Organic Kale and Cereal Rye Grain Production Following a Sunn Hemp Cover Crop

: A four-year ﬁeld experiment was initiated in 2011 at the EV. Smith Research Station, in central Alabama, to determine the e ﬀ ect of sunn hemp ( Crotalaria juncea L.) termination methods on organically grown kale ( Brassica oleracea, var. acephala L.) for fresh market and cereal rye (Secale cereale , L.) for grain. Three di ﬀ erent termination methods for the sunn hemp cover crop were chosen: (1) rolling / crimping with an experimental two-stage roller / crimper, (2) rotary mowing, and (3) rotary mowing with incorporation (disking). Kale plots were harvested in the winter and rye plots were harvested in the following spring. Kale plots were fallow from January to June (kept mowed) until planting sunn hemp again across all plots in late spring of the next growing season. Over four growing seasons, average sunn hemp biomass (dry basis) was 10,981 kg ha − 1 with plant height of 2.4 m. The average C / N ratio of sunn hemp was 23:1. Sunn hemp biomass amounts di ﬀ ered among growing seasons (from 5589 to 14,720 kg ha − 1 ) due to di ﬀ erent weather conditions. Kale yield also varied across growing seasons, with the highest yield of 17,565 kg ha − 1 measured in 2012 and the lowest (3915 kg ha − 1 ) in 2014 due to massive weed pressure. Generally, sunn hemp residue management a ﬀ ected kale yield, with greater yields measured for mowed and incorporated residue (15,054 kg ha − 1 ) compared with lower yields for mowed (6758 kg ha − 1 ) and rolled sunn hemp (5559 kg ha − 1 ). Lower yields were related to poor kale seed-to-soil contact (hair pinning) from large amounts of sunn hemp residue on the soil surface. Over four growing seasons, cereal rye grain yield varied among growing seasons, with an average yield of 1358 kg ha − 1 . Moreover, sunn hemp residue treatments a ﬀ ected grain yield, with greater yields for rolled (1419 kg ha − 1 ) and mowed residue (1467 kg ha − 1 ) compared with a lower yield (1187 kg ha − 1 ) for mowed and incorporated sunn hemp residue.


