Seed distance in relation to row distance: Effect on grain yield and weed biomass in organically grown winter wheat, spring wheat and spring oats

Field Crops Research 134 (2012) 144–152

Contents lists available at SciVerse ScienceDirect

Field Crops Research journal homepage: www.elsevier.com/locate/fcr

Seed distance in relation to row distance: Effect on grain yield and weed biomass in organically grown winter wheat, spring wheat and spring oats Ullalena Boström a,∗ , Lars Eric Anderson b , Ann-Charlotte Wallenhammar b a b

Swedish University of Agricultural Sciences, Department of Crop Production Ecology, P.O. Box 7043, SE 750 07 Uppsala, Sweden Rural Economy and Agricultural Societies/HS Konsult AB, P.O. Box 271, SE 701 45 Örebro, Sweden

a r t i c l e

i n f o

Article history: Received 22 September 2011 Received in revised form 30 May 2012 Accepted 1 June 2012 Keywords: Weed management Protein Cereals Seed rate Quality

a b s t r a c t Inter-row hoeing at wide row distance can form part of a weed management strategy in which mechanical and cultural tools are combined. Crop plant distribution is a key parameter for maintaining high grain yield levels and weed suppressive ability in such a system. The influence of seed distance in relation to row distance was studied in 10 field trials on spring wheat, spring oats and winter wheat. Post-emergence weed harrowing was applied at 12–12.5 cm row distance and inter-row hoeing at 24–25, 36–37.5 and 48–50 cm distance. Three seed distances within rows were applied: (1) Normal for the region, (2) reduced to 2/3 the normal, and (3) reduced to half the normal. To obtain normal seed distance at 12–12.5 cm row distance, 400, 525 and 600 germinable seeds m−2 were sown in winter wheat, spring oats and spring wheat, respectively. The highest grain yields were found after weed harrowing at row distance 12–12.5 cm; 5550 kg ha−1 of winter wheat, 3765 kg ha−1 of spring oats and 3105 kg ha−1 of spring wheat. Inter-row hoeing at the row distance 24–25 cm, while keeping seed distance constant, lowered grain yields of all crops by 12–16%, but had no significant influence on weed biomass. Increasing the row distance from 24–25 to 36–37.5 cm reduced grain yield by 450–460 kg ha−1 in winter wheat and oats, while a further increase to 48–50 cm reduced yield by an additional 450–520 kg ha−1 in all crops when averaged over seed distances. At row distance 24 cm and wider, grain yield of winter wheat and oats increased more when the seed distance was reduced from normal to 2/3 of normal than when reduced from 2/3 to half the normal distance. In inter-row hoed winter wheat and spring oats, weed biomass increased with increasing row and seed distance. Within the range of seed distances used in this study, a reduction to half of normal was considered as the most profitable when these crops are sown at row distances 24 cm or wider. There was no indication that seed distance in spring wheat should be changed when using wide row distances. Increased row and seed distances improved the content of gluten and reduced the content of starch in wheat grain, while protein content was increased in the grain of all crops . The response to treatments in terms of grain content of ergosterol was inconsistent between winter and spring wheat. © 2012 Elsevier B.V. All rights reserved.

1. Introduction A frequent consequence of the transition from conventional to organic farming is increasing weed infestation that develops into a major problem (Cavigelli et al., 2008; Corbin et al., 2010). This constitutes one of the primary constraints on production (Turner et al., 2007). In organic farming the use of herbicides is prohibited, so the weed control tools consist of cultural, biological and mechanical means (cf. Hatcher and Melander, 2003; Melander et al., 2005).

Abbreviations: Row distance, 12.0–12.5 cm (RD12 ), 24.0–25.0 cm (RD24 ), 36.0–37.5 cm (RD36 ) and 48.0–50.0 cm (RD48 ); Seed distance, Normal for the region (SD1 ), 2/3 of the normal (SD2/3 ), half the normal (SD1/2 ). ∗ Corresponding author. Tel.: +46 0 18671449; fax: +46 0 18 67 28 90. E-mail address: [email protected] (U. Boström). 0378-4290/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.fcr.2012.06.001

Combination of various measures into weed management strategies is essential for the economic profitability of organic production. In successful strategies, the crop rotation used is of fundamental significance (Albrecht, 2005), as is the sequence of crops in the rotation (Eyre et al., 2011). In long-term organic rotations, the need for weed management is higher when annual crops are the main component, while inclusion of a perennial forage crop may reduce populations of broadleaved weeds in the seedbank (Wortman et al., 2010). In stockless organic farming systems, the crop rotation is often based on annual crops, primarily cereals that tolerate physical weed management. When sown at narrow row distance, post-emergence weed harrowing may be rather efficient, providing 60–85% weed control after the first pass and 84–98% after three passes (Rasmussen et al., 2010). However the efficiency varies between, e.g. weed species (Lundkvist, 2009) and timing of

U. Boström et al. / Field Crops Research 134 (2012) 144–152

145

Table 1 Location of the experimental sites and range values of characteristics of topsoil (0–30 cm) in the trials of winter wheat, spring oats and spring wheat.

Number of trials Latitude Longitude pH (H2 O) Organic matter (%) Sand (%) Silt (%) Clay (%)

