Introduction
During transition from conventional to organic farming, yields can be limited by nitrogen (N) availabilityReference Berry, Sylvester-Bradley, Philipps, Hatch, Cuttle, Rayns and Gosling1–Reference Scow, Somasco, Gunapala, Lau, Venette, Ferris, Miller and Shennan4. In comparison with synthetic fertilizers, organic N, from cover crops or other sources, is slowly broken down by microbial activityReference Drinkwater, Mosier, Syers and Freney5, Reference Tu, Louws, Creamer, Mueller, Brownie, Fager, Bell and Hu6. As a result, synchronizing crop N demand with microbial N release is a serious challenge for organic growersReference Pang and Letey3, Reference Jackson7. Little information is available about how farmers in transition to organic practices should manage short- and long-term N fertilityReference Pasakdee, Banuelos, Shennan and Cheng8.
Short-term N availability is essential for maximum productivity. Tillage increases available N for plantsReference Calderon, Jackson, Scow and Rolston9. Soil microbial populations are stimulated and immobilized N within soil organic matter is made available to plants after tillageReference Calderon, Jackson, Scow and Rolston9–Reference Calderon, Jackson, Scow and Rolston11. Furthermore, low carbon-to-nitrogen organic residues, such as leguminous cover crops, supply varying amounts of organic N depending on cover crop biomass and qualityReference Drinkwater, Wagoner and Sarrantonio12, Reference Quemada and Cabrera13. Finally, timely applications of precision-placed supplemental N enhance fertilizer use efficiency, which improves yield and reduces N leachingReference Pasakdee, Banuelos, Shennan and Cheng8.
Beyond transition, organic systems depend on long-term N-use efficiency to balance productivity with sustainabilityReference Drinkwater, Mosier, Syers and Freney5. Reduced tillage improves soil aggregation, which augments the capacity of the soil to act as a N reservoirReference Hoyt, Monks and Monaco14. High carbon-to-nitrogen cover crops such as grasses promote soil development with high organic inputs, scavenge and redistribute residual N above ground, and reduce N loss through the soil profileReference Jackson7. As a result, management strategies that integrate cover crop rotations and attempt to reduce tillage may result in more sustainable and productive organic systems.
The objectives of this study were to quantify the amount of supplemental N required to maximize organic broccoli yield, to evaluate the ability of various leguminous cover crops to supply N, and to compare the effect on N availability and broccoli yield when cover crops were incorporated with conventional tillage (CT) or retained as a surface mulch using no-tillage (NT) practice.
Materials and Methods
Experiments were conducted in the spring and fall of 2006 at the Virginia Tech, Kentland Agricultural Research Farm, near Blacksburg, VA, on a Hayter loam soil (fine-loamy mixed, mesic Ultic Hapludaf, pH 6.4 at 4–6 cm soil depth). The field plots were in their first year after a 3-year organic transition period; conventional field corn in 2002 was followed by applying aged chicken manure in 2003 (3 t ha−1) and growing combinations of cover crops and vegetable crops in 2003–2005 (Table 1).
Table 1. Cover and vegetable crop sequences, 2003–2006.
1 On 29 July 2005, annual ryegrass (Lolium multiflorum Lam.) was overseeded in established pepper plots. After final pepper harvest in late September, ryegrass grew well and completely covered all plots during fall 05–spring 06.
CP, cowpea; LL, lablab; SH, sunn hemp.
The experimental design was a randomized split split-block. Spring treatments were cover crops as main plots (3.7 m×130 m), tillage system as subplots (3.7 m×65 m) and nitrogen (N) rate as sub-subplots (3.7 m×16.3 m), with two replications. Fall treatments were arranged differently, with tillage system as main plots (3.7 m×130 m), cover crop as subplots (7.4 m×43.3 m) and sidedress N rate as sub-subplots (7.4 m×7.6 m), with four replications. For both crops, tillage systems were CT, in which the cover crops were flail mowed and incorporated 15-cm deep using a tractor-drawn rotary tiller, and NT, in which the cover crops were flailed mowed and the residues left on the surface as organic mulch. Two weeks after transplanting, subplots received four rates of organic N fertilizer (Tables 2 and 3). Total N was calculated per plot based on the treatment rate and applied uniformly by hand.