Introduction
In conservation systems, cover crops are utilized to improve soil properties and enhance cash crop growth. During the past decade, use of cover crops in conservation agriculture has steadily increased. Benefits associated with cover crops include reduced soil erosion, reduced runoff, and increased infiltration and water holding capacity. In addition to improved soil quality and soil water, legume cover crops can fix nitrogen from the air and release the nitrogen into the soil to be used by subsequent cash crops. Legume cover crops are especially important in organic production systems, where commercial fertilizer options are limited and very expensive. In addition, production of healthy food and vegetables by local small farms is on the rise as US customers recognize a need to consume good-quality food and vegetables. Grocery stores and restaurants have found success in marketing locally grown and organic vegetables to their customers for added value and profit. Kale (Brassica oleracea, var. acephala L.) offers an abundance of vitamins and minerals beneficial to human health. It is one of the most cold hardy of all vegetables, able to withstand temperatures below freezing down to minus 8 • C. Brassicas, including kale, are an excellent source of vitamin K, vitamin A (in the form of carotenoids), manganese, vitamin C, dietary fiber and calcium. In addition, brassicas are a very good source of vitamin B1, vitamin B6 and iron [1].
According to the National Agricultural Statistics Service [2], in the United States, 2500 farms reported harvesting kale in 2012, which is a 262% increase from 954 farms in 2007. Moreover, the harvested area of kale production for fresh market significantly increased from 1616 hectares in 2007 to 2532 hectares in 2012 [2,3] Kale can be cultivated in a wide range of soils. The best production can be achieved on low acid or neutral (pH 6-6.5) deep soils with loamy texture and proper water and air capacity. Brassica plants have a high nutrient uptake demand. Thus, basic soil preparation includes deep plowing with incorporation of high amounts of organic manure [4]. Kale crops respond positively to nitrogen fertilizer by improving vegetative growth and delaying premature bolting, although concern about nitrate (NO3-N) accumulation in plant tissues and environmental pollution should be emphasized [5].
To reduce nitrogen fertilizer use, there is a need to determine the ability of a legume cover crop sunn hemp (Crotalaria juncea L.) to fix and release atmospheric nitrogen that can be available to a cash crop (i.e., kale). Sunn hemp (summer legume) originated in India and Pakistan and has major benefits including protection from soil erosion, building organic matter from its large amount of biomass (from 5560 to 6670 kg ha −1 ) as well as producing from 133 to 156 kg ha −1 of nitrogen [6] in a short period of time. Sunn hemp seeds are produced mainly in India, Columbia, South Africa, and in the USA in Hawaii [7]. According to USDA-NRCS [6], biomass production of sunn hemp as a summer cover crop in Alabama can exceed 5600 kg ha −1 biomass and 134 kg ha −1 of nitrogen. Balkcom and Reeves [8] reported that average sunn hemp biomass production in central Alabama was 7600 kg ha −1 , with a nitrogen content of 144 kg ha −1 .
Sunn hemp management, such as planting date, can affect the performance of a subsequent crop. Balkcom et al. [9] examined how two different sunn hemp planting dates following wheat (Triticum aestivum, L.) harvest (early and late June) and corn (Zea mays L.) harvest (late August and early September) affected biomass production of cereal rye (Secale cereale L.) grown as a winter cover crop. Rye biomass production following sunn hemp planted after wheat averaged 38% greater over two growing seasons compared to fallow. Rye biomass production following sunn hemp planted after corn was equivalent to fallow over the two growing seasons. Despite the ability of sunn hemp to produce significant biomass within short periods of time, Balkcom et al. [9] attributed the lower growing degree accumulation following corn limited sunn hemp biomass production. As a result, the N benefit was minimal compared to sunn hemp following wheat.
Termination is also an important management consideration. In no-till systems, appropriate termination methods of cover crop residue is essential. Planting cash crop seeds into soil with large amounts of cover crop residue on the soil surface can be difficult. Many researchers reported that large amounts of different cover crop residues caused a hair pinning condition where seeds did not have adequate seed-to-soil contact for optimum germination and growth [10]. A previous field study conducted by Kornecki et al. [11,12]) with no-till cotton (Gossypium hirsutum L.) indicated that large amounts of cereal rye residue on the soil surface caused wrapping of cover crop residue on row cleaners and coulters that prevented a good contact between seeds and soil, which created the hair pinning condition.
For centuries, cereal rye has been an important grain largely consumed in different parts of the world-Scandinavia, predominantly with influence in Finland, Russia, Germany, Poland, Ukraine and other regions of the former Soviet Union with harsh, colder climates. Rye consumption is closely associated with regional cuisines. Cereal rye is a weather-resilient cereal grain, capable of being grown in soils and conditions inhospitable to wheat and other grains [13].
Traditionally used in breads, crispbreads, and fermented beverages, rye can be substituted for wheat, brown rice, spelt, or other grains in a wide variety of dishes, both spicy and sweet. In the United States, it is mostly valued as a minor grain; however, its use has increased considerably in recent decades due its perceived perception for improved nutritional quality. The growing interest of rye among consumers as a whole and processed grain is mainly due to the potential health benefits related to bioactive enriched whole grain with the potential reduced risk of chronic diseases such as cardiovascular diseases, certain types of cancer and type 2 diabetes [14][15][16]. Further, the role of high dietary fiber in rye towards positive health benefits has been widely reported [15,17]. Similarly, previous studies have also suggested that high bioactive profiles such as phenolic acids and lignans in cereals, particularly in rye, can contribute to its health promoting functions [18,19].
Cover crops can be managed with a variety of methods. One is mechanical termination utilizing rolling/crimping technology. This requires injuring the plant with the crimping bars without cutting stems [20]. A common practice to accelerate the termination process is to utilize commercial herbicides such as Glyphosate (Roundup TM ) as a supplement to rolling/crimping. This practice, however, is not permitted in organic systems. Because of this restriction, mechanical termination by rollers/crimpers must be as effective as chemical applications. In conservation systems, rolling/crimping is especially important to manage tall cover crops such as sunn hemp. Flattening and crimping of a living cover crop scarifies and damages plant tissue, which accelerates its termination by promoting desiccation. Dried-out residue forms a thick mulch that covers the soil surface; this reduces soil erosion, reduces weed germination and growth, increases infiltration and conserves water for the following cash crop [21]. Other methods are mowing or mowing with incorporating cover crop residues, but these methods can generate problems such as re-growth of cover crops and exposing the soil surface to rainfall causing soil erosion, Additionally, loose residue can interfere with cash crop planting.
Therefore, the objective of this study was to determine the effect of sunn hemp cover crop termination methods (rolling/crimping, mowing only, and mowing with incorporation of sunn hemp) on organically grown kale and cereal rye for grain ( Figure 1).
Agronomy 2020, 10, x 3 of 16 of rye among consumers as a whole and processed grain is mainly due to the potential health benefits related to bioactive enriched whole grain with the potential reduced risk of chronic diseases such as cardiovascular diseases, certain types of cancer and type 2 diabetes [14][15][16]. Further, the role of high dietary fiber in rye towards positive health benefits has been widely reported [15,17]. Similarly, previous studies have also suggested that high bioactive profiles such as phenolic acids and lignans in cereals, particularly in rye, can contribute to its health promoting functions [18,19]. Cover crops can be managed with a variety of methods. One is mechanical termination utilizing rolling/crimping technology. This requires injuring the plant with the crimping bars without cutting stems [20]. A common practice to accelerate the termination process is to utilize commercial herbicides such as Glyphosate (Roundup TM ) as a supplement to rolling/crimping. This practice, however, is not permitted in organic systems. Because of this restriction, mechanical termination by rollers/crimpers must be as effective as chemical applications. In conservation systems, rolling/crimping is especially important to manage tall cover crops such as sunn hemp. Flattening and crimping of a living cover crop scarifies and damages plant tissue, which accelerates its termination by promoting desiccation. Dried-out residue forms a thick mulch that covers the soil surface; this reduces soil erosion, reduces weed germination and growth, increases infiltration and conserves water for the following cash crop [21]. Other methods are mowing or mowing with incorporating cover crop residues, but these methods can generate problems such as re-growth of cover crops and exposing the soil surface to rainfall causing soil erosion, Additionally, loose residue can interfere with cash crop planting.
Therefore, the objective of this study was to determine the effect of sunn hemp cover crop termination methods (rolling/crimping, mowing only, and mowing with incorporation of sunn hemp) on organically grown kale and cereal rye for grain ( Figure 1).

