figs 4 and 5.
f These varieties are of breadmaking quality,
j Homozygous for the major dwarfing gene Rht2.

Genetic improvements in winter wheat

677

contained seven rows drilled 16-8 cm apart. The

distance between the centre rows of adjacent subplots was 1-5 m, and there was a path approximately 1 m wide separating the ends of the plots.
Both experiments were surrounded by at least one
line of plots of winter wheat.
Seed dressed with a benomyl/thiram mixture
was sown on 25 October 1977 at the rate of 160 kg/
ha. In Paternoster Field fertilizer was applied
before sowing to give 50 kg K 2 O/ha and 50 kg
P2O6/ha. No fertilizer was applied before sowing
in Camp Field. The experiments were sprayed
with methabenzthiazuron before emergence to
control weeds, and with benomyl on 28 April,
triadimefon on 12 May and benodanil on 17 May
to control eyespot and foliar diseases. All chemicals
were applied at the manufacturers' recommended
rates. Nitrogenous fertilizer was applied on 1 April
at 38 kg N/ha in Paternoster Field and at 104 kg
N/ha in Camp Field. To prevent lodging in Camp
Field, coarse netting (10 x 10 cm mesh) was placed
over the plots in May and held in position with
stakes. The shoots readily grew through the netting,
which did not appear to have any adverse effect
on growth or yield. At the beginning of June,
following a period of high evapotranspiration the
plants in Paternoster Field 'showed signs of water
stress, and the experiment was irrigated to restore
the soil to field capacity.

data on the growth of the grains in these positions.

For the harvests 3 weeks after anthesis to maturity,
the remaining grains were threshed out of the 10ear samples and the total grain dry weight per
sample determined.
At maturity, two further 10 cm strips of plants
were harvested from the C subplots, and leaf,
stem and sheath, chaff and grain dry weights
determined. Further strips of row were removed
from some subplots to enable edge and neighbour
effects to be estimated (Austin & Blackwell, 1980).
The remaining plants were harvested with a plot
combine. The total yield of dry grain (0 % moisture
content) was determined for each subplot. Elsewhere (Austin & Blackwell, 1980) it has been shown
that yields calculated on the basis of the harvested
area (4*2 x 1-18 m) overestimate the true yield by
24 %. Therefore the yields given in this paper have
been reduced by 24% to correct for edge effects.
The plant material harvested on 22 and 23 May
and at anthesis was milled and analysed for nitrogen by a Kjeldahl procedure. Milled samples of
the grain and of the straw from the samples taken
at maturity were also analysed. From these data,
and from the dry weights at the times of sampling
the amount of nitrogen present in the crops on these
occasions was calculated.

Sampling and harvesting

Plant establishment, numbers of shoots and ears

In each subplot the number of plants in one

50 cm length of row was recorded on 1 February.
On 7 March the total number of shoots in a marked
50 cm length of row in each A and B subplot were
counted. Further counts of the shoots in these
lengths were made at fortnightly intervals until 27
July.
On 22 and 23 May when the penultimate leaves
were fully expanded the shoots in two 10 cm wide
strips across the centre five rows of each C subplot
were cut at ground level. Previous experience
(Austin et al. 1976; Austin & Jones, 1975) indicated
that leaf area index (LAI) was likely to be at or
close to its maximum at this stage. The area and
dry weight of the leaf laminae, and the dry weights
of the stems and sheaths, were determined. Subsequently the plots were inspected each day and the
dates when 50% of the shoots had reached ear
emergence and anthesis were recorded. At anthesis,
and at weekly intervals thereafter until maturity,
samples of ten ear-bearing shoots were taken at
random from each C subplot. From these samples,
leaf lamina area and dry weight and the dry weights
of the stems and sheaths and of the ears were
determined. The grains from florets 1, 2 and 3
(1 = basal) of two central spikelets of each ear
were dissected out, dried and weighed, providing

The average number of plants/m2 determined

from the counts in February was 272. In both
fields, plant densities of nine of the varieties varied
little from this average, though the densities of
Mardler and Holdfast were 310 and 300 and of
Hobbit, 230. In neither field, however, was the
number of ears/m2 correlated with plant density.
Numbers of shoots, counted on the 2 x 50 cm row
lengths, were plotted against time to provide
estimates for each variety in each field of the
maximum shoot number. Shoot survival was calculated as the number of fertile ears expressed as
a percentage of the maximum number of shoots
(Table 2). Number of ears/m2 was estimated by
dividing the corrected grain yield per plot by the
mean grain weight per ear determined at harvest
in August. The accuracy of these calculated values
was verified from counts of the stubble at harvest,
but the latter data are not presented because they
were based on only a small subsample of the harvested area. Maximum number of shoots/m2 was
calculated from the number of ears/m2 at maturity
(Table 4) and the percentage survival of shoots
(Table 2). As expected, fewer shoots were produced
in Paternoster Field than in Camp Field, maximum
numbers being 701 and 937/m2 respectively. In
both fields, Little Joss and Holdfast produced

RESULTS

678

R. B. AUSTIN AND OTHEES

Table 2. Maximum

numbers of shoots, number of ears and percentage survival of

shoots of varieties in two fields

Maximum number
of
shoots/m2
Variety
Little Joss
Holdfast
Cappelle -Desprez
Maris Widgeon
Maris Huntsman
Hobbit
Mardler
370/500
730/3637
989/10
Armada
Benoist 10483
S.E.

