Abstract
Globally, there has been an increase in stringent regulations governing the use of chemical soil fumigants for controlling diseases, pests, and weeds. Grafting has been identified as an effective alternative to soil fumigation for managing soilborne diseases and pests in intensive vegetable cropping systems. The majority of watermelon (Citrullus lanatus) grafting research confirms that selected rootstocks play a role in improving plant resistance or tolerance to common soilborne diseases. Currently, there is a lack of evidence-based literature on the effects of grafting on watermelon fruit quality attributes and yield components. Previous reviews report wide variation in the impact of grafting on watermelon production, depending on rootstock–scion combinations and environmental conditions. This review employed evidence-based synthesis methods to comprehensively and methodically summarize research results of the impact of grafting on watermelon, with a focus on fruit quality and yield. In this systematic review, 548 citations (studies published during 2011–21) were screened against strict inclusion criteria, and data were extracted from 47 studies. Meta-analysis of percent differences between the grafted watermelon treatment and the nongrafted or self-grafted watermelon control was performed using extracted data of yield components and a wide range of fruit quality attributes. Meta-analysis of research data with variance measures was also conducted based on a rather limited number of studies. Our findings showed higher levels of total yield, average fruit weight, fruit length and width, fruit lycopene and soluble solids content, rind thickness, flesh firmness, lightness, chroma, and flesh nitrogen (N) content in grafted watermelon treatments compared with the nongrafted or self-grafted control. In particular, total yield, average fruit weight, and flesh firmness exhibited significant increases of a more than 10% difference. In contrast, grafted plants demonstrated decreases in fruit pH, hue angle, and flesh calcium content, although the reduction was not greater than 10% relative to the control. Meta-analysis of research data with variance measures further confirmed significantly greater total yield and flesh N content in grafted watermelon treatments compared with the nongrafted or self-grafted control. In addition, the meta-analysis results confirmed greater benefits of watermelon grafting in the presence of known soilborne disease pressure in contrast to the production scenarios without soilborne disease problems.
Watermelon (Citrullus lanatus) is one of the 10 most highly cultivated fruits in the Cucurbitaceae family, which originated in Africa and is now widely grown across the globe (Chomicki et al. 2020). As a leading country in watermelon production, China produced over 60.6 million t in total fruit weight in 2019. As a result of the substantial production quantity in China, Asia accounts for 79.5% of all watermelon production in the world. In 2019, Africa produced 7.5%, followed by the Americas, Europe, and Oceania, accounting for 6.9%, 5.8%, and 0.2% of global watermelon production, respectively. The United States was the eighth largest watermelon producer in 2019 (United Nations Food and Agriculture Organization 2021).
Seeded watermelons are commonly consumed in Mediterranean nations, various ethnic groups across Europe, and many Asian and African countries, whereas seedless watermelons, including both miniature and “icebox” sizes, are preferred in the United States and northern Europe (Netherlands Ministry of Foreign Affairs, Centre for the Promotion of Imports from Developing Countries 2018; US National Research Council 2008). Seedless watermelons were first introduced to the US market in 1990. A decade later, seedless triploid cultivars accounted for 50% of watermelon production and, by 2007, 75% of production. As demand for seedless watermelons increased, additional challenges such as increased seed prices and reduced resistance to soilborne pathogens arose, due in part to breeding with susceptible tetraploid inbred lines to produce new triploid hybrids. This increased susceptibility to diseases such as Fusarium wilt (Fusarium oxysporum f. sp. niveum), imparted an immense financial strain on American farmers (Bruton et al. 2007).
Chemical soil fumigants have been widely used by growers to combat increases in soilborne pathogen infestations. As of 1999, before the establishment of regulations for the use of methyl bromide (MB), a broad-spectrum soil fumigant for controlling pathogens, pests, and weeds before planting, ∼60% of all US MB was customarily used for soil fumigation in watermelon, tomato, strawberry, pepper, cucumber, squash, and eggplant fields (Lynch and Carpenter 1999). However, multiple studies have associated MB exposure with increased risk of developing stomach cancer, damage to the central nervous system and respiratory system, and various toxicity injuries (Barry et al. 2012; Park et al. 2020; Rosskopf et al. 2005). Additionally, MB is classified as a Class 1 stratospheric ozone-depleting substance (US Environmental Protection Agency 2021). The Montreal Protocol (United Nations Environment Programme Ozone Secretariat 2020) encouraged the banning of MB, and by 2010, MB use was completely prohibited in Europe and banned with exceptions for critical-use needs in the United States (Public Health England, Centre for Radiation, Chemical and Environmental Hazards 2019; US Environmental Protection Agency 2021). While the search for alternative chemical fumigants continues, other nonchemical approaches have shown their effectiveness in soilborne disease management, including the vegetable grafting technology (King et al. 2008; Lee 1994; Lee and Oda 2003; Lee et al. 2010).
For centuries, grafting has been used to improve productivity in plants of the Cucurbitaceae family. Records from fifth-century China hold the first known mention of cucurbit grafting; gourd plants were self-grafted to help increase root volume and enlarge fruit size (Lee and Oda 2003). The first documented heterografting for vegetable soilborne disease management is far more recent, dating back to Japan and Korea in the late 1920s and early 1930s, when watermelon was grafted onto squash (Cucurbita moschata) rootstocks (Kubota et al. 2008). The rootstock refers to the plant providing the roots of the grafted plant, whereas the scion is the aboveground shoot that produces the stems, leaves, and fruit. The rootstock is chosen based on the biotic and abiotic stress resistance or tolerance of its root system as well as its vigor and input use efficiency. These first heterografting trials resulted in increased soilborne disease resistance and drove the innovation of grafting techniques for many vegetables (Lee 1994). Many rootstocks have been developed for commercial production of cucurbitaceous and solanaceous vegetables (Davis et al. 2008a, 2008b; Lee and Oda 2003). In Japan, Korea, and several surrounding countries, the majority of watermelon, cucumber, and solanaceous vegetable crop production uses grafting (Bekhradi et al. 2011). Grafting technologies eventually spread to Europe in the early 1990s and later to the United States (Lee et al. 2010). In Europe, grafting adoption in watermelon production substantially increased throughout the 1990s and early 2000s (Lee et al. 2010). Conversely, the slow adoption of grafting by American farmers has resulted in delayed development of grafting technology in the United States compared with many European and Asian countries.
Horticultural crop farmers in developing countries experience a yield gap of 60% to 70% of the land’s potential annually (Brown et al. 2015). Grafting with appropriate rootstocks enables farmers to reduce this yield gap and produce a more reliable watermelon harvest (Bigdelo et al. 2017). The predominant commercial watermelon rootstocks are C. moschata, Cucurbita maxima, interspecific squash hybrids C. maxima × C. moschata, and Citrullus amarus (King et al. 2010). Extensive research has illustrated the effectiveness of grafting to overcome soilborne pathogen issues as well as environmental challenges; however, it is unclear how grafting affects crop yield and fruit quality, especially under low levels of disease pressure or abiotic stressors. The body of literature of original research showed that depending on the scion cultivars used and the production systems, grafting with selected rootstocks can result in mixed impacts on fruit characteristics such as color, soluble solids content (SSC), lycopene, carotenoids, titratable acidity (TA), pH, fruit firmness, rind thickness, fruit shape, total yield, marketable yield, and average fruit weight (Alan et al. 2017; Bekhradi et al. 2011; Bertucci et al. 2018; Colla et al. 2011; Davis et al. 2008b; Fallik and Ilic 2014; Ioannou et al. 2000; Kombo and Sari 2019; Kyriacou et al. 2017; Miguel et al. 2004; Petropoulos et al. 2014; Rouphael et al. 2010; Suchoff et al. 2019; Turhan et al. 2012). Similarly, previous literature reviews of watermelon grafting describe broad variations in results attributed, in part, to the rootstock–scion combination, environmental production conditions, and the impact of the research methodology—particularly as it relates to harvest maturity (e.g., Devi et al. 2020a; Fallik and Ilic 2014; Fallik and Ziv 2020; Kumar et al. 2017; Kyriacou et al. 2017, 2018). The effect of grafting on SSC led to variable results influenced by a variety of factors such as the scion cultivars, rootstock selection, rootstock–scion compatibility, and production season (Devi et al. 2020a; Fallik and Ziv 2020). Fruit sugars were lower or more variable in scions grafted onto bottle gourd (Fallik and Ilic 2014; Kyriacou et al. 2017). Fallik and Ilic (2014) suggested that grafting had the largest influence on the SSC-to-TA ratio among the fruit quality changes. However, several reported impacts of grafting with certain rootstocks are based on a single publication (e.g., Fallik and Ilic 2014) or nonsignificant differences (e.g., Devi et al. 2020b). Despite the increasing number of grafted watermelon quality studies and narrative reviews on the topic, there is a lack of systematic reviews to help identify standardized effects of grafting on fruit quality. Furthermore, a statistical approach is currently lacking with respect to quantifying the rootstock effects on watermelon fruit yield and quality across many international studies.
