Haisheng Xie, David R.S. Rourke and Andrew Hargrave
Conservation Tillage Productivity Centre
Minto, Manitoba

INTRODUCTION

Wild oat (Avena fatua L.) is one of the most troublesome weeds in the Canadian prairies. The Canadian prairies are a semiarid region prone to soil erosion, and appears to be an appropriate site to adopt conservation tillage practices. Due to the environmental concerns and the development of herbicide resistance, it would be desirable to integrate all aspects of crop management systems, rather than solely rely on herbicides, to manage wild oat problems. Such an integrated approach has been seldomly studied in Canada, especially under zero tillage conditions. The objective of the present study is to determine whether manipulating agronomic practices, including row spacing, seeding rate and fertilizer/seed placement, in zero tillage system can change the interference of wild oats with wheat and canola

MATERIALS AND METHODS

Except when otherwise indicated, the field experiments were conducted at Minto, MB in 1994 and 1995. The soil belongs to the Ryerson Loam Association. Wheat (cv. Katepwa) and canola (cv. Garrison), and a natural wild oat population from the area of Weyburn, SK, were used. Two tillage systems, conventional tillage (CT) and zero tillage (ZT), have been established since 1990. CT consisted of a light cultivation in the spring preceded crop seeding, and a deep tillage followed by a light cultivation in the fall. There was no tillage operation proceeded in the ZT system except that in 1993 in which all plots were treated with fall banding fertilization.

There were four treatment factors in the first experiment and two treatment factors in the second experiment (Table 1). The plots, 7 m x 3 m, were arranged as a split plot design with 4 replicates. In 1995, the same treatment combinations in wheat were assigned to the same plots where canola was grown in 1994. Also in 1995, the same treatment combinations in canola were assigned to the same plots where wheat was grown in 1994. The field in those two experiments was essentially flee of wild oat infestation prior to 1994. Wild oat at the rate of 62 kg/ha was seeded (half being broadcasted and half being seeded with a disc drill) prior to crop seeding in 1994. No additional wild oat seeds were "planted" into the plots in 1995.

In 1993, a preliminary field study was conducted at Fairfax, MB The soil belongs to the Carroll Clay Loam Association. Wheat (cv. Katepwa) was grown under zero tillage conditions Treatment factors included; A) row spacing (10", 15" and 20"), and B) fertilizer/seed placement (narrow row, paired row and wide row). The plots, 8 m x 3 m, were arranged as a completely randomized block design with 4 replicates.

RESULTS

(1) Wheat

Under zero tillage conditions, 20" row spacing had a much poorer wheat yield compared to those at 10" or 15" spacing (Fig. 1). Thus the 20" treatment was discontinued after 1993. Averaged over three years, 15" row spacing generally resulted in similar yields to 1 0~' spacing. Among the three placement treatments, the highest yield was obtained with paired row placement regardless of row spacing, whereas the relative performance between narrow row and wide row placement differed with year and row spacing.

Averaged over three fenoxaprop rates, wheat yields were similar between the two tillage systems in 1994, but the yield tended to be lower under zero tillage in 1995 even though wheat seedling emergence was not affected by tillage (Table 2; Fig. 2). Row spacing had limited effect on yield and crop density. Paired row placement resulted in the best wheat seedling stand and yield. Wide seed row treatment had poor yield and seedling establishment. Narrow row placement performed poorly in 1994 but performed similarly to paired row placement in 1995.

In 1994, wild oat seedling emergence was not affected by tillage system, but 10" row spacing and wide seed row placement had lower wild oat emergence (Fig. 3). In untreated plots later in the growing season, wild oat panicle density was similar in two tillage systems. 10" row spacing resulted in lower density in wild oat panicles, especially in ZT; while fertilizer/seed placement had a limited effect. In 1995, there were slightly more wild oats emerged in ZT, especially at 10" spacing; whereas in CT wild oat seedling density was higher at 15" spacing (Fig. 3). Narrow row placement tended to stimulate wild oat emergence, and paired row placement often resulted in fewer wild oat seedlings. In untreated plots later in the growing season of 1995, there were more wild oat panicles in ZT, especially at 10" spacing. Paired row placement had the lowest wild oat panicle density, whereas wide row placement resulted in the highest density.

On average, fenoxaprop application at 46 g/ha resulted in similar wheat yield and net return to those at 92 g/ha of fenoxaprop (Table 2; Fig. 4). At 15" row spacing, however, the yield was slightly increased at 92 g/ha of fenoxaprop compared to that at 46 g/ha. There was a marked reduction in economic return when fenoxaprop was not applied.

Wheat yield under ZT was slightly increased at high seeding rate (150 vs. 100 kg/ha) (Table 2; Fig. 5). Wild oat panicle density was much reduced by high seeding rate even though wild oat seedling emergence was not changed.

(2) Canola

Averaged over three clethodim rates, in both years canola yield and seedling emergence were higher in ZT than in CT (Table 2; Fig. 6). The effect of row spacing was generally marginal and varied with year and tillage Paired row placement usually resulted in the best canola yield and seedling stand. The performance of narrow row placement was poor in 1994, but the same treatment in 1995 performed similarly to paired row placement especially at 15" row spacing. Wide row placement, particularly at 15" spacing, had poor yield and crop stand.

In the spring of 1994, there were fewer wild oat seedlings in ZT, 15" row spacing, wide row placement or paired row placement (Fig. 7). In untreated plots later in the growing season, wild oat panicle density was lower in ZT and with paired row placement, while row spacing had little effect. In 1995, CT resulted in less wild oat emergence and the variation among other treatment factors within CT was limited (Fig. 7). Under ZT, the influence of fertilizer/seed placement on wild oat seedling varied with row spacing. Later in the season of 1995, wild oat panicle density in untreated plots was higher in ZT than in CT. Row spacing had little effect on wild oat population. There were more wild oat panicles with wide row placement than with other placements.

Between the two clethodim rates of 23.7 and 47.4 g/ha, canola yield and net return were generally similar in 1995 but there were slightly higher yields and net profit at the 47.4 g/ha rate in 1994 (Table 2). The yield difference between the two herbicide rates was smaller under ZT (Fig. 8). There was a marked yield reduction when clethodim was not applied.

Canola yield in ZT was slightly increased at high seeding rate (9 vs. 6 kg/ha), especially in 1995 (Table 2; Fig. 5). Wild oat panicle density was reduced by the higher canola seeding rate, particularly in 1994, even though wild oat seedling emergence was not affected.

DISCUSSION

The present study demonstrated that wild oat interference with wheat and canola could be affected by various tillage and agronomic practices. The performance of the two tillage systems varied with year. In 1995, the delayed cultivation in CT due to excessive soil moisture level appeared responsible for the lower density of wild oat seedlings and panicles as compared with those in ZT, which could be partially attributed to reduced wheat yield under zero tillage in 1995. Earlier seeding date would result differently. The present results suggested that at least in canola the better crop establishment in ZT had a contribution to decrease wild oat interference. Studies in the USA and Alberta showed that there were more grassy weeds in ZT than in CT (Blackshaw et al., 1994; Buhler, 1992; Schreiber, 1992). Derksen et al. (1993) did not reveal such trend in Saskatchewan. Martin and Felton (1993) found that in Australia wild oat density and seed bank in the soil were lower under ZT than under CT, which was consistent with our previous results (Rourke and Hargrave, 1993). It would be desirable to monitor wild oat populations in the plots of both tillage systems for a longer term.

The influence of row spacing on wild oat competition was generally marginal, especially in canola Barton et al. (1992) indicated that the row spacing had limited effect on wild oat density and biomass in barley. O'Donovan (1994) also found that row spacing had no impact on tartary buckwheat interference with canola However, some studies showed that narrow row spacings resulted in less wild oat competition than wider row spacings (e.g., Kirkland, 1993). It should be noted that in zero-till wheat fields, wild oat infestation between the two row spacings showed opposite trends in the two years. Although the reason for such discrepancy remains to be determined, the poor crop seedling establishment in 1995 due to wet-then-dry weather might be involved. The present study also demonstrated that row spacing in excess of I 5" was infeasible under our experimental conditions, which could be attributed to intensive intraspecific competition and to fertilizer damage in crop seedlings.

