Yield and yield components of oilseed rape grown after different preceding crops and in different crop rotations
Olaf Christen and Klaus Sieling
Institute of Crop Science and Plant Breeding, Christian-Albrechts-University, 24118 Kiel, Germany
christen@pflanzenbau.uni-kiel.de
sieling@pflanzenbau.uni-kiel.de
Abstract
In many areas of the world oilseed rape is grown in short rotations. Information, however, about the response of oilseed rape to different preceding crops and/or crop rotation is scarce. To determine the effect of different preceding crops and crop rotations on the seed yield and yield components of oilseed rape, two experiments were conducted at the Hohenschulen experimental station in Kiel, NW Germany. Additional factors included the nitrogen fertilization, fungicide application and genotypes. The results reported herein are based upon the harvest years 1988 to 1998.
Averaged over the different rotations and husbandry treatments, the seed yields in the 11 experimental years varied between 2.71 t ha-1 and 4.58 t ha-1. In contrast, the effect of the different husbandry treatments was smaller. Averaged over 11 years, the fungicide application caused a yield increase of 0.37 t ha-1. The highest grain yields of 3.60 t ha-1 to 3.86 t ha-1 occurred when oilseed rape was directly following peas. Low yields between 3.18 t ha-1 to 3.24 t ha-1 were obtained when oilseed rape was grown in monoculture or directly after oilseed rape. In rotations with oilseed rape following a preceding cereal crop (wheat or barley), the seed yields averaged between 3.35 t ha-1 and 3.61 t ha-1 depending on the crop grown two years before in the Hohenschulen rotation experiment. In general, the yields of oilseed rape increased with the length of the rotation and the length of the break between two oilseed rape crops. The yield component number of seeds m-2 was affected by the previous cropping accordingly, whereas the thousand seed weight did not respond to the cropping history. Based upon disease assessments in the rotation experiment, we argue that an increase in the incidence of fungal diseases has considerably contributed to the yield decrease of oilseed rape in short rotations. Differences in the mineralisation of nitrogen after the various preceding crops were less important for the explanation of yield differences.
Keywords: Preceding crops, nitrogen fertilization, fungicide application
Introduction
The proportion of arable land cropped with oilseed rape has increased considerably in Germany in recent years. Extensive research has been conducted on the effect of different preceding crops including oilseed rape on the grain yield of a following cereal crop (e.g. Christen et al. 1992, Christen and Sieling 1993). Information, however, on the effect of different crop rotations or preceding crops on the seed yield of oilseed rape is scarce. Based upon survey data from former East Germany, Möller and Makowski (1977) report seed yields of oilseed rape grown after a clover dominated pasture of 2.36 t ha-1. Oilseed rape following a cereal crop yielded 2.19 t ha-1. If the crop rotation is also taken into account, the yield differences between a favourable and an unfavourable preceding crop tend to increase even further. George et al. (1985) report an averaged yield of oilseed rape following directly rapeseed of 1.9 t ha-1, whereas oilseed rape in a four- or five year rotation produced a seed yield of 2.48 t ha-1. The results of this study were again obtained in a survey of practical farms. Contrasting results are reported from rotation experiments with oilseed rape. In an experiment described by Kübler (1988), the seed yield of oilseed rape was unaffected by the crop rotation, regardless of a two, three or four year break. This results concurs with conclusions reached by Gonet and Ploszynska (1987) working in Poland. In their experiments even oilseed rape grown in monoculture did not show a negative yield response compared with oilseed rape grown in a four course rotation.
The objective of this research was to quantify the effect of different preceding crops and crop rotations and the interaction with different fertilizer and fungicide applications on the seed yield of oilseed rape. For this purpose results from different field experiments at the Hohenschulen Experimental station in NW Germany near Kiel are presented. Following factors were tested in the Hohenschulen rotation experiment: Crop rotation (including different preceding crops), fungicide application and nitrogen dressing (Tab. 1).
