Organic Heirloom Tomato Production Using Grafted Transplants, 2007

High Tunnel Harvesting Fruit

CEFS High tunnel tomatoes (May, 2007)

 

Summary
The nutrient uptake and fruit yield of grafted tomatoes was evaluated in controlled university greenhouses, research stations, and on private farms in North Carolina between 2007 and 2008. The physical effects of grafting stimulated plant growth and nutrient accumulation in our greenhouse study, while utilization of commercial rootstocks increased the level of grafting effects. In our two-year systems comparison trial, the total fruit weight of the high tunnel system out-produced the field system by 12-30% and the heirloom tomato Cherokee Purple grafted onto the rootstock Maxifort produced 23% more fruit compared to non-grafted plants. All three collaborative farmers were enthusiastic about incorporating grafting into their future farm practices.

Introduction
The goals of this project were to evaluate the nutrient uptake of grafted and non-grafted tomatoes as well as compare the productivity of grafted tomato crops.  Fertilizer applications are commonly required in low input soil-based tomato systems. Supplying post-planting nutrients to an organic tomato crop is a challenging task for reasons such as matching nutrient release with crop requirements and the expense of soluble materials that meet the National Organic Program (NOP) standards.  Literature from Asia, the Mediterranean, and Morocco suggests that grafted herbaceous plants are more efficient at absorbing certain macro- and micro- nutrients which could be a great advantage in low input systems such as organic operations.  If grafted tomato plants are more efficient at taking up nutrients then a reduction of fertilizer inputs may be possible or perhaps greater yields could be achieved at the same N input rate? If either of these factors are true than using grafted plants could make production systems more economically viable for farmers. This research project included: 1) the evaluation of the nutrient content and plant productivity of grafted tomatoes in a controlled research greenhouse, 2) a multi-year organic, high tunnel and open field systems comparison trial with nested grafting and N input level treatments, and 3) three, organic, on-farm research trials.  

STUDY 1: NUTRIENT UPTAKE AND PLANT GROWTH OF GRAFTED TOMATOES IN A CONTROLLED ENVIRONMENT

Materials & Methods
In the spring of 2007, a greenhouse experiment was conducted at North Carolina State University’s (NCSU) Southeastern Plant Environment Laboratory.  Three tomato cultivars Trust, German Johnson and Maxifort were utilized. Six treatments were included: 1) Trust grafted onto Maxifort, 2) Trust grafted onto Trust (self-grafted), 3) Trust (non-grafted), 4) German Johnson grafted onto Maxifort, 5) German Johnson grafted onto German Johnson (self-grafted) and 6) German Johnson (non-grafted).

The experimental design was a replicated 2 x 3 factorial with a randomized complete block design.  Plants were grafted following the ‘Japanese tube-grafting’ method. Seedlings were fertilized with a modified Steiner solution twice per day.  Plants were distributed within five adjacent blocks in the growth chamber post-grafting. Within each block, groups of five plants, from each of the six treatments were randomly arranged.  One plant per block was selected arbitrarily to be destructively sampled each week, for a total of five weeks. The 1st sampling date represents the 36th day after grafting (DAG) and the last sampling date, the 63rd DAG. 

Leaf area, leaf, stem, and fruit weight, and fruit number were sampled on a weekly basis for five weeks.  Root weight, plant height, and leaf tissue for a nutrient analysis were sampled on the last harvest date only.  Leaf area measurements were taken with a leaf area meter.  Stems were cut at a visible graft union if present or 0.6 in. cm above the potting media surface if this junction was not apparent.  Leaf tissue samples consisted of both leaflets and petioles. Potting media was removed from plant roots by rinsing in water.

All leaf, stem, and root samples were placed in a drying oven for 24 hours at 70ºC, after which their dry weights were measured, ground, and then sent to the North Carolina Department of Agriculture and Consumer Services Agronomic Division (NCDA&CS) for further sample preparation and nutrient analysis. Total N, P, K, Ca, Mg, S, B, Cu, Fe, Mn, Zn, and Na concentrations of the leaf tissue were determined. Leaf tissue samples were also processed for total N by oxygen combustion with an elemental analyzer on the NCSU campus. 

Statistical analysis was carried out using a general linear model for plant growth parameters measured once at the final sampling date (i.e. height and root weight) and nutrient content of the leaf tissue ((Proc GLM, SAS Institute, Cary, NC).   Multivariate analysis of variance (MANOVA) was utilized to protect against high type I errors (SAS Institute, Cary, NC).  Repeated plant growth parameter measurements taken over five weeks (i.e. leaf area, leaf weight, and stem weight) were transformed to their common logarithm to rectify heteroskedacity and then analyzed with a linear mixed model (Proc MIXED, SAS Institute, Cary, NC).  A non-linear mixed model was employed to fit an apparent logistic growth curve for fruit weight and fruit number and also to account for small number counts (0-3) (Proc NLMIXED, SAS Institute, Cary, NC).  Pearson correlation coefficients were used to measure associations between N leaf tissue content and random variables (Proc CORR, SAS Institute, Cary, NC). An alpha level of P<0.05 was used for all statistical tests.

