A database search was conducted for available research articles on tomato grafting from Scopus (https://www.scopus.com/), which has broader coverage (Pranckute, 2021). The query: TITLE-ABS-KEY (“GRAFT*” OR “SCION*” OR “ROOTSTOCK*” OR “STOCK*”) AND TITLE-ABS-KEY (“TOMATO*” OR “LYCOPERSICUM ESCULENTUM” OR “Solanum lycopersicum” OR S. lycopersicum” OR “L. ESCULENTUM) was used for the search. An asterisk was used to replace the plurals and terms derived from the word. The database search using the above query was performed on November 11th, 2021, which resulted in 1285 publications from 1944 to 2020. The metadata including authors, year of publication, journals, citations, country, and author’s keywords, were collected.

All the retrieved data were uploaded on Rayyan (https://www.rayyan.ai/) (Ouzzani et al., 2016), and the duplicates were removed. Initially, the selection of articles was performed independently by two collaborators following the eligibility criteria (Table 1). The second step involved the resolution of conflicts by the third collaborator individually. Finally, 397 publications were obtained at the end of the selection process (Table S1 [suppl]).

The temporal trends in the number of published articles per year were analyzed. Each article was framed based on the central goal: productivity and quality of fruit, economy and technology, abiotic stress, biotic stress, and characterization studies. The temporal trends in tomato grafting were compared with all plant species and two relevant species in vegetable grafting: watermelon and eggplant (Lee et al., 2010). Additional surveys were performed to achieve this by replacing the tomato terms for “PLANT*”, for all plants; “WATERMELON*”, “CITRULLUS LANATUS”, OR “C. LANATUS” for watermelon and “EGGPLANT*”, “Solanum melongena”, OR “S. MELONGENA” for eggplant. T-test was performed to compare the number of publications in tomato and other three groups (plants, watermelon, and eggplant) as well as linear regression between years and number of publications of each group.

The collaboration index (Koseoglu, 2016) and authors index (Aria & Cuccurullo, 2017) were determined. Also, the ten most productive countries, the ten most cited articles, and journals with at least ten published articles were mapped. These analyses were conducted using the Bibliometrix package 3.0.5 (Aria & Cuccurullo, 2017) in software R 4.1.2 (R Core Team, 2020).

Co-authorship network mapping was built for authors and countries with at least five articles having five citations. In the co-occurrence network mapping, the query terms from the author’s keywords were removed, with keywords included in at least five articles. The links and link strength represent the number of connections from one item to others and the strength of those connections, respectively. All the network mappings were prepared using the VOSviewer 1.6.16 (van Eck & Waltman, 2010).

Each article was evaluated separately to obtain technological trends in grafting, identification of the main rootstocks, grafting methods, cultivation environment, and grafting specificity. The relation between grafting methods and countries was visualized using non-metric multidimensional scaling (NMDS) plots based on Bray-Curtis dissimilarity matrices employing vegan package 2.6-2 (Oksanen et al., 2022) in software R 4.1.2 (R Core Team, 2020).

In tomato grafting, the number of research articles increased at an annual rate of about 8.8%. These number differed from the tendency observed for all plants grafting (t = -6.2943; p < 0.001) from 1944-2020 (Fig. 1A). However, while comparing the number of articles in vegetable grafting (Fig. 1A), tomato grafting had similar tendencies as watermelon (t = 0.0745; p = 0.9407), and eggplant (t = 1.8618; p = 0.06522).

Until the end of the 20^th century, only 10% of the articles on tomato grafting were published inconstantly (Fig. 1B). Over this period to 2020, a significant increase has been observed in the number of publications concurrently analyzing the response of grafted tomatoes to abiotic stresses as well as productivity and fruit quality. Since the beginning of this historical series evaluated, studies on the characterization of tomato grafting and its application in response to biotic stresses have been subject to constant accomplishment. The economic and technological aspects of tomato grafting were barely published over the years.

