The Effects of Spring Tillage on Crop Emergence and Soil Moisture

Abstract
In the spring of 1991, the Alberta Farm Machinery Research Centre completed a research project on the effect of pre-seeding tillage on crop emergence. The project was initiated because of a request from Alberta Agriculture's Provincial Farm Machinery Specialist, Murray Green. The objective of the project was to determine the effects of tillage depth and secondary tillage operations on crop emergence. A secondary objective was to determine how tillage affected soil moisture, soil particle size and crop emergence.

While some trends were apparent from the soil moisture, particle size and trash results, no statistically significant relationship between experimental factors resulted.

Introduction
While numerous papers have been written on tillage and soil conservation practices, little specific research on tillage effects has been done successfully in Southern Alberta. Most of the research into the tillage area has concentrated on no-till applications. Generally, results from no-till experiments follow that of Johnson, Lowery and Daniel (1982). No-till farming practices appear to be the most drought resistant and the best moisture conservation tool available to farmers.

The amount of moisture remaining in the soil after conventional tillage practices has, for the most part, not been addressed. Early spring tillage continues to be used in most farming practices to control weeds and prepare a seedbed for seeding operation. However, Lindwald (1984) stated that water is the factor limiting plant growth in the semi-arid regions of Canada. While weed control and seedbed preparation is required, moisture loss in the soil should be limited where possible. In areas where soil moisture is the limiting plant growth factor, wind erosion is normally an additional problem. Boyer (1983) indicated that residue management is very effective in controlling both soil temperature and crop emergence and concluded that maximum moisture loss occurs with conventional tillage.

Based on comments from researchers and literature in the tillage area, a project examining tillage depth and secondary tillage interaction was developed. It should, however, be noted that the primary reason for tillage is still weed control. To examine the soil moisture retained after tillage may not be examining the real effects of tillage practices or addressing all the issues.

Experimental Procedure

Statistical design
Chisel plow depth and secondary tillage operations were the two factors of the experiment. Chisel plow factor levels included 102 mm (4 in) tillage depth, 51 mm (2 in) tillage depth and 0 mm or no chisel plow tillage. No tillage, harrows, packers and a harrow packer combination made up levels of the secondary tillage factor.

The three levels of tillage depth and the four levels of secondary tillage resulted in a 3 x 4 full factorial split block design experiment. The split block design was made of three blocks containing 12 plots each. Blocks were used to eliminate random effects due to differences in initial field moisture content.

Plot layout
Plots were 8.53 m (28 ft) in width by 30.48 m (100 ft) in length. Figure 1 illustrates the 36 plot lay out. A 2.74 m (9 ft) tool bar was used to apply the tillage treatments to the plots. Twelve John Deere heavy duty cultivator shanks were mounted with 22.86 cm (9 in) spacing between shanks in four rows on the tool bar. The shanks were fitted with 30.48 cm (12 in) cultivator sweeps. John Deere specifications indicated a trip force of 658 kg (1450 lbs) on the shanks. Three row tine harrows and spiral packers were mounted on the back of the tool bar.

Three tillage passes at 8 km/h (5 mph) on each plot resulted in 8.23 m (27 ft) width of worked plot. A 2.14 m (7 ft) Cereal Implements 2300 Hoe Drill was used to seed three strips in each plot. After seeding, plots were 6.4 m (21 ft) wide. Laura spring wheat was seeded at 84.08 kg/ha (75 lb/ac) and 29-25-0 fertilizer was applied at 84.08 kg/ha (75 lbs/ac).

Plot outline

.

Plot Number	Treatments cultivator depth (in)	Secondary Tillage
1	0	None
2	0	Harrows
3	0	Packers
4	0	Harrow & Packers
5	2	None
6	2	Harrows
7	2	Packers
8	2	Harrow & Packers
9	4	None
10	4	Harrows
11	4	Packers
12	4	Harrow & Packers

Figure 1. Plot layout

Results and Discussion

Soil moisture
Surface soil moisture content was measured using a 102 mm (4 in) core sampler with a 19 mm (0.75 in) diameter. The oven drying technique was used to determine soil moisture content. Three random samples were taken in each plot. Each day 108 soil moisture samples were taken over the plot. Soil moisture was determined prior to tillage and at 24 hour intervals up to seeding (Graphs 1, 2 and 3). Seeding took place 13 days after tillage of the plots. The same trends occurred in each plot over the 13 day sample period, with increased soil moisture content after precipitation.

Graph 4 outlines the average soil moisture data relating to tillage depth. After precipitation, any trends relating to soil moisture were less apparent. However, at a tillage depth of 102 mm (4 in), soil moisture content fell more rapidly than on plots tilled to a depth of zero or 51 mm (2 in) after the precipitation period.

Four levels of the secondary tillage factor were applied to the plots. Graph 5 illustrates the soil moisture with respect to the secondary tillage levels. Soil moisture decreased with warm, dry weather and increased with precipitation. No correlation between soil moisture content and secondary tillage was apparent.

Graph 6 shows the soil moisture data separated by blocks. No correlation between soil moisture content and block is apparent.

Since no correlation between soil moisture content and experimental factors occurred, soil moisture was graphed using differences between plot moisture just prior to the precipitation period and plot moisture before tillage. Precipitation occurred four days after tillage operations.

Graph 7 describes change in soil moisture related to tillage depth. A relationship between soil moisture and tillage depth appears to exist. The deeper the tillage, the greater the moisture loss. Plots which were not tilled gained soil moisture, while those tilled to depths of 51 mm (2 in) and 102 mm (4 in) lost 0.2 and 1.0 % moisture content, respectively.

Graph 8 compares the change in plot soil moisture content to before tillage moisture. Plots tilled to a 102 mm (4 in) depth experienced greater moisture loss than others plots. The steep slope of the plotted line illustrates a more rapid moisture loss with the 102 mm (4 in) tillage. Any differences between the no-till group and the group tilled at 51 mm (2 in) are less obvious.

Graph 9 reflects the change in soil moisture content related to tillage depth. No correlation is apparent.

Moisture data was also analyzed with respect to tillage depth, excluding any plots which received secondary tillage treatments (Graph 10). Results indicated tilled plots lost more moisture than plots which were not tilled. There was no significant difference between those tilled to 51 mm (2 in) or 102 mm (4 in) in depth. However, no-till plots appeared to retain more moisture than tilled plots.

Graph 1. Soil Moisture Data - Plots 1 to 4.

Graph 2. Soil Moisture Data - Plots 5 to 8.

Graph 3. Soil Moisture Data - Plots 9 to 12.

Graph 4. Soil Moisture Data vs Tillage Depth.

Graph 5. Soil Moisture Data vs SecondaryTillage Treatment.

Graph 6. Soil Moisture Data vs Block.

Graph 7. Soil Moisture Depletion vs Tillage Depth.

Graph 8. Change in Soil Moisture vs Tillage Depth.

Graph 9. Change in Soil Moisture vs Secondary Tillage Treatment.

Graph 10. Change in Soil Moisture vs Tillage Depth (Excluding 2nd Treatment).

An analysis of variance was performed on the soil moisture data (Table 1). Since variations in the results because of the block design were considered random effects, the analysis was performed with the interactions of the blocks and the factors in the error term. All interactions, including the blocks, were non-significant at a P-value above 10%. P-values above 10% indicated little or no relation between the experiment factors and soil moisture content.

Table 1. ANOVA of soil moisture content

Source of Variation	DF	Sum of Squares	Mean Square	F	P-Value
Treatment	11	35.067	3.188	0.774	>0.10
Blocks	2	7.845	3.922	0.953	>0.10
Depth (3)	2	11.012	5.506	1.337	>0.10
Tillage (4)	3	0.877	0.292	0.071	>0.10
Error	22	90.584	4.117
Total	35	133.496

.
Residue cover
Nine random residue cover samples were taken prior to tillage. Residue cover was determined by weighing 0.25 m² (2.291 ft²) samples of laying and standing stubble. Samples were cut, bagged and weighed. Before tillage, average stubble density was 2687.511 kg/ha (2390 lb/ac) with a standard deviation of 875.09 kg/ha (780.84 lb/ac). The coefficient of variation between the samples was 32%. Three random 0.25 m² (2.291 ft²) samples were taken from each plot after tillage. (Table 2 and Graph 11). Samples included both trash laying on the surface and standing stubble.

