INTRODUCTION

The root-lesion nematode, Pratylenchus penetrans (Cobb), is a significant parasite of flue-cured tobacco (Nicotiana tabacum L.) in Ontario (8,9,10,14,19,21). About 97% of the land (>23,000 ha) cropped to tobacco each year is fumigated to control nematodes and black root rot disease, and to improve leaf quality (personal communication, D.L. VanHooren, Tobacco Specialist, Ontario Ministry of Food and Rural Affairs, Delhi, Ontario). The fumigant materials used are predominantly (>96%) broad-spectrum fumigants such as Telone II (94% 1,3-dichloropropene and 6% related C₃ hydrocarbons), Telone C17 (78% 1,3-dichloropropene, 6% related C₃ hydrocarbons, and 17% chloropicrin), Vorlex Plus (80% 1,3-dichloropropene and 20% methylisothiocyanate), and Vorlex + CP (68% 1,3-dichloropropene, 17% methylisothiocyanate, and 15% chloropicrin). It is recommended that these materials be injected with a single shank chisel per row to a depth of 10 to 15 cm about 3 wk before planting. As the zone of soil affected by the fumigant is only about 0.1 m on either side of the point of injection and good nematode control is only within this narrow band, it is important that transplant rows accurately follow the fumigant rows (11,16). The median recommended rates of application for nematode control are 8.8 and 3.7 ml/m of row for the Telone II and Vorlex Plus formulations, respectively, and 9.4 and 7.0 ml/m of row for the Telone C-17 and Vorlex Plus CP formulations to control both nematodes and black root rot (2). Although the black root rot resistant tobacco variety ‘AC Gayed’ is registered and recommended for use in heavy, cold, bottom soils with a history of producing black root rot, only about 4% of growers plant it because of slightly reduced yield (235 kg/ha) (2). About 50% of the fumigant used in Ontario contains chloropicrin and is applied at the higher multi-purpose rates recommended for both nematode and black root rot control (personal communication, P. Goodwin, Spray Application Technology Specialist, Agricultural and Rural Division, Ontario Ministry of Agriculture, Food, and Rural Affairs, Simcoe, Ontario).

Fumigants also are known to affect the populations of many soil microorganisms (20,31,32). A temporary loss of soil nitrifiers and a subsequent increase in soil NH₄ levels was observed by Tu (32) and Ball-Coelho (3). Inhibition of nitrification reduces leaching losses in sandy soils because NH₄-N is less mobile than NO₃-N (5). This preservation of NH₄ has the net effect of storing N in the soil until the tobacco plant has developed sufficiently to make use of it and it is thought to contribute significantly to the overall response to fumigation. Other plants, such as corn (Zea mays L.), also may respond to enhanced ammonium supply because NH₄ absorption is less energy costly to the plant than NO₃ absorption (6). The loss of living microbial biomass, however, is of relatively short duration and populations, including the soil nitrifiers, generally recover in about 7 wk (10,16). Work done by Ball-Coelho (3) has shown that, even with fumigation, tobacco yield and quality respond to enhanced ammonium supply resulting from the use of the nitrification inhibitor dicyandiamide (DCD) and suggests that this response would also occur in non-fumigated tobacco, if nematodes could be controlled by alternative means.

The increase in tobacco yield and quality with fumigation is well documented and virtually universally accepted by Ontario growers and the tobacco industry based on long experience. The vast majority of growers do not consider testing to determine nematode populations prior to applying fumigants necessary, nor is it generally recommended (2).

It is likely that the regulations governing the existing chemical fumigants will severely restrict or perhaps even eliminate their usage in the foreseeable future (17). Fumigant constituents, such 1,3-dichloropropene, are downwardly mobile and have been found in groundwater (15,18,22). Szeto (27) found 1,2,2-trichloropropane, a trace impurity in the soil fumigants Telone and Telone II, in 44% of the domestic water wells sampled in lower Fraser Valley, southwestern British Columbia, and suggests that the most likely origin of this contaminant was from past applications of fumigants to raspberry and strawberry fields. The Ontario tobacco industry is the largest single user of fumigants in Canada and at this time no practical alternatives to chemical fumigation exist.

