Exploring the Feasibility of Integrating Weed and Nitrogen Management with Seed Meals in Organic Vegetables

Abstract

Corn gluten meal (CGM) and soybean meal (SBM) have demonstrated value as bioherbicides and organic fertilizers, but suggested application rates usually target either weed suppression or crop nutrition, not both. The objective of this study was to explore the feasibility of integrating weed and nitrogen management by evaluating effects of increasing seed meal rates within planting holes of plastic mulch film on weed density, soil nitrogen availability, and crop yield in tomato (Solanum lycopersicum) and broccoli (Brassica oleracea). CGM (10% N) or SBM (7% N) were applied at rates of 0.5, 1, 2, 3.5, or 5 g planting hole⁻¹ N (depending on crop and year) after crops were transplanted, and 40 weed seeds per planting hole were seeded. Weed density decreased with increasing seed meal rate, regardless of type, and velvetleaf (Abutilon theophrasti) was more susceptible than the grass weeds tested. Velvetleaf suppression at the 5 g planting hole⁻¹ N rate ranged from 66% to 97%, relative to the weedy control. Soil nitrogen availability increased with the application rate, but ammonium mineralized from seed meals applied at the highest rates were likely phytotoxic to weeds and crops. Seed meals never increased the crop yield and reduced the tomato yield in 2018 by 39% to 64%, relative to the weed-free control. The results suggest that integrating the management of weeds and nitrogen with seed meals in plastic mulch planting holes is not feasible because application rates required for consistent weed suppression are also toxic to crops.

1. Introduction

Managing weeds and soil fertility are among the top four production challenges identified by organic farmers [1]. Weed management on organic farms often includes manual hand weeding, and organic farmers identified accessing labor as the top non-production related challenge [1]. Thus, research is needed to improve weed management and soil fertility and reduce labor needs in organic farms. Integrating the management of nitrogen and weed management using seed meals could potentially reduce weeds, improve crop nutrition, and reduce labor needed for hand weeding. Depending on their application method and rate, seed meals have the potential to control weeds through physical or chemical modes of action [2,3]. Seed meals can be used to physically abrade emerged weed seedlings when applied with compressed air [2] or applied to soil as an herbicide to prevent weed seed germination or stunt seedling growth [3].

Plastic mulch films are a critically important weed management tool for organic specialty crop growers [4], yet weeds can still emerge and compete with crops through planting holes (i.e., the hole created in the mulch to allow for crop planting and growth), along the edges of the mulch, or through the mulch itself. Unmanaged weeds growing through plastic mulch film planting holes have been shown to reduce pepper (Capsicum annuum) and tomato (Solanum lycopersicum) yield by 29% to 45% [5]. Seed meals applied directly into the planting hole offer a unique opportunity for integrated weed and nitrogen management because many seed meals have potential value as organic fertilizers and bioherbicides [6].

Corn gluten meal (CGM) is one of the most studied seed meals for its herbicidal potential and is commercially available as a natural preemergent herbicide. Bingaman and Christians [7] applied CGM rates of 0, 324, 649, and 973 g m⁻² to surface soils in greenhouse pots and found that the growth and survival of many weed species was reduced by the lowest rate of 324 g m⁻², but efficacy varied greatly among species. Green foxtail (Setaria viridis) survival was reduced by 37% at 324 g m⁻² and 100% at the 973 g m⁻² rate, whereas shattercane (Sorghum bicolor) survival was reduced only 42% to 51% across all rates. Velvetleaf (Abutilon theophrasti) was among the least affected weed species, as survival was 100% at the 324 g m⁻² rate and was only reduced by 35% at the highest rate of 973 g m⁻² [7]. By contrast, Wortman [2] found that lower rates of 54 to 269 g m⁻² CGM stimulated weed germination and growth instead of inhibiting them. This highlights the fertilizer potential of CGM, which contains approximately 10% nitrogen by weight. Under field conditions, McDade and Christians [3] tested CGM at rates between 100 and 400 g m⁻² in direct seeded vegetables and found similar reductions in weed growth (50% to 82%), but the CGM had similarly negative effects on crop seedling survival, which was reduced 41% to 73%. Webber III et al. [8] observed similar weed control of 72% after an application of 400 g m⁻² CGM in transplanted onion, and transplanting seemed to mitigate herbicidal effects on the crop.