No correlation between tillage depth or secondary tillage to stubble amount is apparent. No correlation was attributed to the small sample size and the large variation in the initial trash density of the field. In addition, both standing and laying stubble were collected. Tests done where standing stubble was only collected may have caused more variations in the stubble density. Due to the number of samples taken and variation in initial field stubble measurements, an analysis of variance was not performed on the data.

Table 2. Residue Cover
Before Tillage

Average Concentration = 2687.51 Kg/Ha
Standard Deviation = 879.05 Kg/Ha
Coefficient Of Variation = 32.56%

Tillage Depth		Secondary Tillage	Stubble Density
mm	in		kg/ha
0	0	None	2106
0	0	Harrows	1671
0	0	Packers	874
0	0	Harrows & Packers	998
51	2	None	3189
51	2	Harrows	1675
51	2	Packers	3448
51	2	Harrows & Packers	943
102	4	None	2427
102	4	Harrows	2571
102	4	Packers	2258
102	4	Harrows & Packers	977

Soil particle size
Soil particle size samples were taken 24 hours after the tillage operations. A 30 kg (66 lb) sample of soil was taken from the top 102 mm (4 in) of each plot of the first block. Samples were placed in Canadian standard sieve numbers 3.5, 6, 8, 12, 16 and 20. Soil remaining on each sieve and passing through all sieves was weighed. Sieves were shook for 30 seconds in a mechanical shaker. Samples were weighed and percentages of soil in each sieve determined (Table 3).

Table 3. Soil Particle Analysis.

Depth		Secondary Tillage	Percentage of Total in Sieve (%) Sieve Size (mm)
mm	in		5.60 - +	3.35 - 5.60	2.36 - 3.35	1.70 - 2.36	1.18 - 1.70	0.85 - 1.18	0.00 - -.85
0	0	None	18.83	15.39	12.30	11.35	15.19	14.53	12.43
0	0	Harrow	18.94	15.03	11.93	11.09	15.51	14.83	12.66
0	0	Packer	18.62	15.35	12.30	11.28	13.19	13.18	16.08
0	0	Har. + Pac.	20.23	14.76	11.72	10.80	13.54	12.69	16.27
51	2	None	17.18	15.32	12.58	11.66	13.22	16.02	14.03
51	2	Harrow	19.95	15.29	11.97	10.99	14.78	13.84	13.19
51	2	Packer	20.38	14.76	11.53	10.57	16.85	12.43	13.50
51	2	Har. + Pac.	17.17	15.72	12.59	11.58	13.23	16.39	13.32
102	4	None	18.61	14.63	11.56	11.34	18.05	12.99	12.82
102	4	Harrow	16.10	14.73	11.84	12.32	20.12	12.26	12.64
102	4	Packer	18.33	15.61	12.53	11.44	11.96	13.96	16.18
102	4	Har. + Pac.	20.50	16.27	13.09	12.02	12.06	12.87	13.19

.
Only one sample was taken from the first twelve plots. Based on the small sample size, the analysis of variance was only applied to the different sieve sizes with the depth of tillage and secondary tillage being considered in the error term. ANOVA results indicated a significant difference in the mass percentage of contents of the sieves. Table 4 illustrates the ANOVA results. The significant difference in sieve sizes is evident in the results. Graphs 12, 13, and 14 illustrate the trends among the samples. An increase in percentages of soil particles in sieves occurred with the 5.6 and 1.18 mm sieves. No relationship between the experiment factors and particle distribution was apparent.

Table 4. ANOVA of soil particle size.

Source of Variation	DF	Sum of Squares	Mean Square	F	P-Value
Treatment	11	0.00	0.000
Blocks (7)	6	412.649	68.775	31.970	<0.005
Error	66	141.982	2.151
Total	83	554.631

.

Graph 12. Soil particle distribution 0 depth.

Graph 13. Soil particle distribution 51 mm (2 in) depth.

Graph 14. Soil particle distribution 102 mm (4 in) depth.