Although 1240 authors were identified, only sixteen authors produced articles (17) with single authorship (Table S2 [suppl]). The author’s index per article and collaboration index were 3.12 and 3.21, respectively. The co-authorship network resulted in nine clusters, i.e., I (nodes 1–2), II (nodes 3–7), III (nodes 8–12), IV (nodes 13–15), V (nodes 16–23), VI (nodes 24–26), VII (node 27), VIII (nodes 28–30), and IX (nodes 31–32) (Fig. 2). The group of Leonardi – Giuffrida, Savvas – Ntatsi, and Louws – Gunter shared the strongest link. The author with the highest number of articles is Louws F. J. (North Carolina State University, USA), with 14 publications.

Authors from 53 countries published research articles on tomato grafting (Table S2 [suppl]). The USA held the highest number of published articles on tomato grafting (261), followed by Spain (255) (Fig. 3A). These countries also hold the highest citations in 2013 and 2019, respectively. Co-authorship networks among countries resulted in five clusters: I (nodes 1–7), II (nodes 8–10), III (nodes 11–13), IV (nodes 14–18), and V (node 19) (Fig. 3B). The USA had the most significant number of collaborations. The strongest links were found between Spain and the UK, as well as Greece and Germany.

The surveyed research articles were published in 153 journals (Table S2 [suppl]), and the journal Acta Horticulturae (56) published the highest number of articles on tomato grafting (Fig. 4). There were only five journals that published ten or more articles on the subject. Although the whole set of articles has been cited 7367 times (Table S1 [suppl]), 69 articles (16.95%) were not cited by any researcher till 2020. There were only 11 articles (2.7%) with more than 100 citations. The articles with the highest number of citations were Li et al. (2002), Alvarez et al. (2006), Kim et al. (2001), Holbrook et al. (2002), and Estañ et al. (2005) with 301, 300, 291, 253 and 241 citations respectively (Table S3 [suppl]).

The selected articles contained 931 author’s keywords (Table S2 [suppl]), and these keywords constituted eight clusters (Fig. 5), i.e., I (nodes 1-4), II (nodes 5-7), III (nodes 8-11), IV (nodes 12-14), V (nodes 15-16), VI (nodes 17-19) and VII (nodes 20-23) and VIII (nodes 24-26). The terms were related to the classification groups with built-in temporal trends analysis. The most used term was “yield” which has a considerable connection with “quality” and “fruit quality”. Other strong connections were “fruit quality” with “salt stress”, “vitamin C” with “Lycopene”, and “Ralstonia solanacearum” with “bacterial wilt”.

The interspecific grafts (61.83%) and experiments conducted in controlled conditions (82.87%) predominated on tomato grafting publications (Table S4 [suppl]). Although fourteen grafting techniques were found in the selected papers (Table S4 [suppl]), 33.25% of papers did not specify the grafting method used in the experiments. The most used grafting methods in most countries were reported as splice (39.55%) and cleft (22.42%) (Fig. 6). In the studies, the primary commercial rootstocks used were Maxifort (16.88%) and Beaufort (14.61%) hybrids from De Ruiter, Bayer® (Table S5 [suppl]). A total of 285 rootstocks (68.83%) were used only once, and eighty-five were not precisely identified. The most used rootstock (76.83%) was observed as S. lycopersicum × S. habrochaites. In the interspecific grafting, the most used species as rootstock were Solanum melongena (7.37%), Solanum tuberosum (4.18%), Nicotiana tabacum (3.44%), Solanum habrochaites (3.44%), and Solanum pimpinellifolium (3.44%).

Tomato grafting studies are relevant among vegetables, as shown by the comparison of the temporal trends of studies on watermelon and eggplant grafting. There has been an observable increase in the number of publications on tomato grafting, in line with the global growth in scientific research (King, 2004; Nabout et al., 2015). However, the articles on tomato grafting represent only a tiny portion of grafting studies as most articles are focused on fruit trees and other plant species. Grafting has been used in fruit trees for a long time before its application in commercial production. In 1927, this technique was adopted in Korea and Japan for vegetables (Sakata et al., 2007). The use of grafting in tomato plants began in the 1960s (Bie et al., 2017). Three decades later, the use of vegetable grafting arose in Western countries (Lee et al., 2010).