Table 1. Husbandry treatments including different fertilizer treatments and fungicide strategies from the Hohenschulen rotation experiment
|
Abbreviation |
Amount |
Distribution |
Fungicide application |
|
|
(kg N ha-1) |
(kg N ha-1) |
|
|
N-, F+ |
180 |
80/100 |
Yes |
|
N+, F+ |
220 |
120/100 |
Yes |
|
N+, F- |
220 |
120/100 |
No |
The field experiment was originally laid out to compare 15 different crop rotations, of which 13 are reported in this paper, since the two others are without oilseed rape. The remaining 13 rotations included the crops winter wheat, winter barley, rapeseed, peas and oats, and ranged from monoculture to five course rotations, summarised in Table 6. According to the principles of crop rotation experiments each single component of these 15 rotations was present every season, i.e. 45 components randomly arranged in one block. With a total of three blocks this accounted for 135 plots in the entire field trial. The results reported herein are based on a total of 51 plots oilseed rape each year. The design was also suitable for comparing the influence of different preceding crops on a subsequent oilseed rape crop, because oilseed rape either followed wheat (rotations 4, 6, 8, 10, 12 and 15) or rapeseed (rotations 1, 6, 10 and 11) as a direct preceding crop, or wheat, oilseed rape or peas the year before. This analysis is restricted to the above mentioned preceding crop combinations, since it allows a quantification of the interaction between the direct preceding crop (R-1) and the crop which had been grown on that particular plot two years before (R-2).
Table 2. Factors and factor levels of the preceding crop combination experiment
|
1. Preceding crop combination |
|||||||
|
|
Pre-preceding crop |
Preceding crop |
|
||||
|
P1 - |
oilseed rape |
winter barley |
(R-B) |
||||
|
P2 - |
oilseed rape |
winter wheat |
(R-W) |
||||
|
P3 - |
winter wheat |
winter barley |
(W-B) |
||||
|
P4 - |
winter wheat |
winter wheat |
(W-W) |
||||
|
2. Cultivars |
3. Amount and application time of the |
||||||
|
|
|
Date |
Beginning of growth |
Stem elongation |
|||
|
C1 - |
Bristol |
N1 - |
40 |
40 |
80 |
||
|
C2 - |
Express |
N2 - |
80 |
0 |
80 |
||
|
C3 - |
Falcon |
N3 - |
40 |
80 |
120 |
||
|
C4 - |
Ramses |
N4 - |
80 |
40 |
120 |
||
|
C5 - |
Wotan |
N5 - |
80 |
80 |
160 |
||
|
C6 - |
Zeus |
N6 - |
40 |
120 |
160 |
||
|
|
|
N7 - |
120 |
80 |
200 |
||
|
|
|
N8 - |
200 |
0 |
200 |
||
Further details on the experimental site and design have been published elsewhere (Christen et al. 1992; Christen and Sieling 1993, 1995 Sieling and Christen 1997; Sieling et al. 1997).
The main emphasis in the preceding crop combination experiment was to quantify the interaction between different preceding crop combinations and the nitrogen dressing as well as the choice of variety. The factor levels of this field trial are given in table 2. In both experiments standard cultural practices were employed. Important for the interpretation is the fact, that all oilseed rape plots were planted on the same date, regardless of the preceding crop in order to avoid any interactions due to different seed dates.
Hohenschulen Rotation experiment
Highly significant differences in the seed yield of oilseed rape in the Hohenschulen rotation experiment were caused by the weather in the experimental years (Y), the various rotations (R) and the husbandry treamtents (H) as well as by the interaction between years vs. husbandry treatments (Y x H). In contrast, the interaction between years vs. rotations (Y x R) and rotation vs. husbandry treatments (R x H) had no significant effect on the seed yield of oilseed rape (Tab. 3).