Soybean meal (SBM) is another widely available seed meal but is more commonly marketed as an organic fertilizer, and less is known about its potential as a biobased herbicide. Shrestha et al. [9] found that rates of 124 to 448 g m⁻² SBM had limited herbicidal potential, particularly beyond two months after application, but increasing SBM rates did increase spinach yield. The results from Wortman [2] in the greenhouse were similar, as rates between 54 and 269 g m⁻² SBM had no effect on weed emergence and growth. However, An et al. [10] applied 200 g m⁻² SBM in transplanted rice paddies, and weed density was reduced by 40% to 100% (depending on the species), and the fresh weight of rice increased by 53%. Similarly, El-Metwally et al. [11] found that 150 g m⁻² SBM reduced the weed biomass of four different broadleaf weeds and two grass weeds by 80% to 84% compared to weedy controls. SBM was applied and concentrated around the crops to a depth of 4–6 cm, which may have acted more as a mulch than as an herbicide. Similar rates applied within planting holes of plastic mulch films may provide these mulching benefits while concurrently delivering nitrogen to the plant.

Given the potential integrated value of seed meals as biobased herbicides, mulch, and fertilizers, our objective was to evaluate the effects of CGM and SBM on weed density, soil nitrogen availability, and crop yield when applied at different rates in the planting holes of plastic mulch film. The results will help to inform optimum application rates and the potential tradeoffs for integrating the use of seed meals as bioherbicides and organic fertilizers.

2. Materials and Methods

To accomplish our study objectives, five field trials were conducted between May 2017 and November 2018 at two locations on the University of Nebraska East Campus Research Farm (lat. 40°50′12″ N, long. 96°39′48″ W). Three trials were conducted in tomato (Solanum lycopersicum) (one trial in 2017 and two trials in 2018 in fields separated by approximately 1 km) and two trials in broccoli (Brassica oleracea) (one trial in 2017 and 2018 each). The soil type at the farm (in both locations) is a Zook silty clay loam, and pre-plant soil analyses (Ward Laboratories, Keaney, NE, USA) for each field location and year are included in

Table 1. Pre-plant analysis of soils collected to a depth of 0 to 20 cm in 2017 and 2018 from two different experimental fields (north and south) at the University of Nebraska—Lincoln East Campus Research Farm.

	pH	Organic Matter (g kg−1)T	NO3-N (mg kg−1)	Mehlich-P (mg kg−1)	K (mg kg−1)
2017
North field (tomato and broccoli)	6.2	4.0	10.7	90	410
2018
North field (tomato and broccoli)	6.1	4.2	9.6	95	422
South field (tomato)	6.7	3.8	9.5	86	683

. Pre-plant soil samples were analyzed for pH (1:1 soil/water dilution), soil organic matter (loss of weight on ignition), nitrate-N (KCl extraction), phosphorus (Mehlich 3 extraction), and potassium (ammonium acetate extraction) [12].

2.1. Experimental Design and Treatments

Each experiment was a randomized complete block design with two treatment factors. The treatment factors included seed meal type—CGM (Preen Natural Weed Preventer, Lebanon Seaboard Corporation, Lebanon, PA, USA) or SBM (Phyta-Grow Leafy Green Special Fertilizer, California Organic Fertilizers, Inc., Hanford, CA, USA)—and rate. The rates were standardized by an estimated N analysis of the seed meals (10% for CGM based on Webber III et al. [13] and 7% for SBM based on the labeled guaranteed analysis) and included 0.5, 1, 2, or 5 g planting hole⁻¹ N in 2017; and 2, 3.5, or 5 g planting hole⁻¹ N in 2018 (Figure 1). The rates were modified in 2018 based on results from 2017 to better identify an optimum rate for integrated N and weed management. All the treatment combinations were compared to weed-free and weedy controls (0 g planting hole⁻¹ N). Assuming 46 cm in row spacing and 1.8 m between row centers (typical for vegetables in the U.S.), a seed meal rate of 0.5 g planting hole⁻¹ N resulted in the addition of approximately 6 kg ha⁻¹ N, and 5 g planting hole⁻¹ N contributed approximately 60 kg ha⁻¹ N. Experimental plots were 4.6 m long on 1.2 m wide raised beds. The plants were spaced 0.46 m apart within rows for a total of 10 plants plot⁻¹.