Table 2. Calendar of events.
1 Signifies operation in 2005.
Bt, Bacillus thuringiensis.
All plots consisted of either two or four 15-cm-high raised beds on 183 cm centers, and with cover crops, zone-seeded using a 10-row (18.3 cm apart) Tye drill (Agco Corporation, Duluth, GA). In both experiments, legume species were seeded on bed tops (73 cm wide) where broccoli transplants would be later set (grow zones) and grass species were seeded in alleyways (55 cm wide) on both sides of the grow-zone, which included bed edges, sides and bottoms. Cover crop seed was separated inside the hopper, so proper seed was drilled into its respective location.
For the spring crop, lablab (LL) and sunn hemp (SH) were seeded in grow zones and sorghum sudangrass (SSG) in alleyways. For the fall crop, LL, soybean (SB) and an SH and cowpea (CP) mixture (SH–CP) were seeded in grow zones and pearl millet (PM) in alleyways. Seeding rates were 45, 40, 56, 45, 90 and 15–75 kg ha−1 equivalent for SSG, PM, LL, SH, SB and SH–CP, respectively. Each row of the Tye drill was regulated separately to ensure a desirable seeding depth and subsequent uniform stand over the surface of the entire bed.
An early maturing broccoli cultivar (Major) was planted in spring, and a mid-season cultivar (Gypsy) planted in fall (Seedway Seed Co., Elizabethtown, PA). Broccoli transplants were produced at the Virginia Tech Horticulture Greenhouse for 6 weeks in 72-cell trays (55 cc/cell) containing an organic potting mix (McEnroe Lite, Millerton, NY). The plants were transported to the farm and acclimated in the shade with daily watering for 10 days before transplanting.
Prior to transplanting, aboveground cover crop biomass samples (0.25 m) were hand cut at the soil surface, dried at 70°C for 14 days, weighed, ground with a cyclone mill and analyzed for total N content using the Kjeldahl procedureReference Peterson and Chesters15. Broccoli transplants were set in twin rows on raised beds (50-cm apart and 37 cm in row) with minimal disturbance of surface mulch or soil, using the Subsurface Tiller–Transplanter (SST–T)Reference Morse, Vaughan and Belcher16. Harmony organic fertilizer (5N–2.2P–2.5K—Harmony Products, Inc., Harrisonburg, VA) and drip tubing (T-Tape-Rainflo Irrigation, LLC East Earl, PA) were applied at planting alongside each row using the SST–T. The same sources and rates of organic fertilizers were applied to both spring and fall broccoli crops (Table 3). All plots were drip irrigated as needed to supplement rainfall shortages. To ensure an adequate stand establishment, any wilted or dead plants were replaced up to 1 week after transplanting. A single fully expanded broccoli leaf was sampled from eight to ten plants per plot during early heading to compare N accumulation. Leaves were combined, dried and analyzed for N as described above.
Table 3. Delivery method, time of application, and sources of organic N fertilizer spring and fall, 2006.
1 Sodium nitrate (3.5 kg N ha−1) and fish and seaweed extract mix (1.5 kg N ha−1).
Weeds were controlled in both crops by swallow cultivation approximately 5–10 days before canopy closure of the twin broccoli rows. Weeding methods included hand pulling and hoeing by hand or push plow. Cover crop mulch in NT plots suppressed weed growth, resulting in fewer weeds and biomass than in CT (data not shown). In the spring, a Multivator (Mitchell Equipment, Marysville, OH) cultivated alleyways between beds in both NT and CT plots 1 week after transplanting.
Insect pests and diseases were kept low by applying recommended nutrients and irrigation to maximize plant health and vigor, planting beneficial insectary strips adjacent to the broccoli rows, and applying a single application of Dipel (Bacillus thuringiensis, Valent Corporation, Walnut Creek, CA) at 1.12 kg ha−1 3 weeks before the first harvest (Table 2).
Differences in cultivars and growing conditions between spring and fall production seasons affected the maturity rate of broccoli. The harvest began 7 weeks after planting in the spring and 11 weeks in the fall and continued for 10 and 14 days, respectively. Bead tightness and uniformity were the primary selection criteria for harvest. Broccoli heads were cut to a uniform length of 20 cm and checked for hollow stem, counted and weighed for total yield, and sorted into marketable and cull yield. Culls included damaged heads or heads <8 cm in diameter.
Data were analyzed by ANOVA using SAS (SAS Institute Inc., Cary, NC). Correlation coefficients were calculated using PROC REG with yield as the dependent variable and leaf N as the independent variable. Figures were constructed using CoPlot (CoHort Software, Monterey, CA). Significant statistical differences between treatment means and interactions were determined by LSD at P<0.050.
Results and Discussion
Supplemental N effects
Higher rates of sidedress N fertilizer significantly increased yield in both experiments (Table 4). However, the N requirement for broccoli was fulfilled at 112 kg ha−1 and there was no significant difference between 112 and 168 kg ha−1. Higher N rates statistically decreased cull yield, and total yield correlated with average head weight (data not presented). Applied N increased broccoli yield in previous studiesReference Cutcliffe17, Reference Kahn, Shilling, Brusewitz and McNew18. Our findings are consistent with Dufault and Waters who reported that increasing the N rate increased broccoli head weight and yield and decreased cull yieldReference Dufault and Waters19. Incidence of hollow stems increased with higher N rates (Table 4). However, effects were variable as indicated by significant cross-treatment interactions of T×CC×N in spring (P=0.036) and fall (P=0.028) (Table 5). Regardless, stem hollowness has been reported to correlate with increased N rate and is considered an indicator of rapid broccoli growth rateReference Hipp20.
Table 4. Sidedress N effects on broccoli yield, head weight, cull yield, stem hollowness and leaf N in spring and fall, 2006.
1 All heads <7.5 cm diameter.
2 Means within a column followed by different letters are significantly different at the 5% level according to the LSD test.
Higher rates of sidedress N increased both yield and percentage of N accumulation in leaf tissue. Optimum leaf N values for broccoli production range from 3.2 to 5.5Reference Wolf21. For spring, reported leaf N values (Table 4) and the x max value of 5.33 (Fig. 1) fell within the optimum range, while leaf N values for 112 and 168 kg ha−1 N (Table 4) and the x max value of 6.31 (Fig. 1) for fall were higher than optimum. Despite different seasonal responses, higher leaf N accumulation resulted in higher yield for both crops. Other studies identified similar relationships among rate of applied N, leaf tissue N accumulation and broccoli yieldReference Zebarth, Bowen and Toivonen22, Reference Everaarts and De Willigen23. Kowalenko and Hall showed that higher N fertilization rates significantly increased N in leaves and broccoli heads without stimulating vegetative growthReference Kowalenko and Hall24. As a result, broccoli leaf N content can be utilized as an indication of broccoli yield potential.
Figure 1. Quadratic correlation between leaf nitrogen and broccoli yield for spring (P<0.001; y=−9.89+6.18x−0.58x 2; R 2=0.80; x max=5.33) and fall (P<0.001; y=0.59+2.65x−0.21x 2; R 2=0.36; x max=6.31).
Table 5. ANOVA summary of treatment interactions for broccoli yield, head weight, cull yield, stem hollowness and leaf N in spring and fall, 2006.
1 T, tillage; CC, cover crop; N, nitrogen.
2 all heads <7.5 cm diameter.
Values in italics are statistically significant (P value <0.050).
Tillage effects
Yields of organic spring broccoli were not different between NT and CT, although NT decreased stem hollowness relative to CT (Table 6). These results are consistent with broccoli yield data from other NT studiesReference Abdul-Baki, Morse, Devine and Teasdale25–Reference Morse, Kingery and Buehring27. During the fall experiment, NT yields were lower than that in CT; however, NT still decreased stem hollowness. Mulched cover crops in NT systems are known to reduce soil temperature and slow plant growth compared with CT systemsReference Hoyt, Monks and Monaco14. During vegetative growth, a sudden drop to subzero temperatures for 4 consecutive days in mid-October caused uniform leaf damage and subsequent yield loss in all plots (Table 7). Under these conditions, smaller NT plants did not recover to the same extent as larger CT plants; thus, NT yields were affected to the greatest extent.