Compared to the 2,982 journals indexed in Scopus under the category of “Agriculture and Biological Sciences”, the number of journals that published articles on tomato grafting is relatively low at 153. However, this number can be considered satisfactory given the broad scope of the category and the specificity of tomato grafting within agricultural research. Tomato grafting has predominantly been published in horticultural journals. Within the “horticulture” category, 25 journals published a collection of 112 articles, representing 42.21% of the total. This indicates that interest in the grafting technique extends beyond the horticulture area and encompasses more than just production issues. Journals from other fields, such as plant physiology, have also published articles on tomato grafting, indicating interdisciplinary interest in the subject. The collection includes many studies on the movement of substances and signaling between roots and shoots. Regardless of the specific subject area, the peer-review process of specialized journals involves subject matter experts, ensuring a high degree of refinement in the published content.

Network analysis was used to examine the relationships between authors and countries within a specific topic. They showed that although numerous authors and countries published their research findings on tomato grafting, few had research collaborations. Only a few authors are part of the co-authorship network because many of them have published independently or in a restricted group with limited and disconnected partnerships. The clusters represented the research groups of different countries, i.e., clusters I, VI, and V: Spanish groups (cluster V having a British author), clusters II and VIII belong to the USA, cluster IV is an Italian research group, cluster VII composed of the single author from Brazil and cluster IX of Chinese researchers. Cluster III was the most heterogeneous, with researchers from Greece, Germany, and Turkey.

A higher number of articles have been published from countries having collaboration among research groups compared to countries without collaboration. Among European authors, the three main interconnected groups have a high level of cooperation, resulting in a greater number of published articles. In contrast, the low level of connectivity among research groups from other countries is reflected in the author’s network and productivity analyses. Only nineteen countries have integrated co-authorship networks. The USA profile reflects the influence of connectivity between countries and scientific production: it has more collaborations and, consequently, published the highest number of articles. USA, Israel, and Italy have numerous connections suggesting a high level of partnership with other countries. Only USA, Spain, and China contributed to over half of the papers. Co-authorship networks between authors from Spain and UK, as well as Greece and Germany, illustrated the close relationships between authors and strong relationships among these countries.

A significant number of journals publishing articles on grafting are interdisciplinary in nature with multi-authored articles. Such interdisciplinary research requires collaboration among a team of researchers from diverse fields, who work together to integrate findings from their respective areas. This approach facilitates scientific and technological advancement by addressing complex and pertinent problems (Fortunato et al., 2018). According to Wuchty et al. (2007), multi-authored articles generally have more citations. Interdisciplinary partnerships can contribute to developing agricultural practices that increase tomato productivity since the prominent tomato-producing countries have research groups in the co-authorship network. Therefore, expanding partnerships and interdisciplinarity research is essential to achieve advancement in tomato grafting.

In contrast, the research groups with limited collaborations and independent authorship may limit the research development. In the case of tomato grafting, independent authorship or restricted partnerships may be associated with local demands and specific challenges for cultivation (Nkansah et al., 2013; Kakabadze, 2018). However, the sharing of challenges, techniques, equipment, and other tools among researchers, even in local demands, contribute to disseminating science and advanced agricultural practices (Nabout et al., 2015; Parreira et al., 2017; de Bem Oliveira et al., 2019).

Studies on the effects of abiotic stress in grafted tomatoes have stood out. Many investigations focused on the genetic, molecular, and physiological responses to abiotic stresses exploring it from the molecular to the crop level, as well as the use of grafting to minimize the negative impacts on tomato production. Tomato crops are sensitive to various climatic factors. As a result, a decrease in agricultural water supplies poses a significant threat to tomato production, affecting the entire production chain and reducing its availability to consumers (Badawy et al., 2020; Araya et al., 2021). The increased periods of drought and unreliable precipitation result in a water deficit for crop production. Abiotic factors such as drought and salinity have become significant concerns for tomato crops. There are reports of using grafted tomatoes as a relevant strategy to mitigate the effects of abiotic factors with potential applications in tomato production.

Grafting has been reported to be effective in reducing or even avoiding the damage caused by salinity (Zhu, 2016). There is a particular concern because tomato cultivation in protected environments and hydroponics requires intensive use of fertilizers which may increase the salinity (Coban et al., 2020). The accumulation of solutes, mainly Na+ and Cl-, in roots affects the plant’s water status by decreasing osmotic potential, reducing water uptake and stomatal conductance. In most cases, salinity and water deficit are concurrent, which intensifies the impairment of tomato production (Uddin et al., 2016).

Despite a recent focus on abiotic factors, stress from biotic factors remains a crucial area of research. It is of paramount importance in tomato cultivation to deal with pathogenic organisms. Whereas the tomato crop is especially susceptible to soil pathogens, tomato grafting has been successfully used to provide resistance against nematodes, bacteria, and viruses by rootstock characteristics (Rivard et al., 2010; Louws et al., 2018). Mitigating soil pathogens has been a key focus in biotic stress studies, with special attention given to bacterial wilt caused by Ralstonia solanacearum, a critical Solanaceae disease in tropical climates (Carvalho et al., 2020). Of the biotic stress studies evaluated, 31.82% aimed to mitigate damage from bacterial wilt.

The characterization of physiological mechanisms using the grafting technique has been present in investigations on grafted tomato plants. The initial studies focused on understanding the connection between rootstock and scion (Pal & Nath, 1944). The complexity of these studies increased over the years, including genetics and molecular biology. The most cited articles in the collection belonged to this category, as they elucidate the grafting mechanism (Kim et al., 2001; Holbrook et al., 2002; Li et al., 2002; Estañ et al., 2005; Alvarez et al., 2006). Some studies provided information on the taxonomic compatibility between species (Zeist et al., 2017; Guimarães et al., 2019), while others defined the more effective grafting methods (Silva et al., 2016; Zeist et al., 2020). Reports provided information on the scion-rootstock connection, such as the cutting angle and the compatibility between the stem’s diameters that contributes to improving the technique. The ideal cutting angle was reported at 50° and 70°, which provides intimate contact between the rootstock and scion, allowing the greater fixation force (Bausher, 2013; Pardo-Alonso et al., 2019a). After joining the rootstock and scion, adhesion occurs between them at the graft junction. The contact between rootstock and scion must be tight, as larger angles result in greater clamping force.

Plants have two poles, root, and shoot, interconnected by vascular tissues. The cutting disrupts this connection, collapses cells, and activates the plant responses to mechanical stress with a positional effect. The responses above the cut include increased auxin and sugar transport, signaling, and biosynthesis of brassinosteroids. However, the expression of stress response genes predominates below the cut. In contrast, genes for wound response, oxidative detoxification, lignin metabolism, and cell wall organization are co-expressed (Xie et al., 2019).

When the rootstock and scion are compatible, their ability to form new cells and tissues decides the union. The intact parenchymatic cells stick together at the graft interface, strengthening this connection over time. These cells produce callus, a mass of undifferentiated and pluripotent cells which divide intensely and get elongated (Fan et al., 2015). The cell wall becomes thicker at the adhesion area with the deposition of pectic material (Frey et al., 2020). This deposition occurs probably due to the increased secretion of β-1,4-glucanase in the extracellular region to facilitate cell wall reconstruction (Notaguchi et al., 2020). The callus proliferates until new cambial cells dedifferentiate and establish a continuous connection between the scion and the rootstock. The restitution of plasmodesmata enables simplastic movement. The peroxidase activity, an enzyme involved in the lignification process, increases significantly at this stage. Thus, the newly formed cells differentiate into vascular tissues, forming a vascular connection between the graft parts (Fan et al., 2015; Frey et al., 2020).

