Options for increasing seed yield of winter oilseed rape (Brassica napus L.): a simulation study

ELSEVIER

Field Crops Research

Field Crops Research 54 (1997) 109-126

Options for increasing seed yield of winter oilseed rape napus L.) : a simulation study B. Habekott6 Research Institute

(

Brassica

*

for Agrobiology and Soil Fertility CAB-DLO), Centrum de Born, P.O. Box 14, NL-6700 AA Wageningen, The Netherlands’ Received

14 May 1996; revised 28 February

1997; accepted

1 March 1997

Abstract Options for increasing seed yield in winter oilseed rape were evaluated quantitatively by means of sensitivity analyses with a crop growth model. Then, crop ideotypes were designed for high-yielding winter oilseed rape under optimal growth conditions, and their relative effects on seed yield for six different locations in Western Europe were simulated. Exploration of options indicated that higher seed yields may be obtained through (in descending order of importance): (a) delayed maturity; (b) improved seed set: (c) smaller petals or apetalous flowers; (d) increased potential growth rate of individual seeds; (e) earlier flowering with retention of the duration of total growth period and (f) erect clustered pods. The results also showed that increased average seed yields can best be obtained by simultaneously increasing the average sink and source capacity for seed filling. The most promising crop type for high seed yield matures late, combines early flowering with a maximum LAZ of about 3 for almost maximum light absorption and has erect clustered pods for source improvement. To take full advantage of the source, the sink has to be increased through a high rate of seed set, a large sink capacity of

individual seeds, apetalous flowers or a combination of these characteristics. This crop type showed large cumulative light absorption, light-use efficiency, harvest index and potential seed yield at different locations in Western Europe. Without a delay in maturity, the potential increases in seed yield were smaller. 0 1997 Elsevier Science B.V. Keywords:

Brassica

napus; Crop modelling;

Ideotypes;

Oilseed rape; Yield determination

1. Introduction

Increased seed yield of winter oilseed rape is one of a wide range of breeding objectives, that includes seed quality, cold tolerance in winter and early spring,

* Corresponding author. Heerenstraat 1, 6701 DG Wageningen, The Netherlands. Tel.: +31 317 412912. ’Phone: 31.317.475700; Fax: 31.317.423110. 0378.4290/97/$17.00 0 1997 Elsevier Science B.V. All rights reserved. PI2 SO378-4290(97)00041-5

resistance to diseases and lodging and seed retention (Thompson and Hughes, 1986). Since the development of the ideotype concept (Donald, 1968; Thompson, 1983; Sedgley, 1991, Rasmusson, 1991), plant physiologists have become interested in identifying crop characters that influence the physiological processes determining final seed yield. Small petals or apetalous flowers may improve light absorption by the green canopy (Mendham et al., 1981a; Yates and Steven, 1987; Rao et al., 1991); erect pods may

110

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

improve light distribution within the canopy, thus improving its photosynthetic capacity (Thompson and Hughes, 1986; Thurling, 1991); an increase in yield components, i.e. pod density or seeds per pod, bay increase sink capacity and partitioning of assimilates to the seeds (Mendham et al., 1981a; Sedgley, 1991; Thurling, 199 1; Grosse et al., 1992b; HabekottC, 1993). In addition to the agronomic benefits, the usefulness of the characters identified as selection criteria depends on other aspects, such as genetic variability, genetic control and measurability of the characters in large-scale trials (Mahon, 1983). Suitable characters may improve the choice of parental lines in breeding programmes and later the selection of superior genotypes @hurling, 1974; Thurling, 1991; Thompson, 1983; Rasmusson, 1991; Grosse et al., 1992a; Diepenbrock and Becker, 1992; Engqvist and Becker, 1993). So far however the application of knowledge of the crop’s physiology in breeding programmes has been limited, because the proposed strategies are not always unambiguous, quantitative effects of various crop traits on seed yield and their mutual interactions have often not been clearly identified. Moreover, studies are frequently based on a limited number of existing cultivars that do not represent the full range of biological strategies in yield formation (Link et al., 1992) and recording of proposed characters in breeding programmes was considered too laborious and expensive (Diepenbrock and Becker, 1992). The objective of this article was to identify options for yield improvement of winter oilseed rape, bearing in mind the constraints mentioned above. A simple crop growth model (Habekotte, 1996a; Habekotte, 1996~) was used for quantitative evaluation and synthesis of options for improving seed yield of winter oilseed rape under optimum growth conditions. These options were based on crop characters that show genetic variability (or are likely to do so) and most of them can easily be recorded in breeding programmes. The model used calculates total dry matter production on cumulative light absorption and light-use efficiency. Therefore, the effect of improved light distribution within the canopy on canopy photosynthesis and light-use efficiency, was analysed with a detailed model of daily canopy photosynthesis (Goudriaan, 1988; Spitters et al., 1989).

2. Material and methods

2.1. Standard potential seed yield of winter oilseed rape

Average potential seed yield was calculated for winter oilseed rape for a location in the centre of the Netherlands, using climatic data from the meteorological station in Swifterbant (5” 05’ E 52“ 34’ N) for a 15-year period (1974/75-1988/89). Potential seed yield is defined as the yield under optimum growth conditions, i.e. determined by crop genetic characteristics and the climatic factors radiation, temperature and daylength. The calculations were performed with LINTUL-BRASNAP, with parameters mainly derived from the cultivar Jet Neuf (Habekottt, 1997a; Habekotte, 1997~). The average simulated potential seed yield was compared with the average of observed seed yields of the same variety grown under near optimum growth conditions near Swifterbant (1979/801990/91) (HabekottC, 1989; van der Meulen, 1990a; van der Meulen, 1990b; van der Meulen, 1991; van der Meulen, 1993). Mean values of simulated and observed data were compared by the Student t-test at the 0.05 level of probability (t0,05). The average simulated potential seed yield is further used as standard seed yield to evaluate the effect of changes in crop characters on seed yield. 2.2. Options for increasing

seed yield

The scope for increased seed yields of winter oilseed rape was examined using the model LINTUL-BRASNAP (Habekotte, 1997a; Habekotte, 1997~). The options for increasing the seed yield of winter oilseed rape were derived from the literature and classified in three groups, i.e. I) light absorption (options l-4), II) light distribution in the green canopy and canopy photosynthesis (option 5) and III) partitioning of assimilates to the seeds (options 6-8) (Table 1). As the absolute ranges of genetic variability in most crop characters are not known, the effect of relative changes in crop characters on seed yield were analysed. The results indicate which crop traits offer the best opportunities for yield improvement, based on available physiological knowledge. The relative changes in the values of most relevant

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

model parameters were 10, 20 and 30%, except for smaller petals (option 3), in which an absolute range of 100% was introduced to examine the possible effect of apetalous flowers. The physiologically based potential increase in canopy photosynthesis resulting from improved light distribution in the canopy (option 5) was derived using a detailed model of daily canopy photosynthesis. The calculations were carried out for the same location and time period used for calculating the potential seed yield of winter oilseed

111

rape (1974/75-1988/89, Swifterbant). Consequences of the following changes were evaluated: (1) Accelerated leaf area development before full flowering to attain maximum light absorption during reproductive growth (Mendham et al., 1981a; Mendham et al., 1981b; Thompson, 1983; Grosse et al., 1992a; Leon, 1993). This was implemented in LINTUL-BRASNAP by increasing the partitioning of assimilates to the leaves (Table 1). (2) Delayed onset of flowering, so as to attain

Table 1 Options for increasing seed yield in winter oilseed rape, related parameters in LINTUL-BRASNAP (Habekotte, 1996~) and their values relative to the standard values. Symbols used are described in Table 3 (OF: onset of flowering; EF: end of flowering; M: maturity; FP: flowering period; CP: critical period for seed set; SFP: seed-tilling period; TGP: total growth period) Option