Crop emergence
Four random 0.25 m² (2.691 ft²) crop emergence samples were taken on each plot. Samples were taken by counting the number of plants in the 0.25 m² (2.691 ft²) area (Table 5). Emergence samples were taken 15 days after seeding. Graph 12 illustrates the average crop emergence for the plots.

Table 5. Crop emergence results.

Depth		Secondary Tillage	Block Number
mm	in		1				2				3
0	0	None	65	46	48	45	62	43	46	46	51	61	61	49
0	0	Harrow	47	50	40	51	49	44	55	43	37	66	45	51
0	0	Packer	61	50	57	47	45	39	61	49	52	52	56	51
0	0	Harrow + Packer	36	57	47	51	45	39	59	57	44	43	60	41
51	2	None	49	34	36	71	34	36	39	42	55	47	50	34
51	2	Harrow	25	45	56	54	40	76	35	37	62	55	48	42
51	2	Packer	50	46	38	53	51	44	54	55	43	41	58	54
51	2	Harrow + Packer	44	46	37	54	45	44	61	52	37	72	51	58
102	4	None	32	44	36	62	48	35	41	36	33	55	39	58
102	4	Harrow	40	65	55	48	43	48	40	38	54	55	40	42
102	4	Packer	65	47	32	49	52	31	44	45	51	43	50	45
102	4	Harrow + Packer	43	63	62	60	62	51	61	54	60	27	44	48

.

Graph 15. Average crop emergence.

The overall average plant count was 48.42 plants per 0.25 m² (2.691 ft²) sample. ANOVA results (Table 6) indicated no statistical significance among blocks or experimental factors related to crop emergence. No significance was probably due to the large amount of precipitation which brought all soil moisture contents and soil temperatures to approximately the same values.

Table 6. ANOVA of crop emergence.

Source of Variation	DF	Sum of Squares	Mean Square	F	P-Value
Treatments	11	303.440	27.585	1.934	>0.10
Blocks (3)	2	32.056	16.028	1.123	>0.10
Depth, D (3)	2	44.858	22.428	1.572	>0.10
Second, S (4)	3	80.422	26.807	1.879	>0.10
D, S	6	178.156	29.693	2.081	>0.10
Error	22	14.266
Total	35	649.350

.
Conclusions and Recommendations

No relationship between initial and final soil residue cover was apparent. Small sample numbers and large variations in initial residue levels contributed to no significant trends. In addition, both standing and laying stubble was used in the samples. In the future, laying and standing stubble should be separated and taken as two different samples.

When comparing crop emergence to experimental factors, no relationship was apparent. Precipitation on the plots was concluded as the main factor influencing no significant crop emergence results. If no rain fell, plots may have shown a significant difference in emergence.

Soil particle samples showed no significant difference among experimental factors. Difficulty in sampling and length of sampling time resulted in only 12 samples taken. A larger sample size may have resulted in more apparent trends among experimental factors. Statistical differences were found between sieve size, but the same trends occurred regardless of tillage depth or secondary tillage used. An increase in the percentage of total soil in the sieve occurred with the 5.6 mm and 1.18 mm sieves. Sampling techniques currently available for soil particle analysis are inconclusive and extremely time consuming. If future work into soil particle sizing is to be conducted, a new method of sample analysis should be addressed.

No statistically significant relationship was found between soil moisture and experimental factors. Since no relationship between soil moisture and experimental factors occurred, results were compared to pre-tillage moisture contents. When soil moisture was compared to pre-tillage moisture, relationships between moisture content and experimental factors existed. Deeper tillage caused greater moisture loss. Plots which were not tilled gained soil moisture. Those tilled to depths of 51 mm (2 in) and 102 mm (4 in) lost 0.2 and 1.0% moisture content, respectively. Moisture was also analyzed with tillage depth, excluding any plots which received secondary tillage treatments. Tilled plots lost more moisture than plots which were not tilled. There was no significant difference between those tilled to 51 mm (2 in) or 102 mm (4 in). However, no-till plots appeared to retain more moisture than tilled plots.

The Alberta Farm Machinery Research Centre (AFMRC) is at the forefront of machinery evaluations, applied and scientific research, and development of innovative agricultural technologies.