Materials and Methods
A field experiment was initiated in 2011 at the EV. Smith Research Station, Field Crops Unit (32°25′19″ N, 85°53′7″ W), near Shorter, in central Alabama on Marvyn Series soil, having taxonomic class: fine-loamy, kaolinitic, thermic Typic Kanhapludults. The Marvyn series consists of very deep, well drained, moderately permeable soils on uplands of the Southern Coastal Plain. At the top horizon (from 0 to 18 cm), soil is a dark grayish brown loamy sand; weak fine granular structure; very friable; many fine roots; about 5 percent, by volume, rounded pebbles less than one inch in diameter; moderately acid; abrupt smooth boundary. This area is characterized by a humid subtropical climate, with an average annual precipitation during four years of conducting experiment of 1185 mm, with an average maximum temperature of 24.1 °C and minimum temperature of 10.7 °C. [22]. Monthly precipitation, maximum and minimum temperature between May 2011 and June 2015 are presented in Figure 2.

Materials and Methods
A field experiment was initiated in 2011 at the EV. Smith Research Station, Field Crops Unit (32 • 25 19 N, 85 • 53 7 W), near Shorter, in central Alabama on Marvyn Series soil, having taxonomic class: fine-loamy, kaolinitic, thermic Typic Kanhapludults. The Marvyn series consists of very deep, well drained, moderately permeable soils on uplands of the Southern Coastal Plain. At the top horizon (from 0 to 18 cm), soil is a dark grayish brown loamy sand; weak fine granular structure; very friable; many fine roots; about 5 percent, by volume, rounded pebbles less than one inch in diameter; moderately acid; abrupt smooth boundary. This area is characterized by a humid subtropical climate, with an average annual precipitation during four years of conducting experiment of 1185 mm, with an average maximum temperature of 24.1 • C and minimum temperature of 10.7 • C. [22]. Monthly precipitation, maximum and minimum temperature between May 2011 and June 2015 are presented in Figure 2. Three different cover crop management treatments (main effects) were used to study effects of termination methods for sunn hemp (seeds from South Africa, cultivar: not stated) as a cover crop on the yield of two selected organic cash crops cereal rye (for grain) and kale for fresh market consumption. The following treatments to manage sunn hemp were applied: (1) Rolling/crimping sunn hemp using an experimental two-stage roller/crimper ( Figure 3a).  The experiment was a randomized complete block design with four replications for each treatment ( Figure 4). The experiment was initiated in late spring (May 25 2011) by drilling (no-till) sunn hemp (45 kg ha −1 ) using a Tye no-till drill (Agco, Duluth, GA, USA; 2.13 m width) across the experimental area (24 plots), and was repeated each growing season (May-June). Sunn hemp was terminated in early fall (September-October) of each growing season using previously described termination methods (Figure 3). A complete list of field activities for the whole experiment is shown in Table 1. Kale seeds (Siberian Dwarf Variety) were planted into sunn hemp residue (11 kg ha −1 ) with different residue treatments three weeks after its termination (end of September and October) in half of the experimental area using the Tye no-till drill. To obtain a 0.38 m row spacing for kale, every Three different cover crop management treatments (main effects) were used to study effects of termination methods for sunn hemp (seeds from South Africa, cultivar: not stated) as a cover crop on the yield of two selected organic cash crops cereal rye (for grain) and kale for fresh market consumption. The following treatments to manage sunn hemp were applied: (1) Rolling/crimping sunn hemp using an experimental two-stage roller/crimper ( Figure 3a). Three different cover crop management treatments (main effects) were used to study effects of termination methods for sunn hemp (seeds from South Africa, cultivar: not stated) as a cover crop on the yield of two selected organic cash crops cereal rye (for grain) and kale for fresh market consumption. The following treatments to manage sunn hemp were applied: (1) Rolling/crimping sunn hemp using an experimental two-stage roller/crimper ( Figure 3a).  