Winter wheat

Spring oats

Spring wheat

3 56◦ 3.5 N-59◦ 17 N 13◦ 58 E-15◦ 3.6 E 5.9–7.2 2.4–4.2 9–54 19–51 21–40

3 56◦ 11 N-59◦ 13 N 13◦ 45 E-15◦ 11 E 6.1–6.7 2.2–8.4 16–44 42–68 14–16

4 58◦ 21 N-59◦ 26 N 12◦ 38 E-15◦ 29 E 6.1–6.6 3–6.3 8–28 27–51 21–51

application (Rasmussen et al., 2008). Cultivation of cereals at wide row distances that allow inter-row hoeing provides an additional possibility for mechanical weed control. Inter-row hoeing can be carried out over a longer period of time than weed harrowing and is thereby more flexible to use. Besides serving as a weed management tool, hoeing may improve crop conditions by breaking soil crusts and aerating the soil (Leblanc and Cloutier, 2001) and soil cultivation may also enhance nitrogen mineralisation (Becker and Böhrnsen, 1994). The negative effects of physical weed management relate to the risk for damaging crop plants as well as the weeds and use of an optimal seed rate is essential (Rasmussen et al., 2008, 2009). Plant stand design is a key parameter for the outcome of weed suppression and grain yield quantity and quality (Kolb et al., 2010; Kristensen et al., 2008; Olsen and Weiner, 2007; Weiner et al., 2001). Seed rates recommended by seed companies are optimised for the row distance normally used, which in Sweden is 12–12.5 cm. Increasing the row distance without a seed rate reduction decreases the distance between individual seeds, thereby increasing intraspecific and inter-specific competition between plants. Based on common garden experiments, Håkansson (2003) describes the relationship between seed rate and row distance in relation to grain yields of cereals and weed incidence. A common assumption is that seed rate should be reduced when row distance is increased, although the optimal magnitude of the reduction is poorly known. The aims of the present study were to assess the influence of seed rate in relation to row distance on (i) grain yield quantity, (ii) grain quality and (iii) weed biomass and abundance in organically-grown spring wheat, spring oats and winter wheat that had been exposed to different mechanical weed control treatments. Inter-row hoed cropping systems sown with different sowing seed distances at 24–25 cm row distance or wider were mutually compared. In addition, at normal seed distance, a weed harrowed cropping system sown at 12–12.5 cm row-distance was compared with the inter-row hoed system sown at 24–25 cm row distance.

one experiment with oats and one with winter wheat were narrowleafed lupins and onions, respectively. The weed flora was dominated by annual weeds at seven sites while perennials dominated one site with oats and two with spring wheat. Both life-forms were dominated by dicotyledonous species. The organic fertilizer Biofer was applied at a rate of 60–80 kg N ha−1 to one experiment with spring wheat and two with winter wheat. One experiment with oats was fertilised with chicken manure applied at 10 tonnes ha−1 and one was fertilised with solid cattle manure applied at 20 tonnes ha−1 . The remaining four experiments were unfertilised. Eight sites had been organically farmed during 4–15 years and two sites with oats during one year. Each year, seed from the same lot was used for all experiments with each particular cereal species. The field experiments were conducted by the Field Experimental Divisions at the Rural Economy and Agricultural Society in the respective region. All spring-sown trials were ploughed in autumn before the start of the experiment and harrowed in spring just before sowing in April or May. The autumn-sown trials were ploughed and harrowed just before sowing in September or October. At the start of the experiment, 10 soil samples were taken from the 0–20 cm layer, pooled and analysed for soil particle size distribution, soil organic matter content and pH (H2 O). 2.2. Treatments and experimental design Cereals were sown at row distances of 12.0–12.5 cm (RD12 ), 24.0–25.0 cm (RD24 ), 36.0–37.5 cm (RD36 ) and 48.0–50.0 cm (RD48 ) (Table 2). With RD12 , seed rates normal for the region were used, i.e., 400 germinable seeds m−2 for winter wheat, 525 seeds m−2 for spring oats and 600 seeds m−2 for spring wheat. At wider row distances, seed rates were adjusted so that the distance between individual seeds in the row was either (1) the same as at 12 cm row distance (hereafter referred to as SD1 ), or (2) the distance between seeds in the row was reduced to 2/3 the normal (SD2/3 ), or (3) the seed distance in the row was halved (SD1/2 ).

2. Material and methods 2.1. Experimental sites and crop species A total of 10 field experiments were carried out in southern and central parts of Sweden during the period 2006–2008. In spring oats (Avena sativa L., cv. SW Sang), two experiments were performed in 2006 and one in 2007, while in spring-sown wheat (Triticum aestivum L., cv. SW Dacke), two experiments were performed in 2007 and two in 2008. In winter wheat (cv. SW Stava), two experiments were established in autumn 2006 and one in autumn 2007. Each crop species was sown in separate experiments at different locations (Table 1). At one experiment with winter wheat, one with oats and four with spring wheat, the previous crops were cloveror grass-clover leys. One experiment with winter wheat and one with oats were preceded by spring cereals. The previous crops in

Table 2 Percentage of normal seed rate sown at different row distances (RD) and seed distances within the row (SD) in weed harrowed or inter-row hoed cereals. Row distance

Weed management

Seed distance SD1 a %

RD12 RD24 RD36 RD48

c

Weed harrowed Inter-row hoed Inter-row hoed Inter-row hoed

b

100 50 33 25

SD2/3

SD1/2

– 75 50 38

– 100 67 50

a Index denotes seed distance; Normal for the region when seeding at 12–12.5 cm row distance (SD1 ), 2/3 of the normal (SD2/3 ), half the normal (SD1/2 ). b Normal seed rates sown at 12–12.5 cm row distance were 400, 525 and 600 germinable seeds m−2 in winter wheat, spring oats and spring wheat, respectively. c Index denotes row distance; 12–12.5 cm (RD12 ), 24–25 cm (RD24 ), 36–37.5 cm (RD36 ) and 48–50 cm (RD48 ).

146

U. Boström et al. / Field Crops Research 134 (2012) 144–152

Table 3 Range of mean air temperature (Tm ) and total rainfall (R) accumulated by month for each of the experimental years and the long-term average values (1961–1990). Month

January February March April May June July August September October November December a

Yeara 2006

2007

2008

◦

◦

◦

Tm ( C)

R (mm)

Tm ( C)

– – −5.3 4.5–5.3 11.1 15.7–16.3 19.6–20.4 16.5–17.5 14.2–15.6 8.7–11.3 4.2–5.1 4.4–5.4

– – 34 42–66 53–104 28–41 17–49 149–228 26–29 90–127 56–58 40–58

−0.3 −3.6 3.5 7.0–8.4 10.0–11.7 15.5–17.1 14.8–16.0 15.5–16.8 10.7–12.1 6.2–7.5 1.5–3.3 1.0–2.9

R (mm)

Tm ( C)

79 26 38 15–28 22–51 76–124 70–147 29–83 46–77 17–22 22–28 5–57

1.6–3.3 2.4–4.3 1.0–2.8 6.6–6.7 11.8–12.5 14.8–15.4 17.3–17.4 14.5–16.3 10.2–12.4 – – –

1961–1990 R (mm) 74–99 24–40 33–71 25–39 11–28 22–31 26–77 91–146 26–73 – – –

Tm (◦ C) −3.8–(−1.6) −3.6–(−1.6) −0.4–1 4.4–5.1 10.4–10.5 14.55–14.7 15.8–15.9 14.8–15.3 10.7–11.8 6.7–7.95 1.5–3.4 −2.1–(−0.1)

R (mm) 41–63 28–39 31–47 34–44 39–42 49–55 61–74 56–63 62–68 56–67 57–75 45–67

Range of data collected from 2 to 3 weather stations placed within 0–30 km from each field site.

