Grafting Improves Growth and Nitrogen-use E ciency by Enhancing NO3- Uptake, Photosynthesis, and Gene Expression of Nitrate Transporters and Nitrogen Metabolizing Enzymes in Watermelon Under Low Nitrogen


 Background Excessive and insufficient application of N fertilizer can inhibit plant growth, reduce N-use efficiency (NUE) and lead to production reduction. Watermelon is an important crop that often restricted by inappropriate N supply. The study aims to test whether grafting with bottle gourd rootstock can improve NUE and growth performance of watermelon under low N and to clarify the underlying physiological mechanism. Methods Grafted (self-grafted and rootstock-grafted watermelon) and ungrafted (watermelon and bottle gourd) seedlings were tested respectively, and treated with 9 mM (normal condition) and 4 mM (low N condition) NO3--N concentrations for 18 days under hydroponic conditions.Results The growth and NUE of bottle gourd rootstock-grafted watermelon seedlings increased under low-nitrate, while decreased slightly in self-grafted seedlings compared with control. Rootstock-grafted plants had higher root morphological characteristics, NO3- uptake, photosynthesis, and NUE traits than self-grafted plants under low-nitrate. The expression of nitrate transporter genes NRT1.5 and NRT2.1, and N metabolizing enzyme genes NR, NiR, GS1 and GS2 of rootstock-grafted plants were significantly up-regulated at low-nitrate treatment, which may lead to increased NO3- uptake and metabolism.Conclusion The bottle gourd rootstock grafting can improve the growth and NUE of watermelon seedlings under low-nitrate treatment. The improved plant performance is attributed to the vigorous root systems and higher NO3- uptake of rootstock roots, and the enhanced N metabolism and photosynthetic capacity of scion leaves. Grafting with the bottle gourd rootstock may be beneficial to the efficient production of watermelon and economic application of N fertilizer.


Introduction
Grafting is a widely used biotechnological tool in vegetable crops and is considered to be a rapid and effective method to protect plants against the effects of soil-borne pathogens (Rahman et al. 2021), and enhance plant's resistance to various abiotic stresses, such as high temperature (De Oliveira et al. 2020), drought (Luo et al. 2020), salinity (Lo'ay and Abo E-Ezz 2021). In addtion, rootstock grafting can also improve the absorption of potassium (K) and magnesium (Mg) by plants (Huang et al. 2013(Huang et al. , 2016, and rootstock is the decisive factor for grafted plants to absorb mineral nutrients (Huang et al. 2016;Nawaz et al. 2017;Zhen et al. 2010). Thus, it is valuable to discuss the genotype variation of rootstock effect for understanding the grafting e cacy (particularly in nutrient uptake and transport).
Watermelon [Citrullus lanatus (Thunb.) Matsum. and Nakai.] is one of the most important Cucurbitaceous vegetables and the most popular fruit in summer. As its pulp is rich in many nutrients bene cial to human health, such as water, protein, sugar, minerals, vitamin C and various amino acids. Watermelon growth requires a large amount of fertilizer, so fertilization is an important means to improve its yield. However, in recent years, the blind application of a large amount of N-fertilizer in protected watermelon cultivation in China has seriously limited the plant growth and reduced the NUE and yield of watermelon.
Although several reports suggest that rootstock grafting can alleviate growth inhibition caused by abiotic stresses though improving plant N uptake and utilization (Yang et  about gene expression and related underlying mechanisms, particularly under low N condition. Moreover, rootstocks with high N acquisition capacity may enable grafted-plants grow better with less N-fertilizer. However, this possibility needs to be further con rmed. Therefore, the purpose of this study is to investigate 1) whether bottle gourd rootstock grafting can improve the growth and NUE of watermelon seedlings under low N conditions and 2) whether the enhanced watermelon performance is attributed to higher root NO 3 − uptake and the improvement of photosynthesis and N metabolism capacity.