Mean

Paternoster
Field
870
865
664
637
673
598
700
741
584
745
558

Camp
Field
1114
1200

50

925
922
947
794
821
837
922
813
753
1172
35

701

937

773

more shoots/ma than all the varieties except

Benoist 10483. Benoist 10483 produced a similar
number of shoots to Little Joss and Holdfast in
Camp Field, but only a little more than the
average number in Paternoster Field. A smaller
percentage of the shoots survived in Paternoster
than in Camp Field (Table 2), but there was a
strong correlation between the shoot survival of
the varieties in the two fields. In both fields a
much greater percentage of shoots survived in
Armada and in Benoist 10483 than average, while
a lower percentage of those produced by Little
Joss survived.
Numbers of ears/m2 were considerably greater
in Camp than in Paternoster Field. Benoist 10483
produced many more ears than average in both
fields and Armada produced more than average
numbers in Camp Field. With these exceptions
there was no consistent trend in number of ears/m2
with year of introduction.
Dates of anthesis and maturity

Defining anthesis as the date when anthers

were visible on an estimated 50% of the ears,
Little Joss was the latest variety in both fields,
and Benoist 10483 the earliest. For the older
varieties, for Huntsman and for 730/3637 anthesis
occurred 5-7 days later in Camp than in Paternoster Field. For all the other varieties the difference was only 3 days (Table 3). Although dates of
maturity could not be denned with precision,
judged by the date when no further green leaf was
present, all varieties were earlier in Paternoster
than in Camp Field. Benoist 10483 was earlier and

No. of ears/m 2

Survival of
shoots (%)

Paternoster
Field

Camp
Field

Paternoster
Field

Camp
Field

296
346
279
274
249
257
308
289
257
283
296

366
468
435
415
379
429
386
385
424
366
535

34
40
42
43
37
43
44
39

32
39
47
45
40
54
47
46
46
45
71

379

715

49

61

15

22

292

442

42

48

44

38
53

Little Joss and Holdfast were later than the other

varieties in both fields. The plots were combineharvested on 8 August in Paternoster Field and on
15 August in Camp Field. For practical reasons all
plots in an experiment had to be harvested on the
same date.
Growth of vegetative organs and grain

Data are presented graphically for only four

varieties: Little Joss, Huntsman, 989/10 and
Benoist 10483. These genotypes represent the
range observed in pattern of growth.
On 22 and 23 May when leaf area indices (LAI)
were likely to have been close to their maximum,
there were significant differences between varieties
in LAI (Table 3), though no consistent trend with
date of introduction, or correlation with yield.
The two highest-yielding varieties, Hobbit and
989/10, had lower than average LAIs in both
fields. On 22-23 May the mean LAI in Paternoster
Field was less than half that in Camp Field, whereas maximum number of shoots/m2 was only 25 %
less (Table 2).
The changes with time in LAI between anthesis
and maturity are shown in Fig. 1 (a) and (6). From
Table 3 and Fig. 1 it can be seen that there was a
considerable decrease in LAI between 22-23 May
and anthesis, this being most marked for Little
Joss, the latest variety, in Camp Field. In Paternoster Field there was little difference between
varieties in LAI (Fig. la). Mainly for this reason
there was little variation in leaf area duration
(LAD, the integral of leaf area index between
anthesis and maturity with respect to time), much

Genetic improvemevts in winter wheat

679

Table 3. Leaf area indices on 22-23 May, dates of anthesis and loss of dry matter from
vegetative organs between anthesis and maturity for 12 varieties in two fields

Leaf area index

on

22-22(May
Variety
Little Joss
Holdfast
Cappolle-Desprez
Maris Widgeon
Maria Huntsman
Hobbit
Mardler
370/500
730/3637
989/10
Armada
Benoist 10483

Paternoster
Field
3-55
2-67

Camp
Field
6-90

307
310
313
304

7-37

664
810
818

3-32
3-65

6-27
7-64
8-84

307

814

2-82

5-88
7-58
7-46
0-68
7-42

312
314

S.E.