Marigold (Tagetes sp) controls the soil populations of numerous nematode species (7,13,23,25,26). Trials were conducted from 1996 to 1998 at the Agriculture & Agri-Food Canada, Southern Crop Protection & Food Research Centre, Delhi farm, to compare a marigold crop rotation to the traditional tobacco cropping system based on a rye rotation crop and chemical fumigation prior to transplanting tobacco. Previously published results indicated that either Tagetes patula cv. Creole or Tagetes erecta cv. CrackerJack, grown as a rotation crop in place of the traditional common fall rye (Secale cereale L.) reduced nematode populations to < 100/kg of soil within 75 d after seeding (4,21). Populations remained below the very conservative economic threshold level of 500 nematodes/kg soil for the rest of the rotation crop season and for the two following years, even when a host crop such as tobacco was grown on the site. These earlier studies also reported the effects of multiple marigold species, seeding density, fertilization rate and incorporation time on nematode and soil microorganism populations, rotation crop biomass production, biomass decomposition, soil fertility, tobacco root growth, and tobacco yield (4,21,29). In this paper, we compare early season soil inorganic N levels, flue-cured tobacco growth and quality, and the economic viability of a marigold rotation system and a rye rotation with and without fumigation prior to transplanting tobacco.

MATERIALS AND METHODS

Two field plot sites, designated “A” and “B,” were established so that there was a rotation crop and a tobacco crop grown each year at the Agriculture & Agri-Food Canada, Southern Crop Protection and Food Research Centre, Delhi, Ontario. The plots were established on a Fox loamy sand soil with the following characteristics: 85% sand, 7.5% clay, 7.5% silt, 1.2% organic matter, pH 6.0–6.4, and bulk density of 1.4 g/cm³. Plots were cropped with common fall rye, T. patula cv. Creole, or T. erecta cv. CrackerJack grown in the rotation crop year followed by flue-cured tobacco cv. Delfield grown in the field crop year. Field plots were 8 m x 6 m in size and treatments were arranged in a randomized block design with four replications.

Common rye was seeded with a grain drill at a rate of 540 seeds/m² in late September to early October after tobacco harvest was complete and the land was worked. The following spring (1, 3, and 5 May in 19951, 3, and 5 May in 1996, and 1997), 45 kg N/ha (NH₄NO₃) was broadcast over the rye plots. In mid summer (28 July, 8 and 1 August in 199528 July, 8 and 1 August in 1996, and 1997), the rye straw plus grain was disked down and an additional 20 kg N/ha (NH₄NO₃) was broadcast over the residue to accelerate residue decomposition. The grain was mature at this point and self-seeded to produce a winter cover crop to help control soil wind erosion. Rye regrowth was plowed under the following spring prior to transplanting tobacco.

Marigold plots were seeded with an 8-row cone-type seeder (Fabro Ltd., Swift Current, Saskatchewan, Canada) in 32 rows 18 cm apart, centered on the 6 m wide plot, and to a depth of about 3 mm. Prior to seeding, 45 kg N/ha of urea was broadcast over the plots and lightly worked in. Weeds were controlled by hand hoeing in 1995 and by hand hoeing plus the herbicides Dacthal (chlorthal dimethyl), 10 kg product/ha applied pre-emergent, and Fusilade II (fluazifop-p-butyl), 2.0 l product/ha applied post-emergent, in 1996 and 1997. Marigold tops were either cut with a small sickle bar mower about 20 cm above the soil surface or left standing over winter. The cut tops were left on the soil surface with no further cutting or incorporation until the following spring. Tops left standing over winter were flail mowed in the spring to break up the stalks, and then plowed under prior to transplanting tobacco.