Related parameter

Relative value (%)

Standard

all parameters

100

P I,DVS

a: b: c: a: b: c: a: b: c: a: b: c: a: b: c:

Light absorption 1. Accelerated leaf area development

before OF

2a. Delay in OF with retention of the duration (d) of TGP

aT,2;

2b. Delay in OF

aT,2

3. Smaller petals

A,,

4. Delay in maturity

aT,4

Light distribution and canopy photosynthesis 5. Improved canopy photosynthesis through improved light distribution

LUEDVs

Partitioning of assimilates to the seeds 6. Increased response of seed set to crop growth during CP

s2

aT,4

7. Increased daily potential growth rate of the seeds

8a. Extended SFP resulting from earlier OF with retention of the duration (d) of TGP

aT.2; aT,4

8b Extended SFP resulting from earlier OF

aT.2

110 120 130 90; 110 80; 127 70; 160 90 80 70 90 80 0 90 80 70

a: 102.5-105 b: 105-110

a: b: c: a: b: c: a: b: c: a: b: c:

110 120 130 110 120 130 110; 90 120; 82 130; 78 110 120 130

112

B. Habekotte’/ Field Crops Research 54 (1997) 109-126

sufficient leaf area at flowering for maximum light absorption during reproductive growth (l-hurling, 1974; Mendham et al., 1981a, HabekottC, 1997b). This was simulated both with (a) and without (b) retention of the duration of the total growth period (Table 1). For both scenarios, the duration of the critical period of seed set (CP) and of the flowering period (FP) were maintained. Delayed onset of flowering was implemented by reducing the temperature-governed rate of development before flowering by lo%, 20% and 30%. This resulted in the onset of flowering being delayed (options 2a and 2b) by 4, 7 and 11 d, respectively, and maturity being delayed (option 2b) by 6, 11 and 18 d, respectively. The average duration (d) of the total growth period in option 2a was kept constant by increasing the temperature-governed rate of development after flowering. (3) Increased transmission of light through the flower layer by small petals or apetalous flowers (Mendham et al., 1981a; Yates and Steven, 1987; Rao et al., 1991). This was achieved by reducing the area per flower (Table 1). (4) Delayed maturity to optimize the use of available light during the growing season (Richards, 1991; Marshall, 1991). In the model, this was accompanied by prolonging the seed-filling period, while retaining the duration (d) of the preceding stages, i.e. the vegetative growth period (VGP) until onset of flowering, the flowering period and the critical period for seed set. Maturity was delayed by reducing the temperature-governed response rate of development after flowering by lo%, 20% and 30%. This delayed maturity by 5, 11 and 18 d, respectively. (5) Higher canopy photosynthesis rates associated with improved light distribution in the canopy induced by more erect pods (Thompson and Hughes, 1986; Thurling, 1991) and a change in the clustering of pods. (6) Increased seed set for given assimilate production during the critical period of seed set, e.g. through reduced branching (Keiller and Morgan, 1988; Habekottt, 1993). This was implemented in the model by increasing the response of seed set to crop growth during the critical period of seed set (Table 1) (Habekottt, 1997c). (7) Higher potential daily growth rate of the seeds (Habekotte, 1993).

(8) Prolonging the seed-filling period (SFP) through earlier flowering of the crop (Thurling, 1991). In these simulations the duration of FP and CP was kept constant and the duration of the total growth period was either kept constant (8a) or modified (8b) (Table 1). Flowering was advanced by increasing the temperature-induced rate of development before flowering by lo%, 20% and 30%. This advanced the onset of flowering by 6, 12 and 17 d, respectively, and maturity (option 8b) by 5, 10 and 13 d, respectively. The average duration (d) of the total growth period (option 8a) was retained by reducing the temperature-governed rate of development after flowering. 2.3. Location,

weather data and initial conditions

A location in the centre of the Netherlands near Swifterbant was selected for the analysis of options for yield improvement. The selection criteria were: availability of at least 15 years of climatic data and experimental data on seed yields of winter oilseed rape. Sowing date was set at August 25, an average date for the central part of the Netherlands (Habekotte, 1989). Initial values for simulating crop growth, initial total crop dry weight (W,,,,; 170 g m-‘) and leaf area index (LAZ,; 0.80 m2 m-‘) at the onset of crop regrowth in spring, were derived from literature (de Boer and Langenhuysen, 1985; Habekotte and Smid, 1992). The onset of regrowth in spring ( ts> was estimated annually by the method of Mendham et al. (1981a): “the date when mean temperatures rose and were maintained above 5°C”. In years, in which temperatures remained above 5°C for more than 5 days before this estimated onset of regrowth, t, was advanced in accordance with the duration (d) of this period. To evaluate the relative effect of a hypothetical improved crop type on potential seed yield and components of yield, six locations in Western Europe were selected, on the basis of variation in climatic conditions and availability of long-term mean monthly weather data (Stol, 1994). The optimal sowing date choosen per location was based on published data (Habekotte, 1989; Grosse, 1989; Mendham et al., 1981a) and on expert knowledge (M.J.J. Pustjens, Van der Have, pers. comm., 1995). Initial crop conditions were similar to those for the

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

calculations in the Netherlands, except for t, for south France (Toulouse). At that location, mean daily temperatures in winter were continuously above 5°C so t, was estimated from the duration of crop growth in spring necessary to attain an LAZ,,, within the range of the other locations (range: 3.0-3.3). 2.4. LZNTUL-BRASNAP LINTUL-BRASNAP (Habekotte, 1997a; HabekottC, 1997~) is an extended version of LINTUL: a simple crop growth model based on Light INTerception and UtiLization of light for dry matter production and partitioning of dry matter to the seeds (Spitters, 1990). Th e model calculates: (1) light reflection and absorption by the green canopy as af-

113

fected by properties of the flower layer, (2) total dry matter production derived from light absorption and light-use efficiency CLUE) by the green canopy, (3) seed density, (4) partitioning of dry matter to the seeds based on either source or sink limitation (Mendham et al., 1981a; Spitters et al., 1989; Habekottt, 1997~) (5) accumulation and mobilization of reserve carbohydrates and (6) crop phenological development in relation to temperature (including vernalization) and photoperiod (Habekotte, 1997b). LUE is defined as a funtion of the development stage of the crop. Seed density is related to dry matter accumulation during the critical period of seed set (CP), 350”Cd from the onset of flowering. Seed yield and reserve carbohydrates are both expressed in weight units similar to those of the green

Table 2 Standard parameter values used for calculating daily canopy photosynthesis, light - use efficiency and fractional light absorption DAYASS. The calculations were carried out for a pod layer above a leaf layer just after flowering (I) and near maturity (II) Parameter

Value

Description

Unit

with

and source

I: pod layer aboue leaf layer just after flowering Leaves 0.0125 mg CO, J-’ E

3 spherical

Light-use efficiency of individual leaves, at low light intensities (de Boer and Langenhuysen, 1985) mg CO, mm2 SK’ Maximum photosynthesic rate of individual leaves (de Boer and Langenhuysen, Scattering coefficient for single leaves for photosynthetic active light (PAR) (Goudriaan, 1988) Light extinction coefficient for diffuse light, corrected for light interception by stems near onset of flowering (Habekotte and Smid, 1992) m* m-* Leaf area index See text

??

