Sustainable management of high tunnel organic vegetable production with short-season winter cover crops

Final Report for GS14-136

Project Type: Graduate Student
Funds awarded in 2014: $10,951.00
Projected End Date: 12/31/2016
Grant Recipient: University of Arkansas
Region: Southern
State: Arkansas
Graduate Student:
Major Professor:
Dr. Curt Rom
University of Arkansas
Expand All

Project Information

Summary:

This study investigated winter cover crops to improve soil quality and reduce nitrogen fertilizer inputs in organic high tunnel production systems. Cover crops included Austrian winter peas, bell beans, mustard, and Daikon radish, which were followed by summer tomatoes and fall broccoli. Winter peas contributed a greater amount of biomass nitrogen than all other treatments, which led to higher tomato leaf chlorophyll and a 48% increase in mean tomato yield compared to the control. Broccoli early-season leaf chlorophyll was also increased by the winter pea treatment and plant biomass was significantly increased.

Introduction

Although cover crops have been well studied for field production systems, their application in in high tunnel vegetable production has had limited study. Available information on the subject was limited to extension bulletins, newsletters, conference presentations, and production manuals (Baldwin, 2010; Blomgren and Frisch, 2007; Evans et al., 2011; Melendez and Rabin, 2012; Rivard, 2013). Despite negative and dismissive reports on high tunnel cover crops from extension agents and farmers in the Northeast (Blomgren and Frisch, 2007), preliminary studies in New Jersey and Mississippi have shown opportunity for winter cover crops to provide benefits to high tunnel vegetable production, including the uptake of leachable excess nutrients, weed suppression, and spring nitrogen release (Evans et al., 2011; Melendez and Rabin, 2012).

Arguments have been made to dismiss the applicability of cover crops in high tunnel systems (Blomgren and Frisch, 2007; Melendez and Rabin, 2012). The risk of soil erosion is minimized in a high tunnel, so cover crops are not specifically required to prevent soil loss. Cover crops grown throughout the winter in a warm high tunnel could provide habitat for overwintering pests. Season extension in a high tunnel minimizes the windows of time between crops, which eliminates niches that cover crops usually fill. And lastly, the capital investment in a high tunnel makes growers reluctant to use valuable high tunnel ground to grow a crop that does not provide immediate economic return.

Despite these issues, the reported and potential benefits of high tunnel cover crops justifies further research on the subject. To begin with, the timing of cover cropping in a high tunnel can minimize the negative aspects of the practice. Local growers have said that the time-period between mid-November and mid-February was the least productive season for high tunnel vegetable growers in the South due to cold temperatures and lack of light (Dr. C. Rom, personal communication). During this period, tunnels are vacant or idle.  By selecting this time period to grow a winter cover crop that can tolerate such environmental conditions, the lost revenue stream for the grower is minimized. The modified environment of a high tunnel, with increased soil and air temperatures (Wien, 2009), could also speed the growth of winter cover crops during the stated 90-day period in mid-winter, producing more green biomass than would be produced during the same time period in the field. It is thought that if a green manure crop were able to improve soil quality and decrease the requirement for purchased organic matter and fertilizer inputs, the cost savings for the grower could justify the additional management input.

Intensive vegetable production in a high tunnel requires inputs of organic materials to maintain soil organic matter, preserve soil quality, and ensure long-term productivity (Coleman, 1999; Lamont et al., 2003; Milner et al., 2009). Green manure cover crops have been shown to increase soil organic matter and nutrient cycling (Pieters, 1927; Powlson et al., 1987), and could replace purchased organic inputs in high tunnel production. Winter legumes have shown the potential to contribute significant amounts of nitrogen to subsequent vegetable crops in a field setting (Burket et al., 1997; Gaskell and Smith, 2007), which would reduce purchased fertilizer costs for high tunnel growers if similar rates of nitrogen contribution occurred in the high tunnel system. The deterioration of soil quality and the formation of hard pans due to frequent tillage is a documented problem in high tunnel production, which could be ameliorated by the ability of green manure cover crops to improve soil structure and aggregate formation (Hermawan and Bomke, 1997; Roberson et al., 1991; Tisdall and Oades, 1982) and improve the rooting depth of vegetable crops due to the “bio-drilling” effect of certain cover crops (Weil and Kremen, 2007). The ability of legume and brassica winter green manures to suppress soil-borne pathogens in vegetable crops (Monfort et al., 2007; Zhou and Everts, 2007) is another important benefit that winter cover crops could provide to high tunnel systems.

The ability of winter cover crops to improve the yield and performance of vegetable crops in field production systems have been well-documented in the scientific literature. It is the hypothesis of this study that short-season winter cover crops grown in a high tunnel, then mowed and incorporated, will improve the growth and yield of subsequent vegetable crops.

Literature Cited

Project Objectives:
  1. To evaluate four winter cover crop species for their ability to improve soil quality and supplement fertilizer requirements when grown as a green manure before a succession vegetable crops in a high tunnel system.
  2. To evaluate the effect of winter cover crops on vegetable crop growth and yield within a high tunnel production system. Vegetable crops will include tomato and broccoli.

Cooperators

Click linked name(s) to expand
  • Luke Freeman

Research

Materials and methods:

Experimental Parameters:

This study was conducted at the Arkansas Agriculture Research and Extension Center in Fayetteville, Arkansas (Latitude: 36.1N; Longitude: 94.1W; Altitude: 427m/1400ft; USDA Cold Hardiness Zone 6b; AHS Heat Zone 7), within the organic horticulture research block. The site had Captina silt loam soil with a pH between 6.5 and 6.7. Crops were grown inside a ClearSpanTM Quonset-style high tunnel (FarmTek, Dyersville, Iowa) with dimensions of 6 m by 41.5 m (20 ft by 136 ft), covered in a single layer of 6 mil polyethylene plastic glazing (with UV protection). Passive ventilation was provided by rolling down sidewall curtains and opening roll-up endwall doors. The high tunnel (HT) was opened for ventilation when the outside temperature exceeded 10 °C (50 °F) in sunny conditions or exceeded 16 °C (60 °F) in overcast conditions, to prevent the temperature inside the high tunnel from exceeding 32 °C (90 °F). All components of this study were managed in compliance with the USDA National Organic Program (USDA, 2000) requirements for certification, and the site had previously received organic management for 5 years.

Prior to the commencement of this study a summer cover crop of sorghum-sudangrass (Sorghum bicolor x S. bicolor var. sudanese) and cowpeas (Vigna unguiculata cv. Iron and Clay) were grown in the HT in order to create consistent soil conditions throughout the HT. In early November 2013, the cover crops were cut with a string-trimmer and the biomass was removed from the HT to create low-fertility conditions. The remaining stubble was mowed and the ground was irrigated with impact sprinklers to provide soil moisture for tillage. On November 14, the soil within HT was tilled to a depth of 8 cm (3 in) with a Toro® Dingo tiller. The remaining cover crop surface stubble was removed from the tunnel on Nov. 15 and the ground was tilled again to a depth of 8 cm (3 in) to prepare for cover crop planting.

Objective 1

The objective was to evaluate five winter cover crop treatments:  1) nontreated control, 2) Austrian winter peas (Pisum arvense), 3) bell beans (Vicia faba), 4) mustard (Brassica juncea cv. Kodiak), and 5) Daikon radish (Raphanus sativus var. longipinnatus), for their performance in a high tunnel system in addition to their ability to affect soil quality characteristics when grown as a green manure from mid-November to mid-March, before a succession of two vegetable crops in a HT system. Soil tests before and after the cover crop treatments, in addition to biomass measurements and foliar nutrient tests, were used to determine the organic matter and nitrogen contribution of the winter cover crops.

Cover Crop Planting and Management:

Cover crop seeds were planted on 16 November 2013, and 20 November 2014, following the same planting methods for both years. The cover crop treatments were arranged in a randomized complete block design (Figure A.1), with three blocks of each treatment of five cover crop. Therefore, each block (measuring 6 m by 12.2 m) was divided into five plots measuring 6 m by 2.4 m (20 ft by 8 ft). Treatment plots were assigned at random throughout each block using an online random number generator. Treatment 1 was the control, with no cover crops planted. Treatment 2 was planted with Austrian winter peas (Pisum arvense) at 1.46 kg per 100 m2 (3 lbs per 1000 square feet) (Cogger, 1997) or 0.22 kg (0.48 lbs) per plot. Treatment 3 was planted with bell beans (Vicia faba) at 1.46 kg per 100 m2 (3 lbs per 1000 square feet) (Cogger,1997) or 0.22 kg (0.48 lbs) per plot. Treatment 4 was planted with mustard (Brassica juncea cv. Kodiak) (Mighty Mustard®, Davidson Commodities, Spokane, WA) at 0.25 kg per 100 m2 (0.5 lbs per 1000 square feet) or 0.04 kg (0.08 lbs) per plot. Treatment 5 was planted with Daikon radish (Raphanus sativus) at 23 kg per ha (20 lbs per acre) (Clark, 2007) or 0.33 kg (0.73 lbs) per plot.