Table 6. Tillage effects on broccoli yield, head weight, cull yield, stem hollowness and leaf N in spring and fall, 2006.
1 CT, conventional tillage; NT, no tillage.
2 All heads <7.5 cm diameter.
Table 7. Minimum temperatures for October 2006 at the Kentland Research Farm. Days 13–16 are in bold type to emphasize the sudden drop to subzero temperature.
Tillage affects both seasonal and long-term N availability and thus can play an important role in determining total yield. In this study, a significant tillage×sidedress N interaction (P=0.002) occurred for broccoli leaf N (Fig. 2). Without additional N, availability and corresponding N uptake in NT plots were reduced. Research has shown that using tillage to incorporate plant residues releases earlier and higher rates of N compared with NT mulched soilReference Drinkwater, Janke and Rossoni-Longnecker10. Furthermore, as the rate of sidedress N increased, N uptake by NT plants approached levels of CT (Fig. 2). Our study is consistent with past research that showed that supplemental N may be required to increase N availability in NTReference Fox, Bandel, Phillips and Phillips28.
Figure 2. Tillage and supplemental N treatment interaction for broccoli leaf N (P=0.002) fall 2006; *, **, *** mean NT and CT values significantly different at P<0.05, 0.01 or 0.001, respectively.
Cover crop effects
The quantity and quality of cover crop residues have determined N contribution in past studiesReference Quemada and Cabrera13, Reference Puget and Drinkwater29, Reference Wagger30. Cowpea and tall growing SH created an environment where both crops could thrive due to differences in plant architecture, thereby resulting in higher cover crop biomass than the other treatments. In the spring, SH residues were greater than LL and had higher plant N (Table 8). In general, LL and SB had more leafy and succulent residues than SH and SH–CP. The relative proportion of leaves and stems largely determine the quality of cover crop residues, rate of mineralization and release of plant-available NReference Quemada and Cabrera13.
Table 8. Cover crop species, biomass, plant N percentage, and cover crop N contribution and their effects on broccoli leaf N and yield in spring and fall, 2006.
1 LL, lablab; SH, sunn hemp; SB, soybean; SH–CP, sunn hemp and cowpea mix.
2 Cover crop N contribution calculated as aboveground×plant N (%).
* Means within a column followed by different letters are significantly different at the 5% level according to the LSD test.
For both spring and fall, broccoli leaf N and yield were not influenced by cover crop treatment (Table 8). The lack of response could be attributed to the failure of N contained in CC residue to significantly affect available soil-N for the subsequent broccoli crop. This hypothesis is further supported by the lack of CC×N interactions for spring and fall broccoli yield and leaf N, even at lower N regimes of 0 and 56 kg ha−1 N (Table 5). Differences in quality of LL and SB may have compensated for greater biomass and N contribution of SH and SH–CP in terms of N mineralization. However, soil analysis was not conducted to confirm this hypothesis.
Conclusion
This study shows that both indigenous soil organic matter and applied fertilizer are required to maximize crop yield in transition soils. Over time, reliance on N fertilizers may decrease as buildup of soil organic matter and biologically buffered N availability are achieved from growing soil-building cover crops, applying organic waste products and a commitment to conservation agricultureReference Drinkwater, Mosier, Syers and Freney5, Reference Wang, Dalal, Moody and Smith31.
In our experiments, all plots were conducted on flat land and used drip irrigation to minimize soil moisture deficits. Therefore, any potential advantages of NT surface mulch in conserving soil moisture and increasing broccoli yield were minimized. Early season N availability and plant growth were greater in CT than in NT systems in our transition soil. Based on this and similar studiesReference Pasakdee, Banuelos, Shennan and Cheng8, when using organic high-biomass NT systems in transition soils, we recommend applying soluble organic fertilizers at planting and through the drip irrigation system (fertigation) to meet early season demands.