0.0125

mgC0,

P g.max scv

0.4222 0.2

mg CO, m-*

Rdif Pod angle distrib. PAI S,, June 15

0.716 spherical

P&mar scv

1.1111 0.2

Rdir

0.903

LA1 Leaf angle distribut. Pods

1, 2 19

JJ’ s -I

m* me2 MJ mm2 d-r

II: Pod layer above leaf layer near maturity Leaves 1 LAI m* m-* P g.max 0.3944 mgCOs mm2 s-’ Other parameters as in I Pods as in I

Light-use efficiency of individual pods at low light intensity (de Boer and Langenhuysen, 1985) Maximum photosynthetic rate of individual pods (de Boer and Langenhuysen, Scattering coefficient for photosynthetic active light (PAR) for single leaves, also used for pods, (Habekotd, 1996) No clustering of pods (see text) see text Pod area index (based on projected area; Habekotte and Smid, 1992) Incoming global radiation at June 15

Habekottt and Smid, 1992 Maximum photosynthetic rate of old individual leaves (de Boer and Langenhuysen, 1985)

1985)

19851

114

B. Habekotte’/ Field Crops Research 54 (1997) 109-126

canopy (GCU, g m - *: green crop units) to account for differences in chemical composition. Varieties adapted to a specific geographical region, such as Western Europe, are expected to respond similarly to sowing date, environmental conditions (Mendham et al., 1981a) and changes in crop characters. For example, delayed sowing generally results in lower seed yields, low temperatures in spring or summer will slow down phenological development, and apetalous flowers will improve light distribution within the canopy. The model was developed to simulate these general trends in yield formation with the cultivar Jet Neuf as a standard for quantifying most parameter values (Habekotte, 1997a; HabekottC, 1997~). LINTUL-BRASNAP requires input of daily incoming global radiation (S, MJ m-’ d- ‘> and daily minimum and maximum temperatures (r,, and T,,, , respectively). The model is written in FORTRAN 77 (Meissner and Organick, 1984), using the FSE system (FORTRAN Simulation Environment, van

Table 3 Description

Kraalingen (Habekotte,

(1995)) 1997a).

2.5. Light distribution

for

crop

growth

simulation

and canopy photosynthesis

To analyse the effects of spatial distribution and clustering of pods on light absorption, LUE and canopy photosynthesis, the standard routines for calculating daily gross assimilation in SUCROS (Spitters et al., 1989) were extended (Goudriaan, 1988; Habekotte, 1996). The routines, henceforth referred to as routines DAYASS, calculate canopy photosynthesis, light-use efficiency and fractional light absorption for two canopy layers combined, an upper layer comprising pods and a lower layer consisting of leaves. DAYASS calculates daily canopy assimilation from incident photosynthetically active radiation (PAR), the proportion of diffuse light, solar elevation, spatial arrangement of leaves and pods and optical and photosynthetic properties of individual leaves and pods (Table 2). The latter include trans-

of symbols used in Table 1, Table 5 and Table 7

Symbol

Unit

aT,Z; aT.4

10m3 d- ’ “C- ’ Response of development rate to temperature for the period from emergence until flowering (X = 2) and for the period from the end of flowering until makity (X = 4) m2 Area per individual flower -1 Initial fraction of reserve carbohydrates (as fraction of total crop dry matter) g !Z Harvest index ( W,/ WtnP,,, ) MJ m-a Cumulative incoming PAR from April 15 until maturity MJ mm2 Cumulative PAR absorbed by the green crop canopy from April 15 until maturity MJ m-* d-i PAR transmitted daily to the bottom of the canopy Extinction coefficient of leaves for PAR m2 m-2 Leaf area index m2 me2 Maximum leaf area index gMJ-’ Light-use efficiency: total dry matter production (g GCU mm2 d- ‘) per unit absorbed PAR (MJ m-* d- ‘), as a funtion of the development stage (DVS) of the crop g MJ-I Average light-use efficiency from April 15 until maturity Mean seed weight mg mm2 Pod density mm2 Seed density mm2 Number of seeds per pod MJ me2 Photosynthetically active radiation (400-700 nml Partitioning coefficient of daily crop growth to the leaves gg-’ g “Cd Potential growth rate of the seeds (per 1000. seeds (GCU)) m-a; _ Parameters for calculating seed density MJ MJ-’ Transmission coefficient of PAR through the flower layer g mm2 Seed weight (GCU) per unit ground area grn-* Total aboveground crop dry weight at maturity (including dead leaves and roots (GCU))

A,, ks.ovs

= 3.3

HI I &cum I a.c”lll I,

Ii LA1 L.AI

LAI,,, LUEDV

s

LUE,,, msw Np N, NS/P PAR P 1,D”S Ris,rotovs s,; sp TR w, Wtop&i

Description

B. Habekotte’/Field

Crops Research

mission and reflection of PAR, maximum rate of gross photosynthesis (P.,,,,) and light-use efficiency at low light intensities (6). The three types of leaf and pod angle distribution are characterized by fractions (F) of leaves and pods in three angle classes, O-30”, 30-60” and 60-90” (spherical: Fl = 0.134, F2 = 0.366, F3 = 0.5; erectophyl: Fl = 0.076, F2 = 0.303, F3 = 0.621; erect: Fl = 0, F2 = 0, F3 = 1). Leaf and pod clustering is taken into account through a cluster factor: the ratio of measured to theoretical values of the extinction coefficient of diffuse radiation in the canopy: Kdi, = 0.8 {(l: scattering coefficient of individual leaves; Goudriaan, 1988; Spitters et al., 1989). The cluster factor refers to both a clustered distribution of leaves or pods which increases mutual shading, with values of the cluster factor lower than 1, and to preferential formation of leaves and pods in gaps within the canopy, resulting in high values for Kdi, and a cluster factor with values exceeding 1. Maintenance respiration was set to 20% of daily gross photosynthesis (in glucose weight units) (Spitters et al., 1989). Conversion of the remaining assimilates into structural plant material was based on the multiplication factor derived for green crop units (l/ 1.49 = 0.67) (Habekottt, 1997a). Table 3 shows a description of symbols used in Table 1, Tables 5 and 7. Calculations were carried out for two standard situations: (I) a pod layer above a leaf layer just after flowering with a relatively high LAZ value (3 m* mm2 > and (II) a similar situation later during the growth cycle, with a lower LAZ (1 m* m- *) (Table 2). The standard situations were characterized by day number of the year, incoming global radiation, P,,,,X, E, leaf and pod angle distribution, Kdi, (indicating the degree of clustering of leaves and pods) and area index of leaves (LAZ) and pods (PAZ). The value of incoming solar radiation on June 15 (day number 166, 19 MJ m-*, Table 2) corresponds to the average value for the proportion of incoming diffuse light for the growth period of winter oilseed rape from April until maturity, for the years 1974/7588/89, near Swifterbant, i.e. 0.67. Subsequently, light absorption, LUE and canopy photosynthesis were calculated for erectophyl or erect pod angle distributions and for values of Kdif lower than the standard value.

54 (1997) 109-126

115

3. Results 3.1. Standard potential rape

seed yield of winter oilseed

Simulated average potential seed yield (4.66 t ha-‘, GCU, 1974/75-1988/89) was similar to the average of observed seed yields for the central part of the Netherlands (4.99 t ha-‘, GCU, (1979/801990/91) (P > 0.05). These values correspond to respectively 3.12 and 3.34 t ha-’ seed dry matter. 3.2. Light distribution

and canopy photosynhesis

The effects of more erect pods and clustering of pods in a pod layer above a leaf layer are only shown for their spherical and erect angle distribu-

Table 4 Canopy photosynthesis (mg CO, me2 dd ‘), light-use efficiency CLUE, g MJ-‘) and fractional light absorption calculated with DAYASS (Table 2) for a canopy with a pod layer above a leaf layer just after flowering (I) and near maturity (II) with differences in pod angle distribution and clustering of the pods. The latter is indicated by different values for Kdi, (Kdi, = 0.716: no clustering). The results are given for day number 166 (June 15) for different values of LAZ and PAZ Pod angle distribution

K,,,

UZ

3(Z)

Z(ZZ)