To assess systematically how grafting affects fruit yield and quality, a review of the body of available literature is necessary. Narrative reviews remain a common form of literature review, despite their lack of replicability, inability to control bias in the review process, and tendency to include a smaller scope of literature. The use of evidence-based synthesis is well established in the health sciences (Chalmers et al. 2002) and is growing in popularity in agricultural research, allowing for meta-analysis of data extracted from many studies. A systematic review is an evidence-based literature review with the objective of answering a well-formulated research question whose answer is otherwise unclear from the accumulated literature. Systematic reviews are commonly combined with meta-analyses where extracted metadata are statistically analyzed. The entire process involves extensive literature searches, multiple reviewers, detailed protocols, and careful record keeping, with the objective of decreasing bias, and allowing for conclusions to be formed from analyzing the results of a wide range of included studies (Koutsos et al. 2019).
This systematic review and meta-analysis were designed to synthesize fruit quality and yield data from global watermelon grafting experiments in the last decade to seek answers to the question: how does grafting affect watermelon fruit yield components and quality attributes? This review offers the benefit of evidence-based synthesis review methodology to assess the impact of grafting with like comparisons, using clearly aligned comparable traits. Metadata were assessed from studies in multiple countries, using various grafting methods, and hundreds of scion and rootstock combinations. This data synthesis was necessary to understand the efficacy of grafting and to determine the viability of grafting as an alternative to chemical fumigation. This systematic review was also expected to enhance our understanding of grafting benefits and limitations toward improvement of grafting technology and rootstock development for watermelon production.
Materials and Methods
Literature search.
The wildcard symbol (*) was used to broaden the scope of the search terms, allowing for any variations of letters preceding or following the word, depending on placement of the symbol. The search results were then refined by languages to only include English and Spanish, which were chosen based on the language fluency of the review participants. The search was also refined by document types to include articles, meetings, reports, reviews, unspecified, and other. The search results of 1020 references were further filtered to articles published between 2011 and 2021, yielding 601 references (Fig. 1). This 10-year period was chosen because the preliminary search indicated that much of the earlier literature was focused on disease tolerance and resistance parameters, whereas more recent literature reported more fruit yield and quality metrics. Of these articles, 296 were from the Web of Science Core Collection, 199 from CABI CAB Abstracts, 28 from BIOSIS Citation Index, and four from SciELO Citation Index.
Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow diagram of article review process (Page et al. 2021). The reasons for exclusion and number of articles excluded for each reason are illustrated in Table 1.
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Different search terms were used due to the more limited search capacities of grey literature sites. Citations were only included if they were in English or Spanish and published between 2011 and 2021. The numbers of documents added from each grey literature source were as follows: nine references from ProQuest Dissertations & Theses Global, 19 references from FAO AGRIS, 19 references from the World Vegetable Center, and zero from EDIS.
An additional 27 references were added to this study on 5 Sep 2021: two hand-selected articles and 25 references from Google Scholar. Hand-selected articles were discovered outside the searches and included because they fit the scope of the study. Google Scholar was searched from Gainesville, FL, on 21 Aug 2021 for “watermelon graft” and limited to 2011–21 publication dates. The first 50 results from Google Scholar were exported and compared with the 527 references from databases and 47 references from grey literature sources; 25 nonduplicate references from Google Scholar were identified and incorporated into this review. This was done to broaden the scope of the search retrospectively.
All citations from each search were imported into Covidence (Covidence, Melbourne, Australia), which is a primary screening and data extraction tool used in research syntheses. A total of 53 duplicates were removed in the entirety of the systematic review and identified using Excel, Covidence, and manual review.
Inclusion criteria.
A team of five researchers (Jordana, Stapleton, Ray, Anrecio, and Freed) reviewed all 548 studies in Covidence. Inclusion criteria differed for each of the three phases of the review: title and abstract screening, full text review, and data extraction (Fig. 1). In the title and abstract screening, the documents were included if they contained the term “graft” and either “watermelon” or “Citrullus lanatus” anywhere in the title or abstract. In the title and abstract screening phase, 463 studies were included and 85 studies were excluded because they did not include the necessary terms in the title or abstract. An additional 78 citations were further excluded because they were abstracts only and did not include full text (Fig. 1). Full texts for the 385 studies were obtained and attached to each reference in Covidence. In the full text review, each study was assessed by two of the five reviewers for inclusion and a study was only included if all the following criteria were met:
Is in English or Spanish
Reports results from at least one measurement of fruit quality or yield from the specified list:
- a.Fruit color assessment: lightness, chroma, hue angle, a*, or b*
- b.Seed count
- c.Rind thickness
- d.Flesh firmness
- e.Flesh pH
- f.Soluble solids content (°Brix)
- g.Lycopene content
- h.Titratable acidity or malic acid content
- i.Macronutrient content of flesh: nitrogen (N), nitrate-nitrogen (NO3-N), potassium (K), phosphorus (P), sulfur (S), calcium (Ca), or magnesium (Mg)
- j.Micronutrient content of flesh: iron (Fe), manganese (Mn), copper (Cu), zinc (Zn), chlorine (Cl), boron (B), or sodium (Na)
- k.Fruit size: length or width
- l.Fruit number per plant or unit area
- m.Average fruit weight
- n.Total fruit weight or total yield (total yield reported or combined marketable and unmarketable)
Is a field experiment(s) with at least three replications and six plants per experimental unit
Has an appropriate control group, either a nongrafted or self-grafted control used for comparison with the grafted treatments; if both nongrafted and self-grafted treatments were provided in the same study, the nongrafted was used as the control
Includes at least one watermelon scion cultivar
Rootstock is a different cultivar or species from the scion
These fruit quality and yield parameters were selected for their importance in the watermelon industry and their prevalence in published research. In the full text screening phase, 47 studies were included, and 338 documents excluded (Fig. 1). When a study failed to meet more than one inclusion criteria, the first justification for exclusion was recorded. The justification for all excluded articles is reported in Table 1.
Number of articles excluded for each criterion in the full text review stage.
In the data extraction phase, general study design, plant density, and cultivar information were recorded for the 47 studies, as well as target parameter data. The 47 studies were randomly assigned to all five reviewers to extract data (Supplementary Material 1). Data of parameters for the treatment (grafted) and control (nongrafted or self-grafted) groups were recorded for analysis. After data extraction, a different reviewer confirmed the accuracy of the extracted data.
Meta-analyses.
The difference in parameter units between studies was not relevant as data were standardized by using percent difference. A Student’s t test was conducted in data analysis software JMP 16.1 (SAS Institute Inc., Cary, NC, USA) and used to determine whether statistically significant differences exist between grafted and control groups. The percent differences were tested at multiple levels of difference for each parameter, when sufficient data were available. Comparisons were made for the target difference values of 20%, 10%, 5%, 0%, –5%, and –10% between the grafted treatment and the nongrafted or self-grafted control (Jordana et al. 2021). Statistical significance was set at P ≤ 0.05. In addition, studies with reported soilborne disease pressure were separated from those without any soilborne pathogen problem for the analysis of extracted data to compare results in the two contrasting scenarios.
After yield and fruit quality parameters were extracted, additional data on the study design were extracted from each article, where available. This additional data included number of observations, means, and a measure of variance. Data on study design were extracted from eight articles and a meta-analysis random effects model was produced in data analysis software R 4.1.1 (The R Foundation, Vienna, Austria) using an inverse-variance weighting of the aggregate difference between the grafted treatment and the nongrafted or self-grafted control (Jordana et al. 2021).