Materials And Methods
Plant materials and experimental treatments Experiment 1 response of grafted watermelon to low-nitrate The experiment was performed from September to November 2020 in a research greenhouse of Northwest Agricultural and perlite; after 4 days, seeds of scion 'Zaojia 8424' were sown. When the seedlings of rootstock had developed one true leaf, scion was grafted onto the two rootstocks, using the "insertion grafting" procedure as described by Hassell et al. (2008). In order to maintain a relatively high temperature and humidity environment, seedlings were moved to a small arch shed covered with a layer of transparent plastic lm and completely shaded for 72 hours. The plastic lm was removed for a short time during initial days to control relative humidity, and it was completely removed after 7 days of grafting. The plants were irrigated with tap water before grafting and transplanting to hydroponic cultivation, and no nutrient solution was used during this period. The self-grafted 'Zaojia 8424' plants were used as control. When the two true leaves emerged, the grafted plants were transplanted into 30  In this experiment, four treatments were composed of two grafting combinations and two N application levels. All treatments were replicated three times with 12 plants in each replicate, and were conducted using a randomized complete block design. The nutrient solution was regularly renewed every 5 days to avoid the de ciency of any speci c ions and the pH of the solution was maintained at 6.0 ± 0.5. The nutrient solution was aerated with air pumps at 2-hour intervals for 1 hour each time. The temperature of the greenhouse varied from 16 to 30℃, and the relative humidity varied from 50 to 85%. The plants were harvested at 18 days following N application.

Measurement of plant growth
Five randomly selected plants per treatment were collected to measure plant height (from stem base to growing point) and stem diameter (at 2 cm above the graft union). The grafted plants were divided into roots, stems and leaves. For ungrafted plants, the part above the cotyledon node was regarded as the "shoot", and the part below was the "root". These organs were rinsed with deionized water, blotted carefully with tissue paper, weighed for fresh weight and then placed in an oven at 105℃ for 1 hour followed by 80℃ for 48 hours to determine their dry weight. The sound seedling index and root to shoot (R/S) ratio were subsequently calculated as follow: sound seedling index = (stem diameter / plant height + root dry weight / shoot dry weight)×whole dry weight R/S ratio = root dry weight / shoot dry weight

Analysis of root morphology
Roots of ve uniform plants from each treatment were randomly selected and washed with deionized water, then placed in a dedicated tray. The entangled roots were gently separated by hand and stretched at. The root scanning was performed by using Imagery Scan Screen (Epson perfection V700 Photo, Indonesia). Total root length, root surface area, root diameter, root volume, root tips, and root forks were measured using image analysis through WinRHIZO Pro 2012a software.

Measurement of NO 3 − content and NO 3 − uptake
The NO 3 − content was measured by the colorimetric method utilizing salicylic acid according to Wang et al. (2018) and was calculated from the standard curve. Then NO 3 − uptake was calculated as the product of NO 3 − content and fresh weight.

Measurement of NUE traits
The N content was measured using the Kjeldahl method described by Bremner (1965). N accumulation (NA) was obtained as the product of N content and plant dry weight (mg N) (Lawlor 2002

Determination of leaf photosynthesis
The rst fully expanded true leaf from top was selected. The chlorophyll content was determined by 80% acetone extraction method according to Hussain et al. (2019). The net photosynthetic rate (Pn) and transpiration rate (Tr) were measured with a 6800 photosynthesizer apparatus from 9:00 am to 11:00 am on a sunny day.
Total RNA extraction and gene expression analysis The root tips and leaves of grafted plants were selected to determine the relative expression of genes related to nitrate transporters (NRTs) and N metabolizing enzymes, respectively. Total RNA was extracted by using the Omega kit (Norcross, Georgia, USA) according to the manufacturer's instructions and treated with RNase-free DNase to remove contaminated DNA.
First-strand cDNA was synthesized using M-MLV reverse transcriptase, and oligo-(dT) 18  quanti ed by the ABI Step One Plus real time PCR detection system and the data were analyzed by using 2 −ΔΔCt method (Livak and Schmittgen. 2001). Self-grafted and rootstock-grafted watermelon seedlings under 9 mM N treatment were used as control to calculate the relative gene expression.

Statistical analysis
A two-factorial ANOVA was performed to examine the effects of graft combination, N treatment, and their interaction on the plant samples. Signi cance levels were determined at *P<0.05, **P<0.01, and ***P<0.001; ns denoted non-signi cant differences. Tukey HSD (p<0.05) was used for the mean separation. Pearson' s correlation analysis was used to analyze the correlation between NUE and other parameters of the self-grafted and rootstock-grafted watermelon seedlings under 9 mM and 4 mM NO 3 − -N conditions, and the graphical presentation was carried out using OriginPro 2021. All analyses were conducted using SPSS 24.0 software package.