The experiment was a randomized complete block design with four replications for each treatment ( Figure 4). The experiment was initiated in late spring (May 25 2011) by drilling (no-till) sunn hemp (45 kg ha −1 ) using a Tye no-till drill (Agco, Duluth, GA, USA; 2.13 m width) across the experimental area (24 plots), and was repeated each growing season (May-June). Sunn hemp was terminated in early fall (September-October) of each growing season using previously described termination methods ( Figure 3). A complete list of field activities for the whole experiment is shown in Table 1. Kale seeds (Siberian Dwarf Variety) were planted into sunn hemp residue (11 kg ha −1 ) with different residue treatments three weeks after its termination (end of September and October) in half of the experimental area using the Tye no-till drill. To obtain a 0.38 m row spacing for kale, every The experiment was a randomized complete block design with four replications for each treatment ( Figure 4). The experiment was initiated in late spring (25 May 2011) by drilling (no-till) sunn hemp (45 kg ha −1 ) using a Tye no-till drill (Agco, Duluth, GA, USA; 2.13 m width) across the experimental area (24 plots), and was repeated each growing season (May-June). Sunn hemp was terminated in early fall (September-October) of each growing season using previously described termination methods ( Figure 3). A complete list of field activities for the whole experiment is shown in Table 1. Kale seeds (Siberian Dwarf Variety) were planted into sunn hemp residue (11 kg ha −1 ) with different residue treatments three weeks after its termination (end of September and October) in half of the experimental area using the Tye no-till drill. To obtain a 0.38 m row spacing for kale, every other row on the Tye drill was blocked off. At the same time, cereal rye was planted (101 kg ha −1 ) with the same equipment (row spacing 0.18 m) to the other half (12 plots with different sunn hemp residue treatments); in each growing season, rye was harvested for grain using a Massey Ferguson 8XP plot combine (last week of May, first week of June, Figure 5a           Biomass samples were cut from two 0.25 m 2 sampling areas in each plot for both sunn hemp and rye. Spring rye samples were collected (Table 1) when rye was at the early milk growth stage (Zadoks growth stage = 73) [24]. Subsamples of sunn hemp cover crop and cash crop: cereal rye and kale tissues (for kale, during kale harvest, Table 1) were collected, dried (55 • C), and ground to pass through a 2 mm screen with a Wiley mill (Thomas Scientific, Swedesboro, NJ, USA), then ground further to pass through a 1 mm screen with a Cyclone grinder (Thomas Scientific, Swedesboro, NJ, USA).
Cash crop subsamples were analyzed for total N concentration (%) by dry combustion on a LECO TruSpec-CN analyzer (Leco Corp., St Joseph, MI, USA). C and N contents present in the sunn hemp tissue were determined by multiplying total C and N concentrations (%) by the corresponding sunn hemp biomass production.
Soil samples were collected with a Giddings (Windsor, CO, USA) UTV mounted soil probe at sunnhemp termination and again at cash crop planting to monitor soil nitrogen availability during the 3 weeks between termination and planting. Eight soils cores were collected per plot with a 100 cm long stainless-steel probe (3.8 cm diam.) to a depth of 30 cm.
Each soil core was split into two depths of 0-15 cm (shallow) and 15-30 cm (deep), composited, mixed, and placed into appropriate sample boxes. Soil was dried at 55 • C for 24-48 h. NO 3 -N and NH 4 -N concentrations were determined using potassium chloride (KCl) extraction. Ten grams of soil was weighed and extracted with 50 mL of 2 M KCl solution. Cups with the soil-KCL solution were covered with labeled lids and placed in a shaking machine (Lab-Line Instruments, Melrose Park, IL, USA; model 4633) at 180 RPM for 1 hour. After completing shaking, the samples were settled for 1 hour. After that, funnel stands, funnels, filters, and vials were employed to filter the soil sample solution. The Whatman #42 filters were placed to each funnel and the settled soil-KCL solution were filtered through the filter directly to vials. Vials were secured and transferred to Auto Analyzer 3. The NO 3 -N and NH 4 -N amounts in soil were based on procedures from Seal Analytical Inc., (Mequon, WI, USA), formally Bran Luebbe/Technicon using the Auto Analyzer 3 and built-in an extensive methods library for fully automated soil analysis.
Different cover crop sunnhemp residue management treatments and years were considered fixed effects and blocks were considered random [25]. Where differences in each year for dependent variables (cover crop biomass, kale yield and cereal rye grain yield) were significant, and when interactions between treatments and years occurred, data were analyzed separately. Treating year as a fixed effect enabled us to examine treatment differences across growing seasons.
Data were subjected to analysis of variance using SAS, Release 9.2 [26]. The ANOVA GLM procedure and treatment means were separated with Fisher's protected Least Significant Differences (LSD) test at the 10% (α = 0.10) probability level [26].