RESULTS - Plant Growth.  
Both scion-rootstock graft combinations had greater shoot weight, leaf weight, and leaf area compared to non-grafts (Table 1.1).  Shoot weight, leaf weight, and leaf area was greater for scion grafted on ‘Maxifort’ rootstock compared to self-grafts indicating a rootstock selection effect.  Treatments with ‘German Johnson’ had greater growth compared to those with ‘Trust’. Grafted treatments (scion grafted onto ‘Maxifort’ rootstock and self-grafts combined) had higher mean values for plant height and root weight compared to non-grafts.  The plant height and root weight of scions grafted onto ‘Maxifort’ rootstock and the self-grafts were not significantly different from each other indicating that plant height and root weight were a result of the physical grafting process. Treatments with ‘German Johnson’ had greater growth compared to those with ‘Trust’. Growth advantages of grafted plants appeared later in the experimental period in most cases. Scions grafted onto ‘Maxifort’ rootstock had greater shoot weight on the last two out of five sampling events while leaf area was greater the last three out of five sampling events compared to the non-grafted treatments.  Our results indicate that the plant growth responses to grafting can be documented 7-8 weeks after grafting.

RESULTS - Leaf Tissue Nutrient Content.
We compared the nutrient content (leaf tissue nutrient concentration ((% or ppm) x the dry leaf weight (g)) of the total leaf tissue in order to account for the variability of shoot growth among grafting treatments. The leaf nutrient content of each plant was calculated to attain an estimate of the total amount of accumulated nutrients in the leaf tissue on the last sampling date. 

The mean leaf tissue nutrient content was greater in grafted treatments (scion grafted onto ‘Maxifort’ rootstock and self-grafts) compared to non-grafts for the following nutrients: N, P, K, Mg, and B (Table 1.2).  Scions grafted onto ‘Maxifort’ rootstock had greater mean leaf tissue content compared to non-grafts for N, P, K, Ca, Mg, S, Fe, Mn, Zn, Cu, and B.  Scions grafted onto ‘Maxifort’ rootstock had greater mean leaf tissue content compared to self-grafts for: Ca, Fe, Mn, Zn, and Cu.  Sodium content was the only nutrient that did not show a significant difference among treatments. The grafting effect resulted in increased accumulation of selected nutrients in the leaf tissue of grafted plants compared to non-grafts. When the ‘Maxifort’ rootstock was employed in a grafting pair, nutrient accumulations for all nutrients tested (*exception Na) were greater than in non-grafted plants.

Connections between plant growth, fruit yield and nutrient availability are well established among non-grafted plants. Establishing a link between N uptake and plant growth in grafted tomatoes was approached by conducting a Pearson correlation analysis.  The N content of the leaf tissue was positively correlated with shoot weight (87%), leaf area (81%), plant height (76%), and root weight (75%) reaffirming the close relationship between N uptake and plant growth (P<0.0001, data not shown).  These results suggest that N content in the leaf tissue resulted in greater shoot growth in grafted plants compared to non-grafted (Fig. 1.1).

Both ‘Trust’ and ‘German Johnson’ scion grafted onto ‘Maxifort’ rootstock had similar responses to grafting in terms of nutrient accumulation in the leaf tissue. The nutrient content values for N, P, K, Mg, S, Mn, and Zn were statistically similar between scion selections. Greater values of Ca, Fe, and Cu were present for the scion variety ‘Trust’ compared to ‘German Johnson’ when grafted on ‘Maxifort’ rootstock. Our results suggest that while the scion selection influences the uptake of selected nutrients, effects were not seen in most cases. 

Overall our study confirms that both increased plant productivity and greater leaf tissue nutrient content is achieved by grafting tomatoes. Plant response to the physical act of grafting appeared to be the most important factor for stimulating plant growth and nutrient uptake however, utilizing the ‘Maxifort’ rootstock expanded the range and level of grafting effects. Scion selection played a very minor role in treatment effects and scion-rootstock interactions were not present.  The positive relationship between the accumulation of nutrients in the leaf tissue and shoot growth among grafted plants suggests that grafted tomatoes can achieve greater plant productivity at the same nutrient input levels compared to non-grafted plants.

STUDY 2: CEFS: GROWING SYSTEM, NITROGEN INPUT LEVEL, AND GRAFTING EFFECTS ON NUTRIENT UPTAKE, PLANT GROWTH, AND FRUIT YIELD.