0-20

Mean

314

of the variation being associated with later anthesis

in the older varieties, and no correlation between
LAD and grain yield. In Camp Field on the other
hand, in spite of the range of 13 days between
varieties in the date of anthesis, the decline to
low LAI occurred on about the same date in all
varieties. Consequently, in this field, the earlier
anthesis in the newer varieties was associated with
greater LAD, greater post-anthesis increase in crop
dry weight and greater grain yield. LAD varied by
2-4-fold, and grain yield was strongly positively
correlated with LAD (r = +0-79). (Unless otherwise stated correlation coefficients given in this
paper have been calculated using variety means,
and have 10 degrees of freedom.)
The total dry matter in the above-ground organs
increased after anthesis in Paternoster Field until
mid July and in Camp Field until late July. There
were no marked differences between varieties in
the rates of dry-matter accumulation, variation
in crop weight on a given occasion during grain
filling reflecting differences in crop weight at anthesis. The weights of all above-ground vegetative
organs continued to increase for at least 3 weeks
after anthesis, weights declining thereafter. The
maximum weights of the vegetative organs of the
four oldest varieties; Little Joss, Holdfast,
Cappelle-Desprez and Maris Widgeon, were considerably greater than those of four of the modern
varieties (Hobbit, 370/500, 730/3637 and 989/10)
having the major dwarfing gene, Rht2 (Gale, 1979).

Date of anthesis
(date in June)
Paternoster Camp
Field
Field
15
12
12
12
11
9
9
10
11
10
9
4

20
19
18
17
16
12
12
13
17
13
12
7

10-3

14-7

Loss of dry matter

from vegetative
organs between
anthesis and
maturity (g/ma)
Paternoster
Field

Camp
Field

146
59
91
79
156
160
235
136
168
173
58
143

109
83
172
156
200
158
137
140
253
203
166
227

30

46

134

167

This difference was due mainly to the greater stem

and leaf sheath weight of the older varieties: 641
and 848 g/m2 in Paternoster Field and Camp Field
respectively compared with that of the four semidwarf ones, 510 and 668 g/m2 (S.E. 14 and 19).
Leaf weights were almost identical for the two
types.
In Paternoster Field the vegetative organs of
these four semi-dwarf varieties gained less in dry
weight after anthesis and maximum weights were
attained earlier than in the four older varieties
(Fig. 2 a). No such differences were evident for the
same sets of varieties in Camp Field (Fig. 26). The
mean loss in weight from the vegetative organs
between the time of attainment of their maximum
weight and maturity was the same for both sets of
varieties (Table 4). Because the maximum weights
of the vegetative organs were greater in Camp than
in Paternoster Field these losses in weight expressed
as a percentage of the maximum weights were 29
and 33 % respectively. There were only minor but
consistent differences between the four modern and
the four older varieties in this respect (Table 4).
Loss in weight from the leaf laminae may partly
have resulted from leaf-fall, though it has been
argued (Austin et al. 1977) that a substantial part
of the loss in leaf weight over this period is likely
to be due to the export of nitrogenous substances.
In all varieties there was a net loss in dry weight
from the vegetative organs between anthesis and
maturity (Table 3). Generally, the loss was smaller

680

E. B. AUSTIN AND OTHERS

6. viii
7. vii
17. vii
27. vii
Date
Fig. 1. Time course of changes in crop components in two fields, (a) and (6): leaf area index; (c) and
(d): grain dry weight. Lines represent four varieties: Little Joss (), Maris Huntsman (), 989/10 (A),
and Benoist 10483 ( ) . The first point on each line is the value of the given component at anthesis. (a)
and (c) Paternoster Field, (6) and (d) Camp Field.
7.vi

in Paternoster than in Camp Field, and Holdfast

lost the least dry weight in both fields. In Camp
Field grain yield was negatively correlated with
this loss in dry weight (r = 0-77). In Paternoster
Field the correlation (r = 0-50) was not significant.
Growth in grain weight/m 2 (Fig. l c and d)
followed the usual pattern and was essentially linear

between 20 and 45 days after anthesis, the mean

rates being 11-5 and 17-5g/m a /day in Paternoster
and Camp Fields (S.E. 0-3 and 0-5 respectively).
(The grain and crop growth rates given in this
paragraph are the means for all varieties except the
early maturing Benoist 10483. Mean rates were
calculated from joint regression analysis.) In both
fields three varieties, Huntsman, 730/3637 and

681

Genetic improvements in winter wheat

9. vi

19. vi

29. vi

9. vii
Date

19. vii

29. vii

800

7. vi

17. vi

27. vi

7. vii

17. vii

27. vii

6. viii

16. viii

Fig. l(c) and l(d). For legend see opposite.