PAZ

PAZ

1

2

1

2

Canopy photosynthesis Spherical 0.716 Erect >, Spherical 0.60 Erect >>

56.5 58.4 59.5 60.9

54.1 57.2 57.9 60.4

40.5 40.9 41.9 42.1

44.8 45.8 46.7 47.4

Light-use efficiency Spherical 0.716 Erect >> Spherical 0.60 Erect ,>

2.43 2.53 2.57 2.65

2.28 2.43 2.45 2.58

2.12 2.20 2.26 2.33

2.08 2.21 2.24 2.35

0.916 0.908 0.911 0.903

0.737 0.715 0.716 0.697

0.828 0.799 0.805 0.177

Fractional light absorption Spherical 0.716 0.895 Erect ,, 0.890 Spherical 0.60 0.890 Erect 1, 0.886

116

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

tions. The erectophyl angle distributions showed the same trend, albeit less pronounced. For pod area indices of 1 and 2 m* m-*, canopy photosynthesis increased by 10% on average and light-use efficiency by 11% on average for erect and clustered pods compared to the standard situation I, characterizing the canopy just after flowering. The fraction of absorbed light decreased very slightly (on average by 1.2%). Later during the growth cycle (situation II), canopy photosynthesis increased by 5% on average for a PAZ of 1 and 2 m* m-* (Table 4). The decrease in fraction of light absorbed, 6% on average, partly offsets the improved light-use efficiency, which was 11% on average. These results indicate that canopy photosynthesis could increase by at most 10% and 5% just after flowering and near maturity, respectively, if light distribution through the pod layer could be improved. These results were used for further analysis of options for yield increase with LINTUL-BRASNAP. 3.3. Options for increasing

seed yield

Seed yield (g m-2, GCU)

Seed density (lo3 nrz )

Seed yield

b

(g m-2, GCU)

fdm

Mean seed weight

(mg, GCU)

c

16 I4

12

0’

0

3.3.1. Sink and source capacity for seed filling The effects of hypothetical changes in sink and source capacity for seed filling on seed yield were analysed as a basis for further evaluation of options for increasing seed yield. Fig. la showes the simulated response of seed yield, averaged for all sowings (1974/75-1988/89, Swifterbant) to a hypothetical change in sink capacity, in this example obtained through changes in seed density (mm2>. Average seed yield and seed density for the standard parameter set are indicated by the solid symbol in Fig. la. At seed densities below the ‘standard’, about 69 X lo3 m-*, seed yield decreases almost linearly with seed density. Hence, on average for all sowings, seed yield is almost completely sink - limited at these seed densities. With increased seed density, seed yield increases up to a maximum of 29%, indicating a source surplus. However, at increasing seed density the source becomes increasingly limiting for seed filling, as illustrated by the decreasing response of seed weight (= msw; Fig. lc). The hypothetical increases in seed density were calculated without possible limitations by pod density. When pod numbers are taken into account, simulated pod densities

a

100

200

3w

400

500

Seed densxy (IO.’ mu*)

Fig. 1. Simulated response of (a) average seed yield to hypothetical changes in seed density of winter oilseed rape, (b) average seed yield to source availability, obtained through hypothetical relative changes (fdm) in dry matter production (g m-‘1 after flowering and (c) average mean seed weight to seed density. The dotted line in (b) represents the situation of complete compensation of the increase in seed density with a decrease in mean seed weight (msw = 466/seed density). The calculations were carried out for sowings in the period 1974/75-1988/89 for climatic data from Swifterbant. The solid symbols represent the average values obtained with the standard parameter set in LINTUL-BRASNAF’ (Habekotte, 1997a; Habekottt, 1997~).

will not limit seed set up to an average seed density of 150 X lo3 m-*, as the maximum number of 35 seeds per pod (Habekotte, 1997a) is not reached. To attain higher average seed densities without limitation of seed set, pod numbers have to increase. If only source capacity during seed filling is increased, by changing dry matter production after flowering, surplus in sink capacity (in terms of seed density) appears to be 12% on average. Reducing the source capacity leads to an almost linear decline in

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

average seed yield (Fig. lb), indicating that in this range, seed yield is almost fully determined by source capacity. The compensatory effect between seed density and mean seed weight is illustrated in Fig. lc by the simulated response of mean seed weight to seed density. The effect of increased seed density (sink) without a simultaneous increase in source capacity is largely offset by a decrease in mean seed weight. The dotted line in Fig. lc represents complete compensation between mean seed weight and seed density (msw = c/nseed; with c = 466 g rnm2). As source capacity is slightly in excess, the decrease in mean seed weight is slightly less than according to the dotted line. These results suggest that to increase the seed yield of winter oilseed rape, source and sink capacity for seed filling should be increased simultaneously. Simulations with a combined increase in sink and source capacity for seed filling, through an assumed simultaneous increase in seed density (by 10 to 100%) and in dry matter production (by 10 to 100%) after flowering, resulted in a positive interaction effect on seed yield (of 0.8 to 55%). The effect was most pronounced at larger increases in sink and source. This also results in deviations from the line that represents the compensatory effect between seed density and mean seed weight as illustrated in Fig. lc by the black circle, for 100% increase in sink and source. An increase in sink capacity only, through an increase in seed density (by 10 to 100%) combined with a higher daily potential seed growth rate (plus 10 to 100%) resulted in negative interaction effects on seed yield (of 0.6 to 19%). A negative interaction effect (of 0.01 to 3.8%) was also found when only source capacity was increased by increasing both dry matter production after flowering (by 10 to 100%) and the availability of reserve carbohydrates for seed filling (by 10 to 100%). 3.3.2. Single crop traits Promising options were identified as those leading to a yield increase exceeding the (rather arbitrary) limit of 3% (Fig. 2 and Table 5). The options for seed yield increase were evaluated based on their effects on sink and source capacity for seed filling and on their effects on cumulative light absorption, light-use efficiency and harvest index and related crop characters (Table 5).

111

Fig. 2. Simulated response of average seed yield (in percentage decrease/increase of the standard yield) to changes in various canopy traits for sowings in the period 1974/75-1988/89 for climatic data from Swifterbant (see text and Table ITable 6 for further details). The dotted line represents 3% increase in average seed yield.

Higher average seed yields attributable to higher sink capacity were obtained directly through stronger (of 10 to 30%) responses of seed set to dry-matter accumulation during the critical period of seed set (option 61, increases (also of 10 to 30%) in the potential growth rate of individual seeds (option 7) and indirectly through increased transmission of light through the flower layer as a result of reduced (by 20 to 100%) petal size (option 31, a delay in maturity (5 to 18 d) and a concomittant increase in the ratio of the durations of the seed filling period (SFP) and total growth period (TGP) (option 4) and increased duration of SFP (by 6 to 17 d) (option 8a). Greater average seed yields as a result of a larger source capacity were obtained either through a higher dry matter production rate (of 5 to 10%) during seed filling as a result of the introduction of erect clustered pods (option 5), or through an increase in the duration of seed filling in options 4 (by 5-18 d) and 8a (by 6-17 d). Except for option 5, both groups of options with positive effects on sink or source capacity for seed filling resulted in higher average harvest indices. Cumulative light absorption was distinctly increased ( 2 3%) through a delay (4- 11 d) in onset of flowering (option 2b), delayed maturity (5- 18 d) (option 4) and apetalous flowers (option 3), although only the latter two resulted in higher seed yields. Light-use efficiency was increased (> 3%) through delayed onset of flowering (11 d) with retention of the duration of the total growth period (TGP) (option 2a), as this extended the relative duration of the vegetative growth period (VGP) compared to TGP and increased the relative contribution of more productive leaves in the VGP compared to the pods