The experimental design was a completely randomised block with four replicates in winter wheat and three replicates in springsown cereals. Due to the machinery used the plot size varied between 36 and 60 m2 . To avoid edge effects, only the inner plot parts were harvested, i.e., an area varying between 16 and 42 m2 .

starch and ergosterol were analysed by NIT using calibration solutions from Foss (Höganäs, Sweden). Between three and six weeks after the last weeding, biomass of weeds was estimated by cutting the weed plants at the soil surface and weighing them fresh. In each experimental plot, four sub-areas of 0.25 m2 were sampled. The sub-samples were pooled before statistical analyses.

2.3. Weed management 2.5. Meteorological background At sites where soil humidity allowed (i.e., one with winter wheat, two with oats and one with spring wheat) plots were weed harrowed once before crop emergence. After crop emergence, RD12 plots were weed harrowed twice, while plots sown at wider row distances were inter-row hoed twice. In winter wheat the first postemergence weeding was performed in the autumn and the second in the spring. Post-emergence weed management was performed when cereals had reached development stages BBCH 21 and BBCH 22–30. Pre- and post-emergence weed harrowing was performed with flex-tine harrows at a driving speed of 6–8 km h−1 . Inter-row hoeing was carried out with tractor hoes equipped with goosefoot shares. At time of the first hoeing the driving speed was 3 km h−1 and at the second hoeing 4–5 km h−1 .

The region has a cold temperate climate. Based on long-term averages (1961–1990), the mean annual temperature is 12.0 ◦ C and precipitation is 300 mm during the growing period for spring sown cereals, i.e. from April until September. The mean annual precipitation in the area is 600 mm. The ranges of monthly rainfall and air temperature during the experimental years are shown in Table 3, together with the long-term average values. Weather data were collected at 0–30 km distance from each site and was provided by the Swedish Meteorological and Hydrological Institute (SMHI) and The Rural Economy and Agricultural Societies. The long-term average values were obtained from two SMHI weather stations situated in the vicinity of the southern experimental sites and two situated in the vicinity of the northern sites.

2.4. Crop and weed assessments

2.6. Statistical analysis

Between six and eight weeks after the sowing of spring-sown crops, the number of plant shoots (including both main shoot and tillers) in a 2-m section of the row was counted and the exact position of the counted row section was marked. About two months later, the number of spikes or ears in the same marked section of plant row was counted. In winter wheat the shoots were counted in May. Attacks by diseases and pests were recorded separately for each experimental plot during the vegetation period. At time of harvest, the degree of lodging in each plot was visually estimated on a scale from 0 (erect crop) to 100 (crop totally flattened). Grain yields were measured with a plot combine harvester. Immediately after harvesting, a 1000 g subsample of grain was taken from each plot, and was sifted to remove waste materials, i.e., screenings. Grain yields reported are corrected to 15% moisture content. The grain was analysed for thousand kernel weight (TKW). Nitrogen content was analysed according to the near infrared transmittance (NIT) method (InfratecTM 1241 Grain Analyser, Foss, Denmark). The content of grain protein was calculated as N content × 6.25. In all experiments with spring-sown wheat and in two out of the three winter wheat experiments, the contents of gluten,

Regression analyses and analyses of variance for dependent variables were conducted separately for each crop species, with site and block within site as random factors, using the Mixed procedure (SAS Version 9.1, SAS Inst., Cary, NC). For SD1 at RD12 and RD24 , a separate analysis of data was made, with row distance as the single fixed factor. Data for row distances RD24 and wider were analysed with seed distance, row distance and their interaction considered as fixed factors. Since missing values occurred, the least squares means were calculated. Numbers of ears or panicles per shoot were square-root transformed to meet assumptions of homoscedasticity. Where Ftests were significant (P < 0.05), the Pdiff statement in the Mixed procedure of SAS was used for multiple comparisons of least square means. The relationships between the dependent variables grain yields or weed biomass and the independent variables row distance and seed rate were described by polynomial regression models fitted to the data. F-tests were used to test for significance (P < 0.05) of linear and quadratic terms and their interactions and not significant terms were removed from the model. To meet assumptions of homoscedasticity, weed biomass in winter wheat and spring oats

U. Boström et al. / Field Crops Research 134 (2012) 144–152

147

Table 4 Effect of row distance (RD) and seed distance within the row (SD) on number of shoots, number of ears or panicles, and number of ears/panicles shoot−1 in cereals. Response variables of each RD are average across all SD, and response variables of SD are average across all RD. Treatment

Winter wheat Shoots −1

SD1 RD12 a SD1 RD24 SEc RD24 RD36 RD48 SE SD1 SD2/3 SD1/2 SE

Spring oats Ears shoot−1

Ears −1

Shoots −1

Spring wheat Panicles −1

Panicles shoot−1

Shoots −1

Ears shoot−1

Ears −1

No m

No m

Sqrt No

No m

No m

Sqrt No

No m

No m

Sqrt No

100.5a 100.0a 23.2 118.9b 131ab 136.6a 34.2 110.0c 132.7b 143.8a 34.2

52.5b 70.8a 9.0 79.6b 101.5a 111.5a 13.8 90.3a 98.6a 103.7a 14.1

0.73b (0.53)b 0.87a (0.76) 0.04 0.85b (0.72) 0.91a (0.83) 0.94a (0.88) 0.05 0.93a (0.86) 0.89b (0.79) 0.87b (0.76) 0.05