989/10, had the highest rates, differences between
the remaining varieties being smaller and inconsistent between fields. In Camp Field, rates for
Huntsman, 730/3637 and 989/10 were 19-7, 19-8
and 22-1 g/m a /day respectively (S.B. 1-56). Crop
growth rates between 20 and 45 days after
an thesis were much lower than grain growth rates

over the same period. Varietal differences in crop

growth rates were not significant in either field,
the mean rates being 5-5 g/m a /day in Paternoster
Field and 9-6 g/m 2 /day in Camp Field (S.E. 0-6
and 1-0 respectively). Samples were not taken
frequently enough to be able to determine accurately dates of cessation of grain dry-matter

682

R. B. AUSTIN AND OTHERS

600

400 -

200 -

-200

10

20

30
40
Days after anthesis

50

60

70

30
40
Days after anthesis

50

60

70

-200

Fig. 2. Changes in dry weight/m2 of grain (

) and straw (
), i.e. stems, leaves and chaff, after
anthesis in two fields: (a) Paternoster Field and (6) Camp Field. , Mean of the four oldest varieties:
Little Joss, Holdfast, Cappelle-Desprez and Widgeon; %, mean of four semi-dwarf varieties, Hobbit,
370/500, 730/3637 and 989/10.

683

Genetic improvements in winter wheat

Table 4. Less in dry weight of various organs from the four oldest and from four newer, semi-dwarf varieties,
between the time of attainment of maximum weight of the vegetative organs and maturity
Paternoster Field
All

vegetative
organs
Oldest varieties
Semi-dwarf varieties
S.E.

Camp Field

Stems

All

Ear
Leaf
vegetative
sheaths laminae structures organs
Loss in dry weight (g/m2
and

266
275

227
220

25
35

23

19

14
20
2

Stems
and

sheaths

Ear
Leaf
laminae structures

314
344

227
220

56
71

53
55

30

18
16

17
19

Loss in dry weight as a percentage of the' total loss from the vegetative organs
Oldest varieties
Semi-dwarf varieties

85
81

9
9

5
10

65
65

Loss in dry weight of the organs as a percentage of their weight at the time of maximum vegetative weight
Oldest varieties
32
36
25
14
26
24
30
33
Semi-dwarf varieties
35
41
22
23
32
33
34
28

accumulation, though it appeared that these reflected dates of anthesis, the earlier varieties
ceasing grain growth sooner than the later ones.
Growth curves of the grains in the first, second
and third florets of the central spikelets of the ears
showed that in these grains there were varietal
differences in growth rates in both fields (see Fig.
3 for examples). The grains in florets 1 and 2 had
the fastest average growth rate. The mean rates
of growth of the grains in the three florets were
similar in the two fields. In both fields, averaging
over florets, Maris Huntsman and 989/10 had the
fastest rates, and Holdfast, Mardler and Benoist
10483 the slowest rates. When the mean rates of
growth of the grains in florets 13 of the central
spikelets were multiplied by the number of grains/
ma, the resulting rates of growth in g/m 2 /day were
strongly correlated with the rates estimated from
whole-ear data. Although the sampling frequency
was insufficient to quantify varietal differences in
the duration of grain growth, the curves given in
Fig. 3, as well as those for the other varieties,
indicate that the duration was less in Little Joss
and Holdfast than in most modern varieties.

varieties in Paternoster Field, the lower fertility,

lower-yielding site, than in Camp Field (Fig. 4).
The greater yields of the modern varieties were
not associated with changes in any one yield
component (Table 5). Thus, 989/10 produced no
more fertile ears/m2 than Little Joss, and its yield
advantage over the latter depended on its having
heavier and more grains per ear. This held true for
both fields. In Paternoster Field, Armada had a
similar number of ears/m2 to Little Joss, and its
yield advantage depended on the production of
more and heavier grains per ear. In Camp Field
however, Armada had more ears/m2 but fewer and
lighter grains. Benoist 10483 produced many more
ears/m2 than Little Joss in both fields, but was a
small-eared variety with grains of similar weight
to those of Little Joss. Its yield advantage over
Little Joss (15% and 38% in Paternoster and
Camp Field respectively) appeared to depend on
its ability to produce more fertile ears/m2, and
this was expressed most strongly in Camp Field.
The total number of spikelets per ear was less
in Paternoster than in Camp Field in all varieties
(19-0 as compared with 20-2). Benoist had the
fewest, 730/3637 the most and there was no clear
Grain yield, yield components, height and
trend with year of introduction, Little Joss being
harvest index
similar to 989/10 and Armada. Number of grains
Holdfast, a wheat of breadmaking quality, gave per spikelet in the central spikelets was greater in
the lowest yield in both fields, the yields of Little the varieties having the RM2 gene (3-18) than in the
Joss being somewhat greater. The highest-yielding four old varieties lacking it (2-50). Number of grains
varieties were 989/10 and Hobbit (Table 1). per central spikelet was slightly lower in Paternoster
Taking Little Joss as the standard against which than in Camp Field (2-77 and 2-95 respectively).
to measure the yield improvement achieved by
For all varieties, mature plant height to the
breeding for similar use, the mean yields of these base of the ear was 23-35 % greater in Camp than
two highest yielding varieties were 139 and 142% in Paternoster Field, the average increase being
in Paternoster and Camp Fields, respectively. 22 cm or 29 % of the height in Paternoster Field
Benoist 10483 yielded relatively less than the other (Table 1). Weight per stem, both maximum and

