Organic or conventional production system and nutrient rate affect the nematode community in carrot production

To investigate these objectives, trials were conducted at the North Florida Research and Education Center-Suwannee Valley near Live Oak, FL (30.304621, -82.899979). Separate trials were conducted using conventional practices and using organic practices. The organic production system trials were conducted on land certified since 2012 by Quality Certification Services, Gainesville, FL in accordance with the United States Department of Agriculture Organic Standards (United States Code, 2000). As required, all inputs and methods were compliant with USDA National Organic Standards, described in a farming system plan, and approved by the certifying agency. All trials were conducted on deep sand soils typical of the area. For conventional production, the soil was Hurricane-Sandy type (siliceous, Thermic Oxyaquic Alorthod). For organic production, it was a Chipley-Foxworth-Albany soil series complex (Thermic, coated Aquic or Typic Quartzipsamments and Loamy, siliceous, subactive, thermic Aquic Arenic Paleudults). A trial using each production system was conducted in 2016–2017 (Year 1) and again in 2017–2018 (Year 2). Trial locations within the station were moved from Year 1 to Year 2. All locations had a history of spring or fall vegetable production and winter cover crops including rye (Secale cereale) in conventional production. In organic production, a mixture of sorghum-sudan grass (Sorghum x drummondii) and ‘Iron Clay’ cowpea (Vigna unguiculata) was grown as a cover crop in summer 2016 and sunn hemp (Crotalaria juncea) was grown in summer 2017. Each production system was managed with conventional tillage, which was done frequently for weed management and residue incorporation prior to study establishment.

For each trial, the experiment was a randomized complete block design with four replicates. Each plot was 18 m long and consisted of one bed with 102 cm wide top and 1.3 m from bed center to adjacent bed center. Nitrogen rate treatments were 168, 224, 280, 336, and 392 kg N/ha for both conventional and organic trials. Carrot production without nitrogen fertilizer or amendment is not commercially viable in the Southeast, so a control treatment without nitrogen amendment was not included. Nutrient amendment was applied to the bed tops only, and rates were calculated using the linear bed foot method (Hochmuth and Hanlon, 2012). Nutrient amendment rates were chosen based on the prior guideline of 196 kg N/ha from the University of Florida (Liu et al., 2020). Nitrogen amendment application was distributed throughout the year to match plant nitrogen demand and timing varied by rate and production system (Tables 1 and 2).

In conventional production, preplant starter fertilizer was applied to flat ground, rototilled to a depth of 15 cm, and pressed into bare ground beds (10 cm high and 102 cm wide). Starter fertilizer source was 13-1.8-10.8 (N-P-K) and 14-1.8-11.6 in 2017 and 2018, respectively. In-season fertilizer for conventional production was ammonium nitrate (32-0-0) banded on the bed top evenly using a single hopper fertilizer drop spreader with directional spouts (First Products, Tifton, GA). In conventional production, nutrient rates other than N were constant across treatments.

In organic production, all nutrient amendments were approved for use in organic production by the certification agency (Quality Certification Services, Gainesville, FL). For nitrogen rate treatments, poultry litter was used each year. In Year 1, Microstart 60 (Perdue AgriRecycle LLC, Seaford, DE) was used and returned an analysis of 3%-2%-3% (N, P₂O₅, K₂O). In Year 2, locally sourced poultry litter was used and it returned an analysis of 2.9-1.42-3.08. Half of the poultry litter was broadcast and incorporated before planting with the remaining poultry litter applied 5 and 7 weeks after planting (Table 2). Preplant litter was applied evenly by hand to shaped bed tops and lightly incorporated to 4 cm deep with a basket weeder followed by bed re-shaping. Post-plant fertilizer was banded on the bed top between carrot rows and incorporated to 4 cm with a basket weeder. Because axenic phosphorous and potassium fertilizers are not readily available for organic production, PK rates varied along with N rate treatments.

Aside from nutrient rate treatments, each trial was managed uniformly based on standard commercial practices in the area. Specific dates for important maintenance activities are provided in Table 3. For each trial, the cultivar Choctaw was direct seeded in the fall using a sponge-type Seed Spider planter (Sutton Agricultural Enterprises, Inc., Salinas, CA). Seed was planted 0.64 cm deep and the press roller on the planter firmed the soil immediately after seeding. In conventional production, carrots were planted in two sets per bed top, with 30.5 cm between sets. Each set had four rows per bed top spaced 4.76 cm apart, such that there eight rows total on each bed top. In organic production, row spacing was wider with four rows per bed top spaced 17.8 cm apart. Wider row spacing was used to accommodate potential mechanical weeding, although ultimately weeding was done primarily by hand in the organic production system. Carrots were harvested mechanically in spring (Table 3) with yield data and other relevant production results forthcoming in separate reports (unpubl. data).