70.0b 91.8a 11.0 108.0b 123.7a 123.7a 18.2 99.2c 117.4b 138.8a 18.2

72.2a 82.9a 26.0 104.1a 123.8a 127.7b 33.8 92.4c 119.6b 143.6a 33.8

1.00a (1.00) 0.97a (0.94) 0.23 1.00a (1.00) 1.00a (1.00) 1.06a (1.12) 0.23 0.97b (0.94) 1.02a (1.04) 1.03a (1.06) 0.23

61.6b 73.9a 13.4 78.5a 88.7a 86.6a 17.2 74.4c 84.8b 94.4a 17.3

60.6b 74.6a 5.8 78.8b 96.2a 96.0a 9.8 81.9b 86.2b 102.9a 9.8

1.03a (1.06) 1.02a (1.04) 0.06 1.02b (1.04) 1.05ab (1.10) 1.07a (1.14) 0.03 1.06a (1.12) 1.03a (1.06) 1.05a (1.10) 0.03

Values within part of columns grouped as SD1 , RD or SD and marked with different letters are significantly different (PDIFF, P < 0.05). a Abbreviations as in Table 2. b Transformed to square root values (Sqrt) for statistical analysis. Re-transformed means in parentheses. c Standard error of means.

was log10 transformed while weed biomass in spring wheat was transformed as log10 + 1. For the calculation of standard errors (SE) and confidence intervals presented in diagrams and Tables, site and block were considered as fixed factors in the model to avoid the variation between blocks and sites being included. 3. Results No statistical significant interaction (P > 0.05) between row distance and seed distance was found for any of the response variables reported below. This demonstrates that seed distance had a similar influence on all response variables, irrespective of row distance and vice versa. Because of that, all results are reported as mean values averaged over seed distances or over row distances. 3.1. Plant establishment and crop stand characteristics During spring 2007 and 2008, when five out of seven springsown experiments were established, the weather was considerably warmer and drier than the long-term average (Table 3). Despite this, crop plant emergence was satisfactory and at the time of counting, the number of shoots at RD12 was 62 m−1 in spring wheat, 70 m−1 in oats and 100 m−1 in winter wheat (Table 4).

Increasing row distance from RD12 (weed harrowed) to RD24 cm (inter-row hoed) without changing seed distance resulted in 31% more shoots m−1 in oats (P = 0.010) while no significant influence was found on number of panicles m−1 (Table 4). In spring wheat, the number of shoots and ears m−1 increased by 20–23% (P < 0.05). Winter wheat responded by producing 35% more ears m−1 (P = 0.001). When row distance was increased from RD24 to RD48 , the number of shoots m−1 increased by 10–15% in all crops (P < 0.05), while the number of panicles or ears m−1 increased by 20–22% in spring wheat and oats and by 40% in winter wheat (P < 0.001). The number of ears shoot−1 increased by 10% in spring wheat and 20% in winter wheat (P < 0.01), while no significant response was found in oats. Reducing seed distance in the row from SD1 to SD1/2 in spring wheat and winter wheat increased the number of shoots m−1 by 30–40%, and the number of ears m−1 by 15–20%. Oats responded to reduced seed distance by production of 40% more shoots m−1 , 57% more panicles m−1 and 13% more panicles shoot−1 . In contrast, the number of winter wheat spikes shoot−1 declined by 12%. Attacks by plant diseases were observed in some of the trials but were not particularly frequent in any of the crops. Diseases were not significantly influenced by treatments. No attacks by insects were recorded. Lodging was insignificant in all crops and was not influenced by treatments.

Table 5 Effect of row distance (RD) and seed distance within the row (SD) on screenings content, moisture content (Wc) and thousand kernel weight (TKW) in cereals. Response variables of each RD are average across all SD, and response variables of SD are average across all RD. Treatment

SD1 RD12 SD1 RD24 SEb RD24 RD36 RD48 SE SD1 SD2/3 SD1/2 SE

a

Winter wheat

Spring oats

Spring wheat

Screenings %

Wc %

TKW g

Screenings %

Wc %

TKW g

Screenings %

Wc %

TKW g

0.7a 1.4a 0.8 1.0a 1.0a 1.2a 0.7 1.3a 1.0a 0.9a 0.7

19.6a 20.2a 2.6 19.7b 19.8b 20.1a 2.8 20.2a 19.8b 19.6ab 2.7

37.0b 38.8a 1.0 38.4a 37.7ab 37.0b 1.1 37.7a 37.5a 37.9a 1.1

2.5b 4.3a 0.8 3.7c 5.8b 8.0a 1.5 8.4a 4.7b 4.4b 1.5

16.0b 17.8a 1.1 17.0c 18.4b 20.8a 0.8 20.9a 17.9b 17.4b 0.8

34.3a 34.0a 2.3 33.8a 33.9a 33.5a 2.0 33.4a 33.8a 34.0a 2.0

1.4b 1.9a 0.6 1.6b 1.6b 2.2a 0.7 2.0a 1.6b 1.7ab 0.7

21.6b 23.0a 0.9 22.2b 22.2b 23.0a 0.9 23.1a 22.1b 22.1b 0.9

32.1b 33.6a 4.1 33.3b 34.2a 34.0a 3.7 33.8a 33.8a 33.9a 3.7

Values within part of columns grouped as SD1 , RD or SD and marked with different letters are significantly different (PDIFF, P < 0.05). a Abbreviations as in Table 2. b Standard error of means.