684

R. B. AUSTIN AND OTHERS

10

20

30
40
Days after anthesis

10

20

30
40
Days after anthesis

SO

60

70

60

70

Fig. 3. Time course of changes in mean grain dry weight in floret 1 for four varieties: Little Joss (#),
Maris Huntsman (), 989/10 (A) and Benoist 10483 (), in two fields: (a) Paternoster Field, (6) Camp
Field, S.E. of a value is 6%.

final, bore a reasonably close, though not proportional, relation to stem height. Stem weight per
unit height was greater in Paternoster than in
Camp Field. Reduction in stem height by one half,
the varietal range, was accompanied by a reduction
in stem weight by about one third. The loss in
stem weight from the time of maximum vegetative
dry weight bore no relation with height. Harvest
index (the ratio of grain yield to grain + straw
yield at maturity) varied from 34-36 % for Little
Joss and Holdfast to 48-51% for 989/10 and

Hobbit. Harvest index was strongly negatively

correlated with plant height, in both fields.
Nitrogen offtake
The amount of nitrogen in the entire aboveground biomass changed little between 22-23 May
and maturity. As expected, the nitrogen concentrations in the various organs were greater in Camp
than in Paternoster Field and this resulted in the
nitrogen offtake by the crop being 2-76 times

Genetic improvements in winter wheat

685

500 -

500
300
350
400
450
Grain yield in Paternoster Field (g/m2)
Fig. 4. Yields of varieties in Camp Field plotted against those in Paternoster Field. The numbers
refer to those alongside the variety names in Table 1. iji, S.E. of a value.

Table 5. Characteristics of theears of varieties in two fields

Variety
Little Joss
Holdfast
Cappelle-Desprez
Maris Widgeon
Maris Huntsman
Hobbit
Mardler
370/500
730/3637
989/10
Armada
Benoist 10483
S.E.

Mean

Paternoster
Field
370

31-9
48-1
45-7
461
461
41-8
43-6
394
48<3
40-3
37-7
0-7
42-2*

No. of grains/spikelet
in central spikelets

No. o)
grains/ear

Mean grain
weight (mg)
Camp
Field
38-8
32-6
44-8
451
47'7
36-9
38-2
43-2
41-9
47-5
37-8
364
11

40-9

greater in this field than in Paternoster Field. As

commonly found (e.g. Pushman & Bingham, 1976),
the higher-yielding varieties had lower grain nitrogen percentages (the correlation between grain
yield and percentage nitrogen was 0-80 in
Paternoster and 0-72 in Camp Field). However,
since the relative variation in yield was greater
than that in grain nitrogen percentage, the total

Paternoster
Field

Camp
Field

Paternoster
Field

30 1
26-4
27-9
28-5
35-2
391
31-6
33-8
42-9
33-6
35-5
26-6
0-6
32-7

36-8

2-31
2-35
2-39
2-52
305
307
2-85
305
3-26
301
2-67
2-69

Camp
Field
2-77
2-59
2-50
2-61
304
3-59
3-36
316
291
3-52
2-66
2-72

009
2-77

008
2-95

32-5
301

30-3
36-2
461
421
381
398

43 6
339

27-8
0-9

36-5

amount of nitrogen in the grain was greater in the

newer, higher-yielding varieties than in the old
ones, in both fields (Table 6). The amount of nitrogen in the grain at maturity as a percentage of
that in the entire above-ground biomass, the nitrogen harvest index, was greater for the newer than
for the older varieties in Paternoster, but not in
Camp Field.

686

R. B. AUSTIN AND OTHERS

Table 6. Nitrogen present in the crop at maturity, nitrogen harvest, index
and grain nitrogen concentration
N in crop
(g/m2)
Paternoster
Field

Camp
Field

Paternoster
Field

6-4
60
6-8
6-5
6-8
71
6-6

74
75
77
78
79
79
78
82

71

82
78

71
72

7-2

161
190
18-3
18-3
180
19-8
171
171
200
20-1
20-4
21-5

Field
71
69
68
68
72
72
71

S.E.

0-27

Mean

6-8

Variety
Little Joss
Holdfast
Cappelle-Desprez
Maris Widgeon
Maris Huntsman
Hobbit
Mardler
370/500
730/3637
989/10

7-2

6-7
70
7-5

Armada
Benoist 10483

<% N i n
grain

Nitrogen
harvest index (%)
Camp

Paternoster
Field

Camp
Field

1-44
1-54
1-40
1-42
1-32
1-21
1-27
1-39
1-26
119
1-44
1-52

2-20
2-64
213
2-18
1-98
1-95
1-97
1-92

200
1-90
201

81

67

063

80

73

005

006

18-8

78

71

136

2-09

2-18

years when increased amounts of nitrogen fertilizer

were being used. As shown in Fig. 4, excluding
Major differences between varieties in their Benoist 10483, the correlation between the yields
yields were found in both experiments, reflecting of the varieties in the two experiments was high
the physiological component of the improvements (r = +0-97, 9 D.P.), indicating that the yield imachieved by breeding since the early years of this provements shown by the new varieties, expressed
century. Thus the experiments lend support to as percentages of the yield of Little Joss, the oldest
Silvey's (1979) conclusion that variety improve- variety, were the same in both fields, i.e. did not
ment alone has been responsible for an increase in depend on fertility in these circumstances. This
national wheat yields of some 40 % over the period finding is common in variety trials when there are
1947-77. But, in comparing our results with no serious limitations due to variable incidence of
Silvey's calculations, it must be emphasized that disease. Indeed, the system of variety production
our experiments, in which diseases were controlled and evaluation used in England and Wales is likely
and which were done at one centre in one season, to give new varieties which outyield those they
cannot be assumed to measure the average im- replace in all or virtually all situations. If potential
provement that would be found over many sites new varieties do not have this attribute, i.e. a lack
and years. For instance, in. our experiments Hobbit of genotype x environment interaction, their varioutyielded Maris Huntsman by 15 % in Paternoster ability of performance relative to a standard
and by 12% in Camp Field (Table 1) as compared variety will generally cause them to be rejected.
with 2% in national trials (Anon. 1979). Hobbit is
Had the plots not been supported to prevent
known to be more susceptible to mildew (Erisyphe lodging, the yields of the old, tall varieties are
graminis D.C.) than Huntsman and its yield ad- likely to have been considerably lower, increasing
vantage over Huntsman is fully expressed only in the differences in yield between the old and new
the absence of this disease. A further difference ones. Thus, as is well known, with the shorter
between our results and those from national trials varieties now in cultivation, it is possible to use
was that Holdfast yielded 12% and 5% less than larger amounts of nitrogen fertilizer and hence
Little Joss in Paternoster and Camp Fields whereas obtain increases in yield that would not have been
in national trials carried out about 30 years ago it practicable with the old, tall varieties. However,
yielded 28% more (Elliott, 1962). This difference our results show that the yield improvements
may reflect the greater susceptibility of Little Joss achieved by breeding new varieties have resulted
to lodging which, despite its slightly lower physio- from physiological and morphological changes in
logical potential, may have given Holdfast a sub- addition to those associated with increased lodging
stantial advantage over Little Joss in the post-war resistance. These improvements were expressed at
DISCUSSION

687

Genetic improvements in winter wheat

800 700 -

".

7 "

4 600 -

n 12

11

8
3

x:
eo

'\"4

'" 1

I 500 -

10

ft'

I 400.

2
-i-H

300
1

200 -

400

500

600
700
800
Straw dry weight (g/m2)

900

1000

Fig. 5. Grain dry weight plotted against straw dry weight at maturity for Camp Field ( ) and Paternoster
Field (#). The lines show a slope of 1-0. The numbers refer to those alongside the variety names in
Table 1. | - i J-l, S.E. of a value.

both low and high fertility in our experiments, as

is often found in trials with varieties carried out in
different years and sites.
As found in other studies (e.g. Van Dobben,
1962) our experiments also show that in a given
environment shorter varieties had a greater harvest
index and grain yield than the older, taller ones
(Table 1). Fig. 5 shows that the straw yield was
negatively correlated with grain yield, the regression coefficients being 0-69 (S.E. 0-16) in
Paternoster and 0-44 (S.E. 0-26) in Camp
Field. The poor correlation in Camp Field was
partly due to Mardler, the grain yield of which was
lower than expected on the basis of results from
other trials, and to 370/500 and Armada. The
regression coefficient for the Camp Field data,
omitting these varieties, was 1-28 (S.E. 0'25).
None of these coefficients was significantly different
from 1-0. Thus in each field there was a ceiling
to total dry-matter production (8-97 and 14-40 t/ha
in Paternoster and Camp Field respectively) with
little variation among varieties. Within an experiment, variation in grain yield was strongly associated with changed distribution of dry matter, in
particular with reduced stem length and weight/m2.
However, there were other differences between
high- and low-yielding varieties and no single
attribute or yield component was uniquely associated with high yield.
Benoist 10483 was the exception to the generalization that variety yields in Camp Field were
strongly correlated with those in Paternoster Field.

Benoist 10483 was the only variety or advanced

breeding line tested which had not been selected
for high yield in the U.K. and it was included
mainly because it was known to be freely tillering.
It had the highest number of ears in both fields,
partly because it produced more shoots, and partly
because more of these shoots survived than in
most other varieties. Armada also had a high
percentage shoot survival, and this variety and
Benoist 10483 were clearly different in this respect
from the varieties and breeding lines originating
from the Plant Breeding Institute which were used
in these experiments.
Although the straw weight at maturity of modern
varieties was less than that of the older varieties
(Fig. 5), the stems and leaf sheaths at this stage comprised 30-35 % of the straw weight in the modern,
but 45-50% of it in the older varieties. As a result,
at maturity the leaf weights of the two lowest
yielding varieties, Little Joss and Holdfast, were
very little different from those of the two highest
yielding ones, Hobbit and 989/10. The same held
true for leaf area index on 22-23 May (Table 3),
and there was no relationship between leaf area
index at this time and grain yield.
Grain growth rates between 20 and 45 days after
anthesis exceeded crop growth rates by an average
of 6-0 g/m 2 /day in Paternoster and 7-9 g/m 2 /day
in Camp Field. This does not necessarily imply that
the requirements for carbon for grain growth could
not be met from current photosynthesis, because
respiratory losses, particularly from the stems, are