Tomato (cv. ‘Defiant’) and broccoli (cv. ‘Arctic’) seeds (Johnny’s Selected Seeds, Winslow, Maine, USA) were planted 6 to 8 weeks prior to field transplanting in 72-cell greenhouse flats filled with a general use soilless media mix in the greenhouse. The seedlings were watered daily and fertilized as needed with 250 mg dm⁻³ of a water soluble 20-20-20 N-P-K fertilizer (Jack’s Professional General Purpose, J.R. Peter’s, Inc., Allentown, PA, USA). The fields were prepared 1 to 7 days prior to transplanting by roto-tilling the soil and then concurrently shaping the raised beds (RB-448; Nolt’s Produce Supplies; Leola, PA, USA) and laying a single drip tape beneath a black or white-on-black biodegradable plastic mulch film (BIO360; Dubois Agrinovation, Saint-Rémi, QC, Canada). The tomatoes were transplanted on 31 May 2017 and 15 May 2018, and broccoli was transplanted on 18 August 2017 and 10 August 2018. Crop seedlings were transplanted 46 cm apart in single rows on each bed top into punched holes in the mulch film. The crops were drip-irrigated between precipitation events to maintain volumetric soil moisture at or above 0.15 cm³ cm⁻³ in the top 20 cm of the soil.

Weed seeds were planted in each mulch film planting hole within 2 days of transplanting. In 2017, 20 velvetleaf (Abutilon theophrasti) and 20 green foxtail (Setaria viridis) seeds were used as model broadleaf and grass weeds, respectively. Forty total seeds were placed in each planting hole and covered with 50 g of topsoil. In 2018, 20 shattercane (Sorghum bicolor) seeds were used instead of green foxtail due to poor germination in the green foxtail seed lot. Prior to seeding, velvetleaf seeds were submerged in a 70 °C water bath for 1 min to improve germination [14]. Seed meal treatments were applied within planting holes (above the buried weed seeds) immediately after seeding weeds, resulting in a layer of seed meal that covered what is typically bare soil between the crop stem and mulch film (Figure 1). The plots were uniformly irrigated 2 days after seed meal application to ensure weed seed germination.

2.2. Data Collection

Emerged weed seedlings were identified by the species seeded (velvetleaf, green foxtail, or shattercane) and counted in 5 random planting holes per plot between 16 and 33 days after transplanting tomato or broccoli.

In 2018, pairs of cation and anion Plant Root Simulator (PRS) probes (Western Ag Innovations, Saskatoon, SK, Canada) were used to measure changes in total plant-available soil nitrate and ammonium within a subset of treatments in tomato and broccoli planting holes [15]. The units used by Western Ag Innovations are μg nutrient 10 cm⁻² ion-exchange membrane surface area time⁻¹ of burial (2 weeks in this study). In tomato, data were collected for the 2 g and 5 g rates of CGM and the weedy and weed-free controls across both locations. In broccoli, data were collected for the 2 g and 5 g rates for both seed meals, along with the weedy and weed-free controls. Probes remained in the soil for a 2-week incubation, beginning 2 days after seed meal application. Upon removal from the soil, PRS probe pairs were rinsed with deionized water and stored at 4 °C before shipping, extraction, and analysis at Western Ag Innovations. The probes were eluted for 1 h in 17.5 mL of 0.5 mol/L HCl, and nitrate and ammonium in the eluant were quantified colorimetrically using automated flow injection analysis (Skalar San++ Analyzer; Skalar Inc.; Breda, Noord-Brabant, The Netherlands).

The tomato fruits were harvested when ripe every 5–7 days and weighed fresh to determine the total yield. In 2017, there were six tomato harvests, and, in 2018, there were five. Broccoli was harvested by cutting 15 cm below the head and weighed fresh to determine the total yield.

2.3. Data Analysis

Yield, weed density, and soil nitrogen data were analyzed with a generalized linear mixed model analysis of variance (proc GLIMMIX; SAS 9.4; SAS Institute Inc., Cary, NC, USA). Crops and years were analyzed separately due to differences in seed meal rates between years. Fixed effects in the model included seed meal type, application rate, and their interaction. The replicate block was a random effect along with location (in 2018, tomato only). Least square (LS) means and standard errors were calculated for all significant fixed effects and compared using Tukey’s honestly significant difference test at α = 0.05.

3. Results

3.1. Tomato

3.1.1. Weed Density

Velvetleaf (p < 0.0001) and green foxtail (p = 0.02) weed density, 33 days after transplanting in 2017, were both influenced by seed meal rate but not by type or their interaction. Velvetleaf density generally declined with increasing rates of seed meals to a maximum of 5 g planting hole⁻¹ N (Figure 2). The trends were similar for green foxtail, but overall emergence was much lower, and tested rates were not different from the weedy control.

In 2018, velvetleaf density was influenced by the seed meal rate (p < 0.0001) and type (p = 0.0003) but not their interaction. Consistent with 2017, velvetleaf density decreased with increasing seed meal rate (regardless of type) up to 5 g planting hole⁻¹ N (Figure 3). However, in 2018, SBM provided slightly better velvetleaf suppression (0.54 ± 0.13 weeds planting hole⁻¹) compared to CGM (1.39 ± 0.13 weeds planting hole⁻¹) across all the tested rates. Shattercane density was not affected by the seed meal type, rate, or their interaction but was overall much lower than velvetleaf.

3.1.2. Nitrogen Availability

Nitrate availability in tomato planting holes between 2 and 16 days after seed meal application was influenced by the CGM rate (p < 0.0001). Nitrate was greatest in the 5 g treatment, followed by 2 g, and lowest in the weedy and weed-free controls that did not receive any CGM (Figure 4). Ammonium availability was also influenced by the CGM rate (p = 0.02) and was 3.3× greater in the 5 g treatment compared to the 2 g (Figure 4).

3.1.3. Yield

The tomato yield in 2017 was not influenced by the seed meal type (p = 0.54), but the effect of the rate was approaching significance (p = 0.056) because of reduced yield in the 5 g treatment (103.5 ± 6.1 kg 10 m row⁻¹) compared to the weed-free control (123.0 ± 8.7 kg 10 m row⁻¹). In 2018, the tomato yield was influenced by the seed meal type (p = 0.03) but not the rate (p = 0.13) because seed meals reduced the tomato yield by 39% to 64%, regardless of rate (Figure 5). Overall, the tomato yield was much lower in 2018 (3.7 kg 10 m row⁻¹) compared to 2017 (117.4 kg 10 m row⁻¹) because of disease and delayed fruit maturation throughout the field.

3.2. Broccoli

3.2.1. Weed Density

Velvetleaf weed density, 27 days after transplanting in 2017, was influenced by the seed meal rate (p = 0.0002) but not by type or the interaction with the rate. Velvetleaf density generally declined with increasing rates of seed meal to a maximum of 5 g planting hole⁻¹ N (Figure 6). Green foxtail density was not affected by the seed meal rate, type, or their interaction (p > 0.05), and overall emergence was generally low (<2 weeds planting hole⁻¹). In 2018, velvetleaf (p = 0.03) and shattercane (p = 0.0005) density were influenced by the seed meal rate (but not by type or their interaction). Weed density was lowest in the 5 g planting hole⁻¹ N rate, followed by the 3.5 g rate (Figure 7).

3.2.2. Nitrogen Availability

Nitrate (p = 0.03) and ammonium (p = 0.03) availability in broccoli planting holes between 2 and 16 days after seed meal application were influenced by the seed meal rate but not by type or their interaction. Both nitrate and ammonium were greater in the 5 g treatment compared to 2 g and the weed-free control, but the weedy control was not different from any other treatment (Figure 8).

3.2.3. Yield

The broccoli yield was not influenced by seed meal type, rate, or their interaction in 2017 or 2018 (p > 0.05). The mean yield in 2017 was 3.88 kg 10 m row⁻¹ and, in 2018, was 0.59 kg 10 m row⁻¹. The yield was relatively low in 2018 because broccoli was harvested before the heads were fully developed (due to an early forecasted killing frost).

4. Discussion

Seed meal effects on weed density in this study were driven by increasing application rates, regardless of seed meal type, and these effects were more commonly observed on velvetleaf (broadleaf) than green foxtail or shattercane (grass). The highest seed meal rate (5 g planting hole⁻¹ N) reduced velvetleaf density by 66% to 97% compared to the weedy control, depending on the crop and year. This level of weed suppression is consistent with previous studies, though the rates required to observe the effect were much greater in this study. The 5 g planting hole⁻¹ N rate was approximately equivalent to 4840 g m⁻² CGM or 6880 g m⁻² SBM concentrated within the planting hole, whereas other studies have observed similar benefits of CGM as low as 224 to 400 g m⁻² [8,16].