U. Boström et al. / Field Crops Research 134 (2012) 144–152

(a)

grain yield=6796-1052*SD-37*RD log (biomass )=1.810+0.012*RD+0.262*SD

600 -2

-1

Winter wheat yield (kg ha )

6000

Weed biomass (g m )

148

4000

400

2000

200

0

0 24-25

36-37.5

48-50

Row distance (cm)

600 -2

grain yield=5328-1175*SD-38*RD log (biomass )=-0.962+0.115*RD-0.001*RD +0.537*SD

4000

400

2000

200

0

Weed biomass (g m )

6000

-1

Spring oat yield (kg ha )

(b)

0 24-25

36-37.5

48-50

Row distance (cm)

600 -2

grain yield=1093+117*RD-2*RD log (biomass+1)=3.205-0.071*RD+0.001*RD

Weed biomass (g m )

6000

-1

Spring wheat yield (kg ha )

(c)

4000

400

2000

200

0

0 24-25

36-37.5

48-50

Row distance (cm) Fig. 1. Effect of row distance at normal seed distance on grain yields and weed biomass in (a) winter wheat, (b) spring oats and (c) spring wheat. Bars represent the 95% confidence interval. Within response variable, columns sharing the same letter are not significantly different at P < 0.05. Weed biomass is shown as backtransformed means.

3.2. Grain yield quantity In all crops, the highest grain yield was obtained at RD12 (Fig. 1). When row distance increased from RD12 (weed harrowed) to RD24 (inter-row hoed) at SD1 , yield decreased by 855 kg ha−1 in winter wheat (P < 0.006), 595 kg ha−1 in oats (P < 0.007), and 385 kg ha−1 in spring wheat (P = 0.057). In winter wheat and oats there was a negative linear relationship between grain yield and row-distance. Each step in row-spacing increase, from RD24 to RD36 and RD48 , reduced yields in these crops by 450–460 kg ha−1 (Fig. 2). In spring wheat, the quadratic term (P < 0.002) in the relationship indicated a small yield increase (30 kg ha−1 ) when row distance increased from RD24 to RD36 , while a further increase to RD48 lowered yield by 520 kg ha−1 . Reducing seed distance from SD1 to SD2/3 increased grain yields by 350–390 kg ha−1 in winter wheat and oats, while a further

Fig. 2. Effect of row distance (RD) on grain yields (solid line) and weed biomass (dashed line) in (a) winter wheat, (b) oats and (c) spring wheat. Data are average over seed distances (SD). The dotted lines represent the 95% confidence interval. Weed biomass is shown as back-transformed means.

reduction to SD1/2 increased yields by 175–195 kg ha−1 (Fig. 3). Spring wheat yields were not significantly influenced by seed distance. 3.3. Grain quality 3.3.1. Weed harrowed RD12 and inter-row hoed RD24 sown at normal seed distance Doubling the row distance while keeping a constant seed distance within the row increased the amount of screenings and the water content of the grain (P < 0.017) in both springsown crops (Table 5). In both wheat crops, the thousand kernel weight (P < 0.03) and gluten content in the grain increased (spring wheat: P < 0.001, winter wheat: P = 0.16), while the starch content decreased (P < 0.05) (Table 6). Protein content in the grain was improved by 5% in oats, 4.5% in winter wheat

U. Boström et al. / Field Crops Research 134 (2012) 144–152

149

Table 6 Effect of row distance (RD) and seed distance within the row (SD) on contents of starch, protein, gluten and ergosterol in cereal grain. Response variables of each RD are average across all SD, and response variables of SD are average across all RD. Treatment

SD1 RD12 a SD1 RD24 SEb RD24 RD36 RD48 SE SD1 SD2/3 SD1/2 SE

Winter wheat

Spring oats

Spring wheat

Starch % of DM

Protein % of DM

Gluten %

Ergosterol mg kg−1

Protein % of DM

Starch % of DM

Protein % of DM

Gluten %

Ergosterol mg kg−1

71.7a 71.2b 0.4 71.5a 71.4a 71.0b 0.4 71.1b 71.3a 71.4a 0.4

11.1b 11.6a 0.5 11.5b 11.6b 12.0a 0.4 11.9a 11.6b 11.6b 0.4

26.1a 27.5a 1 26.8b 27.2b 28.4a 0.9 28.3a 27.1b 27.0b 0.9

16.9a 15.8a 2.4 15.8b 16.6a 16.6a 2.1 16.3a 16.3a 16.4a 2.1

12.0b 12.6a 0.2 12.6b 12.7b 12.9a 0.3 12.8a 12.7a 12.7a 0.3

70.0a 68.8b 0.4 69.1a 68.6b 68.5b 0.8 68.4b 68.9a 68.9a 0.8

12.4b 13.7a 1.1 13.3b 13.9a 14.0a 1.2 14.0a 13.6b 13.5b 1.2

27.6b 31.9a 2.9 30.3b 32.6a 33.0a 3.0 33.2a 31.5b 31.2b 3.0

17.4a 16.7b 2.7 17.1a 16.5b 16.4b 2.7 16.3b 16.8a 16.9a 2.7

Values within part of columns grouped as SD1 , RD or SD and marked with different letters are significantly different (PDIFF, P < 0.05). a Abbreviations as in Table 2. b Standard error of means.

and 10.5% in spring wheat (P < 0.034). The content of ergosterol decreased in spring wheat (P = 0.033) and winter wheat (P = 0.088).

3.3.2. Row distance 24 cm and wider When the row spacing was widened to 24 cm and greater, the water content of the harvested material, i.e., grain and screenings, increased with increasing row distance in spring wheat and oats (P < 0.05) (winter wheat: P = 0.14) (Table 5). Increased seed rate in the row lowered moisture content in the grain of all crops (P < 0.05). In wheat, the content of screenings in the harvested material was below 2.2%, while in oats, the content was considerably higher and averaged 4.4% at SD1/2 . The amount of screenings in spring wheat and oats increased in response to wider row distance and increased seed distance (P < 0.02). The thousand kernel weight decreased in winter wheat as a response to wider row distance but increased in spring wheat (P < 0.02). At row distance 24 cm and wider, the gluten content increased with increasing row distance and with reduced seed rate in both spring and winter wheat, while the trends for starch were the reverse (P < 0.022) (Table 6). The protein content in wheat increased with increasing row distance and reduced seed rate (P < 0.003), while in oats only row distance influenced protein content (P = 0.049). In winter wheat, the content of ergosterol increased with increasing row distance while the response in spring wheat was the reverse (P < 0.002). Reducing seed distance increased the content of ergosterol in spring wheat grain (P = 0.004).

3.4. Weed biomass In SD1 RD12 , back-transformed means of weed biomasses were 30, 95 and 240 g m−2 in oats, spring wheat and winter wheat, respectively (Fig. 1). Increasing the row distance from RD12 (weed harrowed) to RD24 (inter-row hoed) while keeping the seed distance constant did not significantly influence weed biomass in any of the crops. At RD24 and wider, the relationships between row distance and weed biomass were quadratic in spring wheat (P < 0.008) and oats (P < 0.021) (Fig. 2). Increasing the row distance from RD24 to RD48 increased weed biomass 2–3 folds in winter wheat and spring oats (Fig. 2), while biomass in spring wheat was less influenced. Reduced seed distance hampered weed biomass in spring oats and winter wheat, while weeds in spring wheat were not significantly influenced.

4. Discussion 4.1. Grain yield In recent years there has been intense debate on whether organic cultivation can feed the world (e.g. Badgley et al., 2007). Several studies have reported lower grain yield of organic cereals than conventional cereals (Korsaeth, 2008; Poutala et al., 1994), but comparable crop yield has also been reported (Cavigelli et al., 2008). In our study, the grain yield of wheat sown at the row distance normal to the area, i.e., 12–12.5 cm, was approx. 1000 kg ha−1 lower than the average yield of conventionally cultivated wheat in Sweden during the same period (Anon., 2010). Compared with grain yield of organic wheat reported in official statistics, yield in our study was 2000 kg ha−1 higher in winter wheat and 300 kg ha−1 higher in spring wheat. The grain yield of oats was similar to the average of conventionally cultivated oats and approx. 1500 kg ha−1 higher than organic grain yield according to official statistics. It can be concluded that grain yield at RD12 was comparatively high in all crops, which may have influenced the level of yield reduction when crops were sown at wider row or seed distances. Favourable weather conditions at the time of sowing ensured that the density of established crop plants mimicked what was expected. The higher production of shoots and ears m−1 when row distance increased while keeping the seed distance constant was probably mainly due to reduced intra-specific competition for light, nutrients and water. This allowed more shoots and tillers to survive and develop. It has been shown that winter wheat grown at low plant density increases green area per plant through increased duration of tiller production, green area per shoot and shoot survival (Whaley et al., 2000). The importance of improved nitrogen availability for survival of later developing tillers and production of more ears has been confirmed for spring wheat (Power and Alessi, 1978). In our study, the enhanced number of ears or panicles m−1 , when row distance was changed from SD1 RD12 to SD1 RD24 , was not high enough to compensate for the lower number of crop plants per unit area. This resulted in 12–16% yield reductions in all crops. At constant seed rate per unit area, Hiltbrunner et al. (2005) and Teich et al. (1993) reported no influence of row spacing on grain yield of winter wheat. Hiltbrunner et al. (2005) noted more degeneration of tillers at narrow rows than at wider, which may have been caused by intense intra-row competition. It is likely that yields are less affected by row distance at sites where nutrient availability is high; while at low nutrient availability yields tend to decrease when rows distances increase.

150 6000

600

4000

400

2000

200

-2

Weed biomass (g m )

-1

Winter wheat yield (kg ha )

(a)

U. Boström et al. / Field Crops Research 134 (2012) 144–152

grain yield=6796-1052*SD-37*RD log (biomass)=1.810+0.012*RD+0.262*SD

0

0 1

2/3

1/2

Seed distance

600

-2

grain yield=5328-1175*SD-38*RD log (biomass)=-0.962+0.115*RD-0.001*RD +0.537*SD

Weed biomass (g m )

6000

-1

Spring oat yield (kg ha )

(b)

4000

400

2000

200

0

0 1

2/3

1/2

Seed distance

600

-2

grain yield=1093+117*RD-2*RD log (biomass+1)=3.205-0.071*RD+0.001*RD

Weed biomass (g m )

6000

-1

Spring wheat yield (kg ha )

(c)

4000

400

2000

200

0

0 1

2/3

1/2

Seed distance Fig. 3. Effect of seed distance (SD) on grain yield (solid line) and weed biomass (dashed line) in (a) winter wheat, (b) oats and (c) spring wheat. Data are average over row distances (RD). The dotted lines represent the 95% confidence interval. Weed biomass is shown as back-transformed means.

In our study, crop plant arrangement not only influenced crop stand characteristics and grain yield quantity, but also several grain quality parameters. The protein and gluten content of the grain increased as a response to increased row and seed distances, while the starch content decreased. This response may have resulted from reduced intra-specific competition for soil nitrogen, although increased weed biomass may have constrained nitrogen availability to some extent. Several studies have demonstrated a positive relationship between nitrogen fertilisation and protein content in cereals (e.g. review by Jenner et al., 1991), and a high correlation between the amount of protein and gluten in wheat kernels (Wieser and Seilmeier, 1998). While the deposition of proteins in wheat grain is influenced to a high degree by external factors of supply, the rate of starch deposition during grain filling is influenced mainly by sink-limited factors, i.e., by factors functioning inside or in the proximity of the grain itself (Jenner et al., 1991). Hiltbrunner et al. (2005)

noted increased protein content in winter wheat kernels when row spacing was widened and attributed this response to nutrient availability. In the present study, increased protein content in the kernel as a response to increased row or seed distances resulted in somewhat lower percentage losses of protein yield ha−1 than of grain yield ha−1 . Gooding et al. (2002) also reported increased grain protein content with decreased seed rate in wheat but the response can be expected to be dependent on the base seed rate (Stephen et al., 2005). In spring wheat, we found a negative relationship between ergosterol content in the grain and row distance and a positive relationship between ergosterol and seed distance within the row. Schnürer (1993) concluded that ergosterol level is a suitable marker for quantitatively monitoring fungal growth in solid substrates. Presence of fungi in the grain constitutes a risk factor when the product is used for food or feed, but it is impossible to monitor all the possible mycotoxins present. According to Ng et al. (2008), the content of ergosterol can be used to assess the microbiological status of grain and feed and may be a valuable indicator of the proportion of infected kernels.