Plant growth
For grafted plants, graft combination signi cantly affected the plant growth, and its interaction with N treatment signi cantly affected the plant height, shoot and whole dry weight ( Table 2). The growth performance of bottle gourd rootstock-grafted (Z/J) plants were better than that of self-grafted (Z/Z) plants ( Fig. 1a), correspondingly, the plant height, shoot, root, and whole dry weight, and sound seedling index of Z/J plants were all signi cantly higher than those of Z/Z plants regardless of the N level ( Table 2). Compared with 9 mM N treatment, these growth parameters of Z/Z plants decreased slightly under 4 mM N treatment, while increased in Z/J plants. For example, 4 mM N treatment reduced plant height, root, shoot, and whole dry weight, and sound seedling index of Z/Z plants by 22%, 14%, 27%, 25% and 17%, respectively, whereas increased the plant height, shoot and whole dry weight, and sound seedling index of Z/J plants by 22%, 28%, 23% and 2%, respectively.

Root morphology
For grafted plants, the root morphology traits were signi cantly affected by graft combination (Table 3). Between the two graft combinations, the root system of Z/J plants was larger than that of Z/Z plants ( Fig. 1b), consistently, the total root length, root surface area, root volume, and root forks of Z/J plants were all signi cantly higher than those of Z/Z plants regardless of the N level (Table 3). Compared with 9 mM N treatment, these parameters of Z/Z plants decreased slightly under 4 mM N treatment, but increased in Z/J plants. For example, 4 mM N treatment reduced the total root length, root surface area, root volume, and root forks of Z/Z plants by 11%, 15%, 16%, and 22%, respectively, whereas increased the total root length and root forks of Z/J plants by 4% and 11%, respectively. For ungrafted plants, N treatment affected the total root length, while rootstock genotype and their interaction signi cantly affected all the root morphology traits (except that the interaction did not affect the root volume) (Fig. 4). The root system of J plants was signi cantly stronger than that of Z plants under 9 mM and 4 mM N levels (Fig. 2b). The total root length, root surface area, root volume, and root forks of J plants were about 1.5, 1.7, 2.0, and 2.0 times of Z plants under 9 mM N treatment, respectively, and 1.9, 2.1, 2.4, and 2.5 times under 4 mM N treatment, respectively (Fig. 4).

NO 3 − content and NO 3 − uptake
For grafted plants, the total NO 3 − content was signi cantly affected by graft combination and N treatment (Fig. 5a), and the whole plant NO 3 − uptake was signi cantly affected by graft combination and its interaction with N treatment (Fig. 5b) Gene expression of nitrate transporters and N metabolizing enzymes Relative expression of nitrate transporters and N metabolizing enzymes genes was conducted in self-grafted and rootstockgrafted watermelon plants. According to the Cucurbitaceae Genome database, we identi ed two nitrate transport genes in watermelon and bottle gourd roots, respectively, and four N metabolizing enzymes genes in watermelon leaves ( Table 1). The relative expression of these genes was signi cantly affected by N treatment, graft combination and their interaction (Fig. 7).
The mRNA levels of NRT1.5 and NRT2.1 in the roots, and NR, NiR, GS1 and GS2 genes in the leaves of Z/J plants were upregulated signi cantly under 4 mM N treatment compared to the 9 mM N treatment. However, these genes in Z/Z plants did not change signi cantly under 4 mM N treatment, except for the up-regulated expression of NiR and GS2 genes. Moreover, the gene expression of nitrate transporters and N metabolizing enzymes in rootstock-grafted watermelon seedlings was signi cantly higher than that in self-grafted plants. For example, under 4 mM N treatment, the relative expression of NRT1.5, NRT2.1, NiR, GS1 and GS2 genes in Z/J plants were 27.3, 2.9, 2.6, 1.8 and 2.5 times higher than that in Z/Z plants, respectively.

NA, NUpE, NUtE and NUE
The NA, NUpE, NUtE and NUE were all signi cantly affected by N treatment, graft combination and their interaction (except that the N treatment did not affect NUtE) (Figs. 8 and 9). Obviously, the NA in roots, stems and leaves of Z/J plants was remarkablely higher than that of Z/Z plants, and increased signi cantly under 4 mM N treatment, especially in leaves (Fig. 8).
In addition, compared with 9 mM N treatment, the NUpE, NUtE and NUE of Z/J plants were increased signi cantly under 4 mM N treatment (Fig. 9). These NUE traits were all signi cantly higher than those of Z