Table 3. Analysis of variance of rotation and husbandry effects on oilseed rape (Hohenschulen rotation experiment, 1988 to 1998).
|
Source |
Df |
Mean square |
F value |
|
Year (Y) |
10 |
1332.40 |
63.70*** |
|
Rotation (R) |
16 |
81.07 |
4.35*** |
|
Husbandry (H) |
2 |
867.61 |
41.48*** |
|
Y x R |
160 |
23.31 |
1.11 |
|
Y x H |
20 |
58.83 |
2.81*** |
|
R x H |
32 |
27.31 |
1.31 |
|
Error |
312 |
20.91 |
|
*** significant at the 0.001 level of probability
Table 4. Seed yield (t ha-1) of winter oilseed rape with different husbandry treatments (Hohenschulen rotation experiment, 1988 to 1998)
|
Year |
N-, F+ |
N+, F+ |
N+, F- |
Mean |
|
1988 |
4.04 |
4.03 |
3.90 |
3.99 |
|
1989 |
3.17 |
3.38 |
3.48 |
3.34 |
|
1990 |
3.20 |
3.01 |
2.76 |
2.99 |
|
1991 |
2.73 |
2.62 |
2.77 |
2.71 |
|
1992 |
3.64 |
3.66 |
3.16 |
3.49 |
|
1993 |
3.94 |
3.96 |
3.36 |
3.76 |
|
1994 |
3.76 |
3.94 |
3.30 |
3.67 |
|
1995 |
4.81 |
4.74 |
4.18 |
4.58 |
|
1996 |
3.47 |
3.39 |
2.72 |
3.19 |
|
1997 |
3.35 |
3.39 |
2.91 |
3.22 |
|
1998 |
3.57 |
3.41 |
2.95 |
3.31 |
|
Mean |
3.61 |
3.59 |
3.22 |
|
LSD p=0.05 for years = 0.17
LSD p=0.05 for husbandry treatments = 0.09
LSD p=0.05 for interaction years x husbandry treatments= 0.30
The different weather patterns in the 11 years included in this experiment caused considerable differences in the seed yield (Tab. 4). Averaged over the other treatments the highest yield of almost 4.6 t ha-1was obtain in the harvest year 1995. The lowest yield with 2.71 t ha-1 occurred in 1991. In all 11 experimental years, the differences between the two nitrogen fertilizer levels (N-,F+ and N+,F+) were small and of no agronomic significance. The differences in the seasonal environment also affected the response of oilseed rape to the different husbandry treatments in the Hohenschulen rotation experiment. The effect of the fungicide treatment, which, averaged over the experimental years, caused a yield increase of 0.37 t ha-1, occurred in 8 out of 11 years. The differences were not consistent. Nitrogen dressing, however,was less important, which concurs with results from the preceding crop combination experiment.
The effect of the two directly preceding crops and the preceding crop combination on the seed yields based on the single plant analysis is given in Table 5. Since the interaction year x preceding crop was not significant, only the average figures for the harvest years 1988 and 1989 are given. Oilseed rape following winter wheat yielded slightly more than following oilseed rape, but the difference was not statistically significant. The highest seed yield, 694 g m-2, averaged over the two experimental years, was obtained from the treatment with oilseed rape following peas - wheat (PW). The smaller seed yield, 371 g m-2, was obtained when oilseed rape followed two consecutive years of oilseed rape (RR). A comparison of the oilseed crops grown directly after wheat with those following oilseed rape reveals the effect of the preceding two crops. Regardless of the directly preceding crop, oilseed rape grown after peas two seasons before had a greater seed yield than crops with wheat or oilseed rape two years before.
