| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Agric. & Agri-Food Can., Southern Crop Protection and Food Res. Cent., 1391 Sandford St., London, ON, Canada N5V 4T3
b Agric. & Agri-Food Can., Southern Crop Protection and Food Res. Cent., Box 186, Delhi, ON, Canada, N4B 2W9
* Corresponding author (ballb{at}agr.gc.ca)
Received for publication October 27, 2001.
 |
ABSTRACT |
|---|
 TOPNOTES ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Abbreviations: BD, bulk density • CFPM 101, Canadian Forage Pearl Millet Hybrid 101 • CFSH-17, Canadian Forage Sorghum Hybrid 17 • CGPMH-1, Canadian Grain Pearl Millet Hybrid 1 • CGSH-7, Canadian Grain Sorghum Hybrid 7 • RLN, root-lesion nematode(s) • rye-F, rye rotation with fumigation the subsequent year • rye-NF, rye rotation without fumigation • SOM, soil organic matter
|
INTRODUCTION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Research into alternative rotation crops has shown that rotation with marigold controls RLN effectively (Miller and Ahrens, 1969; Reynolds et al., 2000). However, the costs and/or difficulties associated with seeding and weed control, coupled with a lack of return from the rotation year crop, has limited the adoption of marigold rotation systems. Hence, a rotation crop that is less expensive to grow and that has potential for return would be more readily adopted by growers of crops such as tobacco and potato.
Previously, we reported (Jagdale et al., 2000) that where forage (CFPM 101) and grain (CGPMH-1) pearl millet were grown as rotation crops, RLN populations were reduced in both the rotation and subsequent tobacco crops compared with where rye (Danko) was grown as a rotation crop. These observations, while encouraging, were based on only one field trial (1997–1998) and required validation. Furthermore, because potato cultivation is expanding in the sandy soils of southwestern Ontario, a rotation system to control RLN biologically where either tobacco or potato is the susceptible crop would be advantageous. The Russet Norkotah cultivar in particular is promising due to its potential to obtain a shape associated with high market returns in sandy soils. Such a rotation system should also minimize the need for fumigants, reduce grower input costs, conserve soil and water, and provide a by-product for human or livestock consumption. Currently, little research exists in this area. Chen et al. (1995), for example, found that a 2-yr alfalfa (Medicago sativa L.) rotation improved subsequent potato yields, but the response was more likely due to improved soil N availability than RLN control. This extended rotation system is not suitable for tobacco because N release following 2 yr of alfalfa would exceed crop N requirements and result in poor leaf quality and uneven ripening.
Our objective was to confirm whether a pearl millet rotation crop could consistently control RLN in subsequent potato crops as it had in the prior tobacco experiment. To do this, we measured RLN populations in soil and roots during two rotation potato crop experiments, comparing pearl millet to suppressive marigold (Cracker-Jack) and supportive (rye) RLN rotation crops.
|
MATERIALS AND METHODS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Experimental Design and Cultural Practices
In both experiments, five rotation crop treatments were imposed in a randomized complete block design with four replications. Rye was included as a supportive host with two plots per block so that one plot could be fumigated (rye-F) before planting potato the following year. The rye-F treatment was included to allow comparison to previous results (Jagdale et al., 2000) and because the rotation system was being developed both for crops that are typically fumigated (tobacco) and for crops that are not usually fumigated (potato).