Among the three fertilizer/seed placement regimes, paired seed row placement consistently had the best performance in wheat and canola Although wild oat seedling emergence was usually not affected by paired row placement, such a treatment resulted in better crop establishment. Besides, applied nitrogen fertilizer was likely more adjacent to the crop plants than to the weeds. Thus this placement could result in stronger crop competitive ability, leading to reduced wild oat interference. The relatively poor performance of wide seed row placement, especially in 1995, appears to be associated with fertilizer damage as well as with poor seed bed quality caused during seeding operation. The poor crop seedling establishment, plus easily acquired fertilizer nutrients by wild oat contributed to greater wild oat infestation in the wide row placement treatment, especially in 1995. Narrow seed row placement performed differently in 1994 and 1995. The poor crop yield in this treatment in 1994 might be related to intensive intraspecific competition and to fertilizer damage. In certain instances, narrow row placement tended to stimulate wild oat emergence. Future studies are still needed to further define the role of fertilizer/seed geometry on weed interference with crops, although it becomes clear that in canola and wheat paired row placement has the best agronomic performance in this region among the three fertilizer/seed placement regimes.

In both wheat and canola, high seeding rate resulted in higher crop density and improved crop competitiveness, leading to reduced wild oat panicle density later in the growing season. Such effects, coupled with similar (wheat) or increased (canola) economic return, suggested that relatively high seeding rate should be considered in an integrated weed management program. High seeding rates also reduced the impacts of weed on crops in a number of previous studies (Barton et al., 1992; Carlson and Hill, 1985; Justice et al., 1994; O'Donovan, 1994; Radford et al., 1980).

The present study showed that if there is a severe wild oat infestation, at least in the short term manipulating agronomic practices alone will not decrease wild oat interference to the point where feasible economic yields can be achieved. Justice et al. (1994) also indicated that narrow row spacing and high seeding rate were not economically viable substitutes for herbicides for Italian ryegrass control in winter wheat. Besides, the seed bank of wild oat in untreated fields might possibly increase over the years and thus may cause significant yield loss in the future. Following wild oat herbicide application, wild oat seed production was greatly reduced. In wheat overall economic return with reduced wild oat herbicide rate was similar to that at the full herbicide rate. The return in canola was only slightly higher at the hill herbicide rate. However, the performances of graminicides, especially at low application rates, might be adversely affected by stressful environments. Further study is also required to investigate long-term seed bank change in the soil following the use of reduced herbicide rates.

CONCLUSIONS

This study indicated that agronomic practice and tillage system could affect the competitive ability of wheat and canola with wild oat. The practices which result in better crop establishment will reduce the impact of wild oat on crop. Among the agronomic practices we investigated, relatively high seeding rate and paired row fertilizer/seed placement were the most promising for reducing wild oat interference and should be incorporated into the integrated weed management systems.

REFERENCE

Barton, D.L., D.C. Thill, and B. Shafli. 1992. Integrated wild oat (Avena fatua) management affects spring barley (Hordeum vul~are) yield and economics. Weed Technology 6:129-135.

Blackshaw, R.E., F.O. Larney, C.W. Lindwall, and G.C. Kowb. 1994. Crop rotation and tillage effects on weed populations on the semi-arid Canadian prairies. Weed Technology 8:231-237.

Buhler, D.D. 1992. Population dynamics and control of annual weeds in corn (Zea mays) as influenced by tillage systems. Weed Science 40:241-248.

Carison, H.L. and J.E. Hill. 1985. Wild oats (Avena fatua) competition with spring wheat: Plant density effects. Weed Science 33:176-181.

Derksen, D.A., G.P. Lafond, A.G. Thomas, H.A. Loeppicy, and C.J. Swanton. 1993. Impact of agronomic practices on weed communities: Tillage systems. Weed Science 41:409-407.

Justice, G.G., T.F. Peeper, J.B. Solie, and F.M. Epplin. 1994. Net returns from Italian ryegrass (Lolium multiflorum) control in winter wheat (Triticum aestivum). Weed Technology 8:317-323.

Kirkland, K.J. 1993. Weed management in spring barley (Hordeum vulgare) in the absence of herbicides. Journal of Sustainable Agriculture 3:95-103.

Martin, R.J. and W.L. Felton. 1993. Effect of crop rotation, tillage practice, and herbicides on the population dynamics of wild oats in wheat. Australian Journal of Experimental Agriculture 33:159-165.

O'Donovan, J.T. 1994. Canola (Brassica rapa) plant density influences tartary buckwheat (Fagopyrum tartaricum) interference, biomass, and seed yield. Weed Science 42:385-389.

Radford, B.J., B.J. Wilson, 0. Cartledge, and F.B. Watkins. 1980. Effect of wheat seeding rate on wild oat competition. Australian Journal of Experimental Agriculture and Animal Husbandry 20:77-81.