Table 5. Effect of the interaction of the different preceding crops on the seed yield (g m-2) of oilseed rape (Hohenschulen rotation experiment, average of 1988-1989)
|
|
Preceding crop (R-1) |
|
|
|
Pre-previous crop (R-2) |
Wheat |
Oilseed rape |
Mean |
|
Wheat |
415 |
490 |
453 |
|
Oilseed rape |
423 |
371 |
387 |
|
Peas |
694 |
483 |
589 |
|
Mean |
511 |
448 |
|
|
LSDp=0.05
Preceding crops (R-1)= 89 |
|||
The magnitude of the seed yield response of oilseed rape to the different crop rotations as well as the different preceding crops is given in Table 6. The percentage of rapeseed in the different rotations had a profound effect on the seed yield. Oilseed rape grown in monoculture yielded 3.24 t ha-1. In contrast, rotations with only 20 or 25 % of rapeseed (rotations 10, 11 or 12) had a seed yield of between 3.59 t ha-1 to 3.67 t ha-1, respectively. A similar trend is shown in the rotations with 50 percentage of oilseed rape like rotation 2 and 3. However, the effect of the percentage of oilseed rape was mitigated by the influence of the direct preceding crop. Therefore, the highest yields always occurred in the crop rotations with oilseed rape following directly after peas (rotation 7 and 13). On the other hand, oilseed rape suffered a yield decrease, when directly grown after rapeseed (rotations 4, 8 and 9). The comparison of the first oilseed rape in rotation 9 with the oilseed rape in rotation 7 reveals an interaction between the direct preceding crop and the percentage of rapeseed in the rotation. The positive effect of the peas occurred in the rotation with a lower percentage of rapeseed.
Table 6. Seed yield (t ha-1) of winter oilseed rape grown in different crop rotations (Hohenschulen rotation experiment, average of the harvest years 1988 to 1998)
|
Crop rotation |
Crops in rotation
Seed yield t ha-1 |
Percentage of rapeseed (%) |
||||
|
1. |
Monoculture 3.24 |
|
|
|
|
100 |
|
2. |
Rapeseed 3.35 |
Wheat |
|
|
|
50 |
|
3. |
Rapeseed 3.50 |
Barley |
|
|
|
50 |
|
4. |
Rapeseed 3.24 |
Rapeseed 3.18 |
Wheat |
|
|
66 |
|
5. |
Rapeseed 3.51 |
Wheat |
Barley |
|
|
33 |
|
6. |
Rapeseed 3.41 |
Wheat |
Wheat |
|
|
33 |
|
7. |
Rapeseed 3.86 |
Wheat |
Peas |
|
|
33 |
|
8. |
Rapeseed 3.49 |
Rapeseed 3.42 |
Wheat |
Wheat |
|
50 |
|
9. |
Rapeseed 3.59 |
Rapeseed 3.41 |
Wheat |
Peas |
|
50 |
|
10 |
Rapeseed 3.61 |
Wheat |
Peas |
Wheat |
|
25 |
|
11. |
Rapeseed 3.59 |
Wheat |
Oats |
Barley |
|
25 |
|
12. |
Rapeseed 3.67 |
Wheat |
Peas |
Wheat |
Barley |
20 |
|
13. |
Rapeseed 3.43 |
Wheat |
Peas |
Rapeseed 3.60 |
Wheat |
40 |
LSD p=0.05 for rotations = 0.22
Since the number of plants m-2 was not significantly affected by the direct previous cropping, the differences in single plant yield were responsible for the small differences in seed yield (Table 7). The major reason for the even not significant, but pronounced differences was the seed yield of the branches 4 ff which was greater in plants grown after wheat as compared to oilseed rape. With an almost constant number of seeds per pod as well as mean seed weight in the respective branches, the greater seed yield in higher category branches (4 ff) was associated solely with more pods at these branches. In contrast, in the cases of the different crop combinations, the number of plants m-2 was more affected by previous cropping (Table 8). The fewest plants m-2 were observed in the treatment following peas - wheat (PW) and wheat - wheat (WW), i.e. treatments without oilseed rape in the cropping history for the last two years. All other treatments which had been cropped with oilseed rape at least once during the two previous years, had more plants m-2, perhaps due to volunteers, although the differences were not statistically significant.