The Exp. A site was sod from 1990–1993, fallow in 1994, and rye–tobacco in 1995–1996 followed by a drilled rye cover crop in the fall of 1996 (Table 1). In the spring of 1997, when the Jagdale et al. (2000) experiment was initiated, rye was incorporated by conventional tillage in plots designated for alternative rotation crops {CGPMH-1, CFPM 101, and grain sorghum [‘Canadian Grain Sorghum Hybrid 7’ (CGSH-7)]}. Rye remaining in two plots per block was incorporated (grain and straw) by discing in late July and then left to regrow throughout the fall, winter, and spring. Canadian Forage Pearl Millet Hybrid 101 was mowed twice mechanically with residues returned as green manure each time. Mowing was important because mature pearl millet residues can be difficult to incorporate and slow to decompose, causing problems during planting and variability in subsequent crop growth. In 1997 (and in subsequent experiments), residues were returned because (i) in 1997, it was not known whether residues would contribute to RLN suppression and (ii) on fragile, low organic matter soils or areas within fields, residue returns help to retain moisture and maintain crop productivity. Subsequent greenhouse trials revealed that CFPM 101 residues do not suppress nematodes (G. Jagdale, unpublished data, 1998) and hence are not needed for that purpose. Grain from CGPMH-1 and CGSH-7 was combined and removed as a cash crop in the fall of 1997. Grain crop stalks and remaining CFPM 101 plants were left standing over the winter to conserve soil, N, and water. One of the two sets of rye plots per block was fumigated in the spring of 1998, and tobacco (‘Delfield’) was then planted over the entire site. Following tobacco harvest in the fall of 1998, rye was drilled (5 Oct. 1998) as a winter cover crop. Other experimental details of the 1997–1998 growing seasons are described in Jagdale et al. (2000). In 1999, the rotation crop treatments of 1997 were repeated on the same plots, except that marigold was grown as a suppressive control (Table 1) on plots formerly planted with CGSH-7 (which did not suppress RLN).
|
On 19 Apr. 1999, all rye treatment plots in both experiments received the recommended rate (OMAFRA, 1999) of fertilizer (Table 1). On 20 April, 3000 kg ha-1 lime was broadcast on all plots of both sites, and 50 kg ha-1 K2O as muriate of potash was broadcast on all plots of Exp. A. The winter rye cover crop at both experiments was incorporated (29 Apr. 1999) by conventional tillage in the plots designated for alternative rotation crop treatments [CGPMH-1, CFPM 101, and marigold at Exp. A and forage sorghum (CFSH-17), CFPM 101, and marigold at Exp. B] while rye remaining in two plots per block at each experiment was incorporated (grain and straw) by discing in late July and was then left to regrow throughout the fall, winter, and spring. On 26 May 1999, urea was broadcast and incorporated with a spring tooth and packer in plots to be planted with CFSH-17, millet, and marigold. Fertilizer N rates used for marigold and millet (Table 1) were determined based on previous work demonstrating that these rates produce sufficient crop growth without leaving excess residual inorganic N in soil after the growing season that could potentially leach into ground water over winter (Ball-Coelho et al., 2001; Jagdale et al., 2000). On 27 May 1999, 9 kg ha-1 chlorthal dimethyl (dimethyl tetrachloroterephthalate) was sprayed on marigold plots, and 1.68 kg ha-1 pendimethalin [N-(1-ethylpropyl)-3,4-dimethyl-2,6-dinitrobenzenamine] was sprayed on millet and CFSH-17 plots. Both herbicides were applied in 170 L ha-1 water.
The alternative rotation crops were drilled on 2 to 3 June 1999 at Exp. A and B (Table 1). Distance between millet rows was wider in Exp. A than Exp. B for ease of harvesting the grain crops at this site (grain sorghum in 1997 and grain millet in 1997 and 1999) and to allow for mechanical cultivation of weeds. A row width more typical of forage crops was used at Exp. B because both millet and sorghum cultivars were forage types.
Due to shallow seeding depth (2 cm) and dry weather following planting, about 0.8 cm of water was applied as one single irrigation on marigold plots at both experiments on 8 June to ensure germination. Weed escapes were controlled through periodic hand cultivation in addition to one mechanical cultivation of millet plots in Exp. A on 28 June 1999. Rye (straw and grain) was incorporated by discing on 23 July 1999. Leaving a stubble height of 10 cm, CFPM 101 (Exp. A and B) and CFSH-17 (Exp. B) were mowed twice at approximately optimum forage quality stage: 22 and 23 July (46–47 d after planting) and 23 and 24 August (42–43 d after first cut) for Exp. B and A, respectively, with residues left on plots each time. Marigold was not mowed midseason because mature plants can be easily incorporated into the soil, marigold residues do not contribute to RLN suppression (Reynolds et al., unpublished data, 1998), and marigolds do not regrow after mowing. Grain from CGPMH-1 plots was harvested with a combine on 27 Oct. 1999. Remaining CFPM 101, CFSH-17, and marigold plants and CGPMH-1 stalks were left standing over the winter.
