Weeds, Nitrogen, and Yield: Measuring the Effectiveness of an Organic No-Till System

Weeds, Nitrogen, and Yield: Measuring the Effectiveness of an Organic No-Till System

Summary

A comparison of organic no-till and conventionally tilled vegetable production systems in year one of the study determined that no-till did not negatively impact vegetable yields. Further, reducing nitrogen fertilizer either completely or by half did not significantly reduce vegetable yields in either tillage system. Pre-season soil analyses showed that plant available nitrogen and total nitrogen tended to be higher in no-till compared to tilled plots. Weed management inputs were considerably less using organic no-till compared to conventional tillage practices.

Objectives/Performance Targets

The objectives of our experiment are:
1. To assess two tillage treatments (no-till and tilled) of a rye and crimson clover cover crop for organic tomato and summer squash production;
2. To determine the interactions between two tillage treatments and three fertilization treatments (no nitrogen fertilizer, half the recommended rate of nitrogen fertilizer, and full recommended rate of nitrogen fertilizer) as they correspond to tomato and squash yields; and
3. To evaluate the management costs of a no-till system compared to that of a tilled system.

Accomplishments/Milestones

In March 2014, the research field, part of a 2,000 m² field at the Student Organic Farm (certified USDA Organic) located at the Clemson University Calhoun Field Research Area, was organized in a randomized complete block design comprising three blocks of four 6.1 m x 7.6 m split plots each (12 split plots total). The research field had been seeded to a cover crop mixture of cereal rye (Secale cereale, VNS) and crimson clover (Trifolium incarnatum, VNS) at a rate of 112 kg/ha and 39 kg/ha, respectively, in October 2013. All subplots were fertilized to correct slight P and K deficiencies in February and March 2014. Tomato seedlings, (‘Celebrity’) and summer squash (‘Success’) were started in the farm’s greenhouse complex on 26 March and 18 April, respectively. 

Three 0.5m² cover crop aboveground biomass samples were taken from the research field on 5 May when it was determined the cereal rye had matured adequately for no-till termination with a roller-crimper (Zadoks stage 70, end of flowering/beginning of grain fill). Clover maturity was not noted. The biomass samples were oven dried for 72 hrs at 55°C and then weighed. Additionally, cover crop tissue samples were sent to the Clemson University Ag Services Lab for for total N analysis. Based on the average dry weight and N analysis of samples, cover crop biomass at termination was approximately 8,400 kg/ha and total N an estimated 145 kg/ha. Cover crops in the 6 tilled plots were terminated on 5 May with a high-speed 1.5 m flail mower (Caroni Spa). The residue was subsequently incorporated into the soil with a disk harrow on the same day. Cover crops in the 6 no-till plots were terminated the following day, 6 May, with a rear-mounted 2.4 m roller-crimper (I & J Manufacturing) that had been filled with 225 kg of water for a total weight of 860 kg. The no-till cover crop required additional crimping on 8 May to terminate rye cover crop that had not been crimped adequately on 6 May. Unfortunately, because of tractor maintenance issues, the second round of crimping was accomplished with the farm’s small, 0.7 m roller-crimper with 113 kg of weight added for a total weight of 230 kg. The small roller-crimper was fitted to a 2-wheel walk-behind tractor. Labor hours spent preparing no-till (crimping) and tilled (mowing and disking) plots were recorded. The block design, which involved extra tractor turns, as well as the large amount of biomass to mow and incorporate influenced the total labor (5.5 hrs) spent preparing tilled plots. By comparison, no-till plots required only 1.8 hours to prepare. No-till field prep time would have been considerably less had it not been for the second round of crimping with the smaller implement. (The initial round of crimping with the large roller-crimper took only 20 mins.) A visual assessment of all no-till plots was made 4 weeks after crimping to estimate re-growth of cover crops. Rye re-growth was less than 1% in all plots – no clover re-growth was observed.

A visual assessment of all no-till plots was made 6 weeks after crimping to estimate percent ground coverage by weeds. Average weed coverage was 35% in tomato no-till plots and 25% in squash plots. The biggest weed problem by far in no-till plots was re-growth of a Japanese millet (Echinochloa esculenta) cover crop that had been used in the field the previous summer (ironically for summer/fall vegetable no-till) and had produced viable seed before termination. Tilled plots were weeded as needed (approximately once a week) beginning on 14 May. Weeding in tilled plots consisted of rototilling, flame weeding, and hoeing. In-season weeding in no-till plots was done as needed (approximately every 1-2 weeks) beginning on 28 May and consisted of over-the-top rotary mowing/weed-eating of weeds that had emerged through the cover crop residue. Hand-pulling or “rouging” of weeds in no-till plots was very limited to avoid disturbing the no-till mulch. Number of weeding events and time of each event were recorded for both tilled and no-till plots. Tilled tomato plots required 12 weeding events totaling 6.5 hrs of labor. Tilled squash plots required 11 weeding events totaling 5.7 hrs of labor. By comparison no-till squash and tomatoes required 4 weeding events each, totaling 1.2 and 1.3 hrs of labor, respectively. Tilled vegetable plots, in general, required greater earlier season weeding than no-till plots. For example, in the first month after cover crop termination, tilled plots required 6 hrs of weed management compared to just 1.2 hrs in no-till fields. 