Correlation analysis
Pearson's correlation analysis between the NUE and other parameters of the self-grafted and rootstock-grafted watermelon seedlings under 9 mM and 4 mM NO 3 − -N conditions was conducted (Fig. 11). The NUE was found to be positively correlated with the plant height, shoot dry weight, whole plant dry weight, NA, NUtE, Pn, Tr, and chlorophyll content; and all correlations between these parameters above were positive. NUtE was also positively correlated with the root dry weight, sound seedling index, total root length, root surface area, root volume, and whole plant NO 3 − uptake; meantime, all these parameters were positively correlated with each other, and they were also positively correlated with the plant height, shoot dry weight, whole plant dry weight, NA and Tr (except that RV has no correlation with PH). In addition, the whole plant NO 3 − uptake was also positively correlated with Pn and chlorophyll content. The NUpE was positively correlated with the relative expression of NRT1.5, NRT2.1 and GS1 genes, and the relative expression of NRT1.5 and NRT2.1 genes, as well as the NiR, GS1 and GS2 genes were all positively correlated with each other.

Discussion
Bottle gourd rootstock grafting increases watermelon growth performance under low-nitrate condition Appropriate rootstock grafting can promote plant growth and improve plant tolerance to nutrient de ciency (Huang et al. 2013(Huang et al. , 2016. In this study, the growth performance of bottle gourd rootstock-grafted watermelon seedlings was obviously better than that of self-grafted seedlings, especially under low-nitrate condition (Fig. 1a, Table 2). The analysis of variance showed that graft combination signi cantly affected plant growth, and the growth performance of ungrafted bottle gourd seedlings was also signi cantly better than that of ungrafted watermelon seedlings, regardless of the N level ( This difference implies that suitable rootstocks should be carefully selected according to different nutrient de ciency environments. Moreover, we found that under 4 mM low-nitrate treatment, the growth of self-grafted watermelon plants slightly inhibited, but increased in rootstock-grafted plants relative to the 9 mM nitrate treatment. We suggest that bottle gourd rootstock grafting is expected to reduce N fertilizer application without compromising the plant growth of watermelon seedlings. Bottle gourd rootstock-grafted watermelon plants have better root advantages in response to low-nitrate treatment Similarly, the results of our experiment showed that bottle gourd rootstock-grafted watermelon seedlings had larger root system than self-grafted seedlings, regardless of the N level (Fig. 1b, Table 3). The analysis of variance showed that graft combination signi cantly affected root morphological traits, and these parameters in ungrafted bottle gourd seedlings were signi cantly higher than those of ungrafted watermelon seedlings (Table 3, Figs. 2b and 4). These results indicated that rootstock-grafted watermelon plants had greater root advantage than self-grafted plants, which bene ted from the genetic characteristics of bottle gourd rootstock. This may partly explain the superior growth of rootstock-grafted watermelon plants, since root morphological characteristics determine a plant's ability to acquire N which directly affecting the growth and development of plants (Jiang et al. 2017). The results of correlation analysis also showed that the growth parameters (including sound seedling index, root dry weight and whole dry weight) were positively correlated with root morphological traits (including total root length, root surface area, and root volume) (Fig. 11).
In addition, numerous studies showed that plant roots can signal the plant to alter the root system in response to various levels of N availability, so as to meet their own demand for N fertilizer (Pacheco-Villalobos and Hardtke 2012; Iqbal et al. 2020). In present study, the results showed that at 4 mM low-nitrate level, the total root length and root forks of bottle gourd rootstock-grafted seedlings and the total root length of ungrafted bottle gourd plants were increased, whereas these parameters of self-grafted seedlings decreased slightly (Table 3, Fig. 4a). The results showed that the bottle gourd rootstock could better adapt to low N through morphological variation. In this study, we found that regardless of the N level, the total NO 3 − content and NO 3 − uptake of rootstock-grafted watermelon seedlings were signi cantly higher than those of self-grafted seedlings (Fig. 5), and the NO 3 − uptake of ungrafted bottle gourd plants was also higher than that of ungrafted watermelon plants (Fig. 6), indicating that grafting with bottle gourd rootstock could improve the NO 3 − uptake of watermelon seedlings. Yang et al. (2013) also reported that bottle gourd rootstock-grafting can promote NO 3 − uptake in NaCl-stressed watermelon leaves. Genotypic variation of nitrogen utilization e ciency in oilseed rape showed that N-e cient genotype exhibited higher root N uptake than N-ine cient genotype (He et al. 2021). Furthermore, compared with 9 mM nitrate treatment, 4 mM low-nitrate treatment signi cantly reduced the total NO 3 − content and NO 3 − uptake of self-grafted plants, whereas increased the total NO 3 − uptake of rootstockgrafted plants, especially increased the NO 3 − content and NO 3 − uptake in the leaves of rootstock-grafted plants (Fig. 5). This result shows that bottle gourd rootstock grafting can improve the NO 3 − uptake and root-to-shoot transport of watermelon seedlings under low-nitrate condition, which is in line with the results of Savvas et al. (2017), who found that nutrient uptake and translocation to the shoot were often improved in favourable grafting combinations. Correlation analysis showed that whole plant NO 3 − uptake was positively correlated with certain root morphological traits (Fig. 11), implying that the vigorous root systems of bottle gourd rootstock play an essential role in promoting the NO 3 − absorption and translocation of grafted watermelon seedlings. Martínez-Ballesta et al. (2010) reported that owing to the vigor of rootstocks, grafted plants usually show an increased uptake of water and minerals compared with self-rooted plants under favorable growth conditions. In addition, the total NO 3 − uptake of rootstock-grafted watermelon seedlings increased under 4 mM low-nitrate treatment but decreased in ungrafted bottle gourd seedlings, especially in the leaves (Figs. 5b and 6b). This may be attributed to the scion/rootstock interaction, which can be regulated by root-shoot-root long-distance signaling (Sasaki et al. 2014).