4.2. Weed biomass The weight of screenings is largely dependent on the amount of weed plant parts that contribute to the weight of harvested material. This weed debris also contributed to the higher water content in the grain measured in our study as a response to increased row distance and seed distance within the row. At SD1 , no significant difference in weed biomass was found between RD12 and RD24 . In our study, weed harrowing was performed at RD12 and inter-row hoeing at wider row distances. At RD24 , number of crop plants per unit area was halved in comparison with RD12 . Thus, the competitive effect caused by the crop on weeds was considerably lower, disguising any differences in weed efficacy that may have existed between the two management techniques. Comparisons between the efficiency of weed harrowing and inter-row hoeing indicate that the weed effect varies between sites depending on, e.g. the prevailing weed flora (Boström, 2006). Weed management not only has an effect on weeds, but may also injure or kill crop plants to a varying degree (Rasmussen et al., 2008, 2009). Higher crop yield at RD12 shows that lower intra-specific competition at RD24 did not compensate for lower seed rate per unit area for any of the crop species. The difference between row distances could be at least partly due to the implementation of two different weed management techniques. At RD24 and wider, post-emergence weed management was performed by inter-row hoeing, implying that treatment responses were influenced by crop plant arrangement and not by weed management technique. Although all plots were hoed, crop plant arrangement had a pronounced influence on weed growth. On a relative basis, treatments had a considerably stronger effect on weed biomass than on grain yield quantity of oats and winter wheat. In both these crops, we noted a doubling of weed biomass as a response to wider row distances. In spring wheat, crop plant arrangement had a non-significant influence on weed biomass, which may reflect low weed suppressive ability despite reduced intra-specific competition. It has been shown that modern semidwarf wheat cultivars are more sensitive to weed pressure than older cultivars (Mason et al., 2008). Reid et al. (2011) commented that breeding selection for high grain yield in organic systems should be performed within organically managed systems. This may also be true regarding selection for weed suppressive ability. In relation to crop plant arrangement, the choice of weed suppressive cultivars should also consider plant morphology, as planophile wheat cultivars are reported to be more advantageous in wider row

U. Boström et al. / Field Crops Research 134 (2012) 144–152

stands than those with more erectophile leaf inclination (Drews et al., 2009). 4.3. Crop plant arrangement It should be emphasised that crop and weed responses to increased row distance from RD12 to RD24 , was not only caused by row distance per se in our study. Also, weed management strategies varied; RD12 was weed harrowed and RD24 was hoed. Thus, responses may have been due to variations between strategies in, e.g., weed efficacy, injuries on crop plants, or effects on soil moisture and nitrogen mineralisation. All treatments in the trials reported here, except SD1 RD12 , involved lower crop plant density per unit area than recommended for the region. Provided the cost per kg of sowing seed is twice the income from grain, it can be concluded that using a seed rate that result in a reduction from SD1 to SD1/2 would be profitable in winter wheat and spring oats, while the results for spring wheat do not support a change to smaller seed distance than is normally used. In winter wheat and oats, inter-row hoeing should be implemented at as narrow a row distance as possible, while yield in spring wheat is unaffected by row distance. Based on regression equations in Figs. 2 and 3, it can be predicted that the use of normal seed rates per unit area in winter wheat and spring oats would be cost-effective at all row distances used in this study. However, at RD36 and RD48 this would result in smaller seed distances than we used. Extrapolation should be made with caution since increased intra-specific competition may lead to lower grain yields than predicted. Using small seed distance when cultivating at wide row distances is not only a question of seed cost, as some of the economic losses that may thereby occur are compensated for by superior weed suppressive ability and thereby increased grain yield. In organic crops in particular, high seed rates may to some extent balance out the loss of crop plants damaged by physical weed control measures. 5. Conclusions Doubling the row distance from 12–12.5 cm (weed harrowed) to 24–25 cm (inter-row hoed) with seed distance held constant reduced grain yields of all crops by 12–16%. No statistical interaction between row distance and seed distance was found for any of the response variables, implying that in all crops the effect of seed distance was similar irrespective of row distance. Based on the range of seed distance used in our study, a reduction in spring oats and winter wheat to half of that conventionally used was considered as optimal for obtaining high yield quantity and low weed biomass at row distance 24 cm and wider. However, it is possible that even smaller seed distance would be profitable in these crops. In spring wheat, a more dense plant arrangement than normally used had no statistically significant influence on grain yield or weeds at wide row distances. Increased row and seed distances improved the content of gluten and reduced the content of starch in wheat grain, while protein content was increased in the grain of all crops. Acknowledgement This work was supported by grants from The Swedish Board of Agriculture. References Albrecht, H., 2005. Development of arable weed seedbanks during the 6 years after the change from conventional to organic farming. Weed Res. 45, 339–350.