Materials & Methods

A systems comparison study was conducted at The Center for Environmental Farming Systems (CEFS)/Cherry Research Farm located in Goldsboro, NC for two consecutive years, 2007 and 2008. Plants were cultivated in two separate, 30ft x 96ft high tunnel units and two, adjacent field plots of the same size.  Grafting treatments consisted of two rootstock/scion combinations, a non-grafted control, and a self-grafted control (2008 only). Grafting treatments included: 1) ‘Cherokee Purple’ scions (Solanum lycopersicum L. ‘Cherokee Purple’) grafted on ‘Maxifort’ rootstock (Solanum lycopersicum L. x Solanum habrochaites S. Knapp & D.M. Spooner., ‘Maxifort’), 2) ‘Cherokee Purple’ scions grafted on ‘Beaufort’ rootstock (Solanum lycopersicum L. x Solanum habrochaites, ‘Beaufort’, 3) ‘Cherokee Purple’ tomato plants (non-grafted) and 4) a  ‘Cherokee Purple’ grafted back onto itself (self-grafted) was included in 2008 only due to problems with seedling production in 2007.  Each treatment listed above was subjected to three different levels of nitrogen (N) inputs: low, medium, and high. Therefore, nine treatments were included in the study in 2007 and twelve in 2008. 

The experiment was arranged as a 2 x 3 x 3 (2007) or 4 (2008) factorial with a replicated split-split plot design (Fig. 1.2 & 1.3). Main plots and subplots were randomized for each year separately. The whole plot factor was the growing system type, high tunnel or open field.  The sub-plot factor was the fertilizer treatment (different levels of N), and the sub-sub plot factor was the grafting treatment.  Each sub-sub plot served as an experimental unit which consisted of six plants per row in 2007 and five plants per row in 2008.  Each experimental unit was replicated four times, twice in each of the high tunnels and twice in each of the corresponding fields.  Guard rows were planted on the front and back end of each tunnel and field plot.   

Field and High Tunnel Description.
The project area was located in the CEFS field, ‘C2’.  The soil type at the project site was a Wickham, sandy loam (WhA), characterized as a deep, well-drained, and a slight to moderately acidic soil (Derrick, 1916). During the fall of 2006, this field was fumigated with methyl bromide in order to manage the federally listed noxious weed ‘Tropical Spiderwort’, Commelina benghalensis L.  Two high tunnels (Atlas Greenhouses Inc.: Alapaha, GA) were constructed on-site in February, 2007. The high tunnels are a snow-arch design with an inflated double polyethylene film (6 mil) roof and twin wall polycarbonate end walls. Doors built into the end walls were sized wide enough to allow the passage of a small tractor for spring tillage. Bows were spaced every 4ft and a 6ft Z-Lock drop down curtain system’ with a motorized crank was employed.  Adjacent field plots were also established at this time.

Seedling Production.
Guidelines set forth by the USDA NOP were followed during seedling production.  A dilution of a soluble organic fertilizer (Omega 6N-6P-6K, Petrik Inc., Woodland, CA) was administered to all treatments once per week. Seedlings were grafted using the ‘Japanese tube-grafting’ method described in the NC Cooperative Extension Bulletin # AG-675.  Approximately two weeks after grafting, all grafted and non-grafted seedlings were transplanted into larger propagation flats. 

Cultural Management.
Cultural management practices typical for each system were employed. The high tunnel system was approached as a hybrid between a greenhouse and an open field culture. Planting dates in the high tunnel system were 20 March 2007 and 18 March 2008, which were approximately one month earlier than the field system 19 April 2007 and 17 April 2008, and reflect the typical planting dates of local growers.  In 2007, inner floating polypropylene fabric row covers were used to cover seedlings inside the high tunnels when evening temperatures were predicted to drop below 40°F.  In 2008, the spring weather was milder and inner row covers were not employed based on the same criteria.

Plants in the high tunnel system were pinched to encourage the formation of two leaders, typical of greenhouse production using grafted transplants. These two leaders were trained to a trellis system consisting of vertical strings hanging from horizontal tension cables extending across the width of the high tunnels. The tomato vines were attached to the strings with plastic plant clips.  The lower leaves were pruned up to ‘one leaf below the first fruit cluster’ and suckers were removed on a weekly basis to steer growth towards the main leaders and fruit production. In 2007, suckering was conducted throughout the season. In 2008, in an effort to decrease the amount of sun-scald damage to fruit, suckering was stopped in mid-June. The high tunnel drop down curtain system was programmed to lower the sidewall curtains when ambient air temperatures inside the tunnels reached above 65.0°F. The controller is designed to open the curtains incrementally based on the change in temperature after five minutes of idle time.  At the end of each season, the high tunnels were sealed for a solarization period of approximately four weeks.

In the field system, the stake-and-weave trellis system was utilized. Leaf pruning was conducted up to the first horizontal string in the field. Plants were not pinched nor suckered, resulting in a bushier growth form compared to the high tunnel plants. Each system was irrigated on an as needed basis, evaluated twice daily. All planting rows in both systems were 13.5 ft. long and 4.5 ft. wide; plants were spaced every 22 in. Black polypropylene landscape fabric was utilized as a weed barrier in all plots. Drip tape with emitters every 12 in. and a flow rate of 19 l/hr. were used for irrigation.

Integrated Pest Management.
Integrated pest management practices were utilized for the management of insect pests.  Weekly scouting events were conducted and management decisions based on established thresholds for organic systems when available. In the early spring of 2007, the high-tunnel system experienced a high population of potato aphids, Macrosiphum euphorbia Two spot applications of insecticidal soap were carried out and followed by multiple releases of the aphid midge, Aphidoletes aphidimyza and the parasitic wasp, Aphidius ervi.

In both 2007 and 2008, applications of Bacillus thurigiensis (Bt) were applied to both the high tunnel and field system on two occasions. These sprays targeted tomato hornworms, Manduca quinquemaculata, tomato fruitworms, Helicoverpa zea, and/or armyworms Spodoptera sp.. In 2008, both the high tunnel and field system experienced high populations of stink bugs, Acrosternum hilare. However, no action was taken due as the negative effect of spraying a broad spectrum insecticide to reduce the population was judged to outweigh the benefits from the beneficial insect community present. Hand removal was the only control strategy utilized.  A series of bi-culture cover crop strips were planted in succession around the perimeter of each high tunnel and field plot in Oct. 2007 and June 2008 in order to attract and sustain beneficial insects and pollinators.

Fertilizer Applications and Cover Crops.
Three levels of total N inputs (pre-plant + post-plant applications) were evaluated each growing season. In 2007, the three N levels were 100 lbs.N/A (low), 150 lbs.N/A (medium), and 200 lbs.N/A (high).  The ‘Vegetable Crop Handbook for Southeastern United States (2007)’ recommends a N application rate of 200 lbs.N/A for soils with low potassium, such as ‘C2’; this recommendation matches the 2007, high N treatment.  All three fertilizer levels in 2007 provided more than adequate N as reflected in their sampled leaf concentrations. As a goal of the study was to investigate the critical level of N inputs required by a grafted tomato crop, the three N levels were lowered to 83lbs.N/A (low), 100lbs.N/A (medium), 150lbs.N/A (high) in 2008.

In 2007, a pre-plant application of 9 T/A of compost (‘Leprechaun NOP Compost’ - McGill Environmental Systems: Harrells, NC) and 91 lbs./A of feathermeal (Nutrimax ‘Super’ Natural Organic Fertilizer (12N-1P-0K), Nutrimax Inc.: Greensboro, NC) was applied to all plots, approximately two weeks before the respective system planting dates.  This pre-plant application of compost and feathermeal was estimated to supply 88 lbs./A1 of crop available N. The remaining balance of N was provided to the crop over a series of post-plant, soluble fertilizer solution applications (Phytamin 801 (6N-1P-K1), California Organic Fertilizers, Inc.: Fresno, CA).

Post-plant fertilizer applications were timed to provide N to the crop at critical times during  plant growth. The majority of post-plant N was supplied during the growth stages, ‘1st bloom to early harvest’ to encourage fruit production.  In 2007, post-plant fertilizer was applied over five fertigation events and was dispensed to each plant by hand.  In 2008, post-plant fertilizer schedule was modified to only two fertigation events and was dispensed to each plant via the drip irrigation system with the aid of a fertilizer injector set a ratio of 1:64 (Dosatron Liquid Dispenser, Model DI 16-11GPM: Clearwater, FL). Additions of K and calcium (Ca) were provided evenly across both systems based on NCDA leaf tissue analysis and soil test recommendations.

Between the 2007 and 2008 season, both systems were planted with a mixed winter cover crop.  The mix consisted of hairy vetch, Vicia villosa and winter rye, Secale cereale.  The cover crop was sampled just prior to mowing and soil incorporation. The growth of the cover crop in each system was very different over the course of 4-5 months. The total cover crop biomass was 43% greater in the high tunnel system (4,078 lbs./A) compared to growth in the field system (2,310 lbs./A). 

The cover crop in the high tunnel system was estimated to contribute 83 lbs.N/A of crop available N compared to the field cover crop with 47 lbs.N/A for the 2008 growing season.  As a result, the target amount of pre-plant N for the low fertilizer treatment was set at 83 lbs.N/A in 2008.  In order to bring the field system up to the same pre-plant N input level as the high tunnel system, feathermeal was added to the field but not to the high tunnel system.  Fifty-two pounds of feathermeal, with an estimated 37.0 lbs.N/A of crop available N, was supplied to the field system.  

Sampling Protocols.
Leaf tissue samples consisting of the ‘most recently mature leaves’ (MRML) were collected six or five times over the growing season in 2007 and 2008, respectively. One leaf was sampled from each plant and then combined with leaves from the same sub-plot to comprise a leaf tissue sample.  All leaf tissue samples were sent to the NCDA&CS, Agronomic Division for analysis of N, P, K, Ca, Mg, S, B, Cu, Fe, Mn, Zn, and Na. 

In addition to leaf tissue analysis, fruit harvests were conducted twice per week.  Fruit was picked from the ‘pink to red’ stages.  Fruit was sorted into marketable and non-marketable categories.  Qualitative judgements relating to marketable and non-marketable fruit were based on observations from regional direct sales outlets for organic produce.  Non-marketable fruit was sorted into categories including: cat-facing, blossom end rot, insect damage, fruit cracking, tomato spotted wilt virus (TSWV), sun-scald (2008 only) and ‘other’. Both the fruit number and fruit weight was recorded for each category.

Statistical Analysis.
Statistical analysis was carried out using a multivariate analysis of variance (MANOVA) to protect against high type I errors and a linear model for repeated measures with Proc Mixed (Proc MIXED) (SAS Institute, Cary, NC).  Linear hypotheses were used to test pair wise differences for fixed factors such as grafted versus non-grafted plants, and ‘Beaufort’ versus ‘Maxifort’ rootstock (‘Estimate Statements’, SAS Institute, Cary, NC). An alpha level of P≤0.05 was used for all statistical tests.

RESULTS AND DISCUSSION

System Effects.
In both 2007 and 2008, the high tunnel system produced greater total fruit yields (weight and number of fruit) and hit peak production three weeks earlier compared to the field system (Fig. 1.4 & 1.5). In both 2007 and 2008, the high tunnel system had a higher incidence of fruit with blossom end rot and cat-facing but lower incidence of TSWV and insect damage compared to the field system (Fig. 1.6 & 1.7). The mean leaf tissue concentrations of N, P, K, Ca, Mg, and Fe were higher in the high tunnel system in 2007. The same nutrients listed above were lower in the high tunnel system compared to the field system in 2008. These opposing trends were likely related to the differences of incorporated winter cover crop biomass preceding the spring tomato crop planting (see methods section for more details). The mean leaf tissue concentrations of Mn, Cu, B, and Na were lower in the high tunnel system compared to the field system across both years.

Grafting Effects.
In 2007 and 2008, both the Maxifort-CP grafts and the Beaufort-CP grafts had greater total and marketable fruit yields compared to non-grafted plants (Fig. 1.8 & 1.9). These yields were attained in a low disease pressure environment. Maxifort-CP grafts and the Beaufort-CP grafts also had higher rates of culled fruit, cracked fruit, and insect-damaged fruit compared to non-grafted plants for both seasons. The mean leaf tissue concentrations for grafted plants were greater for N, P, K, Mn, Cu, Zn, and B but lower for Mg and Na compared to non-grafted plants across both years. Self-grafted (2008 only) were similar to non-grafted plants for all leaf tissue nutrient concentrations results. ‘Cherokee Purple’ scion grafted on ‘Maxifort’ rootstock had greater mean leaf tissue nutrient concentrations for P, S, Cu and Zn compared to those grafted on ‘Beaufort’ rootstock across both years, the opposite was true for B and Mg suggesting that rootstock selection can influence nutrient accumulation in the leaf tissue.

Nitrogen Input Level Effects.
The mean leaf tissue N concentration never fell below standard optimum N ranges at any of the N input levels evaluated over the course of the two year study (83-200lbs.N/A). In 2008, both the high and medium N input levels (150lbs.N/A and 100 lbs.N/A, respectively) produced greater total harvest yields compared to the low N level (83lbs.N/A) (Fig. 1.10). A positive correlation between mean leaf tissue N concentrations and total fruit yield (>70.4%) was present in 2008 indicating a strong relationship between % N concentration and grafting onto hybrid rootstocks that in turn boosts fruit yield (Fig. 1.11 & 1.12).

In conclusion, maximum fruit yield were achieved with the combination of the high tunnel system and the ‘Cherokee Purple-Maxifort’ grafts indicating that when specific scion-rootstock combinations are paired with the more controlled environment fruit yields can be maximized (Fig. 1.13). Grafted plants attained higher yields compared to non-grafts even in a low disease pressure environment suggesting that higher seedling costs may be offset by greater yields even when soil borne disease pressure is absent. Grafting resulted in greater nutrient accumulation in the leaf tissue for the majority of essential nutrients. Although the system and fertilizer level effects on leaf tissue concentration were variable across the two study years, grafting effects were consistent, indicating that grafting on hybrid rootstock had the strongest influence on nutrient accumulation in the leaf tissue.

STUDY 3: ON-FARM CASE STUDIES.
Material & Methods.
On-farm trials were conducted at three, private North Carolina farms in 2007 and 2008. Each trial evaluated two rootstocks and one scion which was chosen based on the likely disease pressure (rootstock) and the grower’s market preferences (scion). Yield information was collected by grower cooperators over the course of the season and the crop was managed according to the grower’s standards. Tissue samples were collected over the course of the growing season and submitted to the North Carolina Department of Agriculture and Consumer Services Agronomic Division (NCDA&CS) for nutrient analysis at periodic crop growth stages.

A number of response variables were measured such as yield including total fruit weight, total number of fruit, and nutrient concentration of the most recently mature leaf tissue (MRML) over time with a focus on nitrogen (N).  In collaboration with Frank Louws and Cary Rivard in the Department of Plant Pathology, disease incidence was assessed over the course of the growing seasons.

Statistical Analysis.
For non-repeated measures such as total cumulative yield a linear model was used to compare grafting effects (PROC GLM, SAS, Cary, NC). Repeated measures models were used to evaluate response to yield overtime (PROC MIXED, SAS, Cary, NC).  All variables had a highly nonlinear trend over time, forcing time to be treated as a categorical variable with a compound symmetric covariance model. An alpha level of P≤0.05 was used for all statistical tests.

On-farm Case Study 1.
The first trial was located in Orange County in a high tunnel system with suspected root-knot nematode (Meloidogyne spp.) pressure. The study was randomized block design with repeated measures. Experimental units were groups of six tomato plants, replicated four times. Four treatments were evaluated with one of each group randomly assigned to four adjacent blocks. Seedlings were grafted using the Japanese tube-grafting method at NCSU Southeastern Plant Environment Laboratory. Rootstock was selected to manage root-knot nematodes and the scion, German Johnson, was selected by the grower according to market preferences. Treatments included: 1) non-grafted German Johnson (GJ), 2) self-grafted GJ, 3) Big Power (Rijk Zwaan) - GJ, and 4) Beaufort (DeRuiter) - GJ.

Seedlings were transplanted into twin rows at 21” spacing on raised beds covered in plastic mulch. Drip irrigation was used. Plants were pruned to a single leader and trained to a vertical string trellis. Tomato vines were attached to the strings with plastic plant clips. Lower leaves were pruned up to ‘one leaf below the first fruit cluster’ and suckers were removed on a weekly basis to steer growth towards the main leader and fruit production. In 2007, pre-plant inputs included incorporated: rape cover crop residue, crabshell meal, feathermeal, and sulfate of potash. In 2008, the experiment was conducted in a similar but adjacent high tunnel due to crop rotation practices. Pre-plant inputs in 2008 included: soybean meal, sulfate of potash, and azamite. Disease pressure was very low across both years and no root-knot galling was seen either year.

In 2007, there was no grafting effect on the total cumulative fruit weight (non-grafted (10.9 lbs./plant), self-grafted (10.9 lbs/plant), Big Power – GJ grafts (11.8 lbs./plant), and Beaufort – GJ grafts (10.2 lbs./plant)) or number of fruit. There was no grafting effect on the N concentration in the MRML tissue over time among treatments.  In 2008, there was no grafting effect on the total cumulative fruit weight (non-grafted (14.6 lbs./plant), self-grafted (12.6 lbs/plant), Big Power – GJ grafts (16.0 lbs./plant), and Beaufort – GJ grafts (17.1 lbs./plant)) or the number of fruit. However, the Beaufort-GJ grafting pairs had a greater N concentration in the MRML tissue over time compared to the non-grafted plants.

On-farm Case Study 2.
The second trial was located in Alamance County in a ‘Hay-grove’ high tunnel system with suspected fusarium wilt (Fusarium oxysporum f. sp. lycopersici) pressure. The study was randomized block design with repeated measures. Experimental units were groups of seven tomato plants, replicated four times. Four treatments were evaluated with one of each group randomly assigned to four adjacent blocks. Seedlings were grafted using the Japanese tube-grafting method at NCSU Southeastern Plant Environment Laboratory. Rootstock was selected to manage root knot nematodes and the scion, Cherokee Purple, was selected by the grower according to market preferences. Treatments included: 1) non-grafted Cherokee Purple (CP), 2) self-grafted CP grafts, 3) Maxifort (DeRuiter) – CP grafts, and 4) Beaufort (DeRuiter) – CP grafts. Although fusarium wilt was seen in other areas of the tomato production areas, it was not evident in the research plot. However, incidence of southern blight (caused by Sclerotium rolfsii) was moderate, with 20-45% of the controls with disease at the end of the season, across both years.

Seedlings were transplanted into rows 18” in-row spacing on raised beds covered in plastic mulch. Drip irrigation was used. Lower leaves were pruned up to ‘one leaf below the first fruit cluster’ and the plants were attached to a vertical, wire-mesh trellis system. In 2007, pre-plant inputs included incorporated: oats and crimson clover cover crop residue, rock phosphate, and feathermeal. In 2008, pre-plant inputs included incorporated:  lime, rock phosphate, winter rye cover crop residue, and feathermeal.

