any mass increase is in the growing grains (Porter et al., 1950; Scott et al., 1983). The head growth rate. (GR, mg TU- ~ ) of each genotype was calculated from.
ELSEVIER
Field Crops Research 37 (1994) 63-75
Field Crops Research
Variation in temperate cereals in rainfed environments II. Phasic development and growth C. L6pez-Castafieda a'b'1, R.A. Richards a,, aCSIRO, Division of Plant Industry and Cooperative Research Centre for Plant Science, PO Box 1600, Canberra, A.C.T. 2601, Australia bResearch School of Biological Sciences, Australian National University, PO Box 475, Canberra, A.C.T. 2601, Australia
(Received 27 July 1993;accepted 9 February 1994)
Abstract Barley yields more grain and total biomass than does triticale which in turn yields more biomass than do bread wheat, durum wheat and oats when sown at the same time in rainfed environments in southem Australia. To determine reasons for these differences, cultivars of each species were grown at five field sites and variation in their phenology and both pre- and postanthesis growth was measured. Barley achieved a higher yield of grain and biomass in a shorter duration than the other species. It reached physiological maturity about 10 days (180 thermal units) before the other species, and reached double ridge and anthesis earlier. Triticale was also earlier to reach double ridge and terminal spikelet than the mean for the other species, although it had a similar physiological maturity to the wheats. Barley and triticaie developed a greater leaf area and dry mass faster than the wheats and oats. The difference in leaf area was established from the time the first leaf had fully expanded. Barley also developed mainstem leaves and tillers faster than the other species whereas triticale was slower in this respect. Crop growth rate was greatest in barley and triticale up to anthesis, but no differences between species were found in their relative growth rates. The growth rate of individual grains and of total grain per unit ground area were substantially greater in barley than the other species. Oats and durum wheat had the slowest individual grain and total grain growth rates. Grain growth rate per unit ground area was significantlyassociated with grain yield at one site where this was examined. The change in stem mass between anthesis and physiological maturity, which was determined to assess the possible contributionof stem reserves to grain, was also positively associated with grain yield at the two sites where it was determined, and more so at the drier site. The change in stem mass averaged 76 g m -2 at the two sites and this represented 25% of the total grain yield. However, the range varied from 13 to 39% of grain yield (corrected for husk mas~ in barley and oats). The loss in leaf sheath mass averaged 68 g m -2 at both sites; this was not associated with grain yield. Key words: Barley; Grain growth; Growth analysis;Oats; Phenology;Stem reserves; Triticale;Wheat
I. Introduction In a comparison of bread wheat, durum wheat, barley, triticale and oats, the grain yield of barley was 25% *Correspondingauthor. ~Present address: Centro de Gen6tica, Colegio de Postgraduados, Montecillo,Mexico, 56230, Mexico. 0378-4290/94/$07.00 © 1994ElsevierScienceB.V. All rights reserved SSDI0378-4290(94)00018-8
higher than those of the other species when averaged over five water-limited and rainfed environments in south-eastern Australia (L6pez-Castafieda and Richards, 1994). The highest-yielding barley had 39% more total grain mass than the highest-yielding wheat. After correcting for the mass of the husk coveting the barley grain, the yield advantage of barley was attributed to a higher above-ground dry mass rather than to
64
C. Lrpez-Castatieda, R.A. Richards / Field Crops Research 37 (1994) 63-75
a higher harvest index. Other differences were that barley reached anthesis earlier and had a larger root mass. Triticale, like barley, had a high above-ground dry mass but its yield was less than that of barley because of its lower harvest index. The better performance of barley over wheat in dry areas has been found in other studies (Siddique et al., 1989, 1990a,b; Gregory et al., 1992; Josephides, 1993) and farmer experience with the different temperate cereals would rank them for yield under drought as barley, complete triticale, durum wheat, bread wheat, substituted triticale and oats (Fischer, 1989). The consistently higher above-ground dry mass achieved by barley and triticale compared to wheat has important implications for wheat improvement. Firstly, it provides a clear demonstration that cereal species which are similar morphologically and are sown and harvested at about the same time, can differ substantially in total biomass and thus in yields. Secondly, it implies that if the characteristics responsible for the greater yield of grain and biomass in barley and triticale can be identified, then improvement in these characteristics in the other species may be possible. To identify the most effective ways to improve the yields of temperate cereals, comparing species may in the end be more valuable than contrasting historical sets of cultivars within species. The principal difference between the aforementioned comparisons of species, and other studies comparing old and new wheat or oat cultivars, is that in the latter, despite large differences in grain yield, there were no differences in aboveground dry mass (Austin et al., 1980; Wych and Stuthman, 1983; Perry and D'Antuono, 1989). The increase in yield in both species occurred because of a greater harvest index. It is interesting to note that in barley, newer varieties have both greater biomass and harvest index than older varieties (Riggs et al., 1981; Wych and Rasmusson, 1983). As we approach the limit to the increase in harvest index, ways to increase biomass genetically must be identified and these may come from comparisons of more divergent germplasm. This paper explores variation between the species in the growth of leaf area as well as dry mass through the season. It also examines variation in the timing of reproductive development as this is a major difference between the species and of vital importance to adaptation in water-limited environments (Richards, 1991 ).