151

Anonymous, 2010. Yearbook of Agricultural Statistics. Official Statistics of Sweden, ISSN: 1654-4382. Badgley, C., Moghtader, J., Quintero, E., Zakem, E., Chappell, M.J., Avilés-Vázquez, K., Samulon, A., Perfecto, I., 2007. Organic agriculture and the global food supply. Renew. Agr. Food. Syst. 22, 86–108. Becker, K., Böhrnsen, A., 1994. Wirkungen mechanischer Pflegemassnahmen auf die Unkrautabundanz und die N-mneralisation im Boden. Zeits. Planzenkrankh. Pflanzens 14, 315–324. Boström, U., 2006. Weed management in organically-grown narrow-leafed lupin. In: Santen, E., van, Hill, G.D. (Eds.), Mexico, Where Old and New World Lupins Meet: Proceedings of the 11th International Lupin Conference. Guadalajara Jalisco, Mexico, pp. 89–92. Cavigelli, M.A., Teasdale, J.R., Conklin, A.E., 2008. Long-term agronomic performance of organic and conventional field crops in the mid-Atlantic region. Agron. J. 10 (0), 785–794. Corbin, A.T., Thelen, K.D., Robertson, G.P., Leep, R.H., 2010. Influence of cropping systems on soil aggregate and weed seedbank dynamics during the organic transition period. Agron. J. 102, 1632–1640. Drews, S., Neuhoff, D., Koepke, U., 2009. Weed suppression ability of three winter wheat varieties at different row spacing under organic farming conditions. Weed Res. 49, 526–533. Eyre, M.D., Critchley, C.N.R., Leifert, C., Wilcockson, S.J., 2011. Crop sequence, crop protection and fertility management effects on weed cover in an organic/conventional farm management trial. Europ. J. Agron. 34, 153–162. Gooding, M.J., Pinyosinwat, A., Ellis, R.H., 2002. Responses of wheat grain yield and quality to seed rate. J. Agric. Sci. 138, 317–331. ´˚ Håkansson, S., 2003. Competition in plant stands of short duration. In: Hakansson, S. (Ed.), Weeds and weed management on arable land: an ecological approach. CABI Publishing, Wallingford, UK, pp. 81–117. Hatcher, P.E., Melander, B., 2003. Combining physical, cultural and biological methods: prospects for integrated non-chemical weed management strategies. Weed Res. 43, 303–322. Hiltbrunner, J., Liedgens, M., Stamp, P., Streit, B., 2005. Effects of row spacing and liquid manure on directly drilled winter wheat in organic farming. Europ. J. Agron. 22, 441–447. Jenner, C.F., Ugalde, T.D., Aspinall, D., 1991. The physiology of starch and protein deposition in the endosperm of wheat. Austr. J. Plant Phys. 18, 211–226. Kolb, L.N., Gallandt, E.R., Molloy, T., 2010. Improving weed management in organic spring barley: physical weed control vs. interspecific competition. Weed Res. 50, 597–605. Korsaeth, A., 2008. Relations between nitrogen leaching and food productivity in organic and conventional cropping systems in a long-term field study. Agric. Ecosyst. Environ. 127, 177–188. Kristensen, L., Olsen, J., Weiner, J., 2008. Crop density, sowing pattern, and nitrogen fertilization effects on weed suppression and yield in spring wheat. Weed Sci. 56, 97–102. Leblanc, M.L., Cloutier, D.C., 2001. Susceptibility of dry edible bean (Phaseolus vulgaris, cranberry bean) to the rotary hoe. Weed Technol. 15, 224–228. Lundkvist, A., 2009. Effects of pre- and post-emergence weed harrowing on annual weeds in peas and spring cereals. Weed Res. 49, 409–416. Mason, H., Goonewardene, L., Spaner, D., 2008. Competitive traits and the stability of wheat cultivars in differing natural weed environments on the northern Canadian Prairies. J. Agric. Sci. 146, 21–33. Melander, B., Rasmussen, I.A., Barberi, P., 2005. Integrating physical and cultural methods of weed control – examples from European research. Weed Sci. 53, 369–381. Ng, H.E., Raj, S.S.A., Wong, S.H., Tey, D., Tan, H.M., 2008. Estimation of fungal growth using the ergosterol assay: a rapid tool in assessing the microbiological status of grains and feeds. Lett. Appl. Microbiol. 46, 113–118. Olsen, J., Weiner, J., 2007. The influence of Triticum aestivum density, sowing pattern and nitrogen fertilization on leaf area index and its spatial variation. Bas. Appl. Ecol. 8, 252–257. Poutala, R.T., Kuoppamaki, O., Korva, J., Varis, E., 1994. The performance of ecological, integrated and conventional nutrient management systems in cereal cropping in Finland. Field Crops Res. 37, 3–10. Power, J.F., Alessi, J., 1978. Tiller development and yield of standard and semidwarf spring wheat varieties as affected by nitrogen fertilizer. J. Agric. Sci. 90, 97–108. Rasmussen, J., Bibby, B.M., Schou, A.P., 2008. Investigating the selectivity of weed harrowing with new methods. Weed Res. 48, 523–532. Rasmussen, J., Nielsen, H.H., Gundersen, H., 2009. Tolerance and selectivity of cereal species and cultivars to postemergence weed harrowing. Weed Sci. 57, 338–345. Rasmussen, J., Mathiasen, H., Bibby, B.M., 2010. Timing of post-emergence weed harrowing. Weed Res. 50, 436–446. Reid, T.A., Yang, R.C., Salmon, D.F., Navabi, A., Spaner, D., 2011. Realized gains from selection for spring wheat grain yield are different in conventional and organically managed systems. Euphytica 177, 253–266. Schnürer, J., 1993. Comparison of methods for estimating the biomass of three food-borne fungi with different growth patterns. Appl. Environ. Microbiol. 59, 552–555. Stephen, R.C., Saville, D.J., Drewitt, E.G., 2005. Effects of wheat seed rate and fertiliser nitrogen application practices on populations, grain yield components and grain yields of wheat (Triticum aestivum). New Zeal. J. Crop Horticult. Sci. 33 (2), 125–138.

152

U. Boström et al. / Field Crops Research 134 (2012) 144–152

Teich, A.H., Smid, A., Welacky, T., Hamill, A., 1993. Row-spacing and seed-rate effects on winter wheat in Ontario. Can. J. Plant Sci. 73, 31–35. Turner, R.J., Davies, G., Moore, H., Grundy, A.C., Mead, A., 2007. Organic weed management: a review of the current UK farmer perspective. Crop Prot. 26, 377–382, http://dx.doi.org/10.1016/j.cropro.2006.01.021. Whaley, J.M., Sparkes, D.L., Foulkes, M.J., Spink, J.H., Semere, T., Scott, R.K., 2000. The physiological response of winter wheat to reductions in plant density. Ann. Appl. Biol. 137, 165–177.

Weiner, J., Griepentrog, H.W., Kristensen, L., 2001. Suppression of weeds by spring wheat Triticum aestivum increases with crop density and spatial uniformity. J. Appl. Ecol. 38, 784–790. Wieser, H., Seilmeier, W., 1998. The influence of nitrogen fertilisation on quantities and proportions of different protein types in wheat flour. J. Sci. Food Agric. 76, 49–55. Wortman, S.E., Lindquist, J.L., Haar, M.J., Francis, C.A., 2010. Increased weed diversity, density and above-ground biomass in long-term organic crop rotations. Renew. Agric. Food Syst. 25, 281–295.