688

R. B. AUSTIN AND OTHERS

the older varieties, the nitrogen offtake by the

grain and by the crop was greater for these varieties than for the older ones (Table 6). In Camp
Field this amounted to an average of 25 kg N/ha
for each 1 t/ha increase in grain yield. This figure
provides an indication of the minimum increase
in fertilizer nitrogen application that would be
needed to sustain an increased yield of 1 t/ha in
intensive wheat growing.
Although the genetic gain in grain yield in winter
wheat during this century has been associated with
a corresponding decrease in straw yield, it is clear
that there must be a limit to the improvement in
yield which can be achieved by continuing this
trend. A further reduction in non-grain biomass may
be achievable by selecting still shorter genotypes,
but because of the need for leaves to be arranged
in the canopy to maximize the interception of
light and its uniformity of distribution over their
area, and to provide adequate mechanical support
for the ears, it is unlikely that stem and sheath
dry weights could be reduced much further. The
average stem and sheath dry weight at maturity
of Hobbit, 370/500, 730/3637 and 989/10 in Camp
Field was 453 g/m a . The dry weights of the other
components at harvest were, in g/m2, grain 707,
chaff (rachis, glumes etc.) 139, and leaf laminae 143.
This dry-weight distribution gives a harvest index
of 49%. Assuming constant biomass yield, a reduction in stem and leaf sheath dry weight to half
the observed value, and a pro-rata increase in
chaff weight to accommodate the extra grain, the
organ weights would become: grain 895, chaff 178,
leaf laminae 143, stem and leaf sheath 226. These
values give a harvest index of 62 %, for an increase
in grain yield over that observed for the genotypes
of 26%. It may prove impossible to produce
genotypes having these characteristics: if so breeders are close to exhausting the opportunity for
increasing yield by modifying dry-matter distribution.
This argument suggests that breeders will need
to detect and exploit genetic differences in total
dry-matter production if there is to be a continued
genetic gain in yield. Straw strength and number
of ears will have to be increased, or at least maintained, implying a need to retain a distribution of
dry matter between grain and straw similar to that
in the best of the present varieties. Attempts to
increase dry-matter production by means other
than increasing photosynthetic rate per unit leaf
area may be unproductive because they are likely
to result in an increased water requirement. As
shortage of water frequently limits growth and
yield at present, genotypes with substantially
greater leaf area expansion rates and LAI would
be more prone to drought, particularly during
Despite the generally lower nitrogen concen- grain filling. The obvious alternative, an increase in
tration in the grain of the newer as compared with

known to be high during this period (Rawson &

Evans, 1971; Austin et al. 1977) and they would
reduce net photosynthetic and hence crop growth
rates. However, some of the carbon assimilated
during early grain growth and temporarily stored
as soluble carbohydrate in the stems is subsequently relocated to the grains (Rawson & Evans,
1971; Austin et al. 1977). As the loss in weight
from the vegetative organs between anthesis and
maturity was greater in the high- than in the lowyielding varieties it seems that the margin between the demand for carbon by the grains and
the supply from current assimilation during the
linear phase of grain growth may be smaller in
the newer varieties.
Although grain yield was positively correlated
with leaf area duration in Camp Field, the lack of
such correlation in Paternoster Field suggests that
the relationship in Camp Field was not causal, as
the highest yielding varieties outyielded Little Joss
by a similar percentage in both fields. The greater
LAD of the higher-yielding varieties in Camp Field
was associated to a considerable extent with their
earlier anthesis, since they reached maturity at
the same time as the lower-yielding varieties. As
grain dry matter originates mainly from photosynthesis after anthesis, when leaf area is declining
rapidly, a further increase in post-anthesis photosynthesis may be obtained by selecting genotypes
in which anthesis is closer to the time of maximum
LAI: our results show that there has been a trend
in this direction.
As noted earlier, the yield advantage of the
modern varieties did not depend on a change in
one, or mainly one, yield component. Varieties
possessing the major dwarfing gene Rht2 (Hobbit,
Mardler, 370/500, 730/3637 and 989/10) had, as
a group, more grains per ear but fewer ears/m2
than the other varieties, though their mean grain
weight was close to average. The greater number
of grains per ear was achieved because there were
more grains per spikelet in the Rht2 dwarf varieties
than in the other ones.
These effects of the Rht2 gene are now well
known (Gale, 1979) and may be the consequence
of reduced competition between the stem and the
ear for a limited supply of assimilate, as suggested
by Friend (1965), Rawson & Hofstra (1969),
Bingham (1969) and Patrick (1972). It is evident,
however, from Tables 1, 2 and 5, that there was
variation in yield and its components among the
five varieties having the Rht2 gene. It is not known
whether the negative association between height
and grain yield is a direct result of competition for
assimilates and holds irrespective of which gene
or genes affect stem growth and plant height.

Genetic improvements in winter wheat

the rate of photosynthesis per unit leaf area, may
provide more dry matter for the same, or even
lower, water use, but there is no evidence that
modern varieties of wheat have this characteristic,
as primitive forms have higher photosynthetic
rates, though much smaller leaves (Evans &
Dunstone, 1970). It has yet to be ascertained

689

whether increases in photosynthetic rate can be

achieved without compensating reduction in leaf
size or longevity, and clearly this is a major challenge for wheat breeders and physiologists.
We are grateful to Mrs C. Starr for analysing the
samples of grain and straw for nitrogen.

Note added in proof

The advanced breeding line 989/10 has now been added to the varieties on the National List and
is known as Norman.
REFERENCES
ANON. (1979). Recommended varieties of cereals.
National Institute of Agricultural Botany, Farmers
Leaflet No. 8.
AUSTIN, R. B. (1978). Actual and potential yields of
wheat and barley in the United Kingdom. ADAS
Quarterly Review 29, 76-87.

JAIN, H. K. & KULSHBESTHA, V. P. (1976). Dwarfing

AUSTIN, R. B. & BLACKWELL, R. D. (1980). Edge and

PUSHMAN, F. M. & BINOHAM, J. (1976). The effects of

neighbour effects in cereal yield trials. Journal of

Agricultural Science, Cambridge 94, 731-734.

a granular nitrogen fertilizer and a foliar spray of

urea on the yield and bread-making quality of ten
winter wheats. Journal of Agricultural Science,
Cambridge 87, 281-292.

AUSTIN, R. B., EDBICH, J. A., FORD, M. A. & BLACK-

WELL, R. D. (1977). The fate of the dry matter,

carbohydrates and 14C lost from the leaves and stems
of wheat during grain filling. Annals of Botany 41,
1309-1321.
AUSTIN, R. B., FORD, M. A., EDRICH, J. A. & HOOPER,

B. E. (1976). Some effects of leaf posture on photosynthesis and yield in wheat. Annals of Applied
Biology 83, 425-446.
AUSTIN, R. B. & JONES, H. G. (1975). The physiology

genes and breeding for yield in bread wheat. Zeitschrift fiir Pflanzenziichtung 76, 102-112.
PATRICK, J. W. (1972). Distribution of assimilate
during stem elongation in wheat. Australian Journal
of Biological Sciences 25, 455-467.

RAWSON, H. M. & EVANS, L. T. (1971). The contribu-

tion of stem reserves to grain development in a

range of wheat cultivars of different height. Australian Journal of Agricultural Research 22, 851-863.
RAWSON, H. M. & HOFSTRA, G. (1969). Translocation

and remobilisation of 14C assimilated at different

stages by each leaf of the wheat plant. Australian
Journal of Biological Sciences 22, 321-331.
SILVEY, V. (1979). The contribution of new varieties to
increasing cereal yield in England and Wales. Journal
of the National Institute of Agricultural Botany 14,
367-384.
SYMB, J. R. (1970). A high yielding Mexican semidwarf wheat and the relationship of yield to harvest
index and other varietal characteristics. Australian
Journal of Experimental Agriculture and Animal
Husbandry 10, 350-353.
VAN DOBBEN, W. H. (1962). Influence of temperature
and light conditions on dry matter distribution,
development rate and yield in arable crops. Netherlands Journal of Agricultural Science 10, 377-389.

of wheat. In Report of the Plant Breeding Institute,

Cambridge, for 1974, pp. 20-73.
BINOHAM, J. (1969). Physiological determinants of
grain yield in cereals. Agricultural Progress 44, 30-42.
ELLIOTT, C. S. (1962). The importance of variety testing
in relation to crop production. Journal of the National
Institute of Agricultural Botany 9, 199-206.
EVANS, L. T. & DUNSTONE, R. L. (1970). Some physiological aspects of evolution in wheat. Australian
Journal of Biological Sciences 23, 725741.
FRIEND, D. J. C. (1965). Ear length and spikelet number of wheat grown at different temperatures and
light intensities. Canadian Journal of Botany 43,
345-353.
WATSON, D. J., THORNE, G. N. & FRENCH, S. A. W.
(1963). Analysis of growth and yield of winter and
GALE, M. D. (1979). The effects of Norin 10 dwarfing
spring wheats. Annals of Botany, London 27, 1-22.
genes on yield. In Proceedings of the Fifth International Wheat Genetics Symposium, Delhi, 1978, pp.
978-987.

ACS 94