Table 7. Effect of different preceding crops on yield components and fungal diseases of oilseed rape (Hohenschulen rotation experiment, average of 1988-1989)
|
|
Preceding crop |
|
|
|
|
Wheat |
Oilseed rape |
LSDp=0.05 |
|
Number of plots |
36 |
24 |
|
|
No. of plants m-2 |
73 |
80 |
ns* |
|
No. of fertile branches |
4.5 |
4.0 |
ns |
|
Seed yield main branch (g) |
4.0 |
3.7 |
ns |
|
Seed yield branches 1-3 (g) |
2.9 |
2.3 |
ns |
|
Seed yield branches 4 ff (g) |
3.8 |
2.0 |
ns |
|
Single plant yield (g)
|
10.7 |
8.0 |
ns |
|
No. of pods on main stem |
47.8 |
43.4 |
ns |
|
No. of pods on branches 1-3 |
35.3 |
29.6 |
ns |
|
No. of pods on branches 4ff |
43.9 |
25.6 |
ns |
|
No. of pods per plant
|
127 |
99 |
ns |
|
No.of seeds per pod on main stem |
19.8 |
19.5 |
ns |
|
No.of seeds per pod on branches 1-3 |
17.4 |
16.5 |
ns |
|
No. of seeds per pod on branches 4ff
|
17.3 |
16.0 |
ns |
|
Mean seed weight on main stem (mg) |
4.18 |
4.28 |
ns |
|
Mean seed weight on branches 1-3 (mg) |
4.13 |
4.15 |
ns |
|
Mean seed weight on branches 4ff (mg)
|
4.15 |
4.20 |
ns |
|
Stem canker rating (0-9) |
1.6 |
2.5 |
0.3 |
|
Verticillium rating (0-9) |
1.7 |
2.0 |
ns |
* - not significant at p £ 0.05
Oilseed rape following directly peas - wheat (PW) had the greatest single plant yield, as well as the greatest seed yield on all branch categories. In the treatments following oilseed rape the differences in seed yield were smaller. The oilseed rape following peas - rapeseed (PR) only produced greater seed yields through a greater seed yield in the branches 4 ff category. In the oilseed rape grown after peas - wheat (PW) these differences were associated with an exceptionally large number of pods per plant, which was significantly different to most other treatments. The largest relative difference, however, also occurred on the branches of the higher order categories (4 ff), whereas the differences on the main stem and the branches 1-3 were smaller. A similar trend occurred in the oilseed rape grown directly after oilseed rape. In these treatments the relative yield difference between the rapeseed - rapeseed (RR) cropping sequence compared with the peas - rapeseed (PR) sequence increased in the higher order branch categories. This being the main reason for the differences in the seed yield, the effect of the previous cropping on the number of seed per pod as well as the mean seed weight in the different branch categories was smaller.
Table 8. Effect of different preceding crop combinations on yield components and fungal diseases of oilseed rape (Hohenschulen rotation experiment, average of 1988-1989)
|
|
Preceding crop combination |
||||||
|
|
WW |
RW |
PW |
WR |
RR |
PR |
LSDp=0.05 |
|
Number of plots |
12 |
18 |
6 |
12 |
6 |
6 |
|
|
No. of plants m-2 |
68 |
85 |
66 |
73 |
80 |
86 |
ns |
|
No. of fertile branches
|
3.87 |
3.80 |
5.91 |
4.37 |
3.52 |
4.16 |
1.30 |
|
Seed yield main branch (g) |
3.76 |
3.46 |
4.94 |
3.97 |
3.10 |
3.98 |
ns |
|
Seed yield branches 1-3 (g) |
1.96 |
2.27 |
4.40 |
2.58 |
1.57 |
2.72 |
ns |
|
Seed yield branches 4 ff (g) |
1.16 |
2.77 |