Each of the 12 split plots was divided into 3 rows, spaced 1.5 m apart, for the N fertilizer portion of the experiment. On 9 May, prior to any application of N fertilizer in the field and prior to transplantation of vegetable crops, 0-15 cm composite soil samples were taken for each of the 36 rows. Soil samples were sent to the USDA-ARS Grassland Soil and Water Research Laboratory for soil health analysis (Soil Health Tool ver. 4.4). Soil plant available N and total N data were subjected to analysis of variance (ANOVA) using the “Fit Model” procedure, JMP version 11.0 (SAS Institute, Inc.), to determine the effect of tillage. Both plant available N and total N were significantly higher in no-till split plots (Fisher’s LSD P ≤ 0.05). Similarly, CO₂-C, a measure of microbial activity in the soil that was included in the ARS soil health analysis, was also higher, though not significantly, in no-till plots compared to tilled. Regardless of tillage treatment, there was simply a great deal of plant available N in the soil pre-season – the average plant available N for all 12 split plots was 58 kg/ha, equivalent to a split application of N fertilizer.

Five-week old tomato and two-week old squash seedlings were transplanted by hand on 9 May in the corresponding tilled and no-till tomato and squash plots – 15 tomato plants per row and 12 squash plants per row with standard 0.3 m spacing between all plants. Tomatoes were trellised on 22 May using the “Florida weave” technique. Pelletized N feather meal derived fertilizer (Nature Safe^® 13-0-0) was applied according to 3 treatment groups randomized per plot: 1) 100% (116 kg/ha, 2) 50% (58 kg/ha), and 3) none. N fertilizer was sidedressed in two split applications for both crops according to treatment group on 9 May at transplanting and on 6 June at first flowering. All seedlings had received starter fertilizer (Nature Safe^® 8-5-5) while in the greenhouse. Marketable squash and tomato yield data were collected by row for each harvest beginning on 9 June with the first squash harvest. Marketable vegetable yield data were subjected to analysis of variance (ANOVA) using the “Fit Model” procedure, JMP version 11.0 (SAS Institute, Inc.), to determine the effects of tillage, fertilizer, and tillage x fertilizer interaction. Regarding tomatoes, average row yields in no-till plots were significantly greater than in tilled plots (Fisher’s LSD P ≤ 0.05); there were no significant fertilizer effects, nor were there significant tillage x fertilizer interactions. Higher yields in no-till tomato plots were not a complete surprise. Two of the three tilled tomato plots had suffered plant losses due to a combination of Southern blight (Sclerotium rolfsii) and Pythium root rot (Pythium spp.). Disease was confirmed by a diagnostic analysis of sampled plants. In contrast, no-till tomatoes remained virtually disease free for the entire growing season. Average squash row yields were comparable between the two tillage systems; there were no significant fertilizer effects, nor were there significant tillage x fertilizer interactions. When row yield data for both crops were transformed from yield per row to yield per “unit of effort” per row (to account for cover crop termination, field preparation, and weeding labor) and run through the same analysis, no-till yields for both crops were significantly higher (Fisher’s LSD P ≤ 0.05) when compared to yields in the more labor-intensive tilled plots. 

The growing season ended with a final harvest of tomatoes on 29 July. The research field was subsequently disked and prepared for summer-fall cover cropping. Due to the aforementioned disease issues encountered, another field at the farm was chosen for year two of the study. Cereal rye and crimson clover were planted on 13 September at a rate of 112 kg/ha and 14 kg/ha, respectively. The lesser rate of clover was due to a malfunction in the farm’s overseeder at planting.

Impacts and Contributions/Outcomes

The results from year one of the study are promising, particularly those from yield/unit of effort analysis. We’ve been able to demonstrate weed management labor savings by through organic no-till and yields comparable to conventional tillage. Also important were the results of the fertilizer portion of the experiment, through which we were able to demonstrate N fertilizer savings by using high biomass cover crops.

Collaborators: