Site description and establishment.

The study was conducted at the University of Tennessee Organic Crops Unit near Knoxville, TN (35.88°N, 83.93°W). Establishment occurred in Aug. 2010 and the soil at the research site is a fine, kaolinitic, thermic Typic Paleudult (USDA-NRCS classification), with a clay loam surface texture (30% sand, 41% silt, 29% clay) and an initial soil pH of 6.9. The climate is humid subtropical with annual rainfall averaging 153 cm and mean annual temperature averaging 14 °C, with a January average of 3 °C and a July average of 25 °C. During the years of the study, annual temperatures at the site averaged 14 °C in 2010, 15 °C in 2011, and 17 °C in 2012. Precipitation at the site was below average in all years of the study, with 104 cm in 2010, 138 cm in 2011, and 106 cm in 2012 (data not shown). Additional weather data from the Organic Crops Unit from 2010 to 2012 is presented in Eichler Inwood et al. (2015). Conventional fertilizer, herbicides, and treated seeds had not been applied to the field containing the study site since Mar. 2008. The site was planted to wheat (Triticum aestivum L.) in Autumn 2007, rye (Secale cereale L.) and white clover (Trifolium repens L.) in Autumn 2008, soybean [Glycine max (L.) Merr.] in Spring 2009, and fallow (mowed for weed management) in 2010 before the start of the study.

The study was established as a randomized complete block design with four replicates. Three vegetable production treatments were randomly assigned to plots (each 2.4 m by 9.1 m) in each block. Management system treatments were as follows: 1) Till (+ACC), spring tillage of a winter cover crop with aboveground cover crop biomass (ACC) retained and soil covered by polyethylene mulch; 2) Till (−ACC), spring tillage of a winter cover crop with aboveground cover crop biomass (ACC) removed before tillage and soil covered by polyethylene mulch; and 3) RT, RT system with no spring tillage and mechanically terminated winter cover crop residue on the soil surface (Table 1). Treatments were applied to the same locations in each year.

Table 1.

Management system treatment factors.

View Table

Production system management.

To begin the study and end existing weeds on site, the field area was plowed (moldboard), disked (offset double disk), and harrowed (spring tine with rolling basket). On 14 Sept. 2010, stockpiled poultry (broiler) litter (a mixture of excreta and pine bedding materials) was applied to all treatments [Till (± ACC), RT] at a rate of 9 Mg·ha⁻¹ (28% moisture content; 2.2 Mg C/ha, 68 kg total N/ha, 105 kg P/ha, 118 kg K/ha) and then incorporated to an approximate depth of 20 cm with a rotary tiller (RTC240; Bush Hog, Inc., Selma, AL) to increase low soil P and K and provide nutrients for cover crop growth. Additional broiler litter was applied in Sept. 2011 at half the establishment rate (4.5 Mg·ha⁻¹, 43% moisture content; 0.88 Mg C/ha, 125 kg total N/ha, 56 kg P/ha, 51 kg K/ha) and incorporated. All plots were tilled with the rotary tiller at the time of broiler litter incorporation to facilitate cover crop establishment and standardize fall tillage among systems. In RT or rotational no till systems, this once annual tillage event serves to end perennial weeds, incorporate any remaining residues, incorporate fertility amendments, and create improved conditions for cover crop establishment (Mirsky et al., 2013). Broiler litter rates were calculated to approximate agronomic recommendations P fertilizers for the vegetable crops based on initial soil test P values (11.6 mg Mehlich I P/kg soil; Mehlich, 1953).

Cover crops were broadcast seeded on all treatments on 5 Oct. 2010 and 7 Oct. 2011, and then rolled to establish soil to seed contact. Plot were planted with a mixture of soft red winter wheat [cv. Haas Cover (2010) and cv. ForageMax (2011), 168 kg·ha⁻¹] and crimson clover (Trifolium incarnatum L. cv. Dixie, 11 kg·ha⁻¹). Cover crops in Till (+ACC) were ended at flowering (wheat = late boot stage) with a flail mower (SH74; Alamo Industrial, Seguin, TX; 5 May 2011) and then incorporated using a rotary tiller and a bed formed (0.9-m width) in the center of the plot, a drip irrigation line applied (5.6 L·min⁻¹ flow rate per 100 m; T-Tape, John Deere/T-Systems, San Diego, CA) and covered with black polyethylene mulch (0.032-mm embossed; Pliant-Berry Plastics, Evansville, IN). Till (−ACC) plots were similarly treated, with the exception that all aboveground cover crop biomass was harvested to a 7.5-cm stubble height and removed. RT cover crops were ended by rolling using the disengaged flail mower which has a roller and will flatten cover crops (Morse, 1999) at cover crop flowering (5 May 2011 and 20 Apr. 2012).

Eggplant (Solanum melongena L. cv. Traviata) transplants were produced in the greenhouse in 128-cell (36-cm³ cell volume) trays (Speedling Inc., Sun City, FL) with an organic seedling media (Premium Lite Growing Mix, McEnroe Organic Farm, Millerton, New York, NY). Transplants were planted in all plots on 11 May 2011, in a single 9.1-m row in the center of each bed and 45-cm spacing between plants. Transplants were planted by hand into polyethylene-mulched treatments after punching a small hole in the mulch. In RT treatments, transplants were planted by hand following pulling a coulter and shank through the soil to create a narrow furrow with minimal cover crop residue disturbance. To conform to crop rotation practices for organic production systems, watermelon [Citrullus lanatus (Thunb.) Matsum. et Nakai cv. Crimson Sweet] was produced during the 2012 season. Watermelon transplants were produced as described for eggplant and planted on 25 Apr. 2012 as described for eggplant except for 76-cm spacing between plants (in new plastic mulch, as applicable to treatment). To manage arthropod pests, all treatments used organically approved methods (dusting with diatomaceous earth and physical exclusion with spunbonded polypropylene rowcover in early crop growth stages). Aboveground vegetable crop residues were removed at the end of each season.

Cover crop biomass, crop performance, and weed evaluation.

Before cover crop incorporation, a sample of aboveground biomass was collected from a 0.09-m² area (30 × 30 cm). Samples were then oven-dried (65 °C for at least 48 h), weighed, and ground. Total C and N (and C:N ratio) of cover crop tissue (by species) was determined by combustion (Flash EA 1112 NC Soil Analyzer; Thermo Fisher Scientific Inc., Waltham, MA). Eggplant fruit were harvested 6, 19, and 26 July and 2, 9, 16, and 23 Aug. 2011 from a 5.4-m length of bed and watermelon fruit were harvested 6, 12, 20, and 25 July 2012 from a 7.6-m length of bed. Fruit were graded according to U.S. Department of Agriculture standards (USDA-AMS, 2006, 2013) into marketable (i.e., Fancy, U.S. No. 1 and U.S. No. 2) and nonmarketable categories, then counted and weighed. Harvest data were extrapolated by assuming 5467 m of row per ha on a commercial basis. In the week before first harvest, recently matured leaves were collected from each plot, oven-dried (65 °C for at least 48 h), and total N determined by combustion as described for cover crop tissue samples. Weed density was assessed in each plot ≈1 month following crop planting (7 June 2011, 30 May 2012). In polyethylene-mulched treatments, all weeds emerging through planting holes were counted. In RT, weeds were counted in a random 0.25-m² area (50 × 50 cm) within the crop row/bed (i.e., the plot area correlating to the polyethylene-mulched area in tilled treatments) in each plot. Weeds were removed by hand weeding following assessments.

Soil sampling and analysis.

Soil cores (0 to 15 cm depth; 1.75-cm internal diameter) were collected from each plot before litter application at the beginning of the study (14 Sept. 2010), and on 16 May, 14 July, and 29 Aug. 2011 and 15 May, 23 July, and 27 Aug. 2012. Soil cores from each plot were composited, air-dried, and then sieved (<2 mm). Soil ammonium-N (NH₄-N) and nitrate-N + nitrite-N [(NO₃-N + NO₂-N)] were assayed as described by Sims et al. (1995) and Sims (2006). Briefly, ≈5 g soil was extracted in 40 mL of 1 m KCl for 60 min on a reciprocating shaker, centrifuged (5 min at 3500 rpm) and then the supernatant was filtered (Whatman 42; Whatman Ltd., Kent, UK). Concentration of inorganic N constituents in filtrate was determined using a microplate reduction technique and absorbance measured at 550 nm (Powerwave XS; Biotek, Woosnooski, VT). Total soil N and C was determined by combustion as described for cover crop tissue samples.

Soil particulate organic matter N (POM-N) and C (POM-C) were measured using methods described by Marriott and Wander (2006). Briefly, 20 g of sieved, air-dried soil was weighed into a mesh-covered vial (53-μm mesh; Wildlife Supply Company, Yulee, FL), shaken with 5% sodium hexametaphosphate, then rinsed multiple times with deionized water until the rinse was clear. The retentate was dried (50 °C) and weighed before homogenization in a ball mill (PowerGen; Thermo Fisher Scientific). Homogenized samples were analyzed for total N and total C content (i.e., POM-N and POM-C) by combustion as described previously for cover crop tissue. Soil permanganate oxidizable carbon (Pox-C) was determined using methods described by Weil et al. (2003). In brief, soil was shaken with 0.02 m KMnO₄ for 2 min and then centrifuged (5 min at 3000 rpm). Absorbance of a diluted aliquot of the supernatant was measured at 550 nm (Powerwave XS). Final concentrations of C and N constituents in soils were determined based on exact weights of extracted soil and extract concentrations.

Statistical analysis.

Cover crop, soil quality, vegetable crop, and weed density data for each season were subject to mixed models analysis of variance (ANOVA) using PROC GLIMMIX in SAS software (version 9.4; SAS Institute Inc., Cary, NC) with management system, year, and the interaction of management system and year considered as fixed effects and block and the interaction of block and year considered as random effects. Soil inorganic nitrogen data were similarly analyzed, but with sampling time considered as a repeated variable. Weed data were log₁₀ transformed and soil inorganic N data were square root transformed to homogenize variance and provide a more normal distribution of outcome variable residuals before ANOVA; back-transformed means are reported. Effects and differences between means were considered significant at P ≤ 0.05. Means were separated using Fisher’s protected least significant difference procedure. Differences in soil constituents over time were determined by mean separation of the year effect within each treatment.

Cover crop biomass.

Average aboveground cover crop biomass at termination did not differ among management systems or years (P > 0.05) and averaged 10.3 Mg·ha⁻¹ (Table 2). This level of biomass accumulation is above the 8 to 9 Mg·ha⁻¹ generally reported as needed for weed control in RT cover crop systems (Mirsky et al., 2013; Mohler and Teasdale, 1993; Reberg-Horton et al., 2012; Smith et al., 2011). Although no studies have specifically evaluated wheat and crimson clover bicultures, biomass observed in the present study is above the median of that reported for mixtures of cereal rye and crimson clover in the region (Reberg-Horton et al., 2012). Total C in aboveground cover crop biomass averaged 4204 kg C/ha and total N averaged 216 kg N/ha, for a C:N ratio of 19:1. This ratio at less than 20:1 and would be expected to release N relatively quickly during decomposition (Whitmore, 1996). However, decomposition rates would likely be complicated by the advanced maturity of the residues (and greater lignin, hemicellulose, and phenolic content) and, for the RT system, placement at the soil surface where biological decomposition would proceed more slowly (Ranells and Wagger, 1996; Reberg-Horton et al., 2012). Biomass of wheat averaged 64% (6.6 Mg·ha⁻¹, C:N ratio of 22:1) of total biomass and crimson clover 36% (3.7 Mg·ha⁻¹, C:N ratio of 15:1; Table 2). Biomass of nonsown species in cover crop biomass was minimal (<0.1 Mg·ha⁻¹).

Table 2.

Mean aboveground cover crop dry biomass, biomass total carbon (C), biomass total nitrogen, (N) and biomass C:N ratio in Spring 2011 and 2012.

View Table

Vegetable crop weed density.

Weed density 1 month following vegetable crop planting was similar (P > 0.05) among treatments which used black polyethylene mulch [Till (±ACC)] in 2011, with an average of less than 5 weeds/m² (Table 3). In 2012, weed density was slightly less in the Till (+ACC) (3.5 weeds/m²) than Till (−ACC) (8.4 weeds/m²; Table 3). Nutsedges (Cyperus sp.), which often pierce polyethylene mulch in plasticulture production systems (Webster, 2005), were not present on the site. The primary weed species observed included field bindweed (Convolvulus arvensis), large crabgrass (Digitaria sanguinalis), and carpetweed (Mollugo verticillata). Weed density was more than 10-fold higher in RT in each season. This difference is perhaps not surprising, given the small area of the transplanting hole available for weed growth in the polyethylene-mulched Till treatments as compared with the RT treatment, but it does give an indication of relative differences in weed pressure among systems. Slightly lower weed density in the Till (+ACC) system as compared with Till (−ACC) may be a product of increased biological activity due to biomass incorporation which could potentially increase weed seed decomposition (Liebman and Davis, 2000). Weeds in RT could negatively impact yield when emerging during the critical weed free period, may interfere with harvest, and may complicate future phases of the crop rotation as compared with the lower weed density in the Till systems.

Table 3.

Mean within-bed weed density as affected by management system in late Spring 2011 (eggplant) and late Spring 2012 (watermelon).

View Table

Soil nitrogen and carbon.

Soil inorganic nitrogen generally did not differ among management systems at the sampled time points, with the exception of higher total soil inorganic N at the May 2011 sample (eggplant planting) and July 2011 sample (Table 4). In May 2011, the lowest soil inorganic N was observed in Till (+ACC) and RT systems (16.3 and 25.7 mg N/kg soil, respectively), and highest soil inorganic N from the Till (−ACC) system (50.0 mg N/kg soil). Higher inorganic N in the Till (−ACC) management system as compared with the Till (+ACC) management system suggests that the aboveground cover crop biomass slowed N mineralization in 2011. The C:N ratio of the aboveground cover crop biomass was not at a level (>20:1) expected to induce substantial N immobilization (Table 2) (Quemada and Cabrera, 1995). It is possible that the incorporated aboveground cover crop biomass may have reduced soil temperature and/or altered soil moisture content or soil water potential compared with Till (−ACC) and thus led to slower N mineralization (Cassman and Munns, 1980). It is also possible that due to biochemical composition of residues in combination with the soil matrix, mineralization kinetics of aboveground and belowground residues was slower than that of belowground residues (including rhizodeposition and unharvested stubble) alone (Frankenberger and Abdelmagid, 1985). This pattern was not observed in 2012, potentially due to slightly higher ambient temperatures in early May 2012 vs. 2011 (average 20 °C vs. 18 °C). Although our study did not monitor cumulative N mineralization, our results suggest that manipulation of aboveground cover crop biomass retention rates through forage removal may help to better target N availability to subsequent crops, but research on N mineralization in these crop management systems is needed to understand these potential relationships. At the same time, cover crop biomass removal would remove an appreciable amount of nutrients, which may necessitate changes to nutrient management strategies. Total soil N at the conclusion of the study in Fall 2012 did not differ among systems and ranged from an average of 1065 mg N/kg soil in the Till (+ACC) system to 1213 mg N/kg soil in the RT system (Fig. 1). For the RT system, this represents an increase of 29% from 938 mg N/kg soil in Fall 2010, or 275 mg N/kg soil [95% confidence interval (CI) (44 mg N/kg soil, 504 mg N/kg soil)], the only system to have significantly increased from 2010. Increased storage of total N in surface horizons of RT compared with conventional tillage systems is widely reported (Beare et al., 1994; Havlin et al., 1990), although the accumulation rates in our study were somewhat surprising given the relatively short, 2-year time period. Trends in POM-N were similar; the highest soil POM-N in Fall 2012 was observed in the RT system (409 mg POM-N/kg soil) and the least in tilled systems (Fig. 1; 283 to 310 mg POM-N/kg soil). For the RT system, this represents an average 87% [190 mg POM-N/kg soil, 95% CI (49 mg POM-N/kg soil, 331 mg POM-N/kg soil)] increase over 2 years (219 mg POM-N/kg in Fall 2010). In comparison, Eichler Inwood et al. (2015) reported an average of 460 mg POM-N/kg following 2 years of perennial organic forage management at the same research site, slightly higher than that observed from the RT system (409 mg POM-N/kg). Higher POM-N under no tillage was similarly reported in southeastern U.S. soils, with a 2-fold increase in POM-N at the 0 to 5-cm depth in no till vs. conventionally tilled soils following 13 years of management in Georgia (Beare et al., 1994). However, the authors reported no differences in POM-N at the 5- to 15-cm depth. In a study evaluating labile organic matter in several established farming system trials primarily in the midwestern United States, Marriott and Wander (2006) reported 35% to 40% enrichment in POM-N from organic systems as compared with respective conventional controls vs. just 8% to 10% enrichment in total soil N. This pattern of preferential enrichment of N the POM pool as compared with total N was similarly present in our study when compared with soil N at the beginning of the study.

Table 4.

Mean total soil inorganic nitrogen (N; ammonium-N + nitrate-N + nitrite-N) as affected by treatment at each sampling.

View Table

Total nitrogen (N) and particulate organic matter-N (POM-N) in Aug. 2012 as affected by management system. Within N fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total N and POM-N from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and Aug. 2012 for respective treatment plots. Raw values of replicates for total N and POM-N are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

View Full Size

Total nitrogen (N) and particulate organic matter-N (POM-N) in Aug. 2012 as affected by management system. Within N fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total N and POM-N from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and Aug. 2012 for respective treatment plots. Raw values of replicates for total N and POM-N are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Total nitrogen (N) and particulate organic matter-N (POM-N) in Aug. 2012 as affected by management system. Within N fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total N and POM-N from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and Aug. 2012 for respective treatment plots. Raw values of replicates for total N and POM-N are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Total soil C at the fall sampling in 2012 did not differ among systems and ranged from 12,500 Mg C/kg soil (23.4 Mg C/ha) in Till (+ACC) to 13,900 Mg C/kg soil (26.0 Mg C/ha) in RT (Fig. 2). The RT system was the only system where a significant increase was observed compared with 2010 (Fig. 2; from 10,400 Mg C/kg soil or 19.4 to 26.0 Mg C/ha). This increase of 6.6 Mg C/ha [95% CI (1.7, 11.5 Mg C/ha)] compares to 3.1 Mg C/ha total added over the two seasons in poultry litter, and aboveground cover crop biomass C accumulation rates of 4.2 Mg C/ha in each year (i.e., 8.4 Mg C/ha over the two seasons). Soil POM-C trends in Fall 2012 were similar, but the RT system had significantly greater soil POM-C (6803 mg POM-C/kg soil) than both Till systems and compared with the beginning of the study (4029 mg POM-C/kg soil; Fig. 2). Marriott and Wander (2006) reported that total soil C enrichment in organic systems (as compared with relevant conventional systems) was less than relative enrichment of POM-C. Our results were consistent with these trends, and indicated a 34% increase in total soil C as compared with a 69% increase observed in POM-C for the RT system, which is not surprising given the preferential allocation of newly sequestered C to the POM pool (Coulter et al., 2009; Hernandez-Ramirez et al., 2009).

Total carbon (C) and particulate organic matter-C (POM-C) in Aug. 2012 as affected by management system. Within C fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total C and POM-C from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and 2012 for respective treatment plots. Raw values of replicates for total C and POM-C are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

View Full Size

Total carbon (C) and particulate organic matter-C (POM-C) in Aug. 2012 as affected by management system. Within C fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total C and POM-C from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and 2012 for respective treatment plots. Raw values of replicates for total C and POM-C are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Total carbon (C) and particulate organic matter-C (POM-C) in Aug. 2012 as affected by management system. Within C fraction, means (bars) indicated by the same letter or no letters are not significantly different, P > 0.05. Dashed lines represent mean total C and POM-C from all plots at study initiation in Sept. 2010. Means indicated by an asterisk (*) indicate a significant (P < 0.05) difference in values between Aug. 2010 and 2012 for respective treatment plots. Raw values of replicates for total C and POM-C are represented by “●” and “○” symbols, respectively. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

The C:N ratio of particulate organic matter did not differ among treatments in Fall 2012 (average of 17:1), and did not significantly decrease over the course of the study. Trends in POx-C were similar to that observed with POM-C; the highest POx-C was observed from the RT system (255 mg C/kg soil) at the end of the study, but this was not significantly different from Till systems and no management system differed from POx-C at the beginning of the study (231 mg C/kg soil; Fig. 3). In other work at this site, Eichler Inwood et al. (2015) similarly reported lessened response of POx-C to management practices than POM-C after 2 years, potentially due to the short research time period given that POx-C reflects a more processed fraction of labile C (i.e., denser materials of smaller particle size) as compared with POM-C (Culman et al., 2012).

Permanganate oxidizable-carbon (POx-C) in Aug. 2012 as affected by management system. Means (bars) are not significantly different, P > 0.05. The dashed line represents mean POx-C from all plots at study initiation in Sept. 2010. Raw values of replicates are represented by “●” symbols. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

View Full Size

Permanganate oxidizable-carbon (POx-C) in Aug. 2012 as affected by management system. Means (bars) are not significantly different, P > 0.05. The dashed line represents mean POx-C from all plots at study initiation in Sept. 2010. Raw values of replicates are represented by “●” symbols. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Permanganate oxidizable-carbon (POx-C) in Aug. 2012 as affected by management system. Means (bars) are not significantly different, P > 0.05. The dashed line represents mean POx-C from all plots at study initiation in Sept. 2010. Raw values of replicates are represented by “●” symbols. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Vegetable crop performance.

For the 2011 season, eggplant yield did not differ among management systems (Fig. 4), with similar total yield (mean = 26.7 Mg·ha⁻¹), marketable yield (mean = 17.2 Mg·ha⁻¹), fancy grade yield (mean = 3.8 Mg·ha⁻¹), and culled fruit weight (mean = 9.5 Mg·ha⁻¹) across management systems. Eggplant leaf tissue N was higher in RT (42 mg N/g) than Till treatments (35 mg N/g), potentially indicating more synchronous N availability and plant demand in the RT systems (data not shown). It appears this did not have a significant impact on yield in our study, although it may be important under different environmental conditions. Early season differences in soil N availability (Table 4) did not significantly impact eggplant yields.

Mean total, marketable, fancy grade, and culled eggplant fruit yield in 2011 as affected by management system. Within fruit grade, means are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

View Full Size

Mean total, marketable, fancy grade, and culled eggplant fruit yield in 2011 as affected by management system. Within fruit grade, means are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Mean total, marketable, fancy grade, and culled eggplant fruit yield in 2011 as affected by management system. Within fruit grade, means are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

For the 2012 season, total, marketable, and fancy grade watermelon yield did not significantly differ among systems, although culled fruit yield was higher in the Till (+ACC) system (Fig. 5). Although not significant due to high variability represented by wide CIs (Fig. 5), total and marketable watermelon yields in RT were about half of that observed in Till (+ACC). Watermelon in RT may have been more negatively affected by weed pressure than eggplant, both due to competitive ability and the greater weed density in these systems in 2012 as compared with 2011 (Table 3). Disease symptoms, confirmed to be caused by Fusarium sp. were observed, but live plant counts indicated no significant differences among systems in plant mortality (P > 0.05; data not shown). Watermelon leaf tissue N did not differ among management systems at the time of sampling (P > 0.05; data not shown), averaging from 29 to 30 mg N/g in all systems.

Mean total, marketable, fancy grade, and culled watermelon fruit yield in 2012 as affected by management system. Within fruit grade, means indicated by the same letter or no letters are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

View Full Size

Mean total, marketable, fancy grade, and culled watermelon fruit yield in 2012 as affected by management system. Within fruit grade, means indicated by the same letter or no letters are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Mean total, marketable, fancy grade, and culled watermelon fruit yield in 2012 as affected by management system. Within fruit grade, means indicated by the same letter or no letters are not significantly different, P > 0.05. Error bars are the 95% confidence intervals of the mean. Management systems: Till (+ACC), spring tillage with aboveground cover crop biomass (ACC) retained; Till (−ACC), spring tillage with ACC removed; RT, reduced tillage system with no spring tillage.

Citation: HortScience 51, 8; 10.21273/HORTSCI.51.8.1038

• Download Figure

Over both seasons, the Till (−ACC) system had yields equivalent to the highest yielding treatments, and did not differ from the Till (+ACC) system. Although similar research in systems with aboveground cover crop biomass removal are lacking (e.g., as mechanically harvested or grazed forage), our results suggest that this type of system may be a practical option to provide additional harvestable biomass for organic growers who integrate crop and livestock production. Our results also suggest that soil C building effects of cover crops in these systems may be primarily dependent on root biomass given lack of differences between Till (−ACC) and Till (+ACC) systems, consistent with other reports on the relative contribution to soil C of shoot vs. root carbon (Rasse et al., 2005; Stockmann et al., 2013). Previous studies have indicated that organic, cover crop–based RT systems can, in some cases, be as productive as tilled systems under organic management in the humid southeastern United States {e.g., sweetpotato [Ipomoea batatas (L.) Lam.]} in North Carolina (Treadwell et al., 2007, 2008) and onion (Allium cepa L.) in North Carolina (Vollmer et al., 2010)]. In organic systems in other regions, cover crop–based RT yields were similar to conventional tillage in tomato (Lycopersicon esculentum Mill.) production in Iowa (Delate et al., 2012) and muskmelon (Cucumis melo L.) and zucchini (Cucurbita pepo L.) production in Italy (Canali et al., 2015). However, RT systems were a near complete failure in a California study with a number of crops, including eggplant (Luna et al., 2012), in which the primary causes for poor crop performance in the RT system included failure to completely kill cover crops, reduced soil moisture, and impracticability of in-season mechanical weed control.

Our results suggest that after 2 years of management, the greatest improvement in soil quality was observed from the RT system, which was greater than that observed in either more intensively tilled systems. However, weed densities were much higher in the RT system. Due to clear soil quality benefits, increased research in crop production aspects of organic RT systems for vegetable production is critically needed. Our results also suggest that given no negative effects of aboveground cover crop biomass removal were observed in terms of soil quality or vegetable crop yield in tilled systems, retaining root and stubble biomass of these cover crops may be adequate for providing cropping system services to subsequent crops, at least in the short term.