E ;‘ 2 i? zt 2 P q

*u

-s

E

3

801

LZI 011

EZI

661 901

El1

ZZI 001

86 66 so1

III

611 so1

ZII

OZI M)I

801

PI1

at

101

LO1

911

101

LO1

e

3 9 3 9

so1

9ZI

CZI 011 ET1 co1

001

ZOI

3

F

so1

101

ZOI

ZOI

so1

101

621

911

v

9

86 101

ZOI

68

811

ZOI

PI1 EOI LII

so1

011

ZII ZOI 8PI III

96 901

ZOI

so1

011

101 9ZI

001

601

101 L6

88 66

EOI

101

101

66 E6

96

L8 L6

101

so1

6El S6

9L 86

101

L6

26

LO1 LO1

PZI

E9

ZII

08

E6 LO1

98

LO1 9ZI

8PI 101

001

101

101

ZOI

EOI

3

3 9 e

L6

19

101

b9 28 E6

96

09

ZOI

ZOI

iXl1

6EI

101

ZOI 101

101

101

001

101

e

9

9ZI

101

ZOI

PZI

9L 101

98 16 601

66

101

19PI

LII

LS9

601

628

66

EZ’E

001

96L’O

101

(,-fI4 '"'m_I

m)x~

001

%)IH

(T-Z

61E’O

rOIXSN

(z-“1

0’69 98'1

66

101

so1

ZII

98'9

9L 68

ZOI

PZ'O

Msu~

(“9

P) dE)l/drIS

(,-P

LL 06

E6

‘z-7 N

WIS

ZOI 101

86 66

N

SE1

d/s

C

was implemented

94 87 82

96 89 85

t

98 94 91

102 103

seeds a 106 b 110 c 114 a 105 b 109 c 112 a 105 b 107 c 108

a The increase in canopy photosynthesis canopy (Habekott6 and Smid, 1992).

Partitioning of assimilates to the Increased response 6 of seed set to crop growth during CP Increased potential 7 growth rate of the seeds Extended SFP 8a resulting from earlier OF and retention of TGP 8b Extended SFP resulting from earlier OF

Light distribution and canopy photosynthesis Increase in canopy a 102 5 photosynthesis a b 103

95 90 86

100 100 101 83 72 63

83 72 63 101 102 103

101 102 103

from about halfway through the flowering

93 85 80

99 95 95 102 101 102

106 110 114 105 109 113 108 112 116 93 86 80

93 86 80

111 122 133

101 102

period (DVS = 2.6, Habekott6,

98 97 96

98 96 94

102 104

105 111 116

113 125 135

1997~) with increasing

103 105 108

95 91 86 105 110 114 114 125 136

101 102

105 111 116

105 111 116

111 122 133

100 101

pod area index of the

88 76 67

88 76 67

100 101

2 s 0

ta z g % B 2, \

120

B. Habekott&/Field

Crops Research 54 (1997) 109-126

in the seed-filling period to total assimilate production. Light-use efficiency also increased through improved light distribution in the canopy as a result of erect clustered pods (option 5). Only the latter option, however, resulted in distinctly higher seed yields (Fig. 2 and Table 5). Increased leaf area formation up to the onset of flowering hardly increased light absorption and total dry matter production, as shown in options 1 and 2a (range l-3%, Table 5). The relative decrease in the duration of the seed filling period in option 2a (SFP/TGP, a reduction of lo-40%) reduced the availability of assimilates for seed filling and hence reduced harvest index (HI, by 8-37%) and seed yield (W,, by 7-35%). In option 2b, the delay in onset of flowering also delayed maturity and this increased cumulative light absorption (by 6- 18%). The sink capacity of the seeds increased, because of enhanced seed set, however, the relative availability of assimilates for seed filling decreased as a result of the lower SFP/TGP ratio, which resulted in a lower HI (down by 3-12%). The higher total crop drymatter at final harvest (up by 5-16%) compensated for the lower HI, with consequently a slightly higher seed yield (up by 2%). Earlier onset of flowering (6-17 d) with associated modification of the duration of the total growth period (option 8b) reduced the duration of the total growth period (by 5-13 d) and hence reduced the cumulative light availability (Z0,C”lll~by 5- 14%) and absorption of light (I,,,,, , by 7-20%). The increase in the relative duration of the seed-filling period (by 5-16%) hardly affected the partitioning of dry matter to the seeds (HI, up by l-2%) and did not compensate for the lower total

dry matter production and thus resulted in lower seed yields (down by 5-16%). The simulated effects of relative changes in crop traits all resulted in moderate seed numbers per pod (range in annual values: and 8-23, range in average values: 10.2-18.0, Table 5). Thus, simulated seed densities were not limited by simulated pod densities.

3.3.3. Combined crop traits The crop traits described above were combined to realize a simultaneous increase in sink and source capacity for seed filling and their combined effect on average seed yield (1975/76-1988/89) was calculated (Table 6). A seed yield increase of 2- 18% was attained without a change in the developmental pattern of winter oilseed rape. When flowering was advanced (by 6-17 d), average seed yield increased by 7-25%, and when maturity was delayed (by 5-18 d), by 12-55%. A negative interaction effect ( > 1%) occurred only for the combination of early flowering (option 8a) and delayed maturity (option 4) (Table 6). This was the result of a lower LAZ,,, at early flowering (Table 5) which partly offset the increase in source availability through delayed maturity (option 4). Fig. 3 shows that when average values exceed 3, LAZ,,,,, does not limit average seed yield.

3.3.4. Zdeotypes for high seed yields Based on the insights gained in yield formation, the most promising crop type for high-yielding winter oilseed rape in Western Europe under optimal

Table 6 Simulated seed yield (in percentages of the standard seed yield of 466 g m-*, Table 5) for combined options for seed yield improvement of winter oilseed rape (1974/751988/89, Swifterbant). The ranges in seed yield result from calculations with the lowest and highest values of the relevant parameters described in Table 1. Sink

Source

Later maturity (4) Increased response seed set (6 * ) Petal size, apetalous flowers (3) Higher potential seed growth rate (7 * ) Earlier flowering (8a) ” Combined

110-139 106-114 101-113 105-112 105-108

options without change in the development

Later maturity (4)

Earlier flowering

110-139

105-108 115-142 112-125 107-123 111-124

X

116-155 112-155 115-154 115-142 pattern of the crop.

X

(8a)

Clustered upright pods (5) 102-103 111-144 107-118 102-117 106-116 107-112

” ” ”

B. Habekotte’/ Field Crops Research 54 (1997) 109-126

100 ~~-

0 I

0

2

3

4

Fig. 3. Simulated response of average seed yield to maximum LAZ for sowings in the period 1974/75-1988/89 for climatic data from Swifterbant. The solid symbol represents the average values obtained with the standard parameter set in LINTUL-BRASNAP (Habekottt, 1997a; HabekottC, 1997~).

growth conditions was formulated: a variety that matures late, combines early flowering with an of about 3 and has erect clustered pods to UZnl,,X maximize the source for seed filling. To take full advantage of this source capacity, sink capacity has to be increased concurrently through a high rate of seed set, a large sink capacity of individual seeds, apetalous flowers, or a combination of these characteristics.