Quality and yield are the main keywords in the studies and have a strong connection. The influence of grafting on fruits’ nutritional composition has also been well-studied (Djidonou et al., 2016). Nutrients such as lycopene, vitamin C, and carotenoids have been extensively studied. Consuming fresh tomatoes has numerous health benefits, including preventing cardiovascular diseases and cancer cell growth (Bhowmik et al., 2012). Research focuses on finding rootstocks that address fruit quality without impacting fruit quantity and biomass (Singh, 2017). In summary, studies show that the scion-rootstock combination directly affects tomato fruit quality.

Although studies on tomato grafting have several objectives, some topics, such as technological development, need to be further explored due to their relevance to the area. Only ten studies were found constituting the “economy and technology” topic. Grafting requires attention, precision, and training to ensure seedling survival. Incorrect execution requires additional seeds, increasing production costs and resulting in significant losses (Pardo-Alonso et al., 2019b). To ensure the grafting success, the angle and position of the cut, cleaning of the environment, as well as expertise of the methods are the key points. Reports show the possibility of automating grafting (Zhao et al., 2015; Pardo-Alonso et al., 2019b; 2020), which can maintain and even increase the survival rate of the grafted seedlings.

Cleft and splice methods have been used in more than 50% of the studies. These are the predominant methods used worldwide, as shown in the NMDS analysis. This study considered the cleft method because it is a synonym for the tube, slant-cut, splice, and wedge (Lee & Oda, 2010; Lee et al., 2010). Our analysis suggests the wide use of this technique in Cucurbitaceae and the commonly used method in Solanaceae (Lee et al., 2010). However, commercial production mainly uses tube grafting, which applies the clips to provide mechanical support for a solid connection between scion and rootstock, resulting in a high survival rate (Lee & Oda, 2010). The advantages of tube grafting include reduced cost and labor, and thus it is a straightforward technique. It can be considered for commercial tomato grafting (Rivard & Louws, 2006; Lee et al., 2010).

The rootstock is part of the graft which provides the desired characteristics. In addition, rootstocks need to be compatible with the scion without reducing plant growth and productivity (Albacete et al., 2015). Our analysis revealed that Solanum habrochaites is commonly used in commercial rootstocks and is a prominent species for interspecific grafts. This species is robust and widely used in crosses to obtain resistant traits (Sifres et al., 2011), which increases genetic variability and maintains compatibility as they belong to the same family. Moreover, the use of several cultivated species from the Solanaceae family as rootstocks were also identified in this study.

Interspecific grafts predominated in the analyzed reports. Though intraspecific grafting is recommended for compatibility, the tolerance characteristics may not be sufficiently shared between the same species (Venema et al., 2008; Rivard et al., 2010; Zeist et al., 2017). Using other species as rootstocks expands the diversity and genetic variability, providing tolerance to grafted plants compared to non-grafted tomatoes (Petran & Hoover, 2014). Compatibility tests demonstrate considerable benefits of using wild species for tomato grafting for various uses. Currently, 4020 species of the Solanaceae family have been recognized worldwide, from these, 1946 belong to Solanum (GBIF, 2021), demonstrating a great potential still underexploited for grafting.

This study applied scientometric methods to analyze the scientific production on tomato grafting for the first time. The most significant relations among different countries and research groups and their impact on the scientific production on the topic were presented. The analyses showed that tomato grafting has been used for the purposes of plant production, but it is also a promising technique to elucidate morpho-physiological processes. It was evident that tomato grafting shows potential for mitigating adverse environmental effects and for obtaining fruits of high quality. However, further studies are needed on the automation of processes, analysis in unprotected environments, the economic and environmental impacts in different locations around the globe, and on the testing of different species and genotypes for rootstocks. Finally, it is recommended that special attention be given to establishing collaboration between interdisciplinary groups in different countries to lead to advancements in the technique as well as scientific productivity and the impact of these studies.