Cover crop seed was broadcast by hand, with each plot divided into quarters and broadcast in sections to create an even distribution. Mustard and radish seed were raked in to a depth of approximately 1.2 cm (0.5 in) with a bow rake. Winter pea and bell bean seed were tilled in to the soil at a depth of approximately 5 cm (2 in) with a small, front-tine tiller. The day after planting, the high tunnel was irrigated with overhead impact sprinklers to provide 31 mL of water per m2 (1.25 acre inches) Sprinklers were situated 1.7 m (5.5 ft) high on PVC risers, each sprinkler 3.7 m (12 ft) apart. Irrigation was provided as needed throughout the growth of the cover crops.

The cover crops were covered with floating row cover (Agribon AG-19) to protect from winter kill when the ambient temperature reaches -6.6 °C (20 °F) or below. An additional layer of 6 mm greenhouse plastic was used when temperatures were forecast to reach -12.2 °C (10 °F) or below.

The cover crop treatments were mowed on 8 April 2014 and 7 April 2015, or approximately 143 and 138 days after planting, respectively. A string-trimmer was used to mow down the cover crops with a board used as a shield to prevent the spread of biomass from the plot being mowed into adjacent plots. Following mowing, cover crop residue was incorporated with a rotary tiller mounted on a two-wheel tractor (BCS America, Portland, OR). Each plot was tilled individually with tiller tines cleaned between plots to prevent cross-plot contamination.  The high tunnel was irrigated with 25 L of water per m2 (1 acre inch) using impact sprinklers and closed for 21 days before planting of tomatoes to encourage breakdown of cover crop residues. Soil samples were taken from each treatment plot at the end of the three-week period to measure cover crop effect on soil quality.

Experimental Design:

The experiment was designed as a split-plot within a randomized complete block design, with year serving as the split plot. The high tunnel was blocked by location with three blocks of the five cover crop treatments (Figure A.1), providing three replications of each treatment within each year. Statistical analyses were performed with SAS 9.2 software (SAS Institute, Cary, NC) using PROC MIXED and PROC GLM.

Experimental Variables:

Experiment 1. Cover crop biomass and nutrient content. Height and biomass measurements were collected on the cover crop treatments on 28 March 2014 and 27 March 2015, one week prior to mowing. Three height measurements were taken of the standing cover crop per plot to determine mean height per plot. Biomass samples were then collected from three randomly placed 0.5 m2 quadrats in each treatment plot, dried in a forced-air drying oven at 50°C for one week, and weighed to determine dry weight biomass. A 100 gram sub-samples from each biomass sample was ground for aerial fraction nutrient analysis. Cover crop tissue nutrient concentrations were determined by inductively coupled plasma atomic emission spectrometry (SPECTRO ARCOS ICP, SPECTRO Analytical Instruments Inc, Mahwah, NJ) after HNO3 digestion at the University of Arkansas Division of Agriculture Soil Testing and Research Laboratory, in Fayetteville, Ar. Carbon and nitrogen concentration (percent dry weight basis) of the biomass samples were measured using an Elementar vario EL cube (Elementar Americas, Inc., Philadelphia, PA).  The total biomass N contribution for each treatment was calculated by multiplying the aerial tissue N concentration by the dry weight biomass (g/m2) to arrive at units of g/m2. Estimates of cover crop biomass N (kg) per hectare were calculated using the equation N g/m2 = 10x N kg/ha.

Experiment 2. Effect of winter cover crops on high tunnel soil quality.  Soil samples were collected from treatment plots at one week before cover crop incorporation (28 March 2014 and 28 March 2015) and three weeks after incorporation (26 April 2014 and 28 April 2015), with six 2.0cm (dia.) by 15 cm (depth) cores collected from each treatment plot and combined for each sample. Soil samples were analyzed for organic matter content, pH, electrical conductivity, bulk density, and soil C and N concentration. Soil organic matter was determined by weight loss on ignition (LOI) using the procedures described by Maguire and Heckendorn (2011). Bulk density measurements were collected on the same soil sample dates using metal cylinder rings measuring 600 mm tall and 550 mm in diameter, with three samples collected per plot and analyzed as sub-samples. Soil bulk density samples were dried in a forced-air oven at 55°C for one week and then weighed to determine bulk density values. Soil pH and EC were analyzed at the University of Arkansas Division of Agriculture Soil Testing and Research Laboratory using a 2:1 soil/water ratio. Carbon and nitrogen concentration (percent dry weight basis) of the soil samples were measured using an Elementar vario EL cube.