Rourke D and A. Hargrave 1993. Final report of Conservation Tillage Productivity Centre (1990-1994). Minto, Manitoba.

Schreiber, M.M. 1992. Influence of tillage, crop rotation, and weed management on giant foxtail (Setaria faberi) population dynamics and corn yield. Weed Science 40:645-653.

*FUNDING FOR THIS PROJECT BY NSERC AND DUCKS UNLIMITED IS GRATEFLULLY ACKNOWLEDGED.

Table l. Treatment description

Treatment factor Treatment level in wheat Treatment level in canola

Experiment I

Table 1. Treatment description
Treatment factor	Treatment level in wheat	Treatment level in canola
Experiment I
A. Tillage system	a. CT	a. CT
	b.ZT	b.ZT
B. Row spacing	a. 10"	a. 10"
	b. 15"	b. 15"
C. Fertilizer/seed placement	a. Side banded fertilizer/narrow seed row (narrow row)
	b. Mid-row banded fertilizer/paired seed row (Paired row)
	C. Seed placement fertilizer/wide seed row (Wide row)
D. Grarninicide rate	a. Fenoxaprop at 0 g/ha	a. Clethodim at 0 g/ha
	b. Fenoxaprop at 46 g/ha	b. Clethodim at 23.7 g/ha
	c. Fenoxaprop at 92 giha	c. Clethodim at 47.4 8/ha
Experiment 2 (ZT, 10" spacing, paired row only)
A. Seeding rate (kg/ha)	a. 100	a. 6
	b. 150	b. 9
B. Graminicide rate	a. Fenoxaprop at 0 g/ha	a. Clethodim at 0 g/ha
	b. Fenoxaprop at 46 g/ha	b. Clethodim at 23.7 g/ha
	c. Fenoxaprop at 92 g/ha	c. Clethodim at 47.4 g/ha
*Fenoxaprop at 92 g/ha and clethodim at 47.4 giha are the field recommended application rates.Table 2. Summary of main effects of tillage, row spacing, fertilizer/seed placement, wild oat herbicide application rate, and seeding rate

Table 2. Summary of main effects of tillage, row spacing, fertilizer/seed placement, wild oat herbicide application rate, and seeding rate.
	Yield (kg/ha)			Wild Oat Panicles [#/m(2)]			Net Return ($/ha)
	1994	1995	x	1994	1995	x	1994	1995	x
WHEAT
Tillage
CT	2332	2573	2453	80	27	54	50	91	71
ZT	2365	2231	2298	79	51	65	82	57	70
Row Spacing
10"	2397	2451	2424	68	44	56	73	84	78
15"	2300	2352	2326	91	34	63	55	65	60
Placement
Narrow row	2224	2508	2366	80	40	60	41	95	68
Paired row	2488	2477	2483	76	31	54	89	88	89
Wide row	2336	2220	2278	82	46	64	62	41	52
Fenoxaprop Rate
0 g/ha	2042	1949	1996	79	39	59	26	9	18
46 g/ha	2459	2595	2527	1.4	1.1	1.3	83	109	96
92 g/ha	2546	2661	2603	0.4	0.9	0.7	83	106	95
Seeding Rate
100 kg/ha	2554	2233	2394	174	43	109	116	56	86
150 kg/ha	2620	2376	2498	136	23	80	116	70	93
CANOLA
Tillage
CT	1185	1998	1592	116	30	73	-50	257	104
ZT	1468	2399	1934	64	57	61	87	439	263
Row Spacing
10"	1296	2152	1724	87	45	66	7	330	169
15"	1356	2245	1801	93	43	68	30	367	198
Placement
Narrow row	1202	2196	1699	99	37	68	-29	348	160
Paired row	1426	2316	1871	80	38	59	56	393	224
Wide row	1351	2083	1717	92	56	74	28	306	167
Clethodim Rate
0 g/ha	1077	1851	1464	90	43	67	-54	240	93
23.7 g/ha	1379	2332	1856	0.5	0.5	0.5	37	398	218
47.4 g,ha	1524	2413	1969	0.1	0.4	0.3	72	408	240
Seeding Rate
6 kg/ha	1681	2357	2019	48	34	41	170	425	297
9 kg/ha	1824	2648	2236	24	28	26	214	526	370
*Labour cost was excluded in the net return calculation. **Wild oat panicle density results were from 0 g/ha wild oat herbicide treatment only.