In the 2000 crop year, plowing and secondary tillage (discing followed by spring tooth and packer) of all plots in both experiments was completed on 2 and 3 May, respectively. The rye-F plots in both experiments were knifed (20 cm deep, 76 cm wide) with 92.7 L ha-1 Vorlex CP (Aventis CropScience Canada Co., London, ON) (68% 1,3-dichloropropene and related chlorinated hydrocarbons, 17% methylisothiocyanate, and 15% chloropicrin) on 4 May using a single-row fumigator with disc hillers, which covered the injection trench with about 15 cm of soil. The fumigant rate was equivalent (per unit row length) to that applied in the 1998 experiment (recommended rate for tobacco), allowing comparison with previous results. All other plots were similarly knifed and hilled on 5 May to equalize physical disturbance, but fumigant was not applied. Seed of potato cultivar Russet Norkotah was planted on 24 May in 76-cm rows spaced 33 cm apart (centered on the fumigant band) at a depth of 5 cm using a single-row planter equipped with a fertilizer attachment. As preplant (4 May 2000) differences in soil inorganic N (top 20 cm) between rotation crops were small (<2 mg kg -1 for Exp. B) or were not significant (Exp. A), a single N fertilizer rate was used on all plots. Fertilizer (670 kg ha-1 of 22–4–22) formulated using monoammonium phosphate, sulfate of potash, and urea was banded 5 to 7.5 cm to the side and 2.5 cm below the potato seed piece during planting according to recommended rates (OMAFRA, 1998). Both experiments were cultivated on 7 June, and on 8 June, 0.915 kg ha-1 s-metolachlor {2-chloro-N-(2-ethyl-6-methylphenyl)-N-[(1S)-2-methoxy-1-methylethyl]acetamide} + 0.86 kg ha-1 linuron [N'-(3,4-dichlorophenyl)-N-methoxy-N-methylurea] (tank mix) was applied. Weed escapes were controlled by hand-weeding as necessary.
Summer rainfall in southwestern Ontario is almost always inadequate to meet the needs of most vegetable crops grown on sandy soils (Tan et al., 1983); thus, soil water content at both experiments was maintained during the 2000 crop season using a sensor-based drip irrigation system. Ro-Drip line with 0.9 L flow h-1 emitter-1, 30-cm emitter spacing, 16 mm i.d., and 0.152-mm (6 mil) thickness (Roberts Irrigation Products, San Marcos, CA) was laid on the hilled soil surface of all plots and covered using a disc hiller to a depth of 2 to 3 cm on 31 May 2000. One calibrated reflectometer (CS-615, Campbell Sci., Edmonton, AL) was placed in each of the five treatment plots of one block at both sites on 9 June to monitor soil water content. Reflectometers were installed perpendicular to the soil surface, with rods buried approximately 10 to 25 cm below the surface and approximately 10 cm laterally from the in-line drip emitters as suggested by Coelho and Or (1996). Irrigation was initiated at both experiments when the soil water content of at least one of the 10 sensors fell below approximately 0.12 m3 m-3 (Shae et al., 1999; Waddell et al., 2000). On three occasions (24, 27, and 28 July), water was added to bring the wetted zone (
one-third of the soil volume to 30 cm deep) to approximately field capacity (0.18 m3 m-3). Note that the soil water content fluctuations depicted in Fig. 1B and 1D are averaged over the five sensors buried at each site and that the average value usually did not fall below the 0.12 m3 m-3 trigger.
|
Soil and Root Sampling
Topsoil samples (composite of ten 2-cm-diam. by 20-cm-deep cores per plot) were collected five times in 1999 (along a diagonal at least 1 m apart and at least 1 m away from plot borders) from both experiments to measure the soil RLN population before fertilizing (14 April), before plowing (29 April), before planting (31 May for Exp. A and 1 June for Exp. B), and after the first and second cut of millet (5 August and 30 September). Samples were also collected from all plots using a tulip bulb planter on 10 September to measure root and soil RLN. Each sample was a composite of six 18-cm-long by 6-cm-diam. cores per plot. Three of the six cores per plot were collected immediately adjacent to plants (one from each end of the plot and one from the central area) while the other three were collected adjacent to the first cores but further away from plants to sample populations associated with both large and small roots. Topsoil samples (composite of ten, 2 cm diam., 20 cm deep cores per plot; three from each end of the plot and four from the central area) were also collected, three times in 2000, from all plots at the top of the hill from three of the four rows on 4 May (before fumigating, fertilizing, and planting), 18 July, and 28 August. To measure root RLN populations, roots were collected from all plots with a shovel (composite of roots from three potato plants per plot—one from the central area of the plot and one from each end—from rows not used for yield determination) on 18 July.
To determine soil bulk density (BD), undisturbed cores (4.8 cm diam. by 2.6 cm) were collected on 20 to 23 June 2000 from each plot in two blocks per experiment from four depths per plot: the top of the hill and three 10-cm increments to 30 cm deep, close to the hill but away from wheel tracks.
Laboratory Methods and Data Analyses
Before use, all CS-615 reflectometers were calibrated individually, which involved burying each probe in a separate bucket of soil and gradually (over a 2-wk period) increasing the soil water content by adding water to the soil from each bucket several times using a pressurized sprayer while mixing in a cement mixer. Volumetric soil water content determinations were made repeatedly throughout the calibration procedure by collecting and oven-drying (105°C) undisturbed cores (90.5 cm3) of soil from each of the 10 buckets before each addition of water. Quadratic equations were then calculated for each probe to describe the relationship between actual (volumetric) soil water content (y) and the average period (ms, x) measurement recorded by each probe over several hours before each soil sampling event.
Soil and root samples were stored in plastic bags at 4°C until processed. Root-lesion nematodes were extracted from thoroughly mixed 50-g subsamples of soil within 2 wk of collection using the Baermann pan technique (Townshend, 1963). Other nematodes, when present (infrequently) in samples, were identified to genus (usually Tylenchorhynchus sp.). Soil RLN counts were converted to an air-dry soil weight basis in 1999 and to an oven-dry soil weight basis in 2000. Root-lesion nematodes were extracted from roots for 2 wk in a misting chamber (Seinhorst, 1950) and counted using a stereomicroscope. Following RLN extraction in 2000, roots were oven-dried (65°C) before weighing. Root nematode populations were expressed on a fresh-weight basis in 1999 and a dry-weight basis in 2000. Gravimetric water contents (oven dry) determined from selected soil samples were converted to a volumetric basis using average BD (over 0- to 10- and 10- to 20-cm depths and over all treatments, 1410 and 1370 kg m -3 for Exp. A and B, respectively) to compare water content of collected soil samples to soil water content estimated using reflectometers (Fig. 1B and 1D).
Harvested potato tubers were stored at 10°C, cleaned within 2 wk using a power washer, left to air-dry for at least 10 min, weighed by plot (total weight), and then stored at 3.3°C. Less than one month after harvest, potato was graded at Brenn-B Farms using a commercial image analyzer (Acu-vision, Exeter Engineering, Exeter, CA) programmed to sort potato into three categories: premium (9.5–12.1 cm long by 4.8–7.1 cm wide), small but marketable (8.9–9.5 cm long), and unmarketable (<8.9 cm long or poor shape or color). The premium shape–size combination was chosen based on the specifications for a specialty baking potato market (Duchess Foods, Brantford, ON, Canada). Samples in each category from each plot were then weighed.
The RLN counts from soil and roots were normalized on ln(x + 1) transformation. Transformed soil RLN counts from 1999 through 2000 were analyzed using repeated-measures analysis, with block and rotation crop treatment as the between-subject variables and time (number of days since the first sampling date in 1999) as the repeated or within-subject variable (SAS Inst., 1996). Transformed 1999 and 2000 root RLN counts, as well as 1999 and 2000 soil water content, and 2000 potato yield and quality in both experiments were analyzed statistically using the General Linear Models procedure (SAS Inst., 1996) for analysis of variance (ANOVA) according to the randomized complete block design. When treatment effects were significant, means were compared using the protected LSD at a 0.05 probability level (SAS Inst., 1996). Regression curves were fitted to investigate temporal changes in the soil RLN population for each rotation crop treatment at both experiments in the rotation crop year (1999) but not for 2000 data as there were only three sampling events, nor for 1999–2000 data as a whole due to the winter interruption. Correlations between ln(RLN in soil) and both soil water content and potato yield were determined using PROC CORR (SAS Inst., 1996). Covariate analysis was used to remove the effect of variation in soil water content on soil RLN when soil water content varied between rotation crops.
|
RESULTS AND DISCUSSION |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
Experiment B (First Rotation Cycle)
Before planting the rotation crops on 3 June 1999, there were no differences in soil RLN populations between plots designated for the different rotation crops (Table 2). Therefore, the inherent site variability was acceptable for the experiment, and treatment-specific preplant management practices such as N fertilization and tillage did not significantly influence RLN numbers. From 5 Aug. 1999 until the final sampling event of the potato crop year in 2000, soil RLN counts differed with rotation crop treatment (Table 2). In CFPM 101 plots, soil RLN populations were 34% (5 Aug. 1999) and 35% (18 July 2000) of those measured in rye-NF plots. In marigold plots, the soil RLN population was 35% of that in rye-NF plots by 10 Sept. 1999. In plots fumigated on 4 May 2000, RLN populations remained suppressed on and after 18 July 2000 as numbers in rye-F plots were just 7% (18 July 2000) and 12% (28 Aug. 2000) of those measured in rye-NF plots. On all sampling events in 2000, plots cropped the previous year with marigold contained fewer RLN in soil than all other rotation crop treatments, including rye-F plots. By the final sampling event in 2000, soil RLN counts in plots where marigold was grown were only 7% of those in rye-NF plots (Fig. 1E).
Trends in Soil Root-Lesion Nematode Populations over Time and Relation to Soil Water Content
The soil RLN population varied with time at both experiments (Table 2). At Exp. A, temporal response curves of RLN in the rotation crop year were cubic (Fig. 1C), with soil RLN populations increasing before planting from April to mid-May, decreasing by 5 August, and then increasing again by September 1999. At Exp. B, the relation between soil RLN and time was quadratic for all crops (Fig. 1E). Soil RLN decreased for all except the two rye treatments by 5 August 1999 (Fig. 1E), with the largest decreases noted for marigold (2095 RLN kg soil-1 on 14 Apr. 1999 to 433 RLN kg soil-1 on 5 Aug. 1999) and CFPM 101 (1028 RLN kg soil-1 on 14 Apr. 1999 to 229 RLN kg soil-1 on 5 Aug. 1999). By 10 Sept. 1999, soil RLN had increased in all crops but did not increase further by 30 Sept. 1999.
The main difference between the two experiments with respect to temporal RLN trends was that soil RLN densities were less and tended to increase at the beginning of Exp. A compared with Exp. B for some treatments (rye-F, marigold, and CGPMH-1) but not others (CFPM 101 or rye-NF rotation plots; Fig. 1C and 1E). This pattern of fewer initial RLN at Exp. A than at Exp. B in the spring of 1999 was likely the result of treatment effects from the previous (1997) rotation crop cycle. Temporal RLN trends throughout the remainder of the 1999 growing season at Exp. A were similar to, but in some cases (marigold and CFPM 101) more dramatic than, those noted at Exp. B. On 5 Aug. 1999, larger decreases in soil RLN were noted at Exp. A than at Exp. B for both marigold (1904 RLN kg soil-1 on 29 Apr. 1999 to 234 RLN kg soil-1 on 5 Aug. 1999 at Exp. A vs. 688 RLN kg soil-1 on 29 Apr. 1999 to 433 RLN kg soil-1 on 5 Aug. 1999 at Exp. B) and CFPM 101 (1062 RLN kg soil-1 on 29 Apr. 1999 to 106 RLN kg soil-1 on 5 Aug. 1999 at Exp. A vs. 426 RLN kg soil-1 on 29 Apr. 1999 to 229 RLN kg soil-1 on 5 Aug. 1999 at Exp. B), which may have been due to the cumulative effects of a second cycle of alternative rotation crops in the case of CFPM 101. By 10 Sept. 1999, the soil RLN population had increased in all crops except marigold, and by 30 Sept., soil RLN numbers increased further in all plots except rye-NF.
Treatment-wide seasonal fluctuations in soil RLN may have been related to changes in soil water content as soil RLN population fluctuations in 1999 were similar to the fluctuations noted in soil water content. For most treatments, populations were least on 5 August when soils were driest (Fig. 1B and 1D). In 1999, rainfall (Fig. 1A) totals in both June (76 mm) and July (28 mm) were less than those of the 10-yr (1993–2002) average total rainfall for these months at this site (96 and 81 mm for June and July, respectively), and soil RLN counts were correlated with soil water content over the three sampling events that year at both Exp. A (r = 0.44, P = 0.0004) and B (r = 0.71, P = 0.0001). This was not the case in the wet 2000 crop year (227 and 164 mm in June and July, respectively) when soil RLN numbers and soil water content were not correlated (Exp. A: r = -0.18, P = 0.18; Exp. B: r = -0.13, P = 0.33). In 2000, soil water content was likely greater than a threshold level that may lead to RLN damage during sampling in dry soil. Taylor and Evans (1998) concluded that lower recovery of Pratylenchus spp. in dry soil was due mainly to mechanical disturbance during augering of brittle anhydrobiotic nematodes because they found improved extraction when water was added to soil before sampling.
Soil water content varied with rotation crop on 5 Aug. 1999 at Exp. A and on 5 Aug. 1999 and 4 May 2000 at Exp. B (Table 2). Soil water content was greater in rye plots (0.13 cm3 cm-3) than in the other rotation crop plots (0.09–0.11 cm3 cm-3) at both experiments on 5 Aug. 1999 (Fig. 1B and 1D), likely because the rye plants had been incorporated a few weeks earlier and so were not transpiring while plants in the other treatments were still extracting water from the soil profile. The increased soil water content may have contributed to greater RLN counts in rye plots on 5 Aug. 1999. After removing the effect of soil water content using covariate analysis, the rotation crop effect on soil RLN on 5 August was no longer significant (p > F = 0.13 and 0.56 for Exp. A and B, respectively). A reverse effect was noted on 4 May 2000 when soils were driest in rye rotation plots at Exp. B. The trend was similar for Exp. A on this sample date although differences were not statistically significant (Fig. 1B). The winter rye cover crop was actively transpiring until incorporation (2 May 2000) 2 d before soil sampling, which probably resulted in less water in soils of rye plots. Removal of the soil water content effect on soil RLN count data on this sample date had little effect on the overall rotation crop treatment effect and on treatment rankings at Exp. B, except that counts from CFPM 101 were equivalent to those in rye-F plots (as opposed to greater numbers in CFPM 101 than in rye-F plots when the soil water content effect was not removed).
Crop Rotation Effects on Root Root-Lesion Nematode Populations
Root RLN population did not differ between rotation crops in 1999, but rotation crop influenced RLN populations in potato roots the subsequent year (2000) at both sites (Table 2 and Fig. 2)
. A fumigation response was apparent in the July potato root RLN population, with counts in rye-F plots 30 and 22% (Exp. A and B, respectively) of those in rye-NF plots (Fig. 2A and 2B). Following marigold, potato root RLN counts were just 4% (Exp. B) and 11% (Exp. A) of those in rye-NF rotation plots and 20% (Exp. B) and 36% (Exp. A) of those found in rye-F plots. Following CFPM 101, the potato root RLN population was 39% of that in rye-NF plots at Exp. A (Fig. 2A) but no different than counts in rye-NF plots at Exp. B. Relative to rye-NF, CFSH-17 (Exp. B) did not suppress root RLN. Root RLN populations were not correlated with soil water content (data not shown) in either the rotation or potato crop years.
|
|
Nonetheless, potato yield was negatively correlated with soil RLN counts in the fall of 1999 and into midsummer of 2000, with the strongest correlation occurring on 4 May 2000 (Table 2). Bernard and Laughlin (1976) found that cultivar Russet Burbank was tolerant of RLN (i.e., yields were not reduced as initial RLN populations increased), but RLN populations were lower in their microplots (<1625 RLN kg-1) than in our field plots (<8000 RLN kg-1). In our study, yields were only weakly correlated with potato root RLN populations (Table 2). Bernard and Laughlin (1976) similarly found that potato root RLN density was not correlated with potato tuber yield.
For comparison with regression equations in the literature relating preplant RLN densities to potato yield, we regressed potato yield with soil RLN counts on 4 May 2000 (after conversion to equivalent units). Root-lesion nematode data from rye-F treatment plots were not included in this analysis because soil samples were collected before fumigation and so numbers were not representative of the initial (for potato) RLN population for this treatment. The slope of the relationship between relative yield (with marigold yield = 1) and RLN (converted to a 100-cm3 basis using measured average BD from each experiment) [relative yield = 0.99 - 0.0003(RLN); R2 = 0.40, P = 0.0001] was less than that obtained by Francl et al. (1987) [relative yield = 0.87 - 0.001(RLN); R2 = 0.28] with cultivar Superior grown in inoculated microplots. Differences in regression slopes may be due to greater RLN susceptibility of early maturing Superior compared with late-maturing Russet Norkotah (Kimpinski et al., 2001). The intercept and slope values for the regression of yield (t ha-1) with May soil RLN [yield = 24.43 - 0.0065(RLN); R2 = 0.28, P = 0.0017] were less than those obtained by Chen et al. (1995) [yield = 33.26 - 3.27(RLN); R2 = 0.18] with Superior following various rotation crops, likely because of the high yields that Chen et al. (1995) observed following 2 yr of alfalfa and 2 yr of clover (Melilotus officinalis Lam.). In contrast to legume rotations, potato yield differences in our experiments were not associated with rotation-induced differences in N availability as soil inorganic N (top 20 cm of soil) differed little between rotation crops the following spring (4 May 2000) or summer (28 Aug. 2000, data not shown).
Input costs for seed, herbicide, and N fertilizer for the marigold rotation exceeded those for the millet rotation but were less than those for the rye-F rotation system due to the high cost of fumigant (Table 3). The gross value of marketable baking potatoes (Can$ 0.75 kg-1, K. Marcoux, personal communication, 2002) was similar for the millet–potato, marigold–potato, and rye-F rotation systems (Table 3). Net returns ranked millet > marigold > rye-F > rye-NF at both experiments (Table 3), in spite of the limited yields due to late blight. Reynolds et al. (2003) calculated that marigold–tobacco rotation would increase returns by 12 to 17% over rye–tobacco with fumigation. In our study, returns from marigold–potato rotation were 8 to 10% greater than from rye–potato rotation with fumigation.
|
Mechanical cultivation is presently a good alternative to herbicides for weed control in millet and marigold because there are currently no labeled herbicides for grass control in millet, pendimethalin can injure millet on light-textured soils when heavy rains follow application, and chlorthal dimethyl is very expensive. Planting in 60-cm-wide rows (which allows cultivation) at lower population had little effect on either CFPM 101 dry matter production (data not shown) or RLN suppression. Mechanical cultivation would cost Can$ 17 to 34 ha-1 (for one to two cultivations) in lieu of the herbicide input cost. Net returns from millet rotation using this scenario would change little (increase by Can$ 6 ha-1), but those from marigold would increase by Can$ 459 ha-1, making the marigold rotation system hypothetically more profitable than millet. Our experience however, has been that marigold is difficult to establish and is a poor competitor against weeds. Weed control options for the marigold rotation system are currently being investigated.
|
SUMMARY AND CONCLUSIONS |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
Temporal trends in the soil RLN population in 1999 were similar for both experiments and were correlated with changes in soil water content. Soil RLN dynamics were not correlated with soil water content in 2000 when soil was wetter during sampling. Treatment differences in soil water content contributed to soil RLN rankings only on 5 Aug. 1999 (at both experiments) when greater soil RLN in rye plots may have been related to greater soil water content. Otherwise, rotation crop effects on soil RLN were not altered when soil water content was used as a covariate in the analysis.
Canadian Grain Pearl Millet Hybrid 1 did not prove to be consistent as a rotation crop for biological control of RLN. In contrast to the earlier tobacco rotation at the Exp. A site (Jagdale et al., 2000), the potato crop following CGPMH-1 rotation did not contain fewer soil or root RLN compared with rye-NF. Canadian Forage Sorghum Hybrid 17 (Exp. B) did not suppress RLN in soil or roots in 1999 or 2000 relative to the rye-NF treatment. Both CGPMH-1 and CFSH-17 were selected for field trials based on RLN suppression in short-term container trials (Jagdale et al., 2000), but given the results of our subsequent field experiments, short-term (6–8 wk) studies may not be predictive of a field growing season.
The RLN counts in rotation crop roots were not useful for predicting RLN suppression in the subsequent crop year. Soil RLN counts in the latter half of the rotation crop growing season appear to be the best predictor of the extent of biological control expected in the subsequent growing season. Potato yields (in 2000), although limited by foliar disease, were negatively correlated with soil RLN counts the previous September (1999) as well as throughout 2000. While there was no significant potato yield response to fumigation at Exp. B (first rotation cycle), total yields following marigold and CFPM 101 rotation were greater than yields following the rye-NF treatment. At Exp. A (second rotation cycle), a fumigation response occurred in the small marketable grade category (rye-F > rye-NF), and marigold and CFPM 101 rotation plots yielded more premium and small marketable potato than the rye-NF treatment. Our preliminary economic analysis demonstrates improved profitability with marigold or millet rotations compared with rye-NF or rye-F rotations.
While the mechanism of biological control of RLN by CFPM 101 remains undetermined, it may be related to S-containing compounds as both CFPM 101 and marigold shoots and roots (unpublished data) contained greater concentrations and amounts of S than plants exhibiting little or no suppression (rye, CGPMH-1, and CFSH-17).
|
ACKNOWLEDGMENTS |
|---|
|
NOTES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
|
REFERENCES |
|---|
|
TOPNOTESABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONSUMMARY AND CONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. R. Ball Coelho, R. C. Roy, E. Topp, and D. R. Lapen Tile Water Quality following Liquid Swine Manure Application into Standing Corn J. Environ. Qual., March 1, 2007; 36(2): 580 - 587. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Crop Science | Vadose Zone Journal | |||
| Journal of Natural Resources and Life Sciences Education |
Soil Science Society of America Journal | ||||
| Journal of Plant Registrations | Journal of Environmental Quality |
The Plant Genome | |||
0.05) from one another, and values in the same row followed by the same uppercase letter are not significantly different (P 
