Tomato Grafting: Developing Grower Recommendations for the Great Plains and Enhancing Our Understanding of the “Rhizobiome”

1) Identify rootstocks that improve productivity and reduce disease losses under Great Plains conditions through a series of on-farm and university research station trials: Multiple rootstocks (>5) will be trialed at several on-farm and research station trials to determine the viability of grafting in the Great Plains. These trials will be arranged in a randomized complete block design (RCBD) and will include four replications, with multiple plants per treatment per rep. The trials will be conducted utilizing heirloom (‘Cherokee Purple’) and hybrid (‘BHN 589’) scions. These cultivars are highly utilized in Kansas and throughout the United States, particularly for high tunnel production. A trial utilizing ‘Cherokee Purple’ scion will be conducted at the KSU Student Organic Farm in Manhattan, KS (open-field). Five rootstock treatments will be compared with nongrafted controls to determine effects on crop performance (yield and plant growth). Each plot will contain six to eight plants (determined by location) and will be blocked to reduce bias conditions in the field study. A similar trial will be performed at the John C. Pair Horticultural Research and Extension Center near Wichita, KS, and this trial will utilize ‘BHN 589’ for nongrafted controls and for scions of grafted plants. Additionally, trials at the Olathe Research and Extension Center and at the two on-farm locations (Clark Farm (high tunnel) and Gieringer’s Orchard (high tunnel)) will test three additional rootstocks in addition to the ones being trialed at the other two on-farm locations (both in high tunnels-Gibbs Rd. Farm and Common Harvest, which are supported by CERES project). All trials will be monitored for incidence of any major soilborne diseases through visual assessments and microscopy and culturing methods, as needed. Fruit yield and plant growth data (plant height and shoot biomass) will be collected. All data will be analyzed using the MIXED procedure (SAS 9.1; Cary, NC).

2) Determine optimum grafting/healing conditions that reduce risk of crop failure and increases the successful implementation of grafting for small-scale growers; Various greenhouse experiments will be conducted to determine the effects of healing chamber temperature and humidity on graft survival. Additionally, novel methods for healing chamber design will be developed that cater towards production for small-acreage specialty crop growers. These include the development of chambers that reduce the risk of high temperatures by using alternative materials and various ventilation systems. These experiments will be performed at KSU-Olathe in Olathe, KS as well as the greenhouse complex at Kansas State University in Manhattan, KS. The goal of these studies will be to design a healing chamber that provides a suitable environment for graft union formation while reducing the risk of crop failure due to excessive heat. Preliminary data has shown that high humidity and/or stress in the healing chamber during days 5-8 can lead to significant growth of adventitious roots by the scion. We will determine the role of humidity at promoting adventitious roots as well as transplant fertility. Plants that are fertilized at the end of the healing process may be more prone to growing adventitious roots as the scion tissue is capable of absorbing nutrients through foliar feeding. A detailed evaluation of foliar fertilization and its role in promoting root growth will be performed. Multiple identical batches of grafted plants will be produced over time and analyzed in a RCBD whereby each of the individual batches represent blocked replicates. Four to six replicates will be utilized based on the needs of each study and will be analyzed using the MIXED procedure (SAS 9.1; Cary, NC).

3) Evaluate the effects of rootstocks and grower/location on soil and rhizosphere microbial communities;

Whole plants will be sampled from each plot in three high tunnel locations (KSU-Olathe, On-farm 1, On-farm 2) in order to determine the effect of rootstock genotype rootstocks on the ecology of the rhizobiome. A self-grafted control will be utilized to determine any effects of the grafting process on the microbial ecology of the plant. There will be a minimum of four blocks in a randomized complete block design, where a plot (experimental unit) consists of at least five plants with 18 inch spacing in row. At KSU-Olathe, there will be six replications (1 rep per high tunnel) and plots will include 8 plants each. The scion will be ‘BHN 589’ (a regionally-popular high tunnel cultivar), and this scion will be utilized at the on-farm locations in order to facilitate cross-location comparisons. Plants will be sampled twice at KSU-Olathe (peak vegetative growth and peak harvest) and once at the on-farm locations (peak harvest).

Sampling for microbial community assessment. At sampling, plant roots from each plot will be collected, sectioned into 1cm lengths and 20 pieces selected for analyses. After sonication in 0.1% Triton-X and diH₂O, the sonication liquid will be collected in a 30ml syringe and passed through 0.22µl nylon membrane to collect fungi, bacteria and suspended particles. The remaining tissues will be dried at 50°C and weight recorded. A subsample of plants will be evaluated for total root biomass. The DNA extracted from tissues will be analyzed by quantitative PCR (qPCR) and relative ITS (fungi) and 16S (bacteria) copy numbers compared as described in Fierer et al (2005).

Separation of the superficial and endophytic rhizobiomes. In molecular analyses, it is uncertain whether epi- or endophytes are considered. For bacteria, the tissues are often sonicated to avoid chloroplast contamination and to maximize epiphyllous bacteria (Kadivar & Stapleton 2006). It is our experience that our choice of next-next generation sequencing tools will permit analyses of both superficial microbial communities as well as those embedded into the root matrix. As a result, we will aim to separate epi- and endophytic communities. The root tissues will be transported on ice to the laboratory, where twenty 1cm pieces will be sampled from each of the tomato plants and pooled in sterile extraction buffer (MoBio UltraClean Soil DNA). To separate superficial and endophytic communities, the tissues will be sonicated for 1min to dislodge superficial bacteria and fungi, transferred from supernatant into clean extraction buffer and the supernatant transferred into a second extraction system. The genomic DNA will be extracted from the roots and particles within the supernatant as per manufacturer’s instructions after homogenization in a FastPrep instrument (MP Biomedicals, CA). Each extract will be PCR-amplified and sequenced to determine the epi- and endophytic communities. Some overlap between the fractions will occur, but true endophytes will be more abundant in the homogenized tissues than in the sonicated supernatant.

Rhizobiome community analysis by high throughput sequencing. Our research group has adopted multiple high throughput sequencing platforms, including the Ion Torrent in the fall of 2012. We will adopt these tools to provide superior sequencing depth at reasonable cost. To extract microbial DNA, the tissues will be homogenized in FastPrep instrument (MP Biomedicals, CA), DNA extracted following the MoBio UltraClean (MoBio Laboratories, CA) instructions and eluted in 100µl of the final buffer S5. The fungal ampicons will be produced with ITS1F and ITS4 fusion primers (Jumpponen and Jones 2009): preliminary data indicate that they reliably generate amplicons adequate to provide richness/diversity estimators for fungal communities. The bacterial amplicons will be produced similarly with 9f and 338R fusion primers to estimate richness/diversity and community composition of bacterial communities.

From the derived OTU (Operational Taxonomic Unit) frequency data, we will calculate common estimators of diversity (including richness, Shannon’s diversity, Simpson’s dominance, and evenness). These data will be used to test hypotheses about diversity responses among the treatments. Hypotheses of interest include (a) Lower disease rootstocks will be associated with higher microbial diversity, (b) Lower disease rootstocks will be associated with higher frequency of beneficial microbes, and (c) Lower disease rootstocks will be associated with higher biomass of beneficial microbes. Note also that we will be able to evaluate responses on various taxonomic levels (species, genus, family) over time to optimize the value of our conclusions and to provide a taxonomic framework for the observed molecular OTU responses. We will evaluate the effects of rootstocks on bacterial and fungal community diversity and on the frequency of particular OTUs using generalized linear models in R. We will focus on OTUs classed as beneficial and pathogenic groups.

4) Extension and Outreach Program: Currently, a major barrier for growers attempting to utilize grafted plants is the ability to purchase or produce their own plants. Therefore, a primary objective of this project is to provide grower training on grafted propagation in addition to the disseminating the results of this specific project. The details of the extension plan are described in the “outreach plan” (see below).