Objective 2:

The objective was to evaluate five winter cover crop treatments:  1) nontreated control, 2) Austrian winter peas (Pisum arvense), 3) bell beans (Vicia faba), 4) mustard (Brassica juncea cv. Kodiak), and 5) Daikon radish (Raphanus sativus var. longipinnatus), for their ability to effect the growth and yield of a succession of two vegetable crops in a HT system. Growth parameters including estimated leaf chlorophyll, foliar nitrogen concentration, and plant biomass were measured on the vegetable crops in addition to harvest yield measurements.

High Tunnel Crop Management:

In order to achieve this objective, following the winter cover crop season, a succession of two vegetable crops, including tomatoes (Lycopersicon lycopersicum, cv. ‘Plum Dandy’) and broccoli (Brassica oleracea var. italica, cv. ‘Bay Meadows’) were be grown in the high tunnel and fertilized at a 0.5x rate to determine the ability of the cover crop treatments to supplement fertilizer inputs. Yield and performance of the two vegetable crops were measured to evaluate the effect of the cover crop treatments compared to a no-treatment control.

The cover crop treatments (objective 1) were mowed and incorporated on 8 April 2014 and 7 April 2015, 143 and 138 days after planting, respectively. Approximately 30 days after the incorporation of the cover crops, two raised beds were created in the high tunnel with a “Junior” Model 1721-D bed shaper (Buckeye Tractor Co., Columbus Grove, Ohio). The beds were oriented parallel to the length of the high tunnel and were spaced 1.8 m (6 ft) apart on center, measuring 81 cm (32 in) wide, and raised 10 cm (4 in) above ground level. The beds were fertilized at half the recommended rate for tomato (McCraw, et al., n.d.), which was calculated at 84 kg N/ha (75 lbs N/acre). A pelletized organic fertilizer (Bradfield Organics®, Luscious Lawn & Garden™ 3N-0.4P-4.2K) was applied and incorporated into each bed to provide 0.47 kg (1.04 lbs) N per bed.

After fertilization, drip tape lines were placed, and the beds were covered with black plastic mulch, using a “Junior” Model 1723 mulch layer (Buckeye Tractor Co., Columbus Grove, Ohio). One drip tape line with 30 cm (12 in) emitter spacing was be placed running the length of each raised bed under the plastic mulch.

Vegetable Crop Rotation:  A series of two vegetable crops were grown in a seasonal rotation including tomato (Experiment 1) followed by broccoli (Experiment 2).  The crops did not constitute a treatment and were analyzed as individual experiments within the study to evaluate the residual effects of the preceding cover crops.

Experiment 1:

Tomatoes (Lycopersicon lycopersicum, cv. ‘Plum Dandy’) were started in the greenhouse on 11 March 2014 and 16 March 2015 and grown for 56 and 46 days, respectively, to be transplanted into the HT in early May. The tomato transplants were grown in organic potting media (Sunshine® Natural & Organic Professional Growing Mix, Sun Gro Horticulture, Vancouver, Canada) in 72 cell-packs, and fertilized with an organic fish hydrolysate fertilizer (Neptune’s Harvest Fish-Seaweed Blend 2N-1.3P-0.8K) once a week diluted to 250 ppm N. The tomatoes were transplanted into the raised beds in the high tunnel on 6 May 2014 and 1 May 2015 in single-rows, with 46 cm (18 in) between plants, making 5 plants per treatment plot. The two plants bordering the adjacent treatment plots served as guard plants, leaving three treatment plants per plot (Fig. A.2). Tomatoes were trellised using the stake-and-weave method, with a wooden stake between every three plants in the row and a steel t-post placed every 12 m (40 ft). The beds were fertilized before plating at 84 kg N/ha (75 lbs N/acre) using a pelletized organic fertilizer (Bradfield Organics®, Luscious Lawn & Garden™ 3N-0.4P-4.2K), which only provided half of the recommended N to the tomato crop based on recommendations from McCraw, et al., allowing for the cover crop N contribution to result in detectable differences in plant growth and yield. Transplants were also fertilized with a liquid fertilizer on the day of planting using (Neptune’s Harvest Fish-Seaweed Blend 2N-1.3P-0.8K) diluted to 250 ppm N with 237 mL (1 cup) applied around the base of each plant.

