In the United States and the world in general, the amount of arable land per capita has been decreasing for many decades (World Bank, 2015). Increased restrictions on the use of chemical fumigants, such as the phase-out of methyl-bromide, to manage soilborne pathogens has compounded this problem (Miguel, 2002; Louws, 2010). Growers are struggling to produce the same yields on lower quality and smaller areas of land (Qiao et al., 2015). Grafting presents a unique solution to this problem by allowing growers to select preferred vegetable varieties/scions for grafting onto vigorous and/or disease resistant rootstocks (Harrison and Burgess, 1962; Louws et al., 2010; Rivard et al., 2010a; Grieneisen et al., 2018). The hardiness, disease resistance, or vigor of rootstocks can improve yield of grafted scions, specifically fresh market tomato (Solanum lycopersicum Mill.) cultivars (Lee, 1994; Oztekin et al., 2009; Grieneisen et al., 2018)⁠. The use of grafted plants is common in greenhouse vegetable production but has not been as widely adopted in field production systems in the USA (Kubota et al., 2008).

Although some tomato rootstocks can provide high levels of resistance or tolerance to a wide variety of plant pathogens (Harrison and Burgess, 1962; Grimault and Prior 1994; Ioannou, 2001; Rivero et al., 2003; Louws et al., 2010)⁠, little is known on the yield response of scions with vigorous rootstocks in the presence of low soilborne pathogen pressure (Suchoff et al., 2019). In a recent meta-analysis, grafted plants yielded similarly or less than non-grafted or self-grafted plants 65% of the time (Grieneisen et al., 2018). Additionally, specific conditions affect graft response in yield, such as scion choice with the rootstock ‘Maxifort’ (Loewen et al., 2019). These inconsistent results indicate a need to specify and optimize practices where grafting can increase yields beyond pathogen pressure. As grafted plants cost appreciably more per plant than non-grafted, grafting nursery businesses and growers must realize a positive return on investment to consider adopting the practice (Rysin and Louws, 2015)⁠.

Pinching, the removal of meristems to allow two axillary stems to grow, has been shown to increase yields in grafted and non-grafted tomatoes in hydroponic systems (Rahmatian et al., 2014). Hypothetically, two main stems with one large vigorous root system should allow growers to plant at a decreased density. However, in the field, pinching has not been sufficiently evaluated to understand if this practice affects yield (Suchoff et al., 2014). Currently, pinching is employed by growers without a thorough understanding of the exact parameters under which this practice should be used.

In standard tomato field production in North Carolina and surrounding states, non-grafted tomato plants are spaced at approximately 46 cm between plants. One study in the region suggested grafted plants can be spaced farther apart due to the increased vigor of the rootstocks and associated increase of fruit production (Suchoff et al., 2014). Grafted tomato plants are recommended to be spaced at 56-61 cm (Tri-Hishtil, personal communication), but limited data regarding yield at these spacings is sparse (Kubota et al., 2008). Increased yield can be realized by harvesting more fruit and/or larger fruit. Spacing plants farther apart is a feature of grafted plants which allows for a greater return on investment (Miguel, 2002; Rivard et al., 2010b; Barrett et al., 2012; Suchoff et al., 2014)⁠.

Although increasing yield and fruit sizes is important, fruit quality (color, sweetness, acidity) must also be maintained or improved as compared to non-grafted tomato plants. The effect of grafting on fruit quality in solanaceous and cucurbitaceous plants has been mixed. Reports for specific rootstock-scion combinations have shown slightly better, slightly worse, or no changes in fruit composition compared to non-grafted or self-grafted tomatoes, (Pogonyi et al., 2005; Riga et al., 2016; Kyriacou et al., 2017)⁠. Although overall differences reported are small enough that marketability would not be affected, the variability of effects of grafting warrants evaluation of quality and marketability.

The impact of pinching, spacing, and grafting has been studied either independently or in greenhouse systems. Information on combinations of these cultural practices on fresh market tomato under field production is lacking. The objectives of this study were to evaluate yield effects impacted by (1) grafted Tasti-lee (TL) and Mountain Fresh Plus (MFP) plants onto rootstocks ‘Beaufort’, ‘Arnold’, ‘Shield’, and ‘Fortamino’ and non-grafted; (2) pinched and non-pinched plants; (3) spacing effects; (4) to evaluate fruit quality including lycopene concentration, pH, titratable acidity, and soluble solids content (SSC) of fruit from grafted and non-grafted plants; (5) and to conduct a partial budget analysis to determine the effect of these practices on profits.

Grafting tomatoes onto the rootstocks ‘Arnold’ and ‘Beaufort’ consistently increased marketable yield of fresh market tomatoes produced in the field in this study. This fits into a larger body of evidence that shows certain rootstocks can provide yield and fruit sizing benefits to growers (Rivard et al., 2010b; Grieneisen et al., 2018) even in the absence of known negative abiotic and biotic constraints. The most significant gains in fruit size increase were in TL at the XLPlus size when grafted onto ‘Arnold’ and ‘Beaufort’ at 74% and 94% respectively for TL, and 30% and 25% for MFP. These data give a strong indication that one mechanism of grafting is to increase fruit size for cultivars like TL. The most likely explanation is that TL, a lower yielding but high fruit quality cultivar, naturally has less vigor than MFP and therefore responded proportionately more to the beneficial impact of the rootstock treatments. Other studies have shown that MFP does not appear to benefit from grafting, specifically on the rootstock ‘Maxifort’ (Grieneisen et al., 2018; Lang and Nair, 2019). The grafted tomatoes in our study were planted at a greater spacing than non-grafted plants (56 and 61 cm vs 46 cm) yet still produced significantly higher yields with the two rootstocks. In contrast, the rootstock ‘Shield’ did not confer an advantage in the absence of bacterial wilt pressure ( Table 2 ).

Although the overall effect of spacing had no significant effect on yield, the highest yielding treatment were unpruned TL grafted onto ‘Beaufort’, non-pinched, planted at 61 cm spacing (128121 and 120358 total kg/ha for 2017 and 2018 respectively) ( Supplemental Table S3 ). The identical treatment at 56 cm spacing had lower, but still significantly high total yields in 2017 and 2018 at 114495 and 102324 kg/ha ( Supplemental Table S3 ). This indicates that there may be specific circumstances where a small increase in spacing could result in significant increases in yield, and more detailed research is needed to elucidate this. Research tailored for specific cultivars may define winning combinations of techniques. Also, when using 56 and 61 cm spacing, 18-25% fewer plants per hectare are needed for yields equivalent to using non-grafted plants at 46 cm spacing. Because there were no consistent significant differences in overall yield or fruit sizes in 56 compared to 61 cm spaced grafted plants, it is recommended that growers plant grafted plants at 61 cm. This agrees with Suchoff et al., 2014 that show spacing grafted plants (non-pinched) tends to be optimized around 61 cm.

Pinching grafted tomatoes did not provide a consistent benefit at 56 and 61 cm across years. It is likely that the spacing differences were too small to show a significant benefit to pinching. It should also be mentioned that both pinching and not pruning plants increased yields ~10%, but the effect is not additive (Ingram et al., 2019). Increasing the foliage with either technique may allow the plant to take advantage of the more vigorous root system of the tomato rootstock (Davis and Estes, 1993; Kanyomeka and Shivute, 2005; Gaytán-Mascorro et al., 2008; Maboko et al., 2011)⁠. To see a more consistent effect of pinching or not pruning, future studies should include wider spacings (>61 cm) to accommodate the additional foliage created by grafted pinched or non-pruned plants. Increasing spacing for pinched plants may be the key to offsetting the costs related to pinching. Also, not pruning pinched varieties may be an important component in maximizing foliage while minimizing planting costs. Pinching is an added cost for growers and the optimal parameters under which this practice is used needs to be studied further before widescale adoption by growers in the field. More details on the pruning and pinching interactions are reported in a separate publication (Ingram et al., 2019).

While both ‘Beaufort’ and ‘Arnold’ provided clear return on investments (ROI) in both scions, ‘Beaufort’ consistently provided the highest ROI ( Figures 7 , 8 ). ‘Shield’ provided the lowest ROI, and in some cases had less yield than the non-grafted controls. While pinching did not provide a clear benefit in ‘Beaufort’ or ‘Arnold’, pinched ‘Shield’ plants did provide a clear ROI. The cost-benefit patterns were similar between TL and MFP. Overall, these results demonstrate that there is a clear economic benefit to grafting in field situations where there is low or no disease present. Shield is typically used to combat bacterial wilt diseases, but bacterial wilt can often have an uneven distribution in fields. One method growers might be able to justify the cost of using bacterial wilt resistant cultivars that lack vigor, would be to pinch them, as this can have a clear impact on yield that can provide a higher ROI. While the cost of pruning is negligible compared to the cost of grafted plants, there was no clear benefit to pruning determinate cultivar tomatoes in this field according to Ingram et al., 2019, which is included in our analysis ( Figures 7 , 8 ).

Despite some significant differences, overall changes in fruit quality from grafted and non-grafted plants may not result in perceptible differences in taste. Turhan et al. (2011) found that SSC was consistently higher in fruit from non-grafted plants compared to those from ‘Arnold’ and ‘Beaufort’, but was 0.4%, less than the 2% SSC generally thought to be needed for consumers to detect differences in sweetness. In contrast, other research studies have found that grafting has no significant positive or negative effect on overall tomato fruit quality (Pogonyi et al., 2005; Davis et al., 2008; Gaytán-Mascorro et al., 2008)⁠⁠.

Rootstocks ‘Arnold’, Beaufort’, and ‘Fortamino’ all significantly improved yields and, sometimes, fruit size in ‘Tasti-lee’ and ‘Mountain Fresh Plus’ grafted plants compared to the non-grafted plants. Of the grafted plants, scions grafted onto ‘Shield’ produced the lowest yields, on average, and were not significantly different than the non-grafted plants. This trial was conducted in the absence of known biotic or abiotic stress factors. The 61-cm spaced plants had similar yields to 56-cm spaced plants, indicating that growers can plant fewer grafted plants and achieve similar yields and fruit sizes. Across all combinations, pinching grafted tomato plants increased yield in only one year, so more research ought to be conducted before recommending this practice. Pinching only provided a specific benefit to plants grafted with ‘Shield’, otherwise the effect was marginal. More research needs to be done on wider plant spacings to determine if spacing greater than 61-cm would see benefits from pinching. In the cultivar ‘Tasti-Lee’, fruit quality characteristics such as lycopene, titratable acidity%, SSC%, and pH differed in grafted and non-grafted plants. Grafting with ‘Arnold’ and ‘Beaufort’ consistently improved yields in both years and show a clear economic benefit. There was no specific economic benefit to pinching except with the rootstock ‘Shield’. This study provides a clear outline of best practices for grafting in field conditions under low-disease conditions.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

TI, FL, and IM were responsible for the experimental design of the field study. TI, SS, and IM executed the field experiment and data collection. MT and PP-V were responsible for the fruit quality study. TI conducted the data analysis and wrote the manuscript. All authors contributed to the article and approved the submitted version.