Reduction of bias.
Bias was managed throughout the study to maintain an objective and systematic approach to the literature search, screening, and data extraction. The review was conducted by five reviewers at the University of Florida. These reviewers all had experience with horticultural research and were familiar with the relevant scientific literature. The protocol submission and searches were conducted by Jordana and Stapleton, title and abstract screening was conducted by three of the reviewers (Jordana, Stapleton, and Ray), and the full text review and data extraction were conducted by all five reviewers. In all phases, the number of documents reviewed or extracted was divided among the reviewers and randomly assigned. For the title and abstract screening and full text review phases, each article’s eligibility was determined from independent assessment by two reviewers. If reviewers did not concur, final reviews and decisions were made by Jordana. After all article data were extracted, each reviewer was randomly assigned a list of articles to check that all data were correctly exported, ensuring that the data extraction process from each article was independently verified by two reviewers.
Results and Discussion
Geographic range of studies included in the systematic review.
Data were extracted from 47 studies that were conducted in 16 countries (Fig. 2). Of the 47 studies, nine were conference proceedings, and two were research theses. The remaining 36 studies were all published in scholastic journals. This illustrates that all data were extracted from academic research studies. It needs to be pointed out that the presence of only academic research might present a potential risk of excluding studies showing nonsignificant effect of grafting compared with the control treatment. Table 2 lists the countries and their corresponding number of references. Of the 16 countries from where studies were sourced, eight were leading watermelon-producing countries in 2019 (US Department of Agriculture, National Agricultural Statistics Service 2019). Only four studies were conducted in China, the leading watermelon-producing country in the world. Due to the language limitation of the reviewers on this systematic review and the restricted access to scientific literature databases in Chinese, only Chinese studies published in English were included; this illustrates the need for review of more Chinese-language grafting literature in the future. While all studies focused on the effects of grafting on fruit quality and yield, study objectives and techniques differed by geographic region.
Map of 16 countries from which 47 studies were sourced. Although ~53% of studies were conducted in United States and Turkey, articles were sourced from five continents in total.
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Number of included studies from each country used in the meta-analysis.
Although China is the largest global producer of watermelons, grafted watermelons only represented 20% of all watermelon fruit production in 2008. Korea and Japan had the highest percentage of cultivated area for grafted watermelon at 95% and 92% of total watermelon production acreage, respectively, with each country grafting more than 700 million vegetable seedlings per year (Lee et al. 2010). A more recent report indicated that grafted watermelon production accounted for 40%, 94%, and 99% of total acreage in China, Japan, and Korea, respectively (Bie et al. 2017). Due to the development of and dependence on grafting, research in eastern Asian countries revolves around innovative grafting robots to reduce grafting labor (Lee et al. 2010). Studies in this review from Asia focused on combining the effects of grafting and biochar, dual and 3-fold rootstock grafting, and rootstock-compatibility testing.
Three studies included in the review were from African countries. Those studies focused on disease resistance and watermelon grafting technique efficacy in addition to reporting the relevant fruit yield and quality data. Watermelons are also commonly cultivated in Middle Eastern countries, which has fueled watermelon grafting research as it spread west from Asia. Approximately 60% to 70% of cultivated watermelon land is planted with grafted plants in Israel (Koren and Edelstein 2004). The two studies from the Middle East were conducted in Israel and Iran and focused on which rootstock produced the highest yielding or most disease resistant plants for grafted watermelon production.
In Europe, most watermelons are cultivated in the Mediterranean countries such as Spain, Italy, Greece, and Turkey (Benincasa et al. 2014). Economically, watermelon is one of the most significant cultivated vegetable crops in the Mediterranean region (Proietti et al. 2008). In 1998, more than 70,000 grafted plants were grown in Turkey, and by 2007, there were 51.7 million grafted crops grown there, of which watermelon represented 53% (Yilmaz et al. 2007). In 2009, 48.2 million watermelon seedlings were grafted in Spain and 10 million in Italy. In France, 186 ha of grafted watermelon were grown in 2009 (Lee et al. 2010). Assessment of the European articles included in this review illustrates that grafting research in Europe has been broadly focused on increased nutrient and water uptake, rootstock and scion interaction, and increased sustainability of grafting.
Of the top 10 watermelon producing countries, three are in the Americas. Brazil was the fourth largest watermelon producer, the United States was eighth, and Mexico 10th for total global production in 2019 (US Department of Agriculture, National Agricultural Statistics Service 2019). Grafting is still relatively new in the United States and does not comprise a large percentage of watermelon production despite 75% of watermelon production being at risk of Fusarium wilt (Bruton et al. 2007). In the United States, crop rotation and chemical soil fumigation have been the main strategies used for soilborne disease management, but alternative approaches are being sought given the challenges associated with MB phaseout and limited land for long-term rotation (Taylor et al. 2008). At present, high costs of grafted plants remain the major barrier impeding grafting adoption in the United States. Various Central and South American countries have established grafting practices; according to Camacho (2007), as cited by Davis et al. (2008b) it was estimated that the combined acreage of grafted watermelon production in Honduras and Guatemala reached approximately 3000 ha in 2007. Articles included from North and South American countries focused on aspects of watermelon production combined with grafting, such as the use of plastic mulch, early vs. late season growth, differences in nutrient uptake, control of soilborne pathogens, and a multitude of studies aimed at identifying appropriate rootstocks.
Percent difference comparisons between grafted treatments and nongrafted or self-grafted controls.
Among the 47 studies included in the meta-analysis, C. maxima × C. moschata, Lagenaria siceraria, C. lanatus, and C. amarus rootstocks as the major rootstocks evaluated for grafted watermelon production accounted for 39.2%, 25.9%, 13.7%, and 10.9%, respectively, of all types of rootstocks used for comparing the grafted watermelon treatment with the nongrafted or self-grafted watermelon control (Table 3). In the meta-analysis, each comparison was defined as a parameter of a grafted treatment compared with that parameter in a corresponding nongrafted or self-grafted control. Table 4 shows that yield parameters had substantially more comparisons, as opposed to fruit quality parameters such as lightness and a*, which only had three and nine comparisons, respectively. Student’s t test (P ≤ 0.05) results indicated how the percent difference between the grafted treatment and the nongrafted or self-grafted control for a given parameter compared against the target difference values of 20%, 10%, 5%, 0%, –5%, and –10%. These target difference values indicated whether the percent difference was significant at each target value. If a value is significant at a 0% target value, this means that the grafted treatment was significantly different from the control. If the parameter percent difference was significant at a 5% target value, this means that the grafted treatment was significantly greater than the control by at least 5%, and likewise, significance at –5% would indicate that the grafted treatment was significantly less than the control by at least 5%. Parameters were sorted into Student’s t test target values based on the overall percent difference for each parameter, for example, if a parameter showed a 7% difference from the control treatment, this would be assessed at a 5% target value since it is testing whether the change is significantly greater than a 5% increase from the nongrafted or self-grafted control.
A list of different types of rootstocks evaluated for grafted watermelon production in the studies included in the meta-analysis and their evaluation frequency based on the number of grafted to nongrafted or self-grafted watermelon production comparisons.
Number of grafted to nongrafted or self-grafted watermelon production comparisons by the yield or fruit quality parameter measured.
In the assessment of SSC, TA, pH, hue angle, and flesh Zn, a 0% target value was used because all parameters showed a <5% difference between the grafted and control values. This Student’s t test determined whether the percent treatment difference was significantly different from zero (i.e., whether the grafted treatment is significantly greater than or less than the control; Fig. 3A). In Fig. 3B, fruit length, number of fruit, flesh K, lycopene content, flesh a*, flesh lightness, and flesh chroma were assessed to compare the percent treatment difference to a 5% difference target value. Significance at a 5% probability level would mean that the percent difference of the parameter was significantly greater than the target of 5%. Figure 3C presents the results for comparing the percent treatment difference to a 10% difference target value for seed count, flesh P, total yield, flesh firmness, flesh N, rind thickness, and fruit width. Average fruit weight was also tested at a 20% difference target value (data not shown). The same approach was used for assessing flesh Mg, Fe, Cu, Mn, Ca, and Na; however, those parameters demonstrated decreased levels in the grafted treatments compared with the controls (i.e., the percent treatment difference was negative). In Fig. 3D, Mg was compared with a –5% difference target value and all other parameters were compared with a –10% difference target value.
Percent difference for each yield or fruit quality parameter was calculated as: [(grafted mean – nongrafted or self-grafted mean)/nongrafted or self-grafted mean] × 100. P values of Student’s t tests for each parameter are listed along the y-axis. Significance at P ≤ 0.05 for a given parameter indicates the percent difference is significantly different from target t test comparison value. (A) Student’s t test at a 0% target comparison value. (B) Student’s t test at a 5% target comparison value. (C) Student’s t test at a 10% target comparison value. (D) Student’s t test at –5% (Mg) and –10% target comparison values. SSC = soluble solids content; TA = titratable acidity; a* = red/green coordinate. Error bars represent 95% confidence intervals.
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Yield components.
The parameters with the most comparisons were total yield (n = 382), number of fruit (n = 198), and average fruit weight (n = 214). Combined, those parameters accounted for 45% of the comparisons assessed in this study (Table 4). Fruit size was also assessed as fruit length and fruit width. The t test results presented in Fig. 3B demonstrate that number of fruit and fruit length of grafted treatments did not show a significantly greater increase than 5% difference from the nongrafted or self-grafted control. However, total yield (P = 0.0004) of grafted treatments exhibited a significant increase compared with the nongrafted or self-grafted control by more than 10% (Fig. 3C), and average fruit weight (P = 0.008) by more than 20%. The mean percent difference of 12% increase reported for fruit width in grafted vs. nongrafted or self-grafted control was not significantly different from 10% (Fig. 3C). These meta-analysis results support the overall improvement in total yield of watermelon by grafting with various rootstocks evaluated in studies conducted during 2011–21 included in the systematic review. Moreover, the findings suggested that average fruit weight, but not fruit number, was the main contributing factor to watermelon yield increase by grafting. Fruit number showed an 8% increase, which was not significantly greater than a 5% target value; however, it was significantly greater than the nongrafted treatment at a 0% target percent difference value (Fig. 3B).
The overall yield improvement by grafting is further supported by the meta-analysis using inverse-variance weighting of the aggregate difference between the grafted treatment and the nongrafted or self-grafted control (Fig. 4). By taking into consideration the variance term reported for total yield data that could be found in five of the studies, this more in-depth meta-analysis confirmed that total watermelon yield of grafted treatments was significantly greater (P = 0.01) than the nongrafted or self-grafted control. However, a significant difference in average fruit weight was not detected (P = 0.32). The average fruit weight discrepancy results between the two meta-analysis methods were likely due to the large difference in the number of comparisons included because only a limited number of studies (n = 3) reported a measure of variance for this parameter.
Meta-analysis using research data with mean and variance measures from available studies. A random effects model was produced using an inverse-variance weighting of the aggregate difference between the grafted treatment and the nongrafted or self-grafted control. The solid line represents the value at which there is no difference between the grafted treatment and the nongrafted or self-grafted control. The dotted line represents the overall value of standardized mean difference. The closer the box plots are to the solid line, the less significant the difference between the grafted treatment and the control. The length of the line through each box plot represents the 95% confidence interval of the data within that study. The standardized mean difference (between the grafted treatment and the nongrafted or self-grafted control) was calculated for each study and an overall level of significance was calculated per parameter. Total yield was significant (P = 0.01), while average fruit weight was not significant (P = 0.32). A level of significance was also calculated from both total yield and average fruit weight studies, determining that there was no significant difference for yield parameters overall (P = 0.08).
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Chromaticity qualities.
The chromaticity qualities included in this systematic review were fruit flesh lightness, chroma, hue angle, and a*, which only accounted for 50 comparisons cumulatively (Table 4). Student’s t test results showed that grafted fruit hue angle (P = 0.038) was significantly less than the nongrafted or self-grafted control, although the reduction of 2% was relatively small (Fig. 3A). Hue angle is calculated as tan−1 (b*/a*), where a* is the red/green coordinate, and b* is the yellow/blue coordinate. An increase in a* indicates an increase in redness of the color. The increase in a* was not significantly different from 5% or 0%, whereas the decrease in hue angle indicates that grafting increased flesh redness by a small margin (Fig. 3A and 3B). Although both lightness (P = 0.023) (light vs. dark) and chroma (P = 0.005) (pure vs. dull) of grafted fruit were significantly different from the nongrafted or self-grafted control, they did not show a significantly greater increase than 5% from the control (Fig. 3B). These significant, yet small differences between the grafted watermelon treatment and the nongrafted or self-grafted control may have rather limited practical implications, and consumer sensory evaluation can be included in future studies to determine how consumer perceived fruit flesh color and preference may be impacted.
Flesh mineral nutrient contents.
The nine fruit mineral nutrients included in this review were N (n = 12), K (n = 17), P (n = 14), Ca (n = 17), Mg (n = 17), Fe (n = 17), Mn (n = 17), Cu (n = 12), and Zn (n = 17) (Table 4). Zinc was not significantly different from the nongrafted or self-grafted control and varied greatly with a 95% confidence interval ranging from a decrease of ∼20% to an increase of ∼13% (Fig. 3A). The K content also varied greatly, and t test results indicated that although there was an overall increase of 8% flesh K in grafted watermelon fruit, it was not significantly greater than a 0% or 5% difference from the control (Fig. 3B). The mean percent increase in fruit N in grafted treatments was more than 20%, but it was not significantly greater than 10% difference from the control (Fig. 3C). Fruit P increased by more than 45%, but it was also not significantly greater than a 10% difference from the control (Fig. 3C). Fruit Mg decreased by roughly 7%, but this reduction was not significantly different from 0% or –5% (Fig. 3D). Levels of Na, Ca, Mn, Cu, and Fe in grafted watermelon fruit decreased by 10% or more relative to the nongrafted or self-grafted control; however, the difference in Na, Mn, Cu, and Fe were not significantly greater than a 10% reduction or different from 0% (Fig. 3D). Conversely, the difference in Ca content was significantly different from 0% (P = 0.013), but the decrease was less than 5% (data not shown). In the meta-analysis using available data with variance measures, a significant increase between the grafted treatment and the nongrafted or self-grafted control was detected for fruit K (P < 0.01), Mn (P < 0.01), N (P < 0.01), and Zn (P < 0.01) contents, whereas a significant decrease was identified in Ca (P < 0.01) and Mg (P < 0.01) (Fig. 5A). However, no significant difference was identified for Cu or Fe and the overall flesh mineral P value was not significant (P = 0.58). Again, the variation in results between meta-analyses is likely due to the limited number of studies that included variance data (n = 1). In this systematic review, the use of the meta-analysis using data with variance measures was largely limited by the lack of reported variance data in scientific studies. Future studies that include variance measures are warranted to allow for a more robust meta-analysis.
Meta-analysis using research data with mean and variance measures from available studies. A random effects model was produced using an inverse-variance weighting of the aggregate difference between the grafted treatment and the nongrafted or self-grafted control. The solid line represents the value at which there is no difference between the grafted treatment and the nongrafted or self-grafted control. The dotted line represents the overall value of standardized mean difference. The closer the box plots are to the solid line, the less significant the difference between the grafted treatment and the control. The length of the line through each box plot represents the 95% confidence interval of the data within that study. The standardized mean difference (between the grafted treatment and the nongrafted or self-grafted control) was calculated for each study and an overall level of significance was calculated per parameter. (A) Flesh mineral nutrient content parameters. (B) Rind and flesh quality parameters. Flesh firmness, lycopene content, rind thickness, soluble solids content (SSC), and titratable acidity (TA) were not significantly different from the control with an overall P value of 0.70. K, Mn, N, and Zn were all significantly greater than the control, whereas Mg and Ca showed significant decreases. A level of significance was also calculated from all the flesh mineral nutrient content studies, determining that there was no significant difference for the overall flesh mineral nutrient content (P = 0.58).
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Flesh and rind quality attributes.
Flesh qualities included flesh firmness with 143 comparisons, seed count with 10 comparisons, flesh pH with 100 comparisons, SSC with 241 comparisons, lycopene with 78 comparisons, and TA with 39 comparisons (Table 4). In the Student’s t test results, fruit SSC (P = 0.007), pH (P = 0.0002), and hue angle (P = 0.038) of the grafted treatment were significantly different from the control (0% target difference value) (Fig. 3A); however, all percent differences were less than 5% difference from the nongrafted or self-grafted control. Lycopene increased by 7% on average and although it was not significantly different from a 5% difference from the control (Fig. 3B), it was significantly different from a 0% difference (P < 0.0001) (data not shown). Rind thickness showed an average increase of 13%, which was not significantly more than a 10% difference from the control (Fig. 3C). The average increases in flesh firmness (P < 0.0001) and seed count (P = 0.017) in grafted treatments were significantly greater than a 10% difference from the control, reaching 23% and 54%, respectively (Fig. 3C). Seed count was only reported in studies using seeded scion cultivars. Results from the meta-analysis using data with variance measures showed no significant difference between grafted treatments and the nongrafted control for lycopene content, flesh firmness, rind thickness, SSC, or TA and the overall flesh and rind quality P value was 0.70 (Fig. 5B).
Comparing studies with and without soilborne pathogen problems.
We also separated studies with known soilborne disease pressure reported from those that did not indicate any soilborne pathogen problem to compute the mean percent difference between the grafted treatment and the nongrafted or self-grafted control, compared in a Student’s t test at a target value of 0% (Fig. 6). Studies that intentionally included soilborne pathogens in the experimental design made up 212 comparisons, compared with 309 comparisons that intentionally excluded soilborne pathogens from the study. Figure 6A illustrates the percent difference comparisons of parameters, specifically from studies in which disease was absent, between the grafted treatment and the nongrafted or self-grafted control. Flesh firmness, rind thickness, and lycopene content were all significantly greater in the grafted treatment than the control (P < 0.0001), with mean increases from ∼7% to 24%. A significant decrease was found in flesh pH (P < 0.0001) and number of fruit (P = 0.05), however, only by <10%. All other parameters, including total yield, average fruit weight, fruit length, fruit width, and fruit SSC and TA were not significant and the grafted treatment had less than (±) 5% difference from the nongrafted or self-grafted control. Figure 6B shows the percent difference comparisons of parameters, from studies where disease was present, between the grafted treatment and the nongrafted or self-grafted control. This indicated a significant increase in total yield, average fruit weight, number of fruit, flesh firmness, and rind thickness of ∼10% to 60% for the grafted treatment in comparison with the nongrafted or self-grafted control. Fruit SSC of the grafted treatment also significantly increased by less than 5%, and all other parameters increased but not significantly. This is indicative of the key benefit of grafting for controlling diseases in infested soils to save crop yield and maintain fruit quality. However, this systematic review did not offer strong evidence for increased yield performance by grafting when soilborne pathogen issues are absent. Further analysis may also be conducted to dissect the watermelon grafting benefits under different environmental stress conditions if data can be extracted.
Percent difference for each yield or fruit quality parameter was calculated as: [(grafted mean – nongrafted or self-grafted mean)/nongrafted or self-grafted mean] × 100. P values of Student’s t tests for each parameter are listed along the y-axis. Significance at P ≤ 0.05 for a given parameter indicates the percent difference is significantly different from the target t test comparison value of 0%. (A) t-test results of studies where no soilborne diseases were present. (B) t-test results of studies where soilborne diseases were present. SSC = soluble solids content; TA = titratable acidity. Error bars represent 95% confidence intervals.
Citation: HortScience 58, 8; 10.21273/HORTSCI16857-22
Overall, findings from this systematic review of yield and quality data from grafted watermelon studies in the past decade are in line with a recent meta-analysis report on grafted tomato yield and fruit quality which confirmed the role of resistant rootstocks in soilborne disease management and soil fumigant use reduction and revealed the generally comparable fruit quality between grafted and nongrafted tomato production (Grieneisen et al. 2018). Future research that reports harvest maturity indices, abiotic and/or biotic stress levels, and measures of variance would enable more studies to be included in research syntheses targeting different scenarios of comparisons. Further analyses are also needed to elucidate driving factors that maximize the benefits of grafting as an environmentally friendly technology with optimized economic feasibility.
Conclusions
Grafting with resistant rootstocks has been successfully used in watermelon production for controlling soilborne pathogens. In this systematic review, the effects of grafting on specific fruit quality attributes and yield components were assessed by examining studies conducted during 2011 through 2021. Studies were included from 16 countries on five continents, each with different study objectives and treatments, but with extractable fruit yield and quality results. Results from the meta-analysis based on percent differences showed that, on average, compared with nongrafted or self-grafted watermelon, grafted watermelon treatments exhibited increases in total yield, average fruit weight, fruit length and width, fruit lycopene and SSC, rind thickness, flesh firmness, lightness, and chroma, and flesh N content. Of these parameters, total yield, average fruit weight, and flesh firmness had significant increases of more than a 10% difference in comparison with the nongrafted or self-grafted control. In contrast, grafted plants demonstrated decreases in fruit pH, hue angle, and flesh Ca content, although none of those parameters showed a significant decrease of more than 10% relative to the control. Meta-analysis of research data with variance measures further confirmed significantly greater total yield and flesh N content in grafted watermelon treatments compared with the nongrafted or self-grafted control. Our findings also indicated more pronounced benefits of watermelon grafting in the presence of known soilborne disease pressure in contrast to the production scenarios without soilborne disease problems.
References Cited
Alan O, Duzyaman E, Sen F. 2017. How growing cycles affect plant growth and yield of grafted watermelon combinations. Fresenius Environ Bull. 26(6):4214–4221.
Barry KH, Koutros S, Lubin JH, Coble JB, Barone-Adesi F, Beane Freeman LE, Sandler DP, Hoppin JA, Ma X, Zheng T. 2012. Methyl bromide exposure and cancer risk in the Agricultural Health Study. Cancer Causes Control. 23(6):807–818. https://doi.org/10.1007/s10552- 012-9949-2.
Bekhradi F, Kashi A, Delshad M. 2011. Effect of three cucurbits rootstocks on vegetative and yield of ‘Charleston Gray’ watermelon. Int J Plant Prod. 5(2):105–109. https://doi.org/10.22069/ijpp.2012.724.
Benincasa P, Massoli A, Polegri L, Concezzi L, Onofri A, Tei F. 2014. Optimising the use of plastic protective covers in field grown melon on a farm scale. Ital J Agron. 9(556):8–14. https://doi.org/10.4081/ija.2014.556.
Bertucci MB, Jennings KM, Monks DW, Schultheis JR, Perkins-Veazie P, Louws FJ, Jordan DL. 2018. Early season growth, yield, and fruit quality of standard and mini watermelon grafted onto several commercially available cucurbit rootstocks. HortTechnology. 28(4):459–469. https://doi.org/10.21273/HORTTECH04051-18.
Bie Z, Azher Nawaz M, Huang Y, Lee J-M, Colla G. 2017. Introduction to vegetable grafting, p. 1–21. In: Colla G, Pérez-Alfocea F, Schwarz D (eds.). Vegetable grafting: Principles and practices. CABI, Boston, MA, USA.
Bigdelo M, Hassandokht MR, Rouphael Y, Colla G, Soltani F, Salehi R. 2017. Evaluation of bitter apple [Citrullus colocynthis (L.) Schrad] as potential rootstock for watermelon. Aust J Crop Sci. 11(6):727–732. https://doi.org/10.21475/ajcs.17.11.06.p492.
Brown ME, Antle JM, Backlund P, Carr ER, Easterling WE, Walsh MK, Ammann C, Attavanich W, Barrett CB, Bellemare MF, Dancheck V, Funk C, Grace K, Ingram JSI, Jiang H, Maletta H, Mata T, Murray A, Ngugi M, Ojima D, O’Neill B, Tebaldi C. 2015. Climate change, global food security, and the U.S. food system. U.S. Department of Agriculture, University Corporation for Atmospheric Research, and National Center for Atmospheric Research, Washington, DC, USA. https://doi.org/10.7930/J0862DC7.
Bruton BD, Fish WW, Zhou XG, Everts KL, Roberts PD. 2007. Fusarium wilt in seedless watermelons. Proc 2007 Southeast Reg Veg Conf, p. 93–98.
Chalmers I, Hedges LV, Cooper H. 2002. A brief history of research synthesis. Eval Health Prof. 25(1):12–37. https://doi.org/10.1177/0163278702025001003.
Chomicki G, Schaefer H, Renner SS. 2020. Origin and domestication of Cucurbitaceae crops: Insights from phylogenies, genomics and archaeology. New Phytol. 226(5):1240–1255. https://doi.org/10.1111/nph.16015.
Colla G, Rouphael Y, Mirabelli C, Cardarelli M. 2011. Nitrogen-use efficiency traits of mini-watermelon in response to grafting and nitrogen-fertilization doses. J Plant Nutr Soil Sci. 174(6):933–941. https://doi.org/10.1002/jpln.201000325.
Davis AR, Perkins-Veazie P, Hassell R, Levi A, King SR, Zhang XP. 2008a. Grafting effects on vegetable quality. HortScience. 43:1670–1672. https://doi.org/10.21273/hortsci.43.6.1670.
Davis AR, Perkins-Veazie P, Sakata Y, Lopez-Galarza S, Maroto JV, Lee SG, Huh YC, Sun ZY, Miguel A, King SR, Cohen R, Lee JM. 2008b. Cucurbit grafting. Crit Rev Plant Sci. 27(1):50–74. https://doi.org/10.1080/07352680802053940.
Devi P, Perkins-Veazie P, Miles C. 2020a. Impact of grafting on watermelon fruit maturity and quality. Horticulturae. 6(4):97. https://doi.org/10.3390/horticulturae6040097.
Devi P, Perkins-Veazie P, Miles CA. 2020b. Rootstock and plastic mulch effect on watermelon flowering and fruit maturity in a Verticillium dahliae-infested field. HortScience. 55(9):1438–1445. https://doi.org/10.21273/HORTSCI15134-20.
Fallik E, Ilic Z. 2014. Grafted vegetables - the influence of rootstock and scion on postharvest quality. Folia Hortic. 26(2):79–90. https://doi.org/10.2478/fhort-2014-0008.
Fallik E, Ziv C. 2020. How rootstock/scion combinations affect watermelon fruit quality after harvest? J Sci Food Agr. 100(8):3275–3282. https://doi.org/10.1002/jsfa.10325.
Grieneisen ML, Aegerter BJ, Stoddard CS, Zhang M. 2018. Yield and fruit quality of grafted tomatoes, and their potential for soil fumigant use reduction. A meta-analysis. Agron Sustain Dev 38:29. https://doi.org/10.1007/s13593-018-0507-5.
Ioannou N, Poullis CA, Heale JB. 2000. Fusarium wilt of watermelon in Cyprus and its management with soil solarization combined with fumigation or ammonium fertilizers. Bull OEPP. 30(2):223–230. https://doi.org/10.1111/j.1365-2338.2000.tb00884.x.
Jordana C, Stapleton SC, Zhao X. 2021. The effects of watermelon grafting on fruit quality and yield: A systematic literature review [research protocol registration]. Center for Open Science, Charlottesville, VA. https://doi.org/10.17605/OSF.IO/9DK2W.
King SR, Davis AR, Liu WG, Levi A. 2008. Grafting for disease resistance. HortScience. 43(6):1673–1676. https://doi.org/10.21273/HORTSCI.43.6.1673.
King SR, Davis AR, Zhang X, Crosby K. 2010. Genetics, breeding and selection of rootstocks for Solanaceae and Cucurbitaceae. Scientia Hortic. 127(2):106–111. https://doi.org/10.1016/j.scienta.2010.08.001.
Kombo MD, Sari N. 2019. Rootstock effects on seed yield and quality in watermelon. Hortic Environ Biotechnol. 60(3):303–312. https://doi.org/10.1007/s13580-019-00131-x.
Koren A, Edelstein M. 2004. Advantages and limitations of grafted vegetable transplants in Israel. HortScience. 39(4):873C. https://doi.org/10.21273/hortsci.39.4.873c.
Koutsos TM, Menexes GC, Dordas CA. 2019. An efficient framework for conducting systematic literature reviews in agricultural sciences. Sci Total Environ. 682:106–117. https://doi.org/10.1016/j.scitotenv.2019.04.354.
Kubota C, McClure MA, Kokalis-Burelle N, Bausher MG, Rosskopf EN. 2008. Vegetable grafting: History, use, and current technology status in North America. HortScience. 43(6):1664–1669. https://doi.org/10.21273/hortsci.43.6.1664.
Kumar P, Rouphael Y, Cardarelli M, Colla G. 2017. Vegetable grafting as a tool to improve drought resistance and water use efficiency. Front Plant Sci. 8:1130. https://doi.org/10.3389/fpls.2017.01130.
Kyriacou MC, Leskovar DI, Colla G, Rouphael Y. 2018. Watermelon and melon fruit quality: The genotypic and agro-environmental factors implicated. Scientia Hortic. 234:393–408. https://doi.org/10.1016/j.scienta.2018.01.032.
Kyriacou MC, Rouphael Y, Colla G, Zrenner R, Schwarz D. 2017. Vegetable grafting: The implications of a growing agronomic imperative for vegetable fruit quality and nutritive value. Front Plant Sci. 8:741. https://doi.org/10.3389/fpls.2017.00741.
Lee J-M. 1994. Cultivation of grafted vegetables I. Current status, grafting methods, and benefits. HortScience. 29:235–239. https://doi.org/10.21273/HORTSCI.29.4.235.
Lee J-M, Kubota C, Tsao SJ, Bie Z, Hoyos Echevarria P, Morra L, Oda M. 2010. Current status of vegetable grafting: Diffusion, grafting techniques, automation. Scientia Hortic. 127(2):93–105. https://doi.org/10.1016/j.scienta.2010.08.003.
Lee J-M, Oda M. 2003. Grafting of herbaceous vegetable and ornamental crops. Hortic Rev. 28:61–124. https://doi.org/10.1002/9780470650851.ch2.
Lynch L, Carpenter J. 1999. The economic impacts of banning methyl bromide: Where do we need more research? Univ Florida Inst Food and Agric Sci 1999 Annual Res Conf on Methyl Bromide Alternatives and Emissions Reductions. Paper no. 55.
Miguel A, Maroto JV, San Bautista A, Baixauli C, Cebolla V, Pascual B, López S, Guardiola JL. 2004. The grafting of triploid watermelon is an advantageous alternative to soil fumigation by methyl bromide for control of Fusarium wilt. Scientia Hortic. 103(1):9–17. https://doi.org/10.1016/j.scienta.2004.04.007.
Netherlands Ministry of Foreign Affairs, Centre for Promotion of Imports from Developing Countries. 2018. Exporting fresh melons to Europe. https://www.cbi.eu/market-information/fresh-fruit-vegetables/fresh-melons. [accessed 1 Oct 2021].
Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brenna SE, Chou R, Glanville J, Grimshaw JM, Hrobjartsson A, Lalu MM, Li TJ, Loder EW, Mayo-Wilson E, McDonald S, McGuinness LA, Stewart LA, Thomas J, Tricco AC, Welch VA, Whiting P, Moher D. 2021. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ 372:n71. http://dx.doi.org/10.1136/bmj.n71.
Park MG, Choi J, Hong YS, Park CG, Kim BG, Lee SY, Lim HJ, Mo HH, Lim E, Cha W. 2020. Negative effect of methyl bromide fumigation work on the central nervous system. PLoS One. 15(8):e0236694. https://doi.org/10.1371/journal.pone.0236694.
Petropoulos SA, Olympios C, Ropokis A, Vlachou G, Ntatsi G, Paraskevopoulos A, Passam HC. 2014. Fruit volatiles, quality, and yield of watermelon as affected by grafting. J Agric Sci Technol. 16(4):873–885.
Proietti S, Rouphael Y, Colla G, Cardarelli M, De Agazio M, Zacchini M, Rea E, Moscatellol S, Battistelli A. 2008. Fruit quality of mini-watermelon as affected by grafting and irrigation regimes. J Sci Food Agr. 88(6):1107–1114. https://doi.org/10.1002/jsfa.3207.
Public Health England, Centre for Radiation, Chemical and Environmental Hazards. 2019. Methyl bromide: General information. Public Health England Centre for Radiation, Chemical and Environmental Hazards. Public Health England Publ. Gateway 2014790.
Rosskopf EN, Chellemi DO, Kokalis-Burelle N, Church GT. 2005. Alternatives to methyl bromide: A Florida perspective. Plant Health Prog. 6(1):19. https://doi.org/10.1094/php-2005-1027-01-rv.
Rouphael Y, Schwarz D, Krumbein A, Colla G. 2010. Impact of grafting on product quality of fruit vegetables. Scientia Hortic. 127(2):172–179. https://doi.org/10.1016/j.scienta.2010.09.001.
Schöpfel J, Farace DJ. 2018. Grey literature [ELIS Classic], p 1746–1756. In: JD McDonald, M Levine-Clark (eds). Encyclopedia of Library and Information Sciences (4th ed). CRC Press, Boca Raton, FL, USA.
Suchoff DH, Schultheis JR, Gunter CC, Hassell RL, Louws FJ. 2019. Effect of rootstock and nitrogen fertilizer on growth and yield in watermelon. J Hortic Sci Biotechnol. 94(6):798–804. https://doi.org/10.1080/14620316.2019.1624629.
Taylor M, Bruton B, Fish W, Roberts W. 2008. Cost benefit analyses of using grafted watermelon transplants for Fusarium Wilt disease control. Acta Hortic. 782:343–350. https://doi.org/10.17660/ActaHortic.2008. 782.43.
Turhan A, Ozmen N, Kuscu H, Serbeci MS, Seniz V. 2012. Influence of rootstocks on yield and fruit characteristics and quality of watermelon. Hortic Environ Biotechnol. 53(4):336–341. https://doi.org/10.1007/s13580-012-0034-2.
United Nations Environment Programme Ozone Secretariat. 2020. Handbook for the Montreal Protocol on Substances that Deplete the Ozone Layer, 14th ed. https://ozone.unep.org/sites/default/files/Handbooks/MP-Handbook-2020-English.pdf. [accessed 2 Sep 2021].
United Nations Food and Agriculture Organization. 2021. FAOSTAT [database]. https://www.fao.org/faostat/en/#home. [accessed 2 Sep 2021].
US Department of Agriculture, National Agricultural Statistics Service. 2019. Data and Statistics. https://www.nass.usda.gov/Data_and_Statistics/. [accessed 2 Sep 2021].
US Environmental Protection Agency. 7 October 2021. Phaseout of ozone-depleting substances: Methyl bromide. US Government, Washington, DC, USA.
US National Research Council. 2008. Lost crops of Africa. Volume III Fruits. National Academies Press, Washington, DC, USA.
Yilmaz SM, Göçmen M, Ünlu A, First AF, Aydinşakir K, Cetinkaya S, Kuzgun M, Çelikyurt MA, Sayin B, Celik I. 2007. Grafting as an alternative to MB in vegetable production in Turkey. Ann Int Conf Methyl Bromide Alternatives and Emission Reductions.
Supplementary Material 1
List of the 47 studies included in the systematic review.
Abd El-Wanis M, El-Eslamboly AASA, Salama MA. 2013. Impact of different grafting methods on yield and quality of watermelon. Res J Agric Biol Sci. 9(6):330–340.
Alan Ö, Ceylan S, Özge M. 2014. Evaluation of hybrid stocks on vigor, yield and quality of watermelon, p. 369–374. Fifth Int Sci Agric Symp “Agrosym 2014,” Jahorina, Bosnia and Herzegovina, 23–26 Oct 2014. University of East Sarajevo.
Alan O, Duzyaman E, Sen F. 2017. How growing cycles affect plant growth and yield of grafted watermelon combinations. Fresenius Environ Bull. 26(6):4214–4221.
Alan O, Sen F, Duzyaman E. 2018. The effectiveness of growth cycles on improving fruit quality for grafted watermelon combinations. Food Sci Tech. 38:270–277. https://doi.org/10.1590/1678-457X.20817.
Aslam A, Zhao S, Azam M, Lu X, He N, Li B, Dou J, Zhu H, Liu W. 2020. Comparative analysis of primary metabolites and transcriptome changes between ungrafted and pumpkin-grafted watermelon during fruit development. Peer J. 8:e8259. https://doi.org/10.7717/peerj.8259.
Bekhradi F, Kashi A, Delshad M. 2011. Effect of three cucurbits rootstocks on vegetative and yield of ‘Charleston Gray’ watermelon. Int J Plant Prod. 5(2):105–109.
Bertucci MB, Jennings KM, Monks DW, Schultheis JR, Perkins-Veazie P, Louws FJ, Jordan DL. 2018. Early season growth, yield, and fruit quality of standard and mini watermelon grafted onto several commercially available cucurbit rootstocks. HortTechnology. 28(4):459–469. https://doi.org/10.21273/HORTTECH04051-18.
Bigdelo M, Hassandokht MR, Rouphael Y, Colla G, Soltani F, Salehi R. 2017. Evaluation of bitter apple [Citrullus colocynthis (L.) Schrad] as potential rootstock for watermelon. Aust J Crop Sci. 11(6):727–732. https://doi.org/10.21475/ajcs.17.11.06.p492.
Ceylan S, Alan O, Elmaci OL. 2018. Effects of grafting on nutrient element content and yield in watermelon. Ege Univ Ziraat Fak Derg. 55(1):67–74. https://doi.org/10.20289/zfdergi.390891.
Colla G, Cardarelli M, Fiorillo A, Rouphael Y, Rea E. 2012. Enhancing nitrogen use efficiency in Cucurbitaceae crops by grafting. Acta Hortic. 952:863–869. https://doi.org/10.17660/ActaHortic.2012.952.109.
Colla G, Rouphael Y, Mirabelli C, Cardarelli M. 2011. Nitrogen-use efficiency traits of mini-watermelon in response to grafting and nitrogen-fertilization doses. J Plant Nutr Soil Sci. 174(6):933–941. https://doi.org/10.1002/jpln.201000325.
Dabirian S, Inglis D, Miles CA. 2017. Grafting watermelon and using plastic mulch to control verticillium wilt caused by Verticillium dahliae in Washington. HortScience. 52(3):349–356. https://doi.org/10.21273/HORTSCI11403-16.
Daley SLR. 2014. Chemical control of rootstock regrowth in grafted watermelon and its effects on plant growth and development (MS Thesis). Clemson University, Clemson, SC, USA.
Devi P, Perkins-Veazie P, Miles CA. 2020. Rootstock and plastic mulch effect on watermelon flowering and fruit maturity in a Verticillium dahliae-infested field. HortScience. 55(9):1438–1445. https://doi.org/10.21273/HORTSCI15134-20.
Edelstein M, Tyutyunik J, Fallik E, Meir A, Tadmor Y, Cohen R. 2014. Horticultural evaluation of exotic watermelon germplasm as potential rootstocks. Scientia Hortic. 165:196–202. https://doi.org/10.1016/j.scienta.2013.11.010.
Freeman J. 2015. Managing root-knot nematode (Meloidogyne spp.) in grafted watermelon. Proc Florida State Hortic Soc. 128:156–158.
Hassell RL, Pat W, Levi A. 2018. Bulldog - Citrullus lanatus var. citroides line, a potential rootstock for watermelon grafting. Acta Hortic. 1204:219–222. https://doi.org/10.17660/ActaHortic.2018.1204.28.
Huh YC, Lee WM, Ko HC, Park DK, Park KS, Lee HJ, Lee SG, Ko KD. 2012. Development of Fusarium wilt-resistant F-1 hybrids of bottle gourd (Lagenaria siceraria Standl.) for watermelon rootstocks, p. 74–82. Proc. of the Xth EUCARPIA Meeting on Genetics and Breeding of Cucurbitaceae, Antalya, Turkey, 15–18 Oct 2012.
Khankahdani HH, Zakeri E, Saeedi G, Shakerdargah G. 2012. Evaluation of different rootstocks and grafting techniques on graft union percent, yield and yield components of watermelon cv. ‘Crimson Sweet’. World Appl Sci J. 18(5):645–651.
Kombo MD, Sari N. 2019. Rootstock effects on seed yield and quality in watermelon. Hortic Environ Biotechnol. 60(3):303–312. https://doi.org/10.1007/s13580-019-00131-x.
Kurum R, Celik I, Eren A. 2017. Effects of rootstocks on fruit yield and some quality traits of watermelon (Citrullus lanatus). Derim. 34(2):91–98. https://doi.org/10.16882/derim.2017.283012.
Liu Q, Zhao X, Black Z, Brecht JK, Sims CA. 2016. Fruit quality of seedless watermelon cultivars grafted onto an interspecific hybrid squash rootstock. Proc Florida State Hortic Soc. 129:130–133.
Moreno B, Jacob C, Rosales M, Krarup C, Contreras S. 2016. Yield and quality of grafted watermelon grown in a field naturally infested with Fusarium Wilt. HortTechnology. 26(4):453–459. https://doi.org/10.21273/horttech.26.4.453.
Ozmen S, Kanber R, Sari N, Unlu M. 2015. The effects of deficit irrigation on nitrogen consumption, yield, and quality in drip irrigated grafted and ungrafted watermelon. J Integr Agric. 14(5):966–976. https://doi.org/10.1016/s2095-3119(14)60870-4.
Pal S, Rao ES, Hebbar SS, Sriram S, Pitchaimuthu M, Rao VK. 2020. Assessment of Fusarium wilt resistant Citrullus sp. rootstocks for yield and quality traits of grafted watermelon. Scientia Hortic. 272:109497. https://doi.org/10.1016/j.scienta.2020.109497.
Petropoulos SA, Olympios C, Ropokis A, Vlachou G, Ntatsi G, Paraskevopoulos A, Passam HC. 2014. Fruit volatiles, quality, and yield of watermelon as affected by grafting. J Agric Sci Technol. 16(4):873–885.
Pofu KM, Mashela PW, Mokgalong NM, Mafeo TP. 2011. Flowering and productivity of watermelon cultivars in inter-generic grafting on nematode-resistant Cucumis seedling rootstocks in Meloidogyne-infested fields. 10th African Crop Sci Conf, 421.
Qin YG, Yang CQ, Xia JL, He J, Ma XL, Yang CY, Zheng YX, Lin X, He ZQ, Huang Z, Yan ZS. 2014. Effects of dual/threefold rootstock grafting on the plant growth, yield and quality of watermelon. Not Bot Horti Agrobot Cluj-Napoca. 42(2):495–500. https://doi.org/10. 15835/nbha4229701.
Salem IB, Armengol J, Boughalleb-Mhamdi N. 2015. Soil fungicide application in combination with grafting for the control of monosporascus root rot and vine decline on cucurbits. Int J Curr Microbiol Appl Sci. 4(9):511–527.
Santos JS Dos, Dias RCS, Grangeiro LC, Simoes WL, Deon MD. 2016. Accumulation of nutrients and agronomic performance of grafted seedless watermelon. Pesqui Agropecu Trop. 46(3):311–320. https://doi.org/10.1590/1983-40632016v4639610.
Seymen M, Yavuz D, Ercan M, Akbulut M, Coklar H, Kurtar ES, Yavuz N, Suheri S, Turkmen O. 2021. Effect of wild watermelon rootstocks and water stress on chemical properties of watermelon fruit. Hortic Environ Biotechnol. 62(3):411–422. https://doi.org/10.1007/s13580-020-00329-4.
Suarez-Hernandez AM, Grimaldo-Juarez O, Garcia-Lopez AM, Gonzalez-Mendoza D, Huitron-Ramirez MV. 2017. Evaluation of creoles rootstocks Lagenaria siceraria on yield in grafted watermelon. Idesia. 35(1):39–44. https://doi.org/10.4067/s0718-34292017005000002.
Suarez-Hernandez AM, Vazquez-Angulo JC, Grimaldo-Juarez O, Cecena Duran C, Gonzalez-Mendoza D, Bazante-Gonzalez I, Mendoza-Gomez A. 2019. Production and quality of grafted watermelon in saline soil. Hortic Bras. 37(2):215–220. https://doi.org/10.1590/s0102-053620190213.
Suchoff DH, Schultheis JR, Gunter CC, Hassell RL, Louws FJ. 2019. Effect of rootstock and nitrogen fertilizer on growth and yield in watermelon. J Hortic Sci Biotechnol. 94(6):798–804. https://doi.org/10.1080/14620316.2019.1624629.
Thies JA, Ariss JJ, Hassell RL, Buckner S, Levi A. 2015. Accessions of Citrullus lanatus var. citroides are valuable rootstocks for grafted watermelon in fields infested with Root-knot Nematodes. HortScience. 50(1):4–8. https://doi.org/10.21273/hortsci.50.1.4.
Thies JA, Ariss JJ, Hassell RL, Levi A. 2012. Resistant rootstocks for managing Root-Knot Nematodes (Meloidogyne incognita) in grafted watermelon and melon, p. 202–211. Cucurbitaceae 2012: Proc Xth EUCARPIA Meeting Genetics Breeding Cucurbits.
Thies JA, Buckner S, Horry M, Hassell R, Levi A. 2015. Influence of Citrullus lanatus var. citroides rootstocks and their F1 hybrids on yield and response to Root-knot Nematode, Meloidogyne incognita, in grafted watermelon. HortScience. 50(1):9–12. https://doi.org/10.21273/hortsci.50.1.9.
Toporek SM. 2020. Characterizing the economically important species of Pythium affecting Cucurbits in South Carolina and grafting to control Pythium disease in watermelon (MS Thesis). Clemson University, Clemson, SC, USA.
Turhan A, Ozmen N, Kuscu H, Serbeci MS, Seniz V. 2012. Influence of rootstocks on yield and fruit characteristics and quality of watermelon. Hortic Environ Biotechnol. 53(4):336–341. https://doi.org/10.1007/s13580-012-0034-2.
Villocino SB Jr , Quevedo MA. 2015. Effects of biochar on physicochemical and sensory quality of watermelon (Citrullus lanatus Thunb.) fruit from grafted and non-grafted plants. Acta Hortic. 1088:481–484. https://doi.org/10.17660/ActaHortic.2015.1088.87.
Wimer J, Inglis D, Miles C. 2015. Evaluating grafted watermelon for Verticillium Wilt severity, yield, and fruit quality in Washington State. HortScience. 50(9):1332–1337. https://doi.org/10.21273/hortsci.50.9.1332.
Yang PM, He ST, Jiang LN, Chen XJ, Li YF, Zhou JG. 2020. The effects of pumpkin rootstock on photosynthesis, fruit mass, and sucrose content of different ploidy watermelon (Citrullus lanatus). Photosynthetica. 58(5):1150–1159. https://doi.org/10.32615/ps.2020. 068.
Yetisir H, Ozdemir AE, Aras V, Candir E, Aslan O. 2013. Rootstocks effect on plant nutrition concentration in different organ of grafted watermelon. Agric Sci. 4(5):230–237. https://doi.org/10.4236/as.2013.45033.
Yetisir H, Sari N. 2019. Effects of bottle gourd (Lagenaria siceraria) rootstocks on plant nutrient content of watermelon Citrullus lanatus (Thunb.) Mats. & Nak. leaf and nitrogen use efficiency. Derim. 36(1):24–32. https://doi.org/10.16882/derim.2019.443162.
Zaaroor-Presman M, Alkalai-Tuvia S, Chalupowicz D, Beniches M, Gamliel A, Fallik E. 2020. Watermelon rootstock/scion relationships and the effects of fruit-thinning and stem-pruning on yield and postharvest fruit quality. Agriculture. 10(9):366. https://doi.org/10.3390/agriculture10090366.
Zhang J, Wang PC, Tian HM, Wang Y, Jiang HK. 2019. Using a new hybrid rootstock significantly increases the grafted plant rate and watermelon yield. Int Agrophys. 33(1):97–106. https://doi.org/10.31545/intagr/104402.
Zhao X, Liu QR, Sanchez MT, Dufault NS. 2018. Performance of grafted seedless watermelon plants with and without root excision under inoculation with Fusarium oxysporum f. sp niveum Race 2. HortScience. 53(9):1340–1346. https://doi.org/10.21273/HORTSCI12842-18.