In 2007, the total cumulative fruit weight was greater for the Maxifort-CP grafts (20.7 lbs./plant) and Beaufort-CP grafts (19.8 lbs./plant) compared to the self-grafted (12.1 lbs./plant) and non-grafted (14.6 lbs./plant) controls. The cumulative number of fruit was greater for the Maxifort-CP (29 fruit/plant) and Beaufort-CP grafts (29 fruit/plant) compared to the self-grafted plants (20 fruit/plant). The Maxifort-CP grafts had higher N concentrations in the MRML tissue compared to the self-grafted and non-grafted plants.

In 2008, the total cumulative fruit weight was greater for the Maxifort-CP grafts (14.8 lbs./plant) compared to the self-grafted (8.05 lbs./plant) and non-grafted (9.3 lbs./plant) controls. The Beaufort-CP grafts (12.3 lbs./plant) had greater total cumulative fruit weight compared to the self-grafted plants. The cumulative number of fruit was greater for the Maxifort-CP grafts (24 fruit/plant) compared to the self-grafted (16 fruit/plant) and non-grafted (18 fruit/plant). Both the Maxifort-CP and Beaufort-CP grafts had greater N concentration in the MRML tissue compared to the self-grafted and non-grafted plants.

On-farm Case Study 3.
The third trial was located in Sampson County in an open field system with suspected bacterial wilt (Ralstonia solanacearum) pressure. The study was randomized block design with repeated measures. Experimental units were groups of seven tomato plants, replicated four times. Four treatments were evaluated in 2007 and six treatments in 2008 with one of each group randomly assigned to four adjacent blocks. Seedlings were grafted using the Japanese tube-grafting method at NCSU Southeastern Plant Environment Laboratory. Rootstock was selected to manage bacterial wilt and the scion, Celebrity, was selected by the grower according to market preferences and early fruiting. Treatments included: 1) non-grafted Celebrity, 2) self-grafted Celebrity grafts, 3) #04-105-Celebrity (D. Palmer) grafts, 4) Dai Honmei-Celebrity (Asahi) grafts, 5) BWR-NCS2-Celebrity (DeRuiter) grafts (2008 only), and 6) Sweet Olive-Celebrity grafts. Disease pressure was severe with 80-100% of the controls with disease by the end of the season. In 2007, the soil borne disease present was primarily bacterial wilt, while in 2008 southern stem blight (Sclerotium rolfsii) dominated. In 2007, data was normalized for tomato spotted wilt virus (TSWV) incidence to reflect injury to individual plant. 

Seedlings were transplanted into 21” in-row spacing on raised beds covered in plastic mulch. Drip irrigation was used. Lower leaves were pruned up to ‘one leaf below the first fruit and trellised with the stake and weave system. In 2007, pre-plant inputs included: incorporated winter rye and hairy vetch cover crop residue, feathermeal, sulfate of potash, and lime. In 2008, pre-plant inputs included: incorporated rape and crimson clover cover crop residue, feathermeal, sulfate of potash, NatureSafe pellets, and boron.

In 2007, the #04-105-Celebrity grafts (18 lbs./plant) had a greater total fruit weight compared to the self-grafted (11 lbs./plant) and non-grafted controls (10 lbs./plant). The Dai Honmei-Celebrity grafts (14 lbs./plant) had intermediate yields that were not significantly different than any other treatment.  There was no grafting effect on the N concentration in the MRML tissue over time among treatments. The Dai Honmei-Celebrity grafts had an intermediate yield (14 lbs./plant) that was not significantly different than any other treatment.

In 2008, the #04-105-Celebrity grafts (11 lbs./plant) had greater total cumulative fruit weight compared to non-grafted (3 lbs./plant), self-grafted (3 lbs./plant), and Sweet Olive-Celebrity (5lbs./plant) treatments. Similarly, the #04-105-Celebrity grafts had a greater total number of fruit compared to non-grafted (20 fruit/plant), self-grafted (6 fruit/plant), and Sweet Olive-Celebrity (12 fruit/plant) treatments. The Dai Honmei-Celebrity grafts (8 lbs./plant), and BWR-NCS2-Celebrity grafts (8 lbs./plant) had intermediate yields that were not significantly different than any other treatment. Differences were not present among treatments for N concentration in the MRML tissue over time.

Publications/Outreach.

M.S. Thesis: Grafted Tomato Performance in Organic Production Systems: Nutrient Uptake, Plant Growth and Yield by Suzanne O’Connell (12/08) <http://www.lib.ncsu.edu/theses/available/etd-11072008-152636/>.