7.48 |
1.80 |
1.46 |
2.86 |
ns |
|
Single plant yield (g)
|
6.88 |
8.50 |
16.8 |
8.35 |
6.13 |
9.58 |
ns |
|
No. of pods on main stem |
49 |
41 |
54 |
47 |
38 |
45 |
ns |
|
No. of pods on branches 1-3 |
29 |
28 |
49 |
36 |
21 |
32 |
15 |
|
No. of pods on branches 4ff |
17 |
31 |
83 |
26 |
17 |
33 |
ns |
|
No. of pods per plant
|
95 |
100 |
186 |
109 |
76 |
110 |
62 |
|
No. of seeds per pod on main stem |
19.3 |
19.7 |
20.3 |
19.8 |
18.5 |
20.2 |
ns |
|
No. of seeds per pod on branches 1-3 |
17.1 |
17.0 |
17.9 |
16.4 |
15.9 |
17.2 |
ns |
|
No. of seeds per pod on branches 4ff
|
16.4 |
18.4 |
17.0 |
15.9 |
15.8 |
16.3 |
ns |
|
Mean seed weight on main stem (mg) |
3.99 |
4.08 |
4.46 |
4.24 |
4.30 |
4.29 |
ns |
|
Mean seed weight on branches 1-3 (mg) |
3.93 |
4.08 |
4.37 |
4.07 |
4.19 |
4.18 |
ns |
|
Mean seed weight on branches 4ff (mg)
|
3.96 |
4.16 |
4.31 |
4.09 |
4.39 |
4.10 |
ns |
|
Stem canker rating (0-9) |
1.3 |
1.9 |
1.5 |
2.1 |
3.8 |
1.7 |
0.9 |
|
Verticullium rating (0-9) |
1.8 |
2.1 |
1.3 |
1.6 |
2.8 |
1.7 |
ns |
* - not significant at p £ 0.05
Preceding crop combination experiment
Similar yield relations as in the rotation field trial were obtained in the preceding combination crop experiment. If winter wheat was grown two years before, the yield of oilseed rape was similar after the direct preceding crops wheat or barley (Table 9). However, growing oilseed rape after only one break crop reduced the seed yield significantly. The extent of losses was strongly affected by the preceding crop. After the cropping sequence oilseed rape - wheat (R-W) only 3.12 t ha-1 and after oilseed rape - barley (R-B) 3.43 t ha-1 were observed compared with 3.77 t ha-1 following wheat - barley (W-B) and 3.71 t ha-1 following wheat - wheat (W-W). The number of seeds m-2 showed a similar pattern, whereas the thousand seed weight partly compensated for the reduced seed number. It was highest if oilseed rape was grown two years ago.
Table 9. Effect of the preceding crop combination on yield (t ha-1), thousand seed weight (g) and numbers of seeds m-2 of oilseed rape (Preceding crop combination experiment, 1993 and 1995)
|
Preceding crop combination |
Seed yield (t ha-1) |
Thousand seed weight (g) |
Number of |
|
Oilseed rape - winter barley |
3.43 |
4.95 |
69460 |
|
Oilseed rape - winter wheat |
3.12 |
4.93 |
63640 |
|
Winter wheat - winter barley |
3.77 |
4.76 |
79990 |
|
Winter wheat - winter wheat |
3.71 |
4.71 |
79450 |
|
LSD0.05 |
0.15 |
0.06 |
3290 |
The cultivars differed significantly in its yield potential (Table 10). Express (3.79 t ha-1) yielded 0.6 t ha-1 more than Falcon (3.18 t ha-1). Although no significant interaction between the cultivars and the preceding crop combination could be observed, some cultivars (e.g. Zeus) seemed to be more suitable to be grown after only a one year break than other (e.g. Ramses, Wotan). Comparing the effects of the directly preceding crops barley and wheat following oilseed rape (R-B, R-W), all cultivars yielded 0.4 t ha-1 less when following wheat, except in Zeus and Ramses which showed only negligible losses.
Table 10. Effect of the preceding crop combination on yield (t ha-1) of six oilseed rape varieties (Preceding crop combination experiment, 1993 and 1995)
|
|
Preceding crop combination |
|
|||
|
|
rape-barley |
rape-wheat |
Wheat-barley |
wheat-wheat |
Mean |
|
Bristol |
3.48 |
3.06 |
3.65 |
3.68 |
3.47 |
|
Express |
3.76 |
3.31 |
3.95 |
4.14 |
3.79 |
|
Falcon |
3.22 |
2.81 |
3.49 |
3.21 |
3.18 |
|
Ramses |
3.16 |
3.10 |
3.95 |
3.74 |
3.49 |
|
Wotan |
3.42 |
2.95 |
3.83 |
3.56 |
3.44 |
|
Zeus |
3.53 |
3.51 |
3.76 |
3.95 |
3.69 |
LSD0.05 for varieties = 0.19
LSD0.05 for variety by preceding crop interaction = n.s.
Table 11. Interaction between the preceding crop combination and N-fertilization on yield (t ha-1) of oilseed rape (Preceding crop combination experiment, 1993 and 1995)
|
|
Preceding crop combination |
|
|||
|
N fertilizer |
rape-barley |
Rape-wheat |
Wheat-barley |
Wheat-wheat |
Mean |
|
40/40 |
3.09 |
2.82 |
3.49 |
3.43 |
3.21 |
|
80/0 |
3.41 |
2.71 |
3.68 |
3.51 |
3.33 |
|
40/80 |
3.38 |
2.93 |
3.55 |
3.76 |
3.41 |
|
80/40 |
3.31 |
2.93 |
3.71 |
3.69 |
3.41 |
|
80/80 |
3.54 |
3.02 |
3.77 |
3.88 |
3.55 |
|
40/120 |
3.23 |
3.40 |
3.77 |
3.81 |
3.55 |
|
120/80 |
3.57 |
3.60 |
4.17 |
3.75 |
3.77 |
|
200/0 |
3.91 |
3.57 |
4.02 |
3.87 |
3.84 |
LSD0.05 for N fertilization = 0.22;
LSD0.05 for N fertilization by preceding crop interaction = 1.02
Averaged over all other treatments, increasing amounts of fertilizer-N increased the seed yield from 3.21 t ha-1 up to 3.84 t ha-1 (Table 11). Varying the distribution pattern within one amount had no effect. The lowest yields occurred after oilseed rape - wheat (R-W) when only 80 kg N ha-1 were applied. Significant interactions between N fertilization and the preceding crop combination or the cultivar did not occur. However, regarding again the seed yields of oilseed rape after oilseed rape - barley (R-B) and after oilseed rape - wheat (R-W), in some cases the timing of N application was important. Following wheat, applying the amount of 160 kg N ha-1 as 40/120 kg N ha-1 instead of 80/80 kg N ha-1, increased seed yield by 0.4 t ha-1, whereas following barley the reverse was observed. After barley, the first dressing should be increased in the 80 kg N ha-1 and 200 kg N ha-1 treatment.
The principal aim of this study was to quantify the effect of various crop rotations and different preceding crops o the seed yield of oilseed rape. A decrease in seed yield of oilseed rape grown in short rotations as well as following an unfavourable preceding crop confirms reports by Möller and Makowski (1977) and George et al. (1985). The relative magnitude of the response is comparable with results obtain for wheat and barley in the same experiment which have been already reported (Christen et al. 1992, Christen and Sieling 1993). However, compared with the survey data given by Möller and Makowski (1977) as well as George et al. (1985), the yield response is slightly lower in our experiment. One reason for this discrepancy might be the high input level used in the Hohenschulen rotation experiment. Additionally the structure of survey data does normally not allow to clearly distinguish between treatment factors. For example Möller and Makowski (1977) explain the differences in their experiments between a preceding clover pasture and a preceding cereal crop solely with differences in the seed date. Another factor which explains the small yield reponse of oilseed rape to different rotations is the fact, that we did not have club root caused by Plasmodiophora brassicae in our experiment (Jakobsen and Olson 1992). Our data, however, support the results by Kübler (1988), that oilseed rape only shows a only small response to the percentage in the crop rotation as long as no extreme treatments like monocultures or five course rotations are included. The opinion expressed by Gonet and Ploszynska (1987) that oilseed rape does not show a yield decrease when grown in monoculture is not supported by our results.
The results of the three compared husbandry treatments indicate that given the level of external input used in our experiments, mineralised nitrogen from crop residues did only contribute little to the yield differences. Based upon detailed analyses of the incidence and severity of diseases that were taken in two years of the experiment reported herein, we argue that the causes of the observed yield differencesare are a slight increase of stem canker (Phoma lingam syn. Leptosphaeria maculans) and verticillium wilt (Verticilium dahlie). However, it remains unclear whether the small differences are sufficient to explain the yield response. Kübler (1988) described an increase in the nematode population, but could not relate this increase to differences in the seed yield. Other experiments indicate, that diseases like Rhizoctonia solani might also contribute to yield differences in short oilseed rotations (Svensson and Lerenius 1987, Tahvonen et al. 1984).
Conclusions
Oilseed rape responses with a comparable magnitude to different preceding crops and crop rotations like cereal crops. As long as club root does not occur, it is possible to grow oilseed rape every third or even every second year without severe yield losses in seed yield compared to longer rotations. Since the nitrogen dressing did have only small effects on the yield response of oilseed rape to different preceding crops, diseases seem to be more important for the explanation of yield differences. Further studies should concentrate on the effect of different preceding crops and crop rotations on the incidence and severity of pests and diseases in short oilseed rotations.
Acknowledgement
The Research was funded by the Deutsche Forschungsgemeinschaft.
References
Christen, O., K. Sieling and H. Hanus, 1992: The effect of different preceding crops on the development, growth and yield of winter wheat. European Journal of Agronomy, 1, 21-28
Christen, O. and K. Sieling, 1993: The effect of different preceding crops on the development, growth and yield of winter barley. Journal of Agronomy and Crop Science, 171, 114-123
Christen, O. and K. Sieling, 1995: Effect of different preceding crops and crop rotations on yield of oilseed rape (Brassica napus L.). Journal of Agronomy and Crop Science 174, 265-271
George, W., W. Heidel and V. Meitzner, 1985: Zum Auftreten der Halsnekrose des Rapses im Bezirk Neubrandenburg unter besondere Berücksichtigung der Fruchtfolge. Nachrichtenblatt für den Pflanzenschutz in der DDR, 39, 237-239
Gonet, Z. and W. Ploszynska, 1987: Reaktionen von Raps auf periodischen und Daueranbau in Monokultur. Proceedings of the 7th Int. Rapeseed Congress, Poznan, Poland, 11-14 May 1987, 899-906
Jakobsen, H.L. and C.C. Olson, 1992: Progress in diseases in trials with different intervals between oilseed rape in crop rotations. 9th Danish plant protection conference, 27-37
Kübler, E., 1988: Wirkung enger und weiter Winterrapsfruchtfolgen. Berichte der Gesellschaft für Pflanzenbauwissenschaften, 1, 83-104
Möller, W. und N. Makowski, 1977: Ergebnisse beim Anbau von Winterraps nach Getreidevorfrüchten. Tag. Ber. Akad. Landwirtsch. Wiss, DDR, Berlin, 149, 69-76
Sieling, K. , O. Christen, B., Nemati and H. Hanus, 1997: Effects of previous cropping on seed yield and yield components of oilseed rape (Brassica napus L.). European Journal of Agronomy 6, 215-223.
Sieling, K., and O. Christen, 1997: Effect of preceding crop combination and N fertilization on yield of six oilseed rape cultivars (Brassica napus L.) European Journal of Agronomy 7, 301-306.
Svensson, C. and C. Lerenius, 1987: An investigation of verticillium wilt (Verticillium dahliae Kleb.) on oilseed rape. WPRS Bulletin, Working group "integrated control in oilseed rape". 30-34
Tahvonen, R., j. Hollo, A. Hannukkala and A. Kurppa, 1984: Rhizoctonia solani damping-off on spring turnip rape and spring rape (Brassica spp.) in Finland. J. of Agric. Science in Finland, 56, 143-154