Pest Management: In 2014 tomato fruitworm (Helicoverpa zea) were first detected on 12 June with applications of Javelin® WG (active ingredient Bacillus thuringiensis, subspecies kurstaki) (Certis USA, L.L.C., Columbia, MD) mixed at 10 mL/L applied on 12 June, 26 June, and 15 July. Symptoms of the disease Southern blight (Sclerotium rolfsii) were fist observed on 5 July with plant collapse from the disease first observed on 7 July.  In 2015 applications of Javelin® WG began on 1 May to prevent seedling damage from the granulate cutworm (Feltia subterranean), which had been observed on broccoli during the previous season in 2014. Javelin® WG was reapplied on 1 June, 6 July, and 21 July to control tomato fruitworm (Helicoverpa zea). Contans® WG (active ingredient Coniothyrium minitans) was applied as a biological control for white mold (Sclerotinia sclerotiorum), which had been observed on the previous cover crop of Austrian winter pea, and for Southern blight (Sclerotium rolfsii), which had been diagnosed on the 2014 tomato crop. Contans® WG was mixed in water at 2.64 g/L and applied to the base of each tomato plant on 15 May at 236 mL (1 cup) per plant to achieve an application rate of 4.5 kg/ha. Despite the preventative application of Contans® WG, Southern blight infection was observed on tomatoes on 6 June with wilting and plant collapse occurring on 17 July. Green peach aphids (Myzus persicae) were observed in early June with applications of Aza-Direct® (active ingredient Azadirachtin [1.2%]) made on 9 June and 6 July at an application rate of 2.3 L/ha. Spider mites (Tetranychus spp.) were observed on 6 June on tomato plants in plots 1, 2, 10, and 15. JMS Style-Oil (active ingredient paraffinic oil 97.1%, JMS Flower Farms, Inc.) was applied to control spider mites on 10 July mixed in water at 10 mL/L. Due to persistence of spider mites, M-Pede® (active ingredient potassium salts of fatty acids 49%, Dow AgroSciences LLC) was applied on 21 July, tank mixed with water at 20 mL/L.

Experimental Variables: Weekly plant chlorophyll measurements were taken throughout the season from three tomato plants per treatment plot with a SPAD-502Plus Chlorophyll Meter (Konica Minolta, Inc.). Chlorophyll measurements were collected from 22 May to 24 July in 2014 and from 26 May to 7 July in 2015. The tomato crop was harvested on a weekly basis beginning on 26 June 2014 and 2 July 2015, which was 51 and 62 days after transplanting, respectively. Harvest continued until 22 July 2014 and 6 Aug. 2015. Ripe fruit was harvested from the treatment plants once a week until all fruit was harvested. Fruit was counted and weighed to determine yield per plot, yield per plant, and average fruit weight. Foliar samples were collected from three treatment plants per plot, dried, ground, and analyzed for nitrogen content using an Elementar vario EL cube. In 2014 foliar samples were collected on 24 July at the end of harvest. In 2015 foliar samples were collected on a weekly basis from 26 May (25 days after planting) to 30 June (60 days after planting) analyzed for N content. Plants were removed following the final harvest, with above-ground plant biomass dried at 63°C for 7 days and weighed to measure dry weight biomass.

Experimental Design:  The experiment was designed as a randomized complete block with three blocks of the five cover crop treatments within the HT. Statistical analysis was performed with SAS 9.2 software (SAS Institute, Cary, NC) using PROC MIXED and PROC GLM.

Experiment 2:

Broccoli (Brassica oleracea var. italica, cv. ‘Bay Meadows’) was started from seed in the greenhouse on 10 July 2014 and 15 July 2015. Transplants were grown in organic potting media (Sunshine® Natural & Organic Professional Growing Mix) in 72 cell-packs and fertilized with liquid fish hydrolysate (Neptune’s Harvest Fish-Seaweed Blend 2N-1.3P-0.8K) once a week diluted to 250 ppm N. Broccoli was transplanted into the high tunnel on 11 August 2014 and 12 August 2015 with 30 cm (12 in) between plants in double-rows spaced 46 cm (18 in) apart in each bed. There were 14 plants per treatment plot, with seven plants in each of two rows. The four plants bordering the adjacent treatment plots were guard plants, leaving ten treatment plants per plot (Fig A.2). Transplants were fertilized with Nitron Liquid Fish (2.6N-0.87P-0.22K, Nitron Industries, Fayetteville, AR) on the day of planting with fertilizer diluted to 250 ppm N and applied at 237 mL (1 cup) per plant. During the season broccoli plants were fertigated with liquid fish fertilizer (2.6N-0.87P-0.22K) injected through the drip tape lines on a weekly basis beginning in mid-Aug. to supply 56 kg N/ha (50 lbs N/acre) calculated as half of the N fertilizer rate recommended by Kahn, et al.