Sunn Hemp Production
Statistical analyses showed that "YEAR" was significantly different (p-value < 0.0001) with respect to both variables: height and biomass for sunn hemp. Height and biomass amount for sunn hemp during each growing season are shown in Table 2. Corresponding average plant height for sunn hemp was 2.40 m, but sunn hemp plants were shorter in 2013 and 2014 compared with 2011 and 2012 (Table 2). Sunn hemp biomass production averaged over four growing seasons was 10,981 kg ha −1 , which was in the upper range of 3362-11,208 kg ha −1 for biomass production reported in Alabama [27].

Nitrogen and Carbon Content for Sunn Hemp Cover Crop
Analysis of variance results presented in Table 3, indicate that factor YEAR had a significant effect on both variables: N and C content in sunn hemp plant (p-value < 0.0001). The average carbon percentage for sunn hemp was 46.5% (Table 4) although small differences were observed across growing seasons. The range of carbon percentage was 46.1 in 2012 and 2013, to 47.4% in 2014 (data not collected in 2011). The nitrogen content averaged over three growing seasons was 2.1% and ranged from 1.8% in 2013, to 2.6% in 2012. Subsequently, the C to N ratio was between 18.1:1 in 2012 to 25.9:1 in 2013 with an average C:N ratio of 23.2:1 over three growing seasons (Table 4). A report by USDA-NRCS [28] indicated that a C:N ratio of plant residues near 24:1 is ideal for microorganisms to process and release nutrients for subsequent cash crops and maintain soil health.  Results from this study indicate that a C:N ratio of 23.4:1 in sunn hemp was in this desirable range. Moreover, a study conducted by Gan et al. [29] indicated that legumes such as chickpea, dry pea and lentil (Cicer arietinum L.) had C:N ratios of 18.8, 14.8 and 19.3, respectively. There are different legume choices that can be used as a summer cover crop with a desirable C:N ratio for optimal microbial activity in soil. Sunn hemp was chosen for this experiment based on its ability to produce significant biomass in a short period of time and reach plant maturity suitable for mechanical termination without sacrificing benefits of a cover crop. Furthermore, the termination period coincided with the optimum planting window for kale and a favorable window for rye.

Kale Yield Harvested for Fresh Market
Variable "BLOCK" was highly significant for kale yield (p-value = 0.0010). Moreover, kale yield was affected by both "YEAR" (p-value < 0.0001) and "TRT" (p-value < 0.0001) ( Table 5. There were significant YEAR*TRT interactions for kale yield, thus, data for these variables were analyzed separately for each year. Different sunn hemp residue management (rolling, mowing, and mowing with incorporation) had a strong effect (p-value < 0.0001) on kale yield (Table 5). The "YEAR" differences were related to different weather conditions and growing conditions (weed pressure) in each growing season. The kale yield in 2011 for mowed and incorporated residue was significantly higher (17,846 kg ha −1 , p-value < 0.0001) compared with mowed only (4792 kg ha −1 ) and rolled only treatments (2792 kg ha −1 ), although no significant difference in kale yield was observed between rolled and mowed cover crop residue ( Figure 6). When incorporating mowed residue (a similar method of managing a green manure) it would promote greater N release, which would be available for plant uptake to increase kale yield, compared with mowed or rolled with no soil incorporation that would release N much slower.
Thavarajah et al. [30] examined effects of different cover crops on organic kale yield. They concluded that ryegrass was most suitable of the cover crops examined by producing high kale yield (42,846 kg ha −1 ) followed by faba bean (33,624 kg ha −1 ). In contrast, Lynex, a winter pea cover crop was associated with the lowest kale yield (16,224 kg ha −1 ), which was comparable to kale yield from this study in 2011. Possibly selecting a legume in this experiment was not the best choice to promote the highest kale yield. In 2012, no differences in kale yield were reported among rolled (17,095 kg ha −1 ), mowed (18,358 kg ha −1 ) and mowed with incorporation (17,243 kg ha −1 ) residue treatments (p-value = 0.494). Same as for 2011, in 2013, significant differences in kale yield occurred among residue management treatments (p-value < 0.0001), with significantly higher kale yield observed for mowed and incorporated residue (13,728 kg ha −1 ), followed by mowed residue (3545 kg ha −1 ) and the lowest (2336 kg ha −1 ) for rolled residue. Extreme amounts of residue left on the surface of both mowed and rolled methods prevented the cutting coulters on the no-till drill to completely cut and allow adequate soil penetration for the disc openers to place seeds which can lead to "hair pinning" often seen in no-till farming systems. Likewise, in 2014, significant differences among sunn hemp residue management treatments were reported (p-value < 0.0001), with significantly higher kale yield for mowed and incorporated residue (11,397 kg ha −1 ), followed by unusually low kale yield for mowed residue (337 kg ha −1 ) and rolled residue (12 kg ha −1 ). The main reason for these differences was unusually high weed pressure on rolled and mowed sunn hemp residue plots that almost entirely inhibited kale growth. Weed pressure increased from year to year in this organic system as no commercial chemicals were applied to manage tough weeds like pigweed (Amaranthus palmeri, L.) and nutsedge (Cyperus esculentus, L.). One effective weed control measure, particularly for pigweed, is tillage. Burying this small-seeded weed has been promoted as a successful weed control strategy, but this is obviously not preferred from a soil health standpoint. It is possible that despite the surface residue present, disking was more effective because seed were buried that prevented germination and subsequent emergence. It is also possible that disking prevented any weed escapes. Regardless of the benefits of surface cover, if any weeds emerged and formed seed, the problem would be much worse for the subsequent years. Actually, in 2011, there was no significant weed pressure. However, starting in 2012, and continuing through 2013, and 2014, there was an increasing weed population present. The prevalence of weeds starting in 2012 with no control likely increased the weed seed bank each year that caused the continued proliferation of weeds the following years.
Agronomy 2020, 10, x 9 of 16 and nutsedge (Cyperus esculentus, L.). One effective weed control measure, particularly for pigweed, is tillage. Burying this small-seeded weed has been promoted as a successful weed control strategy, but this is obviously not preferred from a soil health standpoint. It is possible that despite the surface residue present, disking was more effective because seed were buried that prevented germination and subsequent emergence. It is also possible that disking prevented any weed escapes. Regardless of the benefits of surface cover, if any weeds emerged and formed seed, the problem would be much worse for the subsequent years. Actually, in 2011, there was no significant weed pressure. However, starting in 2012, and continuing through 2013, and 2014, there was an increasing weed population present. The prevalence of weeds starting in 2012 with no control likely increased the weed seed bank each year that caused the continued proliferation of weeds the following years.

Soil NO3-N and NH4-N Amounts for Kale
Based on analysis of variance results presented in Table 6, the amounts of NO3-N and NH4-N in soil for kale were influenced by factors: TRT DEPTH and TIME. In addition, NO3-N amounts were dependent on factor YEAR. Amounts of NO3-N and NH4-N in soil for kale with respect to growing seasons, sampling time, sampling depth and sunn hemp residue management treatment are shown in Table 7. Table 6. Analysis of variance results for soil NO3-N and NH4-N amounts (kg ha −1 ) for kale. TRT (cover crop sunn hemp mowed, rolled or incorporated), DEPTH (sampling depths: 0-15 cm and 15-30 cm) and TIME (sampling at cover crop termination or at planting cash crops).

Soil NO 3 -N and NH 4 -N Amounts for Kale
Based on analysis of variance results presented in Table 6, the amounts of NO 3 -N and NH 4 -N in soil for kale were influenced by factors: TRT DEPTH and TIME. In addition, NO 3 -N amounts were dependent on factor YEAR. Amounts of NO 3 -N and NH 4 -N in soil for kale with respect to growing seasons, sampling time, sampling depth and sunn hemp residue management treatment are shown in Table 7. Table 6. Analysis of variance results for soil NO 3 -N and NH 4 -N amounts (kg ha −1 ) for kale. TRT (cover crop sunn hemp mowed, rolled or incorporated), DEPTH (sampling depths: 0-15 cm and 15-30 cm) and TIME (sampling at cover crop termination or at planting cash crops).  * Same lower-case letters in each row for each chemical (i.e., NO3-N) and variable (i.e., growing season) represents no difference in the soil amount of the particular nutrient.

Soil NO 3 -N Amount
Results indicate that the NO 3 -N amounts differed among all four growing seasons (p-value < 0.0001) ( Table 6), and NO 3 -N amounts in soil increased consistently to the next growing season from 15.0 kg ha −1 in 2011 to 40.3 in 2014 ( Table 7). The measured NO 3 -N amounts over all four growing seasons was significantly higher (p-value < 0.0001) at the shallower depth (0-15 cm) (38.9 kg ha −1 ) compared to the deeper depth (15-30 cm) (14.9 kg ha −1 ). The amount of NO 3 -N at sunn hemp termination was significantly lower (13.4 kg ha −1 ) compared to NO 3 -N amounts at kale planting (40.3 kg ha −1 ) indicating that the NO 3 -N increase was related to nitrogen release during the sunn hemp decomposition period. In fact, sunn hemp residue management treatments had a significant effect on NO 3 -N amounts in soil (p-value = 0.0239), where mowing with incorporation treatment released significantly higher amounts of NO 3 -N (30.5 kg ha −1 ) compared to rolled residue only (23.5 kg ha −1 ). The NO 3 -N amount measured for the mowing only treatment was 26.5 kg ha −1 , which was not statistically different from the rolling and mowing with incorporation treatments.

Soil NH 4 -N Amount
Amounts of NH 4 -N in soil were not significantly different among growing seasons (p-value = 0.7144) with an average amount of 8.8 kg ha −1 . In contrast, a significant difference in NH 4 -N amounts was observed between depths (p-value < 0.0001) with greater amounts (13.7 kg ha −1 ) in the 0-15 cm depth compared to the 15-30 cm depth (3.9 kg ha −1 ). The amount of NH 4 -N varied between sampling times (p-value = 0.0501) with a lower NH 4 -N amount (8.1 kg ha −1 ) present at sunn hemp termination compared to kale planting (9.5 kg ha −1 ). Likewise, NH 4 -N amounts among sunn hemp management treatments were different (p-value = 0.0809) with greater NH 4 -N amounts (9.8 kg ha −1 ) for rolled sunn hemp residue compared to lower NH 4 -N amount (8.0 kg ha −1 ) for mowed with incorporation. The NH 4 -N amount for mowed residue was 8.6 kg ha −1 and was not different from rolled and mowed with incorporation residue treatments (Table 7).

Nitrogen Content in Kale
Significant differences for N contents in kale occurred for Year and were also dependent on sunn hemp management treatments. In addition, there were significant interactions between Year and TRT for N contents (Table 8); therefore, data were analyzed separately for each year. The N content in kale tissue with respect to treatment effects for each growing season is presented in Table 9. In 2011, significantly higher N amounts (126 kg ha −1 ) were associated with mowed and incorporated sunn hemp residue (p-value = 0.0002), compared with lower N-content values for mowed (37 kg ha −1 ) and rolled (21 kg ha −1 ) sunn hemp residue (without significant difference between these treatments). In 2012, N content amounts in kale were not different (p-value = 0.9408) among sunn hemp residue treatments with an average kale N content of 194 kg ha −1 . However, in the 2013 and 2014 growing seasons, significantly higher kale N contents were observed for mowed residue with incorporation (p-value < 0.0001) that corresponded to N contents of 352 kg ha −1 in 2013 and 121 kg ha −1 in 2014. Similar to 2011 growing season, lower kale N contents were associated with rolled and mowed sunn hemp residue treatments with no difference between these treatments.

Cereal Rye Cash Crop for Grain
Cereal rye grain yield averaged over four growing seasons and treatments was 1358 kg ha −1 . However, grain yield was significantly different (Table 5) with respect to variable BLOCK (p-value =0.0171) and the variable YEAR (p-value < 0.0001). In addition, there was a significant interaction between YEAR*TRT (p-value = 0.0683). Therefore, data for cereal rye grain were analyzed separately for each year.
In the 2011, 2012 and 2014 growing seasons, there was no difference in rye grain yield between rolled, mowed, and mowed with incorporated sunn hemp residue. In contrast, treatments had an effect on cereal rye grain yield (p-value = 0.0003) in 2013 with significantly higher cereal rye grain yield measured following rolling only (2193 kg ha −1 ) and mowed sunn hemp (2098 kg ha −1 ) treatments compared to mowed with incorporation (1553 kg ha −1 ) sunn hemp residue (Figure 7). These differences were likely weather related. In February 2013, there were several rainfall events with unusually high cumulative rainfall amounts of 315 mm. It is possible that cereal rye planted into rolled or mowed sunn hemp surface residue across an undisturbed soil was protected from rainfall energy that reduced soil erosion and runoff, which lead to better cereal rye development. In contrast, mowed residue with incorporation did not protect the soil from rainfall energy that caused soil sealing, which may have suppressed cereal rye growth [31,32]. Moreover, in 2014, the experimental area had high weed pressure (especially for plots with mowed and incorporated residue) and severe lodging of cereal rye plants.
Organic practices were followed across the experimental area that did not allow commercial herbicide use for weed control. These combined aspects likely caused the decrease in rye grain yield observed. Cereal rye grain yield for Wrens Abruzzi variety obtained in this experiment were similar to yields reported in Georgia (1993-1995 data) averaging 1476 kg ha −1 (23.7 bushels/acre) [33].