Rye controls 1 and 2 were plowed in early May. Rye control 1 was row fumigated with Vorlex Plus CP at 67 l product/ha. The fumigant material was applied to the plot rows by single point injection to a depth of about 20 cm and hilled over the injection trench with an additional 15 cm of soil to give an effective injection depth of about 35 cm. The fumigant hill was knocked down at transplanting. Marigold plots were flail mowed to break up stalks and plowed on 16 May 1996 (site A), 15 May 1997 (site B), and 12 May 1998 (site A) and the land prepared for transplanting. Tobacco seedlings produced in an unheated greenhouse were transplanted into field plots on 30 May 1996 (site A), 28 May 1997 (site B), and 19 May 1998 (sites A and B). Field plots consisted of 5 rows centered on the 6-m-wide plot. Rows were 1.07 m apart with a spacing of 61 cm between plants in the row. All data were collected from the middle three rows. Normal cultural practices for this region were followed, including application of 830 kg 3-6-18/ha applied in bands at transplanting plus a sidedress application of 300 kg 9-0-27/ha applied in late June.

Population densities of P. penetrans were determined from a bulk sample consisting of 10 soil cores taken 20 cm in depth. Cores, 2.5 cm diameter, were taken from arbitrarily selected locations not less than 1 m from the edge of plots for rotation crops or taken in-row not less than 1 m from the ends of the middle 3 rows of the plot for tobacco crops. Samples were collected in the spring at time of transplanting, mid-season, and in the fall. Nematodes were extracted from a 50 g sub-sample by the Baermann pan method and counted, and then the population densities per kg of soil were calculated (30).

Population densities of P. penetrans in tobacco roots were determined by digging up the main root mass of two tobacco plants from the middle three rows in each plot. The roots were washed to remove soil, allowed to drain, air dried to remove the excess surface water, and weighed to determine the fresh root weight. An arbitrarily selected sub-sample of small fibrous roots from each root system was taken and transferred to a misting chamber for 14 d to extract the living nematodes into tap water (24). The nematodes were counted and the number per g of dry root was calculated. After extraction was complete the root sample was oven dried at 65°C to a constant weight for DM determinations.

Soil inorganic nitrogen (N_inorg) was measured in June to help predict the approximate fertilizer N sidedress rate for tobacco and in July as an indicator of N release from previous crop residues. Soil NH₄ and NO₃ levels were determined from a composite sample of 10 soil cores per plot, 2.5 cm in diameter x 20 cm deep, taken in-row, between plants (P1) and 0.2 m from the plant between rows (P2) on 4 June and 11 July 1996 and 4 June and 10 July 1997. Soil N_inorg was extracted by shaking 25 g field-moist soil in 25 ml 2N KCl for 1 hr. Soil NO₃ and NH₄ concentrations were determined from the filtered extract by continuous flow colorimetry (28). Soil water content was determined gravimetrically to convert concentrations to a dry-weight basis. Tobacco plant height was determined on 17 July 1996, 7 July 1997, and 13 July 1998 by measuring the distance from ground level to top of leaf canopy of the 7th plant in rows 2, 3, and 4 of each field plot. Days to flowering was calculated as the time from transplanting until 50% of the plants in plot rows 2, 3, and 4 had at least one open floret.

Tobacco was harvested in five primings and cured in a downdraft stick kiln. The cured leaves were sorted into grades according to the Ontario Farm Products and Sales Act (1), and yield, grade index (average value in $/kg), and percent A - grades calculated for each plot. Yield is the average of calculations based on area harvested and on number of plants harvested. A - Grades are defined as those grades under the Farm Products Marketing Act for which a minimum grade price has been established by mutual agreement between the producers and buyers. The A - Grades are lemon, orange, tan, or mahogany in colour, relatively free of serious defects, and have been identified by the buyers as the most desirable leaf produced by Ontario growers. The Pearson product moment correlation coefficients (r) were determined between the transformed soil nematode populations in the spring and mid-season, and tobacco root nematode populations at the end of harvest with plant height in July, days to flowering, and yield. The transformation used was log x + y, where x=raw data and y=1, 10, or 100 as required to produce a normal distribution in the data set.

Analysis of results indicated no significant differences related to marigold cultivar, incorporation time, seeding rate, or nitrogen fertilization rate in the tobacco crop yield following the marigold rotation crop (4,21). Allowing the marigold crop to stand over winter, rather than mowing in August or September, slowed release of N from marigold residues and conserved it until the tobacco plant was in place and absorbing nutrients (4). Therefore, combined results for marigold cultivars Creole (T. patula) and CrackerJack (T. erecta) (45 kg N/ha broadcast prior to seeding, 100 viable seeds/m², tops left standing over winter, stalks flail mowed in spring followed by plow and disc tillage in mid-May prior to transplanting tobacco) were compared to rye, with and without fumigation, using the General Linear Model ANOVA procedure of CoHort Software (Version 5.0, Box 19272, Minneapolis, MN.). Where significant treatment effects were observed, means were compared at the 0.05% probability level using Protected Least Significant Difference (LSD).

Data from a survey of custom farm-work rates charged in Ontario (12) were used to estimate the difference in cost of field operations between the traditional rye plus chemical fumigation vs. marigold rotation, under two marigold management scenarios: one - as tested in this trial, and the second - a less costly option based on assumptions about alternative herbicides and cropping practices. Mean yield and grade indices for 1996 site A, 1997 site B, and 1998 site A were used to estimate the difference in total tobacco crop returns following the traditional rye rotation and the marigold rotation.

RESULTS AND DISCUSSION

Relationships Between Soil Inorganic Nitrogen and Crop Response

The mean soil NH₄ concentration in early June (Table 1) was greater with rye + fumigation (9.5 mg/kg) than the mean of rye alone (1.7 mg/kg) and marigold (2.2 mg/kg) treatments. These results agree with previous observations by Tu (32), Elliot et al. (9), and Ball-Coelho (3) comparing fumigated and non-fumigated plots after rye. This is partially due to the temporary reduction in the populations of soil nitrifiers (8,9,10,32). In 3 of 4 trials, early season soil NO₃ in rye plots was not reduced by fumigation. Increased NH₄ without a concurrent decrease in NO₃ suggests that mineralization of the decaying microbial biomass may have been replenishing the soil N_inorg pool. We observed a small increase (mean of 4 trials) in total early season N_inorg concentration in-row (P1), which is the only zone of soil affected by fumigation, in fumigated vs. non-fumigated plots. The increase, 6.7 mg/kg of soil, translates to about 3.6 kg N/ha. This relatively small increase in NH₄ and total inorganic N does not offer much support to the theory by Elliot et al. (9) that the tobacco yield and quality response to fumigation is not entirely due to nematode control, but also due to improved N fertility early in the growing season.

Table 1. Pre- and post-sidedress NH₄ and NO₃ concentrations in top 0.2 m of soil in-row (P1) and 0.2 m from plant between rows (P2) in tobacco grown after rye or after marigold 1996 to 1998

View Fullscreen

By the second wk of July the fumigant effect on soil N_inorg concentration was gone. Additional N availability in the zone of soil containing the new transplant and its developing root system could result in more early season growth, and, based on our field observations of these and other past trials, plants in the fumigated plots appeared larger and somewhat darker green in mid-June than those from non-fumigated plots. However, the differences in NH₄, NO₃, and total N_inorg concentrations between treatments were not large enough to produce statistically significant correlation values with any tobacco growth or quality parameters in this trial (data not shown), and mean tobacco yield and quality from plots previously cropped to marigold was equal to or greater than tobacco from plots previously cropped to rye followed by chemical fumigation (Table 2).

Table 2. Effects of chemical fumigation and marigold rotation crop on flue-cured tobacco growth, yield, and quality for 1996 to 1998.

View Fullscreen

Differences in root growth may help to explain this contrary result. In previous trials (3) root growth between rows in late July was greater in fumigated tobacco than for tobacco grown in a non-fumigated rye check plot. It was speculated that this may be the effect of nematode injury to the roots restricting growth and proliferation into the soil between rows. In turn, this reduced root growth was probably responsible for the observed reduction in yield (436 kg/ha, mean 1994 and 1995) due both to reduced access by roots to soil water and nutrients, and to reduced availability of N in the non-fumigated plots.

In the present study, in-row root lengths were greater for marigold and fumigated rye plots than for non-fumigated rye, and between-row root length was greater for marigold than for either rye plots in 1997 (data not shown) (21). The trends were the same in 1998 site A, but not statistically significant. We also observed differences in fresh weight of root measured after harvest (21). After marigold, tobacco fresh root weight was consistently greater than the non-fumigated rye check and, in two of four trials (1997 and 1998, site B), greater than the fumigated rye check. The virtual elimination of root-lesion nematodes from both the in-row and between-row areas of the marigold-cropped plots probably resulted in a larger and more laterally distributed tobacco root system by midsummer. In turn, this root system would have greater access to soil water and nutrients, especially from outside the zone of influence of the fumigant. This may have more than compensated for the slightly lower N availability in early June as compared with the fumigated plots. In support of this theory, we observed that soil gravimetric moisture content (mean ± SE) was lower after marigold (7.8±0.4 and 8.0±0.2% in 1996 and 1997 respectively) or after rye + fumigation (7.8±0.4 and 7.5±0.4% in 1996 and 1997 respectively) than after rye without fumigation (8.7±0.4 and 8.7± 0.4% in 1996 and 1997 respectively); indicating that both marigold rotation and fumigation had similar effects on improving root growth and, thereby, water uptake.

Relation Between Nematode Populations and Crop Response

Early season growth rate, as measured by plant height in July, was consistently less for non-fumigated rye plots than for either fumigated rye or marigold plots (Table 2).

The effects of early and mid-season nematode feeding injury and reduced nutrients available to the non-fumigated rye plots might be responsible for these results. Spring and mid-season nematode populations were negatively correlated to plant height in 1996 and 1997, mainly due to differences between the fumigated and non-fumigated rye treatments (Table 3). The lack of correlation with spring counts at both sites in 1998 was due to relatively low numbers of nematodes that year (21). On site B, 1998, the mid-season nematode populations, while still relatively low at about 1500/kg of soil, were significantly correlated with plant height (Table 3).

Table 3. Correlation (r) of root-lesion nematode populations in soil and tobacco roots with plant height in July, days to 50% flower, and yield for 1996 to 1998

View Fullscreen

Another measure of early and mid-season plant growth rate, days to flowering, was consistently longer for the non-fumigated rye treatments indicating that tobacco in these plots was slower growing, again, probably due to the combined effects of more nematode feeding injury due to greater nematode populations and reduced N_inorg (Tables 1,2). In 1996 site A and 1998 site B, tobacco in the marigold rotation plots flowered later than in the fumigated rye plots and, although not significant, the trends in 1997 site B and 1998 site A were similar. In these plots the in-row (P1) spring nematode populations for marigold and fumigated rye treatments were the same (21). A possible explanation for this response may be that the slightly lower N availability in early June, although not correlated to days to flower to a degree to be statistically significant, delays plant maturation and, thereby, increase days to flowering. Although tobacco after marigold was slower to flower than fumigated tobacco, final crop yield was not reduced. This is probably due to development of a more efficient root system in the zone of soil outside the effects of fumigation (P2), later in the growing season, which, in turn, allowed the tobacco to “catch-up” in terms of yield and quality. Days to flowering were positively correlated with spring nematode populations in 1996 and 1997, but not consistently to either mid-season or tobacco root nematode populations. Again the lack of correlation with spring nematode populations in 1998 is probably due to the relatively low soil nematode populations observed that year.

Grade index and percentage of A-grades are both measures of leaf quality. The non-fumigated rye check had a lower grade index in three of four trials (mean of $0.30/kg) and consistently produced a lower percentage of A-grades (mean of 14.1%) (Table 2). This result is consistent with virtually all prior research trials and grower experience in Ontario. The reduced rate of growth due to nematode feeding injury observed in these plots delays crop maturity which, when combined with a leaf grading system that severely penalizes immaturity, has a direct negative impact on leaf quality.

Yield in non-fumigated rye plots was reduced by a mean of 629 kg/ha compared to the fumigated rye check (Table 2). Again this is a response consistent with many research trials and grower experience in Ontario. In 1996 site A and 1997 site B, marigold plots yielded more than the traditional fumigated rye plots. Excluding 1998 Site B (double cropped tobacco), the mean difference in yield between the fumigated rye check and marigold was 163 kg/ha which is a relatively low increase in yield, but nevertheless, could be economically significant to growers. Yield was consistently correlated to root nematode populations as expected, which supports the hypothesis that damage to the roots results in reduced uptake of soil moisture and nutrients over the course of the entire crop growing season.

Cost of Production

Mr. Carl Fletcher, Business Management Advisor, Ontario Ministry of Agriculture, Food, and Rural Affairs, Komoka, Ontario, conducted a survey of custom farm-work rates charged in Ontario, in 1997 (12). An estimate of the differences in cost of the marigold rotation crop vs. fall rye followed by chemical fumigation can be made based on 1997 mean costs for custom farm work. Only a few field operations are changed and the major differences are related to costs for marigold seed, herbicides for the marigold crop, land preparation and seeding, fumigation, and gross crop returns (yield x grade index, data not shown) (Table 4).

Table 4. Estimated cost of production, based on mean 1997 custom farmwork rates charged in Ontario, for field operations relating to rye and marigold rotation crops and chemical fumigation

View Fullscreen

The most costly marigold rotation system, (Marigold, scenario 1, Table 4) involves the usual fall tobacco stalk cutting and tillage, seeding rye in the fall as winter cover, spring tillage to work down the rye and prepare a seedbed, spring seeding of T. erecta cv. CrackerJack marigold purchased in small quantities from the flower seed industry, and an expensive herbicide such as Dacthal. This system results in a decrease in cost of field operations related to rotation crop management and fumigation of $627 plus an increase in gross returns of $764 (mean of $548, $1740, and $3 for 1996 Site A, 1997 Site B, and 1998 Site A respectively) for a net gain of $1391/ha in favour of the marigold system which is about a 12% increase in gross crop returns.

A less costly crop management system (Marigold, scenario 2, Table 4) might involve allowing the tobacco stalks to stand over winter as a ground cover thereby saving the cost of the fall land preparation and seeding down the rye winter cover crop, minimal spring tillage, no-till drill seeding a cheaper marigold cultivar purchased in bulk (eg. T. minuta cv. Mexican, $110/kg), and a cheaper herbicide such as Treflan + Fusilade II plus row cultivation to control weeds. This system would result in a decrease in cost of field operations related to rotation crop management and fumigation of $1121 plus an increase in gross returns of $764 for a net gain of $1885/ha in favour of the marigold system which is about a 17% increase in gross returns. The true reduction in net cost of production for most growers would be somewhat less than the $1885 to $1391/ha calculations based on custom farm-work rates. Most growers in Ontario own and operate the majority of the equipment required to grow the rotation and tobacco crops and their costs of production might be expected to be lower than the custom rates. While these numbers are only estimates, the range in values calculated certainly indicate that Ontario growers could adopt this crop production system without suffering any economic loss. In addition, growers could greatly reduce their personal exposure to toxic chemicals and also reduce the potential for groundwater contamination by fumigants or contaminants contained within fumigant formulations.

In conclusion, these trials indicate that the marigold rotation crop is functionally and economically a viable alternative to chemical fumigation for the control of root-lesion nematodes in flue-cured tobacco and, if coupled with a black root rot resistant tobacco cultivar, can eliminate the need for high fumigant application rates and for formulations containing chloropicrin. The marigold system slightly reduces early season soil N availability, but does not require any major changes to tobacco crop management. The marigold rotation results in a small increase in tobacco yield with no reduction in cured leaf quality and may result in a net increase in gross returns to the grower when compared to fall rye followed by chemical fumigation.