Bingaman and Christians [7] found that velvetleaf was particularly difficult to control, with CGM at the highest rate (973 g m⁻²) reducing velvetleaf survival by only 35%. At a similar rate tested in this study in 2017 (968 g m⁻²; 1 g planting hole⁻¹ N), there was no observed reduction in velvetleaf density. In most cases, at least 2 g planting hole⁻¹ N (1936 g m⁻² CGM or 2752 g m⁻² SBM) was required to achieve significant reductions in weed density in this study. At these relatively high rates, it is possible that the weed suppressive mechanism included a mulching effect whereby the thick layer of seed meal (Figure 1) occluded light prior to decomposition. The level of weed suppression observed at the highest rates are consistent with those reported by El-Metwally et al. [11] (80–84%) who used SBM to create a mulch around crop plants.

Green foxtail density was rarely affected by seed meal rate or type, and shattercane density was reduced by 85% in the 5 g planting hole⁻¹ N treatment only in 2018 broccoli. Bingaman and Christians [7] observed a better control of green foxtail with CGM as mortality reached 100% at 973 g m⁻². However, shattercane proved difficult to control in both studies. The highest rate of 973 g m⁻² provided only a 51% reduction in shattercane survival in Bingaman and Christians [7], and shattercane was only affected by seed meals in this study in 2018 broccoli. A rate of 3.5 g planting hole⁻¹ N (approximately 3388 g m⁻² CGM or 4816 g m⁻² SBM) was required to achieve 46% reduction in shattercane density—far greater than the rates required in Bingaman and Christians [7].

Nitrogen availability between 2 and 16 days after seed meal application was generally proportional to the seed meal rate (Figure 4 and Figure 8). Based on the typical guaranteed analyses, we would expect 10% N from CGM and 7% N from SBM by weight. Once fully mineralized, the seed meal rates applied in this study could provide between 0.5 and 5 g/plant N or approximately 6 to 60 kg ha⁻¹ N (assuming 1.83 m between-row spacing and 46 cm in-row plant spacing). However, depending on soil conditions, several months is often required before even 50% of organic fertilizer N is mineralized [17].

Of particular interest in this study were the relatively high rates of ammonium observed following the application of the highest seed meal rate (5 g planting hole⁻¹ N). While increased nitrate near the crop is potentially beneficial, excessive ammonium has been shown to reduce weed seed germination and is also potentially harmful to the crop [18,19,20]. The elevated ammonium and potential for toxicity to germinating weed seedlings seems to be the most likely explanation for the weed suppression consistently observed at the 5 g planting hole⁻¹ N rate in this study. Where ammonium levels decreased at the 2 g planting hole⁻¹ N rate (Figure 4 and Figure 8), weed suppression was often reduced proportionately (Figure 3, Figure 6 and Figure 7) or not detectable (Figure 2).

The potentially toxic effects of elevated ammonium were perhaps most evident in the yield data because, despite improved weed control, the seed meals either had no effect (broccoli in both years and tomato in 2017) or reduced yield (tomato in 2018; Figure 5). While other studies have reported yield benefits from seed meal applications [9,10,11], none required the high rates to achieve weed control benefits that were observed in this study (>1936 g m⁻² CGM or >2752 g m⁻² SBM). McDade and Christians [3] demonstrated the potentially phytotoxic effects of CGM on direct seeded crops at rates of only 100 to 400 g m⁻². Transplanting vegetables was suggested as an alternative strategy for avoiding these phytotoxic conditions, but the results of this study suggest that broccoli and tomato transplants may be susceptible to ammonium toxicity even if the plants are not killed by the seed meals.

5. Conclusions

The concept of integrating the use of seed meals as organic fertilizers and biobased herbicides is attractive because it has the potential to mitigate the greatest challenges facing organic farmers, including weed management, soil fertility, and labor availability. However, the results of this study in tomato and broccoli suggest that the seed meal application rates required to achieve consistent weed suppression are potentially toxic to crops. Despite improved weed control and greater nitrogen availability at the highest application rates, the seed meals either had no effect on or reduced crop yields. Ammonium toxicity was the most likely mechanism of toxicity to weeds and crops, and this limits the potential for integrating weed and nitrogen management with seed meals.