Presentations/Workshops:
Center for Environmental Farming Systems (CEFS) 2009 Seasons of Sustainability Agriculture Workshop Series on High Tunnels. ‘Organic Nutrient Management’. Presenter:  S. O'Connell. Goldsboro, NC. February, 2009. ~ 50 Attendees. Presentation online at: <http://www.cefs.ncsu.edu/calendar2009.htm#hightunnel>


National Association of County Agriculture Agents Conference, 'SARE-funded Research at CEFS: Grafting Heirloom Tomatoes for Disease Resistance in Intensive Farming Systems'.  Presenters: S. O'Connell, C. Rivard, S. Hartmann (participating farmer). Greensboro, NC. July 15, 2008. Pre-registration ~75.


Growing Small Farms: Enhancing Sustainability Workshops 'Heirloom Tomato Grafting Workshop'. Presenters: S. O'Connell, M. M. Peet, C. Rivard, A. Hitt (participating farmer). Pittsboro, NC. Nov. 12, 2008. ~ 50 attendees. Presentation online at: <http://www.ces.ncsu.edu/chatham/ag/SustAg/2008tomatografting.html>


Seasons of Sustainable Agriculture Series 'High Tunnel Workshop and Tour'. O'Berry Center, Goldsboro, NC. 'Nutrient Management'. Presenter: S. O’Connell. Feb. 17, 2008. ~60 attendees.   

Posters:
‘The Grafted Heirloom Tomato System for Organic Production in High Tunnels: Are There Advantages in the Absence of Diseases’? O’Connell, S., Peet, M., Rivard, C., Louws, F., and Harlow, C. American Society for Horticultural Science Annual Conference. St. Louis, MI. July 24-28, 2009.


‘Physiological Disorders in Grafted Heirloom Tomatoes Grown in High Tunnels Using Organic Production Practices’. Peet, M., O’Connell, S., Rivard, C., Louws, F., and Harlow, C. American Society for Horticultural Science Annual Conference. St. Louis, MI. July 24-28, 2009.


‘The Grafted Tomato System: Are There Advantages in the Presence of Soil Borne Diseases’? Rivard, C., Louws, F., O’Connell, S., Harlow, C., and Peet, M. American Society for Horticultural Science Annual Conference. St. Louis, MI. July 24-28, 2009.


‘The Performance of Grafted Heirloom Tomatoes in Organic Production Systems: High Tunnels and the Open Field’. O’Connell, S., Peet, M., Harlow, C., Louws, F., and C. Rivard. American Society for Horticultural Science Annual Conference. Orlando, FL July 21-24 2008. **Winner of National Graduate Student Poster Competition.


‘Nutrient Uptake Efficiency and Plant Growth Indicators of Grafted Tomatoes (Suzanne O'Connell*) O'Connell, S. and M. M. Peet 68th Annual Meeting Southern Region American Society for Horticultural Science. Dallas, TX. Feb. 2-4 2008. **Winner of Southern Region Graduate Student Poster Competition. 


‘Grafting Rootstocks onto Heirloom and Locally Adapted Tomato Selections to Confer Resistance to Soil Borne Diseases and Increase Nutrient Uptake for Market Gardeners’ New American Farm Conference. Peet, M., S. O’Connell, F. Louws, C.Rivard, and C. Harlow. Advancing the Frontier of Sustainable Agriculture. Kansas City, Missouri. March 25-27, 2008. 


‘Improving Performance of Organic Heirloom Tomatoes Using High Tunnels and Grafting’. ISHS Symposium on Tomato in the Tropics. Peet, M., O’Connell, S., Rivard, C., Louws, F., and C. Harlow. 2008. Villa de Leyva, Colombia. September 9-12, 2008.


Use of High Tunnels and Grafting for Organic Production of Heirloom Tomatoes in North Carolina. ISHS International Workshop on greenhouse Environmental Control and Crop Production in Semi-Arid Regions. Peet, M., Rivard, C., O’Connell, S., Louws, F., and C. Harlow. Tucson, Arizona. October 20-24, 2008. 


Abstracts:

High Tunnels and Grafting For Disease Management in Organic Tomato Production. 2008. Rivard CL, Louws FJ, Peet MM, and O’Connell, S. Phytopathology 98:S133-S133 American Phytopathological Society Centennial Meeting. Minneapolis, Minnesota July 26-30, 2008.

Nutrient Uptake Efficiency and Plant Growth Indicators of Grafted Tomatoes. 2008. O'Connell, S. and Peet M.M. HortScience 43:3. American Society of Horticultural Science Southern Region 68TH Annual Meeting. Dallas, Texas. February 2-4, 2008. 

Informational Websites:
<http://www.ces.ncsu.edu/depts/hort/greenhouse_veg/>
<http://www4.ncsu.edu/~soconne/>

 

On-Farm Collaborators:

 

Research Team Members: