Yield and fruit quality of grafted tomatoes, and their potential for soil fumigant use reduction. A meta-analysis

Grafted fresh market tomatoes are widely used in commercial production throughout Europe and Asia, and interest among commercial producers in the Americas has increased in recent years. Many field trials have found dramatic net economic return increases relative to non-grafted scion cultivars. However, optimal yields require growing conditions that satisfy the agronomic needs of both rootstock and scion cultivars. Most commercial rootstocks are resistant to multiple soilborne pathogens, allowing grafted plants to maintain high yields in pathogen-infested fields without the use of soil pesticides, including fumigants. Here we comprehensively and quantitatively review, for the first time, all available published trial data on fruit quality and yield of grafted tomatoes. Collectively, 159 publications included 202 different rootstocks, 126 geographic locations, and 1023 experimental treatments. Yield performance varies with the specific rootstock/scion combinations and with the conditions of a given production system. Among 949 heterograft treatments (rootstock/scion of different cultivars), grafted plant yields were not significantly higher in 65% of the cases, yet they averaged a 37% yield increase for all data. In addition, grafted/non-grafted yield ratios in 105 experimental treatments with rootstock ‘Maxifort’ varied dramatically by scion. However, European trials used completely different scions than US trials, so the roles of scion and geographical differences remain unclear. Concerns that grafting might contribute to inferior fruit quality (pH, titratable acidity, total soluble solids, lycopene, vitamin C, firmness, “taste”) seem unfounded in general, though isolated cases show dramatic differences. Grafted tomatoes show promise to reduce the usage of various soilborne pathogen treatments, with 33% of commercial tomato rootstocks either resistant or highly resistant to seven or more common soilborne pathogens. Our approach integrated trial data from around the world, though limitations in available data complicated our analysis of relationships between some experimental variables and fruit yields and quality.