In conventional production, the entire trial was fumigated for plant-parasitic nematode management before planting using 1,3-dichloropropene (Telone II, Dow Agrosciences, Wilmington, DE) at 168 L/ha. Fumigation was conducted using a broadcast shank rig with 30 cm spacing between shanks and fumigant released approximately 30–35.6 cm deep in the soil profile. In organic production, for nematode control, a commercial formulation of live Purpureocillium lilacinum fungi (Melocon WG, Certis USA LLC, Columbia, MD) was used for nematode biocontrol. The P. lilacinum formulation was applied to the soil using a CO₂-powered backpack sprayer at labelled rates on 19 October, 12 January, and 27 January (‒26, 59, and 74 DAP, respectively) in Year 1. In Year 2, P. lilacinum was applied on 1 December and 1 February (16 and 78 DAP).

Supplemental fertilizer applications were made in both organic and conventional systems and were uniform across N rate treatments. In both systems, dolomitic lime and boron were broadcast preplant each season at 2,242 and 1.12 kg/ha, respectively. In conventional production, 112 kg/ha potash was applied in the form of two midseason broadcast applications of Sul-Po-Mag (0N-0P-18.3K-22S-11Mg) per season. In conventional production, phosphate was applied at 52 kg/ha in Year 2, but not Year 1.

In-season weed management was accomplished primarily by hand weeding in organic production. In conventional production, weeds were well-managed with two post-emergence applications of linuron chemical herbicide. Alternaria leaf blight was the main disease of concern in both production systems. In both systems, this disease was managed using weekly fungicide applications when weather was wet and conducive to this pathogen. In conventional production, a variety of chemical fungicides were applied, rotating chemistries by FRAC (Fungicide Resistance Action Committee) code to minimize pathogen resistance. In organic production, organic compliant fungicides were used including Strepomyces bacteria (Actinovate AG, Valent BioSciences LLC, Walnut Creek, CA) and copper (Nordox 75WG, Nordox, Oslo, Norway) in Year 1. In Year 2, copper and a hydrogen peroxide-peroxyacetic acid formulation (OxiDate 2.0, BioSafe Systems, LLC., East Hartford, CT) were rotated. In both production systems, cereal rye cover crop was grown in trial alleys to protect plants from blowing sand and serve as a barrier to wind-borne pathogens.

Nematode soil populations in each plot were quantified at midseason and near harvest each year. Using an Oakfield tube, 12 soil cores to 25 cm depth were collected from the carrot rooting zone. Soil was homogenized manually, and nematodes were extracted using the sucrose-centrifugation method (Jenkins, 1964). The nematode community (plant-parasitic and free-living nematodes) were quantified morphologically to genera level by microscope.

Following nematode community quantification, abundances of individual trophic groups – including bacterivores, fungivores, herbivores, and omnivore-predators – were calculated based on published groupings for individual nematode families (Yeates et al., 1993). The most abundant genera (Table 4) were also subjected to analysis. This included bacterivores Cephalobus spp. and Rhabditis spp., fungivores Aphelenchus spp. and Aphelenchoides spp., and plant-parasitic nematodes Mesocriconema spp. and Meloidogyne incognita. Paratylenchus spp. was also abundant at harvest 2017 in the organic system (228 nematodes/100 cm³ soil), but was not included in analysis as it was not abundant in any other season-system.

Key nematode community indices were also calculated based on nematode abundances. Maturity index, structure index, enrichment index, channel index, and Hills N1 diversity were calculated. Briefly, the maturity index is a measure of system disturbance based on the average nematode colonizer-persister (cp) value in a sample (Bongers, 1990; Bongers and Bongers, 1998). The cp value is a 1–5 ranking of the life strategy of nematodes with 1 indicating an extreme colonizer (short life cycle, short generation time, and high reproductive capacity) and 5 indicating an extreme persister (longer life cycle, longer generation time, and low reproductive capacity). The structure index is a measure of the number of trophic links in a system based on the abundance of extreme colonizers relative to nematodes common in most environments – namely cp2 fungivores and bacterivores (Ferris et al., 2001). The enrichment index is a measure of resource enrichment in a system based on relative abundances of enrichment opportunists (Ferris et al., 2001). The channel index is a measure of decomposition pathways with higher values indicating predominantly fungal decomposition channels and lower values indicating bacterial decomposition channels (Ferris et al., 2001). Hills N1 diversity was calculated using either nematode genera or nematode guilds – trophic group and cp value combinations such as cp2 bacterivores. Hills N1 diversity is derived from the Shannon diversity index and values are interpreted as a measure of the number of common genera or guilds in a sample (Neher and Darby, 2006). Trophic guild abundances were calculated, but not included in analyses due to similarity to either the most abundant genera in a particular guild or total abundance for a particular trophic group.