121

The performance of this crop ideotype in terms of seed yield, light absorption, light-use efficiency and harvest index was calculated for six locations in Western Europe (Table 7). In these calculations an Z‘AZnl,Xof at least 2.8 was considered sufficient for almost maximum light absorption (Fig. 3). Flowering was advanced by 3-6 d (ur,* increased by 5-lo%, Table 1) and maturity was delayed by 6-9 d Car,4 reduced by 20%, Table 1) or by 13-17 d Cur.4 reduced by 35%, Table l), respectively. With a further increase in sink size through apetalous flowers, a stronger response of seed set to dry matter accumulation during the period of seed set (up by 15%) and a higher sink capacity of individual seeds (an increase of 15%), the source capacity was almost completely used (2 95%) for both maturity types. When maturity was delayed by about one week, cumulative absorption of PAR (I,,,,,) increased by 13-19%, LUE,,, by 2-4%, HI by 29-31% and potential seed yields by 43-51% (Table 7). Delaying maturity by about two weeks increased cumulative absorption of PAR (I,,,,,) by 21-30%, LUE,,, by 2%, HI by 29-39% and potential seed yields by 59-70% (Table 7). With the same modifications in sink size and source capacity, and without the changes by a delay

Table 7 Simulated potential seed yields and related cumulative PAR absorbed (I,,,,, ) , light-use efficiency CLUE,,,) and harvest index (HI) at different locations in Western Europe for a standard variety (St), and for hypothetical high-yielding varieties (hv) without and with delay in maturity (range: about one week - about two weeks; see text). The results of the hypothetical varieties are given in percentages of the standard values per location. Symbols used are described in Table 3. All weights are given in GCU Location

Delay of maturity Copenhagen Hamburg de Bilt London Paris Toulouse

I.%,C”ITl

Y

st (t ha- ‘) No delay of maturity Copenhagen 4.63 Hamburg 4.56 4.50 de Bilt London 5.13 4.72 Paris 4.66 Toulouse

st (MJ mm*)

hv (%I

st(gMJ-

130 130 130 127 132 132

665 638 578 594 551 480

106 105 105 110 108 106

1.81 1.84 1.87 1.90 1.86 1.80

148-167 148-166 147-164 143-159 151-170 151-169

HI

LW,,

hv (%I

113-122 113-121 114-122 119-127 118-128 118-130

1)

hv (%)

st (tt- ‘1

hv (%)

104 104 104 103 103 105

0.34 0.33 0.34 0.36 0.34 0.34

120 121 120 114 121 120

104-102 102-102 103-102 102-102 102-102 103-102

129-137 131-139 128-136 122-129 130-139 130-138

B. Habekotte’/Field

122

Crops Research 54 (1997) 109-126

in maturity, the increases in potential seed yield varied between 27-32%. This crop type also shows higher values for cumulative absorption of PAR ($lO%), LUE,, (3-5%) and HI (14-21%) (Table The values of LUE,,, were only slightly higher, because increased canopy photosynthesis (5-IO%, Table 4) was only introduced during the seed-filling period and the delay in maturity, accompanied by a prolongation of the seed-filling period, partly offsets this increase, because capacity of the pods is lower than that of the leaves. Irrespective of maturity class, similar relative yield increases were simulated for given measures to increase source capacity and sink capacity solely by increasing seed set by 45%. For the relevant combinations of crop types and locations the number of seeds per pod varied from 12 to 21 and simulated pod density did not limit seed set.

4. Discussion

and conclusions

4.1. Standard potential rape

seed yield of winter oilseed

This paper discusses the options for increasing the potential seed yield of winter oilseed rape which has to be attained by introducting higher-yielding varieties. On locations with non-optimum growth conditions, yield increase has to be obtained in the first place by improvement of agronomic measures. The model with the cultivar Jet Neuf as a standard for quantifying most parameter values, reproduced well the observed average yield level of this cultivar for the central part of the Netherlands, grown under near optimal growth conditions. 4.2. Light distribution

and canopy photosynhesis

Pod angle distribution and the clustering of pods were analysed in this study, as they affect light distribution within the canopy and canopy photosynthesis. As the spatial arrangement of pods of current varieties are not well known, the spherical pod angle distribution is used as a reference and no clustering of pods was assumed (Thompson and Hughes, 1986; Spitters et al., 1989).

The results indicate the potential increase in canopy photosynthesis and LUE that can be attained by a change in light distribution in the canopy resulting from more erect pods and increased clustering. Erect clustered pods above a leaf layer may increase canopy phtosynthesis by maximally lo%, compared to non-clustered pods with a spherical angle distribution. 4.3. Options for increasing

seed yield

4.3.1. Sink and source capacity for seed filling Yield components such as pod density, seeds per pod and mean seed weight have been studied extensively in relation to seed yield (Geisler and Henning, 1981; Mendham et al., 1981a; Thompson and Hughes, 1986; Sierts et al., 1987; Pouzet et al., 1988; Grosse et al., 1992a). Because of the compensatory effects among these yield components, their usefulness in breeding programmes as selection criteria for yield improvement is limited (Thurling, 1991; Grosse et al., 1992a). This study suggests that the compensatory effect between seed density and mean seed weight can be broken by simultaneously increasing sink and source capacity for seed filling (Fig. lb and c). This provides the best prospects for increasing seed yield. 4.3.2. Single crop traits Various options for increasing seed yield potential of winter oilseed rape have been suggested (Table I). In this study a number of options were ranked according to their effect on seed yield, by analysing the effect of relative changes in crop characters on potential seed yield. Higher simulated seed yields resulted from (in descending order of improtance): later maturity (Richards, 1991; Marshall, 1991); increased seed set (Keiller and Morgan, 1988; Habekotte, 1993); small petals or apetalous flowers (Mendham et al., 1981a; Yates and Steven, 1987; Rao et al., 1991); increased potential growth rate of individual pods (Habekotte, 1993); early flowering with retention of the duration of the total growth period @hurling, 1991); and clustered erect pods (Mendham et al., 1984; Thompson and Hughes, 1986) (Fig. 2 and Table 5). In the simulations, pod densities did not limit seed set and will not be a limiting factor until the maxi-

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

mum number of seeds per pod is reached (in this study 35, Habekotte, 1996a), which varies from 18 to 35 depending on variety (Pechan and Morgan, 1983; Mendham et al., 1984). This cultivar-specific maximum determines whether increased seed densities can be obtained with moderate numbers of pods, as has been shown by Mendham et al. (1981a>, or whether pod densities have to increase concomittant1y. Accelerated leaf area development until onset of flowering (Thurling, 1974; Mendham et al., 1981a; Mendham et al., 1981b; Thompson, 1983; Grosse et al., 1992a; Leon, 1993) hardly increased light absorption, total dry matter production and seed yield (range l-2%, option 1, Table 6). A simulated maximum LAZ of about 3 m2 m-’ sufficed for near-maximum light absorption during the critical period for seed set and seed filling (Fig. 31, assuming that light absorption during the seed filling phase is similar to that at maximum leaf area (Mendham et al., 198 la; Habekotte, 1996b). However, data from Grosse (1989) for ten cultivars grown in three growing seasons, showed that LAI,,,,, (range: 2.4-5.4 correlated positively with leaf area index duration after flowering (LAID), which in turn was positively correlated with total dry matter production from the onset of flowering until maturity and with seed yield. This could indicate that at higher LA&,,,, leaves are productive longer, which may increase total dry matter production and seed yield. The possible physiological effects of LAZ,,, on leaf area index duration and of the latter on canopy photosynthesis were not included in LINTUL-BRASNAP. Thus, this study allows no conclusions on the relative effect of LA&,, on seed yield via its effect on LAID. The explorations indicate that a modified development pattern is highly desirable to increase seed yields. If maturity of winter oilseed rape is delayed, harvest may overlap with that of cereals and some farmers may lose interest in growing it (Mendham et al., 1981a; Almond et al., 1986; M.J.J. Pustjens, Van der Have, pers. comm., 1995). Farmers of large farms however, find it an advantage to have differences in maturity to match the harvesting capacity of the farm, and to reduce yield losses by harvesting crops close to maturity (F. Grosse, Saaten Union, pers. comm, 1995). In terms of climate, late maturity is only desirable

123

at locations with minimal risk of water shortage during late stages of crop growth (Thurling, 1991). Early flowering may result in flowers being damaged by frost in early spring (Merrien and Pouzet, 1988; Thurling, 1991; Richards, 1991). However, some damage to early flowers may be compensated by later flowers, therefore, early flowering seems to be a feasible option at locations with occasional frosts in late spring and will probably not lead to lower average seed yields (Richards, 1991; M.J.J. Pustjens, Van der Have, pers. comm., 1995). Crop characters such as petal size or apetalous flowers, time of maturity, clustering and pod-angle distribution of pods and time of flowering are easily recognizable and directly applicable in breeding programmes (M.J.J. Pustjens, Van der Have, pers. comm., 1995). Other crop characters such as the response of seed set to the availability of assimilates during the critical period of seed set and potential growth rate of the seeds cannot be derived directly from field observations and are not recorded in current breeding programmes (M.J.J. Pustjens, Van der Have, pers. comm., 1995). Seed set is related to seed density, which is easily calculated from seed yield and mean seed weight. Selection for seed density excludes the compensation effect between pod density and seed set per pod and is therefore more suitable as a selection criterion. It can only be used however during the late stages of selection when lines are sown in plots (Thompson and Hughes, 1986). Studies in spring rape (Chay and Thurling, 1989) suggest that pod length may be a suitable character for early-generation selection for seed density and seed yield. Promising results were obtained crossing a selection with long pods with an inbred line with short pods, to combine the long pod character with an increase in seeds per pod and with minimum reduction in pod density (Chay and Thurling, 1989; Leon and Becker, 1992). Potential seedgrowth rate determines final weight of seeds under sink-limited growth conditions in combination with the duration of the seed-filling period, which may be approximated by the duration of the period from the end of flowering until maturity. Thus, indications of potential seed growth rates may be obtained under conditions with limited seed set, which can be created experimentally. This paper contributed to the evaluation of the

124

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

effect of apetaly and reduced petal size on final seed yield in terms of improved availability of incoming light for crop growth and seed set as is also discussed by others based on experimental research (Yates and Steven, 1987; Rao et al., 1991; Fray et al., 1996). However, the change in floral morphology may reduce cross pollination and needs further attention in hybrid breeding (Buzza, 1995; M.J.J. Pustjens, Van der Have, pers. comm., 1995). Genetic variation has been observed or is probable for the crop characters considered (Grosse et al., 1992a; Leon and Becker, 1992; A.P. Sorensen, Cebeco, pers. comm., 1995). Moreover, there is an enormous reservoir of genetic diversity for the improvement of winter oilseed rape, in its own gene pool and in that of genetically related Brassica species and therefore genetic variability will not easily limit the scope for change in these crop characters. 4.3.3. Ideotypes for high seed yields Combination of physiological insight and experimental evidence may lead to a model of the ideal crop morphology, often called the ‘ideotype’ (Sedgley, 1991; Diepenbrock and Becker, 1992). This study suggests that ideotype characters be combined with the aim of simultaneously increasing sink and source capacity for seed filling, provides best prospects for yield increase. The effect of proposed crop types on potential seed yield was quantified for six locations in Western Europe and showed higher values of seed yield, light absorption, light-use efficiency and harvest index. The latter three were indicated as weak yield determinants of winter oilseed rape compared to winter wheat (Habekotte, 1997d). The breeding value of the formulated crop types further depends on ranges of genetic variability and genetic control of single or combined characters, which need further research (Diepenbrock and Becker, 1992; Mahon, 1983). Moreover, higheryielding varieties should also include other characteristics related to crop management, e.g. higher seed yields may demand increased nitrogen application and varieties should be resistant to lodging as was found for winter wheat (Austin et al., 1980; Spiertz and van Keulen, 1980; de Vos and Sinke, 1981). Cold tolerance, resistance to diseases and seed retention also need further attention, if full advantage is to be taken of the improved yield potential.

Acknowledgements The author wishes to thank Prof. R. Rabbinge, Prof. H. van Keulen, Ir. W.J.M. Meijer, Dr. A. Haverkort, Dr. F. Grosse, Dr. A. Elings, Dr. M. van Ooijen, Prof. F.W.T. Penning de Vries for valuable comments on the manuscript, Ir. M.J.J. Pustjens and A.P. Sorensen MSc. for the valuable discussions concerning this study and J. Burrough-Boenisch for correcting the English.

References Almond, J.A., Dawkins, T.C.K., Askew, M.F., 1986. Aspects of crop husbandry. In: Scarisbrick, D.H., Daniels, R.W. (Eds.), Oilseed rape. Collins, London, 127-175. Austin, R.B., Bingham, J., Blackwell, R.D., Evans, L.T., Ford, M.A., Morgan, CL., Taylor, M., 1980. Genetic improvements in winter wheat yields since 1900 and associated physiological changes. J. Agric. Sci., Camb. 94, 675-689. Buzza, G.C., 1995. Plant breeding. In: Kimber, D.S., McGregor, D.I. (Eds.1, Brassica Oilseeds, Production and Utilization, CAB International, Wallingford, 153-175. Boer, D.J. de, Langenhuysen, L., 1985. Toetsing en detaillering van een simulatiemodel van het productiepatroon van winterkoolzaad (Brassica napus L., Jet Neuf). Doctoraal verslag Theoretische Productie Ecologic, Landbouwuniversiteit Wageningen, 75 pp. Chay, P., Thurling, N., 1989. Identification of genes controlling pod length in spring rapeseed, Brassica napus L., and their utilization for yield improvement. Plant Breed 103, 54-62. Diepenbrock, W., Becker, H.C., 1992. General introduction. In: Diepenbrock, W., Becker, H.C. (Eds.), Physiological Potentials for Yield Improvement of Annual Oil and Protein Crops. Adv. Plant Breed 17 (1992) 1-18. Donald, C.M., 1968. The breeding of crop ideotypes. Euphytica 17, 385-403. Engqvist, G.M., Becker, H.C., 1993. Correlation studies for agronomic characters in segregating families of spring oilseed rape (Brassica napus). Heriditas 118, 211-216. Fray, M.J., Evans, E.J., Lydiate, D.J., Arthur, A.E., 1996. Physiological assesment of apetalous flowers and erectophile pods in oilseed rape (Brassica napus). J. Agric. Sci., Camb. 127, 193-200. Geisler, G., Henning, K., 1981. Untersuchungen zur Ertragsstruktur von Raps (Brassica napus L.1. II Die generative Entwicklung der Rapspflanze in Abhangigkeit von der Bestandesdichte. Bayer. Landw. Jahrb. 58, 203-211. Goudriaan, J., 1988. The bare bones of leaf-angle distribution in radiation models for canopy photosynthesis and energy exchange. Agric. and Forest Meteorol. 43, 155-169. Grosse, F., 1989. Untersuchungen zur Ertragsbildung und Ertragsstruktur in einem Winterrapssortiment. Dissertation zur

B. Habekotte’/Field

Crops Research 54 (1997) 109-126

Erlangung des Doktorgrades des Agranvissenschaftlichen Fakultlt der Christian-Albrechts-Universifat zu Kiel, 141 pp. Grosse, F., Leon, J., Diepenbrock, W., 1992a. Ertragsbildung und Ertragsstruktur bei Winterraps (Brassica napus L.). 1. Genotypische Variabilidt. J. Agron. and Crop Sci. 169, 70-93. Grosse, F., Leon, J., Diepenbrock, W., 1992b. Ertragsbildung und Ertragsstruktur bei Winterraps (Brassica napus L.). 2. Vergleich zwischen Eltemlinien und deren Fl- und F2Generationen J. Agron. and Crop Sci. 169, 94-103. Habekotte, A., 1989. Invloed van het zaaitijdstip op opbrengst en ontwikkeling van winterkoolzaad en granen. Flevobericht 302, 85 PP. Habekottt, B., 1993. Quantitative analysis of pod formation, seed set and seed filling in winter oilseed rape (Brassica napus L.) under field conditions. Field Crops Res. 35, 21-33. Habekotte, B., 1996. Winter oilseed rape: analysis of yield formation and crop type design for higher yield potential. Ph.D. Thesis, Wageningen Agricultural University, 156 pp. Habekotte, B., 1997a. Description, parameterization and user guide of LINTUL-BRASNAP 1.1. A crop growth model of winter oilseed rape (Brussica napus L.). Quant. Appr. Syst. Anal, 9, AB-DLO/PE, Wageningen, 66 pp. (in press). Habekottt, B., 1997b. A model of the phenological development of winter oilseed rape (Brassica napus L.) (submitted to Field Crops Res.). Habekotte, B., 1997~. Evaluation of seed yield determining factors in winter oilseed rape (Brassica napus L.) under optimal growth conditions by means of crop growth modelling (submitted to Field Crops Res.). Habekotte, B., 1997d. Identification of strong and weak yield determining components of winter oilseed rape compared to winter wheat (submitted to Eur. J. Agron.). Habekotte, B., Smid, H.G., 1992. Growth analysis and pod and seed set of winter oilseed rape (Brassica napus L.). Experimental results. CABO-report 166, Wageningen, 94 pp. Keiller, D.R., Morgan, D.G., 1988. Distribution of 14carbon labelled assimilates in flowering plants of oilseed rape (Brassica napus L.). J. Agric. Sci., Camb. 111, 347-355. Kraalingen, D.W.G. van, 1995. The FSE system for crop simulation, version 2.1. Quant. Appr. Syst. Anal., 1, AB-DLO/PE, Wageningen, 77 pp. Leon, J., 1993. Bedeutung der Ertragsphysiologie fur die Zllchtung von Raps. Fat Sci. Technol. 95, 283-287. Leon, J., Becker, H.C., 1992. Rapeseed (Brassica napus L.) Genetics. In: Diepenbrock, W., Becker, H.C. (Eds.), Physiological Potentials for Yield Improvement of Annual Oil and Protein Crops. Advances in Plant Breeding 17, 54-81. Link, W., Sttitzel, H., Leon, J., 1992. General conclusions. In: Diepenbrock, W., Becker, H.C. (Eds.), Physiological Potentials for Yield Improvement of Annual Oil and Protein Crops. Advances in Plant Breeding 17, 279-287. Mahon, J.D., 1983. Limitations to the use of physiological variability in plant breeding. Can. J. Plant Sci. 63, 11-21. Marshall, D.R., 1991. Alternative approaches and perspectives in breeding for higher yields. Field Crops Res. 26, 171- 190. Meissner, L.P., Organick, E.I., 1984. FORTRAN-77, featuring

structured

programming.

125 Addison-Wesley

Publishing

Com-

pany, 500 PP. Mendham, N.J., Shipway, P.A., Scott, R.K., 1981a. The effects of delayed sowing and weather on growth, development and yield of winter oilseed rape (Brassica napus L.). J. Agric. Sci., Camb. 96, 389-416. Mendham, N.J., Shipway, P.A., Scott, R.K., 1981b. The effects of seed size, autumn nitrogen and plant population on the response to delayed sowing in winter oilseed rape (Brassica napus L.). J. Agric. Sci., Camb. 96, 417-428. Mendham, N.J., Russell, J., Buzza, G.C., 1984. The contribution of seed survival to yield in new Australian cultivars of oil-seed rape (Brussica napus L.). J. Agric. Sci., Camb. 103, 303-316. Merrien, A., Pouzet, A., 1988. Principaux facteurs limitant les rendements du colza d’hiver dans les conditions fraqaises. Document de CETIOM, INRA, ENITA et CNRS, Paris, 16- 19. Meulen, A. van der, 1990a. Resultaten rassen- en bemestingsonderzoek 1989. Rapport Rijkswaterstaat, directie Flevoland, 1990 - 11 liw, 51 pp. Meulen, A., van der, 1990b. Resultaten rassen- en bemestingsonderzoek 1990. Rapport Rijkswaterstaat, directie Flevoland, 1990 - 33 liw, 53 pp. Meulen, A., van der, 1991. Resultaten rassen- en bemestingsonderzoek 1991. Rapport Rijkswaterstaat, directie Flevoland, 1991 - 22 lio, 48 pp. Meulen, A., van der, 1993. Resultaten rassen- en bemestingsonderzoek 1992. Rapport Rijkswaterstaat, directie Flevoland, 1993 - 8 lio, 48 pp. Pechan, P.M., Morgan, D.G., 1983. The use of radiography in studies of plant development in vivo. Planta 159, 476-481. Pouzet, A., Sauzet, J., Raimbault, J., Estragnat, A., Devineau, J., Ruer, C., Jung, L., Nesme, D., 1988. La conduite des cultures de colza d’hiver. Document de CETIOM, INRA, ENITA et CNRS, Paris, 143-158. Rao, M.S.S., Mendham, N.J., Buzza, G.C., 1991. Effect of the apetalous flower character on radiation distribution in the crop canopy, yield and its components in oilseed rape (Brassica napus). J. Agric. Sci., Camb. 117, 189-196. Rasmusson, D.C., 1991. A plant breeder’s experience with ideotype breeding. Field Crops Res. 26, 191-200. Richards, R.A., 1991. Crop improvement for temperate Australia: Future opportunities. Field Crops Res. 26, 141-169. Sedgley, R.H., 1991. An appraisal of the Donald ideotype after 21 years. Field Crops Res. 26, 93- 112. Sierts, H.P., Geisler, G., Leon, J., Diepenbrock, W., 1987. Stability of yield components from winter oilseed rape (Brassica napus L.1. J. Agron. Crop Sci. 158, 107-113. Spiertz, J.H.J., Keulen, H. van, 1980. Effects of nitrogen and water supply on growth and grain yield of wheat. Proceedings 3rd International Wheat Conference, May 22-June 3, 1980, Madrid: 595-610. Spitters, C.J.T., 1990. Crop growth models: Their usefulness and limitations. Acta Hart. 267, 349-368. Spitters, C.J.T., Keulen, H. van, D.W.G. van Kraalingen, 1989. A simple and universal crop growth simulator: SUCROS87. In: Rabbinge, R., Ward, S.A., Laar, H.H. van (Eds.), Simulation

126

B. Habekottt!/Field

Crops Research 54 (1997) 109-126

and system management in crop protection, Simulation Monographs 32, Pudoc, Wageningen, pp. 147-181. Stol, W., 1994. Synoptic and climatic data for agro-ecological research, the AB-MET database. Simulation Reports CABO‘IT, 37. AB-DLO, Wageningen, the Netherlands, 103 pp. Thompson, K.F., 1983. Breeding winter oilseed rape Brassica napus. In: Coaker, T.H. (Ed.), Advances in Applied Biology, Vol. VII, Academic Press, London, 2-86. Thompson, K.F., Hughes, W.G., 1986. Breeding and varieties. In: Scarisbrick, D.H., Daniels, R.W. (Eds.), Oilseed rape. Collins, London, 32-82. Thurling, N., 1974. Morphophysiological determinants of yield in

rapeseed (Brassica camprestris L. and Brassica napus L.). 1. Growth and morphological characters. Aust. J. Agric. Res. 25, 697-710. Thurling, N., 1991. Application of the ideotype concept in breeding for higher yield in the oilseed brassicas. Field Crops Res. 26, 201-219. Vos, N.M. de, Sinke, J., 1981. De vooruitgang in opbrengst by wintertarwe in de periode 1940-1980. Bedrijfsontwikkeling, 12: 615-618. Yates, D.J., Steven, M.D., 1987. Reflexion and absorption of solar radiation by flowering canopies of oil-seed rape (Brassica napus L.). J. Agric. Sci., Camb. 109, 495-502.