Pest management: In 2014, plant damage and plant loss from the granulate cutworm (Feltia subterranean) was severe after broccoli was transplanted into the high tunnel. Javelin® WG (Bacillus thuringiensis, subspecies kurstaki) was applied at 1.12 kg/ha on the day of planting (11 Aug.) with repeat applications on 13 Aug., 20 Aug., and 4 Sept. Replanting was required due to plant loss from cutworm damage, with 80 broccoli plants replaced on 19 Aug. using extra transplants grown in the greenhouse. Heat stress was also an issue, with daily high temperatures ranging from 32 to 36°C in the week after planting. To mitigate heat stress to the broccoli plants, Surround® WP (Kaolin clay) was sprayed on the broccoli foliage and sprayed on the black plastic mulch to reduce heat retention. Surround® was mixed in water at 60 g/L and sprayed on 14 Aug. and 20 Aug.

In 2015 prior to broccoli planting OxiDate® 2.0 (hydrogen dioxide 27.1% and peroxyacetic acid 2.0%) was injected through the drip tape lines at a 1:200 ratio for 1 hour to control the Sclerotinia sclerotiorum and Sclerotium rolfsii that had been identified in the high tunnel earlier in the season. Broccoli transplants were sprayed with Javelin® WG and Entrust® (spinosad, Dow AgroSciences LLC) on the day of planting (12 Aug.) to prevent cutworm damage, with repeat applications of Javelin® WG made on 17 Aug., 21 Aug., 9 Sept., and 28 Sept. Surround® WP was sprayed to minimize heat stress on the broccoli seedlings, with applications on 12 Aug., 17 Aug., and 21 Aug. Despite the preventative insecticide applications, cutworm damage still sustained with 29 plants replaced on 20 Aug. and another 8 plants replaced on 27 Aug with additional broccoli transplants grown in the greenhouse.

Experimental Variables: Weekly plant chlorophyll measurements were taken from three selected broccoli plants per treatment plot with a SPAD-502Plus Chlorophyll Meter (Konica Minolta, Inc.) on a weekly basis throughout the season. In 2014 chlorophyll measurements were collected from 16 Sept. to 4 Oct. and in 2015 measurements were collected from 27 Aug. to 2 Oct. The broccoli crop was harvested on a weekly basis throughout the month of October, with both primary heads and side shoots harvested and weighed. The harvested portion from each plot was weighed to determine yield per plot, average yield per plant, and average head weight. In 2015 each harvested head was also measured for head diameter (cm) and stem diameter (mm). Total above-ground biomass was also measured per plot on a fresh weight basis following the final harvest. Foliar samples were collected at harvest to determine nitrogen content of the treatment plants.

Experimental Design:  The experimental design was a randomized complete block with three blocks of the five cover crop treatments within the HT. Statistical analysis was be performed with SAS 9.2 software (SAS Institute, Cary, NC) using PROC MIXED and PROC GLM.

Research results and discussion:

Objective 1

In horticultural cropping systems, the long-term productivity of the soil is determined in part by soil quality characteristics including SOM, BD, pH and EC. The presence and availability of soil N is also an important factor in the productivity of horticultural crops. It has been reported that the intensive cultivation of crops within high tunnel systems may lead to soil quality decline, with decreases in SOM and increases in BD (Blomgren and Frisch, 2007). Winter cover crops were tested for their growth and production within a high tunnel system and for their effect on soil quality and soil fertility.

Within this study, four winter cover crop species were assessed for their growth and biomass production in a high tunnel production system. Differences were observed for cover crop height and dry biomass production (Figure 1, Figure 2) between years 2014 and 2015 of the study, which can be attributed to a change in soil N status. Across all treatments there was an increase in mean soil N by 402 mg/kg from the beginning of the vegetable crop production cycle (March) to the end of the production cycle (Nov.) in 2014, resulting in a higher soil N available to the cover crop treatments grown in the second year of the study. While the mustard and radish cover crops showed an increase in both height and biomass from 2014 to 2015 due to elevated soil N, Austrian winter pea produced a similar dry biomass between the two years (Figure 1, Figure 2). The consistency in winter pea biomass production despite changes in soil N status can be attributed its ability to fix atmospheric N as a legume (Ladd et al., 1981; Snapp et. al., 2005).

We found a positive correlation between cover crop height and dry biomass production (Figure 3) and a negative correlation between cover crop biomass and weedy biomass (Figure 5). The crops winter pea, mustard, and radish all resulted in suppressed weed growth when compared to the non-treated control (Figure 4).

Winter pea and bell bean cover crops contained the highest N concentration in aerial tissue (Figure 6), which is supported by the literature on leguminous cover crops (Ladd et al., 1981; Snapp et. al., 2005; Gaskell and Smith, 2007). This elevated tissue N for the legume cover crops resulted in greater biomass N that was incorporated into the soil after cover crop termination. In 2014 winter peas contributed a significantly greater amount of N compared to all other cover crops and in 2015 winter peas and bell beans contributed a similar amount of N through biomass incorporation (Figure 8). This did not result in a statistically significant increase in soil N as measured 30 days after incorporation, however (Table 5), despite the fact that measured soil N after winter pea cover crop incorporation was slightly greater than the control (78 mg/kg in 2014 and 33 mg/kg in 2015), with bell beans resulting in a similar elevated soil N in 2015.

The incorporation of N-rich biomass did have an effect on the soil C/N ratio. When comparing pre-incorporation soil C/N to post-incorporation C/N, both winter peas and bell beans showed a significant reduction in 2014 and 2015. Radish also resulted in a significant decrease in soil C/N in 2015 (Figure 11). Cover crop biomass N showed a negative correlation with soil C/N, confirming the observation that high-N cover crop biomass results in decreased soil C/N after biomass incorporation (Figure 12). Studies have linked decreased soil C/N ratios to increased availability of N for plant uptake.

A decrease in soil C/N can also be caused by decreased soil C. We measured a mean decrease in soil C from pre-incorporation to 30 days after cover crop incorporation by 0.8 g/kg in 2014 and 1.8 g/kg in 2015. Soil C did not increase significantly over the course of the study and actually decreased slightly when comparing measurements taken in April 2014 to April 2015 (mean decrease of 7 g/kg).

The decrease in soil C coincides with a decrease in soil organic matter from April 2014 to April 2015. Despite the fact that the cover crops grown contained a significant amount of C in their biomass, which was returned to the soil through incorporation, we did not measure increases in either soil C or soil OM. This conflicts with studies that have shown cover crops to lead to net increases in soil OM in field conditions (Pieters, 1927; Powlson et al., 1987). It is possible that the winter production of cover crops in the high tunnel did not produce enough biomass C to offset the soil C lost from tillage.

We also found a general trend in decreasing soil pH over time, though bell beans, mustard, and radish resulted in slightly elevated soil pH measurements compared to the control (Figure 13). This points to an ability for these three cover crop species to buffer the decrease in soil pH caused by annual cropping, but additions of liming amendments would still be required to keep soil pH within the 6.5 to 6.8 ideal range for vegetable crops.

The increase in soil EC over time in high tunnel cropping systems has been a concern for growers (Blomgren and Frisch, 2007). The results of this study did not point to cover crops as being able to prevent the increase in soil EC within high tunnel vegetable production. For most dates cover crop treatments did not result in soil EC that was significantly different than the control (Figure 14). In fact, on some sampling dates (April 2014 and Nov 2015) cover crop treatments resulted in soil EC levels greater than the control.

While cover crop treatment did not have a significant effect on soil bulk density (BD), BD decreased from 2014 to 2015, from 1.61 to 1.43 g/cm3 during the two-year cropping cycle. Because no differences were observed between the BD measured in the cover crop plots compared to the untreated control plots, it can be assumed that the decrease in soil BD can be attributed to the production practices, which included annual tillage and the formation of raised plasticulture beds for vegetable production. Cover crops have resulted in improved soil aggregation and porosity in prior studies (Clark ed., 2007; Snapp et al., 2005), which will result in lower soil BD. In the case of this study, however, any effect that the cover crop treatments would have had on soil BD was negated by the mechanical practices of bed shaping and cultivation.

Objective 2

Cover crop treatment had a greater effect on parameters measured on the tomato crop compared to broccoli parameters. On tomato the winter pea cover crop treatment resulted in greater foliar chlorophyll estimates compared to the control and mustard treatments across both years of the study. Complimenting this finding, tomato foliar N levels were increased by the winter pea treatment in 2015, pointing towards an increase in available N to the plant from this treatment which is being transported to the leaves and contributing to increased chlorophyll production in the leaf. Tomato dry weight biomass was also greatest following the winter pea treatment further supporting the idea that this cover crop treatment is resulting in increased plant-available N that is allowing for greater vegetative growth.

Looking at the tomato yield parameters there were not significant effects due to cover crop treatment. Instead, year was a main effect leading to increased total yield, marketable yield, marketable percent, average fruit weight, and yield per plant. This points to the effect that 2014 crop management had on 2015 tomato production in the high tunnel. It is possible that organic fertilizer residue had a long-term effect on the following cropping season, with residual fertilizer leading to increased plant available nutrients from 2014 to 2015. Also, the residual effect of cover crop treatments could provide an explanation for the increase in yield parameters for tomato from 2014 to 2015, but the lack of difference between cover crop treatments and the non-treated control does not make that explanation seem plausible. Within 2015 the winter pea cover crop resulted in numerically greater plant yield and fruit number, although statistical differences between treatments were not detected. The tomato fruit from the winter pea treatment were significantly smaller than fruit from the bell bean treatment and similar to the average fruit weight of the control treatment. This all points to the winter pea cover crop leading to a tomato crop that had greater N uptake, a greater amount of leaf chlorophyll, greater biomass, and a higher number of tomato fruit produced albeit smaller fruit that other treatments.

The broccoli crop also showed an increase in estimated leaf chlorophyll content from 2014 to 2015 and the winter pea cover crop treatment resulted in foliar N concentrations that were significantly greater than all other treatments in 2015. The winter pea cover crop also led to larger broccoli plants compared to the non-treated control over the two years of the study, although biomass of broccoli grown after the radish cover crop were similar in size. Average broccoli head weight showed an increase from 2014 to 2015, following the trend of increased leaf chlorophyll in the second year of the study. No differences were detected in yield parameters due to cover crop treatment, however.

In conclusion it appears that the leguminous cover crops winter pea and bell bean are resulting in increased N uptake by tomato and broccoli crops grown afterward in a high tunnel, with winter pea having more consistent results from 2014 to 2015. In tomato this increase N uptake is shown though increase foliar N concentrations and estimated leaf chlorophyll following the winter pea cover crop. Broccoli also showed a response to the winter pea cover crop in foliar N with the cover crop resulting in greater foliar N concentrations that all other treatments in 2015. The increased N uptake in tomato and broccoli from the winter pea cover crop resulted in increased plant biomass, with the tomato dry weight biomass following winter peas being significantly higher than all other treatments over the two years and the broccoli fresh weight biomass following winter peas being significantly greater than the broccoli biomass grown in the control plots. Yield results were not clear for either crop, though winter pea did lead to numerical increases in tomato yield and tomato fruit number. Austrian winter peas as a winter cover crop for high tunnel tomato and broccoli production does appear to have a benefit in plant growth, though continued research is needed to understand the full effects on crop yield.

Tables and Figures

Participation Summary

Educational & Outreach Activities

Participation Summary

Education/outreach description:

Publication of this research will be pursued in horticulture journals including HortTechnology and HortScience. Outreach activities included presentation of research findings at the Southern Region ASHS meeting in January 2015 and the national ASHS meeting in August 2016.

Project Outcomes

Project outcomes:

This research has demonstrated that winter cover crops can play a role improving the sustainability of intensive high tunnel vegetable production. Austrian winter peas were recommended based upon their performance in a high tunnel environment, being evaluated for biomass production and nitrogen fixation. However, within the time limit of the study no cover crop species had a significant improvement on soil organic matter and other soil quality parameters. The potential benefits for vegetable production were reported, including a slight increase in yield for tomato with fewer N fertilizer inputs. Results of the study will give high tunnel vegetable growers in the Southern Region biological management tools for more sustainable production options. Research results have been shared with horticultural scientists and students at the Southern Region ASHS meeting in January 2015 and the national ASHS meeting in August 2016.

Economic Analysis

None

Farmer Adoption

Farmer adoption has not yet been observed or documented.

Recommendations:

Areas needing additional study

Areas of additional study include the investigation of other winter cover crop species for high tunnel application. Because the legume and brassica cover crops investigated in the study did not lead to significant gains in soil OM, grass cover crop species should be investigated for C contributions and the increase of soil OM.

The rate of N release from green manure cover crops within the high tunnel and the rate of uptake by subsequent vegetable crops should be investigated to determine if the rate of N mineralization from N-rich cover crop residue matches the rate of uptake from vegetable crops.

Summer cover crops for high tunnels in southern locations should be investigated an another option for growers wanting to utilize biological tools to reduce fertilizer inputs and improve soil quality with a high tunnel cropping system. The growth of summer cover crops could coincide with plastic removal to allow the summer cover crops to receive rainfall and to prevent heat buildup with the high tunnel.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.