2. Materials and methods Full experimental details were provided in the first paper of this series (L6pez-Castafieda and Richards, 1994). Briefly, five experiments were sown in each of three years (1988, 1989, 1990) in the central (Condobolin, C) and south-eastern (Moombooldool, M) Australian wheatbelt. Experiments are referred to as C88, C89, M88, M89 and M90, which denote the site and year of the experiment. Fourteen cultivars (four bread wheat, two durum wheat, four barley, two triticale and two oats), chosen for their high yield and adaptation to this region as well as variation in flowering time, were evaluated. They were sown at the recommended time in late May or early June at each site after a pasture phase. There were three replications arranged in a neighbour design at each site. Seeding rate was adjusted to give 140 plants m -2 and phosphorus and nitrogen fertiliser was banded with the seed at a rate of 15 kg ha -1 each. Plots were 10 rows wide at both sites with 18 cm between rows, and were 15 m long at Condobolin and 10 m at Moombooldool. Weeds and diseases were controlled by chemicals as required. Monthly weather data were provided in the companion paper ( L6pez-Castafieda and Richards, 1994). Rainfall between April and October ranged from 250 to 370 mm. There was a terminal drought in each experiment.
2.1. Phasic development In C89 and M89, apex development was monitored regularly on ten plants of each cultivar. The times of appearance of the first double ridge (DR) and of terminal spikelet were determined. Terminal spikelet (TS) formation is unambiguous in wheat and triticale. However, for barley the first appearance of awn primordia was taken to indicate the completion of the spikelet initiation phase (Kirby and Appleyard, 1984) whereas stamen differentiation in the earliest formed spikelets was used in oats (Moncur, 1981). Anthesis (A) was recorded as the time when 50% of culms reached anthesis in all experiments. Physiological maturity (PM), recorded as the time when no green parts were present in a plot, was noted in experiments C89, M89 and M90. Intervals between successive phenological stages (DR, TS, A and PM) were calculated in both calendar days and thermal time. Thermal time was calculated in thermal units (TU) using 0°C as the
C. L6pez-Castatieda, R.A. Richards / Field Crops Research 37 (1994) 63-75
base temperature and summing the average daily temperatures at each site. Photothermal time was calculated using the same base temperature but summing the daily product of average temperature during the light period and the light period as a proportion of a day (Masle et al., 1989).
2.2. Crop growth Two quadrat samples (10.17 m 2 each) of aboveground plant parts were taken from each plot at regular intervals in all experiments. The two quadrat samples from each plot were bulked. A subsample was taken, and in the pre-anthesis harvests, separated into leaves and stems (including leaf sheaths). The number of tillers per sample was recorded as well as the leaf number on the main stem. At the first harvest at M89 and C89, leaves from 30 plants were removed and the areas of the first, second and third main stem leaves were determined as well as the total leaf area. When ears or panicles emerged, a subsample of 30 stems was taken from each sample and separated into heads, dead leaves, leaf blades, leaf sheaths and stems. The lengths of the stems were measured and the areas of leaf blades in samples were determined using a Delta-T Devices area measurement system. The ovendry mass of plant parts and of the bulk sample were also determined. Leaf area index (LAI) was computed from the product of the sample specific leaf area (m 2 kg ~) and the total leaf mass per unit ground area. Stem density was calculated from the ratio of stem mass to stem length.
65
CGR=(W2-W1)/(T2-T1),
(2)
where W1 and W2 represent the dry mass (g m -2) at T1 and 72. Between anthesis and physiological maturity at M90, 20 heads were randomly harvested from each plot on a regular basis. These were oven-dried and then weighed. As there is no further increase in chaff mass (i.e. the head or spike without grains) after anthesis, any mass increase is in the growing grains (Porter et al., 1950; Scott et al., 1983). The head growth rate (GR, mg T U - ~) of each genotype was calculated from the slope of the relationship between head dry mass and thermal time in successive harvests during the linear phase of grain growth. The total grain growth rate, on a per unit area basis (mg m - 2 T U - I ), was calculated from the product of head GR and the average number of spikes m -2 in M90, whereas individual grain growth rate ( mg T U - ~) was calculated by dividing head GR by the number of grains sp;ke- ~ at M90 for each cultivar.
3. Results
3.1. Development Thermal time to reach each development stage in each cultivar was very similar at both C89 and M89 (Fig. 1 ). Minor exceptions occurred in that anthesis was somewhat earlier at M89 whereas physiological 2500
2.3. Crop growth indices
0~2000
,.~
:E
Relative growth rate (RGR) is defined as the rate of dry mass accumulation per unit of existing dry mass and crop growth rate (CGR) as the rate of dry mass accumulation per unit ground area (Warren Wilson, 1981 ). These indices were calculated, as suggested by Radford ( 1967 ), by substituting thermal time for calendar time. RGR was calculated from time TI to 72 as:
1500 0
._E "~ E LJ~
I--
1000
500
0
0
R G R = (loge W2-1oge WI)/(T2-T1),
(1)
where logeW1 and logcW2 are the natural log values of dry mass (g m -2) at TI and 72 respectively. CGR was calculated from TI to 72 as:
I
~
i
1
i
500
1000
1500
2000
2500
Thermal time at C89 Fig. 1. Thermal time to reach each developmental stage in each genotype of barley ( I l L bread wheat ( • ), triticale (rq) and oats (e) at M89 and C89. The line shows the 1:1 relationship.
66
c. Ldpez-Casmfieda, R.A. Richards / Field Crops Research 37 (1994) 63-75
maturity was earlier at C89. Also times to reach double ridge and terminal spikelet in triticale were longer at C89 than M89. Variation in reproductive development within species was substantial and tended to mask major differences between species (Table 1 ). Nevertheless, several consistent patterns emerged. Barley reached physiological maturity about 180 TU (ca. 10 days) before bread wheat, durum wheat and triticale and about 90 TU before oats (Table 1 ). The earlier maturity in barley arose because intervals between all developmental stages were less than in the other species. The exception was between double ridge and terminal spikelet (taken as the appearance of awn primordia in barley) where it tended to be longer in barley. This could be because the reproductive apex of barley does not have a terminal spikelet and the appearance of awn primordia, used as a surrogate for terminal spikelet, may over-estimate the equivalent of terminal spikelet in barley. Other consistent differences between species were that triticale reached both double ridge and terminal spikelet quickly, particularly at M89, but had the longest interval from terminal spikelet to anthesis and to physiologTable 1 Thermal time and calendar days (in parentheses) from sowing to double ridge (DR), terminal spikelet (TS), anthesis (A) and physiologicalmaturity (PM). Values are averagedover C89 and M89 Cultivar
DR
TS
A
ical maturity. A sample of the variation evident within and between species is given in Fig. 2 where the intervals between each of the developmental stages are presented for the earliest and latest flowering wheat and barley and the mean value for triticale. Values averaged for M89 and C89 are placed in order oftheir time to reach double ridge. Little resemblance to this order was found in the intervals between the subsequent stages. Most of the variation in the time to reach anthesis was due to the interval between sowing and DR formation. The relationship between time to DR and time to flowering (Fig. 3a) suggests that the period between DR and flowering was shortened as floral initiation became later. An explanation to the apparent accelerated development to anthesis of plants with a long time to double ridge is that the longer daylength experienced by late flowering plants may also be important. Fig. 3b shows that expressing the data in photothermal time does not alter the relationship between time to DR and time to flowering; thus it was likely that photoperiod sensitivity of the later-developing lines accelerated their development after DR formation. 3.2. Pre-anthesis growth
Characteristics of each species at the first detailed harvest in each experiment are given in Table 2. With
PM 800
Barley
Galleon O'Connor Ulandra Malebo
442(45) 447(46) 600(65) 477(50)
758(83) 1110(117) 1642(154) 670(73) 1140(120) 1740(160) 885(97) 1400(140) 1925(170) 840(9~) 1246(128) 1772(161)
466(48) 635(68) 635(68) 679(73)
835(92) 862(94) 918(100) 942(102)
Bread wheat
Kulin Me,or Rosella M3344
1256(129) 1376(138) 1376(138) 1396(138)
1920(170) 1965(172) 2048(176) 2024(174)
M3 600
"i
400
E 0
J¢
i.-200
Durum wheat
Altar84 Carcomun
530(56) 835(91) 1357(137) 1913(170) 598(64) 835(91) 1347(135) 1926(170)
Tntica~
Dua Currency
S-DR
493(52) 698(76) 1238(127) 1916(168) 529(56) 754(82) 1307(133) 1959(171)
Oa~
Echidna Hakea LSD(P=0.05)
496(52) 838(91) 1246(128) 1843(164) 646(70) 881(97) 1350(136) 1910(169) 18(2)
15(2)
13(1)
50(3)
DR-TS
TS-A
A-PM
Fig. 2. The averagetime interval from sowing to double ridge (SDR), double ridge to terminal spikelet (DR-TS), terminal spikelet to anthesis (TS-A) and anthesisto physiologicalmaturity (A-PM) for the earliest-and latest-floweringbarley(Galleon,G, and Ulandra, U), bread wheat (Kulin, K, and M3344, M) and the mean value for triticale, T, at C89 and M89. Cultivars are placed in order of their time to reach the doubleridge stage.
C. Lrpez-Castaaeda, R.A. Richards / Field Crops Research 37 (1994) 63-75 1400
o m
(a)
1300
o m
._E 1200
E (D
.C I--
1100 400
I
I
i
500
600
700
Thermal time to double ridge 800
- (b)
,m L. o
]= 0
750
o 7O0
E "~
650
E Q eoo o 55O
200
I
I
I
250
300
35O
Photothermal time to double ridge
Fig. 3. Relationshipbetween (a) thermaltime and (b) photothermal time to reach double ridge and anthesis for each genotypeof barley ( • ) , bread wheat ( • ), durum wheat ( ,a ), triticale ([]) and oats (o). Values are averagedover C89 and M89. The fitted line for (a) is y=-0.006x2+7.71x-1089 and for (b) y= -0.016x 2 + 9.73 Ix- 738. the exception of M90, this harvest was made when the number of main stem leaves was about 4.5 and leaf area index (LAI) was below 1.0. In almost every experiment barley had the highest LAI, developed main stem leaves fastest (and hence had the shortest phyllochron interval), produced the most tillers, and with the exception of oats, had the highest specific leaf area (SLA). The above-ground dry mass ( A G D M ) of barley was also higher than that of the other species except triticale in all experiments. Averaged over all experiments barley and triticale had the same A G D M and this was about 50% greater than in the other species. The LAI of triticale was 17% less than for barley when averaged over all sites. Triticale contrasted with barley in that it had
67
a longer phyllochron interval and thus fewer main stem leaves than did bread wheat and barley. Also, triticale generally had fewer tillers than the other species at this first harvest. There was evidence that triticale grew better at M88 and M89 than in the other experiments. Variation between species in the plant characteristics to the beginning of reproductive development was much greater than the variation within species. Table 3 shows the same plant characteristics as in Table 2 for all cultivars at C89. Although there was some variation within species, notably Ulandra barley, for most characteristics, variation within species was small. Differences among species in leaf area were establlshed from the time the first leaf had fully expanded and these differences were maintained as later leaves appeared. Fig. 4 shows the cumulative leaf area of each species in 1989 (average of C89 and M89) after emergence. The leaf area at harvest (330 thermal units) comprises both main stem leaves and tillers whereas the earlier points represent leaf area of the main stem leaves only as the contribution from tiller leaves was negligible. The first leaf of barley emerged earlier and it was larger than in the other species. Although the cumulative area of the first two leaves of triticale and oats was greater than for barley, triticale and oats have a slower leaf appearance rate and barley maintained a higher leaf area. The difference in leaf area between barley and triticale on the one hand, and the two wheat species on the other, was large and begins from the appearance of the first leaf and continued through the tillering phase. Oat leaves were larger than bread and durum wheats whereas bread wheat had a significantly greater leaf area than durum wheat beginning from when leaf 2 had fully expanded. The most intensive sampling for LAI was made in 1988 (C88 and M88) and the development of leaf area for C88, which was typical for all experiments, is given in Fig. 5. The LAI of barley was highest at all harvests up to anthesis. There was little difference between the two wheat species. Both triticale and oats had higher LAI than wheat and there was evidence that triticale maintained its green leaf area longer than the other species. The most intensive sampling for AGDM was made in C89 and M89 and the results at these two sites, shown in Fig. 6a and b, were representative of the other experiments. The AGDM was greatest at M89 and growth continued to physiological maturity. This contrasts
68
C. L6pez-Castatieda, R.A. Richards / Field Crops Research 37 (1994) 63-75
Table 2 Above-ground dry mass (AGDM), leaf area index (LAI), main stem leaf number (LNO), phyllochron interval (PI), tiller number (TNO) and specific leaf area (SLA) at the first detailed harvest in each experiment Exp. species
C88 (47 DAS) a Barley Bread wheat Triticale Oats LSD (P = 0.05) C89 (53 DAS) a Barley Bread wheat Durum wheat Triticale Oats LSD ( P = 0.05) M88 (56 DAS) a Barley Bread wheat Durum wheat Triticale LSD ( P = 0.05) M89 (54 DAS) ~ Barley Bread wheat Durum wheat Triticale Oats LSD ( P = 0.05) M90 (68 DAS) a Barley Bread wheat Durum wheat Triticale Oats LSD (P = 0.05)
AGDM (g m -2)
LAI
LNO (plant -1 )
PI (TU leaf -I )
TNO (m -2)
SLA (m 2 k g - ' )
49 31 44 27 6
0.8 0.4 0.6 0.5 0.1
5.8 4.9 4.6 4.3 0.1
63 74 79 84 2
801 488 352 521 65
26.2 22.5 23.3 27.5 1.2
39 23 16 30 23 5
0.6 0.3 0.2 0.4 0.4 0.1
5.0 4.2 4.0 3.9 3.9 0.3
69 82 87 88 88 5
779 544 582 545 658 79
26.1 20.3 18.0 22.6 26.0 1.8
48 39 37 58 5
0.7 0.5 0.5 0.7 0.1
5.9 5.4 5.0 5.0 0.1
73 79 85 85 2
799 585 462 450 46
23.2 21.3 22.3 18.9 0.8
19 16 13 23 14 3
0.2 0.2 0.1 0.3 0.2 0.05
4.7 4.0 3.8 4. I 3.9 0.2
68 80 85 79 82 5
712 556 502 558 550 57
22.5 19.8 17.9 21.4 23.3 1.6
151 103 59 124 98 19
2.4 1.6 0.9 1.8 1.8 0.3
7.2 6.5 6.0 6.1 6.0 0.3
76 84 91 89 91 3
878 710 570 572 723 95
24.6 24.0 25.0 23. I 26.0 2. I
aDays after sowing. w i t h C89, w h i c h e x p e r i e n c e d a m o r e s e v e r e t e r m i n a l d r o u g h t as a r e s u l t o f w h i c h t h e r e w a s n o a c c u m u l a t i o n in dry m a s s after a n t h e s i s . B a r l e y a n d triticale h a d the h i g h e s t A G D M at all h a r v e s t s at b o t h sites a n d this w a s also true in the o t h e r e x p e r i m e n t s . B r e a d a n d d u r u m w h e a t s h a d the l o w e s t A G D M at m o s t h a r v e s t s at b o t h sites; this w a s m o s t a p p a r e n t at M 8 9 . T h e dry m a s s o f oats w a s c o n s i s t e n t l y h i g h e r t h a n for the w h e a t s b e f o r e a n t h e s i s b u t b y m a t u r i t y little d i f f e r e n c e in total A G D M b e t w e e n oats a n d w h e a t w a s found. T h e A G D M at a n t h e s i s v a r i e d b e t w e e n species. A l t h o u g h b r e a d w h e a t , d u r u m w h e a t a n d oats r e a c h e d a n t h e s i s later
t h a n d i d b a r l e y a n d triticale, t h e i r A G D M at a n t h e s i s w a s less. C r o p g r o w t h rates ( C G R ) at M 8 9 a n d C 8 9 w e r e g r e a t e s t in barley, triticale a n d oats up to a n t h e s i s ( F i g . 7 a a n d b ) ; the C G R o f oats d e c l i n e d first at b o t h sites. D i f f e r e n c e s after a n t h e s i s w e r e less at the d r i e r site ( C 8 9 ) a n d w e r e less in oats a n d triticale at M 8 9 . D e s p i t e the d i f f e r e n c e s in A G D M a n d C G R , differe n c e s in relative g r o w t h rate ( R G R ) b e t w e e n species w e r e n o t d e t e c t e d ( d a t a n o t s h o w n ) . T h u s , the differe n c e s in m a s s e s t a b l i s h e d early w e r e m a i n t a i n e d until p h y s i o l o g i c a l maturity.
C. L6pez-Castaaeda, R.A. Richards / Field Crops Research 37 (1994) 63-75
69
Table 3 Above-ground dry mass (AGDM), leaf area index (LAI), main stem leaf number (LNO), phyllochron interval (PI), tiller number (TNO) and specific leaf area (SLA) 53 days after sowing at C89 Genotype
AGDM ( g m -2)
LAI
LNO (leaf -~ )
PI (TU leaf -L )
TNO (m -2 )
SLA (mZkg - ' )
42 44 31 40
0.7 0.7 0.4 0.6
5.0 5.0 4.7 5.1
68 69 73 67
770 793 777 777
28.2 25.4 23.8 26.9
20 20 24 27
0.2 0.3 0.3 0.3
4.0 4.2 4.5 4.1
87 81 77 83
457 567 557 597
24.3 18.2 19.9 18.6
15 16
0.2 0.2
3.7 4.3
93 81
553 610
18.0 17.9
28 33
0.3 0.4
4.0 3.9
87 89
603 487
22.8 22.5
23 22
0.3 0.4
3.8 4.1
91 85
673 643
26.6 25.4
6
0.1
0.4
8
126
2.3
BaHey Galleon O'Connor Ulandra Malebo
Bread wheat Kulin Meteor Rosella M3344
Durum wheat Altar 84 Carcomun
Triticale Dua Cu~ency
Oa~ Echidna Hakea LSD ( P = 0 . 0 5 )
~"
3O
!
X
0 •o
4
Ill
Q
z. (g
o
10 (P .--I
(.~
O-
I 0
100
200
300
I
400 300
600
900
Thermal time Fig. 4. Cumulative main stem leaf area for leaves 1,2 and 3 and total leaf area per plant at 330 TU after sowing for barley ( I I ) , bread wheat ( • ), durum wheat (zx), triticale (•) and oats (o). Values are averaged over C89, M89 and cultivars, and are calculated from the leaf area of each fully expanded leaf and the phyllochron interval. The standard error for leaf 1, leaf 2, leaf 3 and total cumulative area is 0.21, 0.25, 0.44 and 1.06 cm 2 respectively.
Thermal
1200
1500
time
Fig. 5. Leaf area index for barley (me), bread wheat ( A ), triticale ( [] ) and oats ( • ) at C 89 in relation to thermal time. Error bars show the standard error at each harvest. Values are averaged over cultivars within species.
70
C. Lrpez-Castatieda, R.A. Richards/ Field Crops Research 37 (1994) 63-75 1200
?
"(a)
(a}
E
I
v
g ¢1 E
[
I
I
900
-""
[ 1.0
600
,
== o o)
I I
1.5
(9 u
300
I
I
0.5
I o
DA