Variables were analyzed separately for each sampling date. Initially, data were subject to one-way ANOVA for combining experiments. In that analysis, year was treated as a random effect, not of scientific interest, whereas production system was treated as a fixed effect of interest. Effect of production system was determined using year by production system interaction as the error term (Carmer et al., 1989). Due to significant interactions, nutrient rate effects were subsequently analyzed separately by production system as well as date. Before completing ANOVA, response variables were transformed, if needed, to meet assumptions of homogeneity of variance using Levene’s test (Levene, 1960) and normality of residuals based on graphing (Cook and Weisburg, 1999). For combining experiments ANOVA, all nematode populations were transformed by natural log(x+1) for both midseason and harvest. Additionally, guild and genera diversity were transformed by x^2 at midseason for combining experiment analysis. No other variables were transformed at midseason or harvest for combining experiments analysis. For one-way ANOVA within production system, bacterivore abundances were transformed by natural log(x+1) for conventional production midseason Year 1 and organic production Year 1 midseason and harvest. Rhabditis spp. abundance was square-root transformed for organic Year 1 harvest. Aphelenchoides spp. abundance was transformed by x ^{^(3/2)} for organic harvest Year 2 and conventional production midseason Year 1 and harvest Year 2. Cephalobus spp. populations were transformed by natural log(x+1) for conventional production midseason Year 1. Variables for all other analyses were not transformed. Production system effects were considered significant at α=0.05 in ANOVA. For nutrient rate treatment, means were separated by Fisher’s protected LSD (α=0.05) if main effects were significant in ANOVA. Analyses were conducted in R statistical software (version 3.4.4, The R Foundation for Statistical Computing, Vienna, Austria).

The nematode community was consistently different between conventional and organic production systems. Soil abundances of each free-living nematode trophic group – bacterivores, fungivores, and omnivore-predators – as well as herbivores were significantly greater in organic than conventional production at both midseason and harvest (Fig. 1). Sensitivity did vary by genera as Rhabditis spp. (cp1 bacterivore), Aphelenchoides spp. (cp2 fungivore), Mesocriconema spp. (minor plant-parasite), and M. incognita (major plant-parasite) were greater in organic than conventional production at both midseason and harvest (Fig. 2). Cephalobus spp. (cp2 bacterivore) was not affected by production system at midseason and was greater in conventional than organic production at harvest. Aphelenchus spp. (cp2 fungivore) was not affected by production system.

Figure 1:

Figure 2:

Similarly, all nematode community indices were affected by production system at both midseason and harvest. The maturity and channel indices were each decreased in organic relative to conventional production (Fig. 3). The enrichment index, structure index, and Hills N1 diversity – derived from both nematode genera and nematode trophic guilds – were each decreased in conventional production relative to organic production.

Figure 3:

Nutrient rate impacts on nematode abundances were relatively inconsistent, and bacterivores and fungivores were the groups most often affected. In organic production, bacterivore and fungivore abundances were greater at higher nutrient rates in Year 1, but unaffected in Year 2 (Table 5). Overall bacterivore abundances in organic production were greater at 336 kg N/ha than certain lower rates at midseason and harvest Year 1, but were unaffected in either season in Year 2 (Table 5). This was driven primarily by extreme enrichment opportunists as Rhabditis spp. (cp1 bacterivore) had similar trends to overall bacterivores. In contrast, the more basal enrichment opportunist Cephalobus spp. (cp2 bacterivore) was not significantly affected by nutrient rate in organic production (Table 5). At harvest in Year 1 in organic production, fungivore abundances were greater at 392 kg N/ha than 224 or 280 kg N/ha, but were not affected at any other date. Both common fungivore genera (Aphelenchus spp. and Aphelenchoides spp.) followed similar trends to overall fungivore abundances in organic production, although Aphelenchoides spp. was affected at both midseason and harvest of Year 1 whereas Aphelenchus spp. was only affected at harvest Year 1.

Conventional nutrient rates had inconsistent effects on fungivores and bacterivores (Table 6). At harvest Year 1, greater nutrient rates (336 and 392 kg N/ha) increased total bacterivores and Cephalobus spp. relative to the lower rate of 224 kg N/ha. In contrast, in Year 2 at harvest, a moderate nutrient rate (224 kg N/ha) had the greatest total bacterivore abundance, but Cephalobus spp. was not affected by conventional nutrient rates in Year 2. Rhabditis spp. was not significantly affected by conventional nutrient rate. Moderate nutrient rates (224 or 280 kg N/ha in Year 2 or Year 1, respectively) increased total fungivore abundances in Year 1 and Year 2 harvest relative to extreme rates. Aphelenchoides spp. followed a similar trend in Year 2, but Aphelenchus spp. was unaffected in either year. Omnivore-predators were unaffected by nutrient rate in either organic or conventional production in any season (data not shown).

Total herbivores and Mesocriconema spp. were not significantly affected by nutrient rate in either conventional or organic production in any season (Table 7). Soil abundances of M. incognita were significantly greater for a moderate nutrient rate (280 kg N/ha) than extreme rates in midseason of Year 2 of organic production. For Year 1 of organic production and both years of conventional production, M. incognita soil abundances were relatively low and unaffected by nutrient amendment application.

Nutrient rates had minimal effect on nematode community indices in either conventional or organic production. In midseason Year 1 of organic production (Table 8), the enrichment index was greater at high nutrient rates (336 or 392 kg N/ha) than a lower rate (224 kg N/ha). There were no significant nutrient rate effects on structure index, or Hill’s N1 diversity based on either genera or guild (Table 8). Similarly, maturity index and channel index were not significantly affected by nutrient rate in any season or production system (data not shown).