It is well known that the absorption of NO mediates root high-a nity NO 3 − in ux of Arabidopsis roots and affects the transport of nitrate nitrogen to leaves (Li et al. 2007). In addition, NRT2 can regulate lateral root initiation in nitrate signaling pathway. In addition, NRT2.1 plays a direct stimulating role in the speci c step of lateral root development, and can coordinate root development and external NO 3 − availability (Remans et al. 2006). In this study, the response of NRT gene expression in roots of self-grafted and rootstockgrafted watermelon plants to low nitrogen was quite different, the expression of NRT1.5 and NRT2.1 genes in roots of rootstock-grafted plants was signi cantly up-regulated under low-nitrate treatment, but there was no change in self-grafted plants ( Fig. 7a and b). The bottle gourd rootstock-grafted plants had a higher ability to absorb NO 3 − than self-grafted plants.
Therefore, the different expression levels of NO 3 − transporter gene may be part of the reason for the differences of NO 3 − absorption between self-grafted and rootstock-grafted watermelon plants, the NRT genes (NRT1.5 and NRT2.1) may related in the higher NO 3 − uptake by rootstock-grafted plant roots. However, other NO 3 − transporter genes associated with high uptake of NO 3 − in rootstock-grafted plant roots need to be further studied.
N metabolism ability of watermelon was enhanced by grafting onto bottle gourd rootstock under low-nitrate condition The results showed that NR, NiR, GS1 and GS2 genes in the leaves of bottle gourd rootstock-grafted plants were signi cantly up-regulated under 4 mM low-nitrate treatment relative to the 9 mM nitrate treatment, and were generally higher than those of self-grafted plants (Fig. 7c-f). The results proved that under low-nitrate condition, bottle gourd rootstock grafting could enhance N metabolism ability of watermelon seedlings by regulating the transcriptional expression of N metabolizing enzyme genes. The correlation analysis showed that the relative expression of NRT1.5 and NRT2.1 genes, as well as the NiR, GS1 and GS2 genes were all positively correlated with each other (Fig. 11), implying that the NO 3 − metabolism in rootstock-grafted watermelon leaves was closely related to its absorption by roots and root-to-shoot transport.
Bottle gourd rootstock grafted plants have higher NUE and photosynthetic capacity under low-nitrate condition After assimilation and metabolism, NO 3 − is gradually become incorporated into organic compounds, such as amino acids, proteins, and other N compounds necessary for plant growth and development (Pratelli and Pilot 2014). Consistent with the change of plant NO 3 − uptake, the N accumulation in roots, stems and leaves of bottle gourd rootstock-grafted plants was signi cantly higher than that of self-grafted plants, and increased under low-nitrate treatment, especially in the leaves (Fig. 8).
The results indicated that watermelon grafted onto bottle gourd rootstock can not only absorb more N, but also transfer more N to leaves, so as to maintain high N accumulation and nally promote growth, which is agree with the view of Savvas et al.
(2017). The correlation analysis also showed that there was a signi cant positive correlation between NO 3 − uptake, N accumulation and plant biomass (Fig. 11).
The ability of plants to absorb and utilization nitrogen can be evaluated according to two representative indexes: NUpE and NUtE. The NUpE and NUtE further jointly determine the overall NUE, and this role is in uenced by the N supply (Garnett et al. 2015). Increasing the NUE is important to maintain a high productivity under low N supply (Xu et al. 2012 (Fig. 9). These results suggest the 4 mM nitrate concentrations in this experiment provided an appropriate level of N fertilization for bottle gourd rootstock-grafted watermelon seedlings, and appropriate N de ciency supply will not cause stress inhibition to the growth of watermelon seedlings, but may help to improve the NUE and promote the growth of watermelon seedlings ).
In addition, photosynthesis is the main source of plant organic matter and participates in the synthesis of N-containing substances in chloroplasts. In turn, photosynthesis is also very sensitive to the change of N availability Xu et al. 2015), because 57% of the N in the leaves is located in the chloroplasts, which is used to synthesize photosynthetic components and related enzymes (Xu et al. 2012). Therefore, there is a close coupling relationship exists between photosynthesis and N metabolism. In this study, we found that the Pn, Tr and chlorophyll contents of rootstock-grafted plants were all higher than those of self-grafted plants, and increased under 4 mM low-nitrate treatment, but decreased slightly in selfgrafted plants (Fig. 10). The correlation analysis showed that the photosynthetic characteristics were positively correlated with N accumulation and NUtE (Fig. 11). The results indicated that watermelon grafted onto gourd rootstock can accumulate more N to enhance photosynthetic capacity. Furthermore, photosynthesis is related to the dry matter accumulation of plants, which is the basis for increasing crop growth, productivity, and NUE (Huang et al. 2016;Iqbal et al. 2020). The correlation analysis in this study proved that the photosynthetic characteristics were signi cant positive correlated with NUE and plant biomass (including plant height, shoot dry weight and whole dry weight) (Fig. 11).
Overall, correlation analysis showed that there were inextricably linked positive correlations among plant growth, root morphology traits, plant NO 3 − uptake, photosynthetic characteristics, NUtE and NUE (Fig. 11). Combined with the changes of the above parameters in the text, we infer that grafting with bottle gourd rootstock may increase plant growth and NUE of watermelon seedlings by making better use of their developed roots and enhancing N metabolism potential and photosynthetic capacity to promote NO 3 − absorption, assimilation and metabolism. The signi cant positive correlation between the gene expression of nitrate transporters and N metabolizing enzymes further revealed that the enhanced NO 3 − uptake and N metabolism of rootstock-grafted plants were regulated at the transcriptional level. The results of this study provide key insights into how NUE could be improved under reduced N fertilizer application without compromising plant growth in the context of watermelon seedlings growth. As for mature plants, could be further studied in later experiments.
Moreover, most previous studies have showed that there is chemical signaling from the root to the shoot, which plays an important role in regulating a plant's morphology and physiology (Tsutsui and otaguchi 2017). Therefore, it is necessary to further study the signal regulation pathway of grafting to promote watermelon plant growth and nitrogen use e ciency.

Conclusions
In conclusion, bottle gourd rootstock grafting can effectively promote the plant growth and NUE of watermelon seedlings, and the rootstock characteristics played a crucial role. The physiological mechanism of the improved performance of grafted watermelon under low N is related to the vigorous root systems and improved NO 3 − absorption capacity bene ted from rootstock, and enhanced N metabolism potential in scion leaves. Signi cant up-regulation of nitrate transporter genes (NRT1.5 and NRT2.1) and N metabolizing enzyme genes (NR, NiR, GS1 and GS2) may be partially responsible for the improvements in    Pearson's correlation analysis between the NUE and the other parameters of the self-grafted and rootstock-grafted watermelon seedlings under 9 mM and 4 mM N conditions. *, ** and ***; Correlation is signi cant at the 0.05, 0.01 and 0.001 levels, respectively. NUE, nitrogen-use e ciency; SSI, sound seedling index; PH, plant height; rDW, root dry weight; sDW, shoot dry weight; wDW, whole plant dry weight; RL, total root length; RS, root surface area; RV, root volume; RF, root forks; NA, N accumulation; NUpE, nitrogen-uptake e ciency; NUtE, nitrogen-utilizition e ciency; NO 3 -, whole plant NO 3 uptake; Pn, net photosynthetic rate; Tr, transpiration rate; Chl, chlorophyll content; NRT1.5 and NRT2.1, nitrate transporters genes; NR, nitrate reductase gene; NiR, nitrite reductase gene; GS1 and GS2, glutamine synthetase genes