Soil NO3-N and NH4-N Amounts for Cereal Rye
Analysis of variance regarding soil NO3-N and NH4-N with respect to cereal rye are shown in Table 10, were dependent on factors: YEAR DEPTH and TIME. Moreover, NO3-N amounts in soil were dependent on treatments (TRT) for sunn hemp. Amounts of NO3-N, NH4-N, in soil for cereal rye at each growing season, two depths, sampling time and sunn hemp termination treatments are presented in Table 11.  Cereal rye grain yield for Wrens Abruzzi variety obtained in this experiment were similar to yields reported in Georgia (1993-1995 data) averaging 1476 kg ha −1 (23.7 bushels/acre) [33].

Soil NO 3 -N and NH 4 -N Amounts for Cereal Rye
Analysis of variance regarding soil NO 3 -N and NH 4 -N with respect to cereal rye are shown in Table 10, were dependent on factors: YEAR DEPTH and TIME. Moreover, NO 3 -N amounts in soil were dependent on treatments (TRT) for sunn hemp. Amounts of NO 3 -N, NH 4 -N, in soil for cereal rye at each growing season, two depths, sampling time and sunn hemp termination treatments are presented in Table 11.  (Table 11). Over four growing seasons, soil NO3-N amounts were significantly higher the 0-15 cm soil depth (32.6 kg ha −1 ) compared to the 15-30 cm soil depth (14.4 kg ha −1 ) (p-value < 0.0001). The amount of soil NO 3 -N measured at sunn hemp termination was significantly lower (10.4 kg ha −1 ) compared to the soil NO 3 -N amount measured at cereal rye planting (36.6 kg ha −1 ) that indicated that the soil NO 3 -N increase was related to N being released during the sunn hemp decomposition process following termination (Table 13). In addition, sunn hemp residue management treatments affected soil NO 3 -N amounts in soil (p-value = 0.0609). Mowing with incorporation released significantly greater amounts of soil NO 3 -N (27.1 kg ha −1 ) compared to rolled residue (22.4 kg ha −1 ) and the mowing only treatment (21.0 kg ha −1 ). No differences were observed between the rolled and mowing only treatments for soil NO 3 -N amounts. The process of incorporation increased the amount of exposure of soil microbes to sunn hemp residue through the mixing of residue and the soil, which enables the soil microbes to break the residue down faster compared to leaving it on the soil surface. Same lower-case letters in each column (Year) represents no difference in plant nitrogen contents with respect to rolling treatments. ** Same upper-case letters in last row represents no difference in plant nitrogen with respect growing seasons (averaged over treatments).

Nitrogen Content in Cereal Rye
The variables YEAR and YEAR*TRT did not affect plant N contents in cereal rye (p-values = 0.7052 and 0.1776; Table 12). In contrast, sunn hemp management treatments did affect cereal rye plant N-contents (p-value = 0.0549). The average cereal rye plant N content across all growing seasons and treatments was 91 kg N ha −1 (Table 13). Greater cereal rye N contents were measured for rolled sunn hemp residue (101 kg ha −1 ) compared to mowed residue N contents (82 kg ha −1 ). Mowed sunn hemp residue with incorporation N contents (89 kg ha −1 ) were equivalent to rolled and mowed sunn hemp residue treatments.

Conclusions
Sunn hemp biomass averaged over the growing seasons was 10,981 kg ha −1 , producing an average total C percentage of 46.5% with an average total N percentage of 2.1%. The average C:N ratio was 23:1. Except in 2012, higher kale yield was observed for mowed and incorporated sunn hemp, compared to lower yields for the mowed and rolled residue. These differences were related to weather condition (extensive rainfall) and the different sunn hemp residue managements methods for large amounts of sunn hemp residue. In three growing seasons (2011, 2012 and 2014), grain yield from cereal rye was not affected by sunn hemp termination methods, although in 2013, cereal rye grain yield was higher for rolled and mowed sunn hemp residue compared to lower grain yield for mowed with incorporation treatment. It appears that better soil planting conditions due to rolling and mowing sunn hemp compared with mowed and incorporated residue resulted in higher cereal rye grain yield. Soil NO 3 -N and NH 4 -N for kale plots varied with sampling depth. Greater amounts were measured in the 0-15 cm layer. Likewise, soil NO 3 -N amounts varied across each growing season. In contrast, there were no differences in amounts of NH4-N among growing seasons. Higher NO 3 -N and NH 4 -N amounts were measured at kale planting compared to amounts measured at sunn hemp termination. For cereal rye plots, soil NO 3 -N and NH 4 -N varied significantly across each growing season and with sampling depth. Sunn hemp management treatments did not affect soil NH 4 -N, in contrast, sunn hemp management treatments did effect soil NO 3 -N. Sampling time at cereal rye planting resulted in higher amounts of NO 3 -N and NH 4 -N compared to amounts measured at sunn hemp termination.
Author Contributions: T.S.K. and K.S.B. collaborated on experimental conceptualization, experimental investigation, contributing to resources, statistical analysis, and writing-original draft preparation. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by the USDA-ARS, and received no external funding.