Abstract
A seedling population from hybrids between ‘Volkamer’ lemon (Citrus volkameriana) and ‘Rubidoux’ trifoliate orange (Poncirus trifoliata) was grafted with ‘Hashimoto’ Satsuma mandarin (Citrus unshiu) to study the inheritance of rootstock effects on salt tolerance in terms of fruit yield. Trees were maintained in a screenhouse, and a salt treatment (25 mm NaCl) was applied to 32 genotypes from June to September every year for 5 years. Rootstocks were genotyped for five salt tolerance candidate genes. Significant effects of rootstock genotype (G) and treatment (E) were found for most traits. Salinity decreased yield and juice volume but improved soluble solids concentration (TSS) and rind thickness. Year effects were highly significant in most cases. G × E interactions were found for fruit weight, total fruit weight, juice volume (JV), leaf water content (LWC), and leaf [Na+]. Therefore, rootstocks that induce early fruit maturation under salinity (by increasing TSS and maintaining JV) can be selected to expand the harvesting calendar of mandarin cultivars. Salt tolerance candidate genes SOS1 and NHX1 were associated with fruit yield traits under normal conditions (1.4 dS·m−1), and SOS1 and CCC were associated with LWC under salinity conditions (4 dS·m−1). Only 5% progeny induced higher accumulated yield than ‘Volkamer’ lemon under salinity. Given the low heritability of rootstock effects on fruit yield under salinity conditions (0.18 at most), marker-assisted selection might be useful.
More than 800 million hectares of land throughout the world are affected by salinity (Food and Agriculture Organization of the United Nations, 2008), which can decrease yield and lead to increased poverty and reliance on imports (Witcombe et al., 2008). Citrus species are among the most salt-sensitive of horticultural crops, although differences in tolerance exist among species in these genera (Maas, 1993). Tree growth and fruit yield of citrus species are impaired at a soil salinity of ≈2 dS·m−1 soil saturation without the concomitant expression of leaf symptoms (Bingham et al., 1974; Cerdá et al., 1990). Several lines of evidence indicate that citrus species are predominantly chloride-sensitive (Bañuls et al., 1997; Cooper, 1961; Ruiz et al., 1999; Walker et al., 1982). Cooper et al. (1956) classified citrus rootstocks into three groups: 1) good salt tolerance (‘Cleopatra’ mandarin Citrus reshni); 2) medium salt tolerance [‘Volkamer’ lemon and sour orange (Citrus aurantium)]; and 3) poor salt tolerance [trifoliate orange and ‘Troyer’ or ‘Carrizo’ citrange (Citrus sinensis × Poncirus trifoliata)]. This classification agrees with the order of these genotypes according to their ability to restrict Cl− transport to the scion on grafted trees under field conditions (irrigation with saline) and in an experiment involving the culture of Satsuma mandarin (Citrus unshiu) rootstock combinations, which evaluated salt tolerance according to fruit and leaf characteristics (García et al., 2002). Chloride concentrations in the leaves of seedlings under salinity conditions have been used by most rootstock breeders to select salt-tolerant genotypes (Sykes, 1985). However, there are many examples of contradictory results related to the ability to exclude Cl− (Grieve and Walker, 1983); in most cases, it is not possible to determine whether experimental or genetic differences are responsible for these results. Thus, Sykes (2011) showed the presence of within-species diversity for chloride and sodium excluding capacities recommending progeny testing to investigate their inheritance for future rootstock breeding. Quantitative trait loci (QTL) analysis of these traits has been carried out in citrus seedlings (Tozlu et al., 1999).
Salt tolerance is a quantitative trait in plants. The identification of QTL that control salinity tolerance is of great importance for breeding salt-tolerant crops (Cuartero et al., 2006; Flowers, 2004; Koyama et al., 2001). However, QTL analysis of rootstock effects takes a long time as a result of the long juvenility period of citrus species and is very expensive because of the required time and huge cultivation costs for maintaining large segregating progenies. Besides, rootstock may affect numerous citrus attributes (Castle, 2010) making interpretation of data analysis very complicated. An alternative strategy is the analysis of candidate genes functionally involved in the salt tolerance response. These genes could serve as useful markers for early selection of rootstocks that confer salt tolerance to the grafted citrus cultivar. The regulation of ion homeostasis is one of the main strategies used by glycophytes for salinity–stress adaptation (Munns and Tester, 2008). It involves a network of transport processes that regulate the uptake, extrusion through the plasma membrane, compartmentalization of salts into cell vacuoles, and recirculation of ions through plant organs. This facilitates plant osmotic adjustment and maintenance of high K+/Na+ ratios in the cytosol (Apse and Blumwald, 2007; Pardo and Rubio, 2011). Genes coding for cation transporters such as HKT (high-affinity K+ transporter), SOS1 (salt overly sensitive), and NHX family (Na+/H+ exchangers) are considered candidate genes for salt tolerance because they regulate the internal concentrations of Na+ in various tissues and indirectly regulate K+ homeostasis (Munns and Tester, 2008; Yamaguchi et al., 2013). Porat et al. (2002) isolated NHX1 from citrus and found that it was markedly induced by salt stress in a salt-tolerant cell line form ‘Shamouti’ sweet orange (C. sinensis). Brumós et al. (2009) found that a cation-chloride cotransporter (AtCCC ortholog) was induced in ‘Carrizo’ citrange but not in ‘Cleopatra’ mandarin seedlings under salt treatment.
Following Shannon and Grieve (1999), salt tolerance values should take into consideration the portion of the plant to be marketed. Salinity effect on fruit yield has been reported almost exclusively by studies that used a single-scion/single-rootstock combination (reviewed by Grieve et al., 2007; Prior et al., 2007; Storey and Walker, 1999) or one cultivar grafted onto two genetically unrelated rootstocks (García-Sánchez et al., 2003, 2006; Navarro et al., 2010; Syvertsen et al., 2010). These studies used ‘Cleopatra’ mandarin and ‘Carrizo’ or ‘Troyer’ citrange. Therefore, the salinity response has been extensively studied in these rootstocks; however, the inheritance of the rootstock-mediated salt tolerance has been hardly approached in citrus. We report an inheritance study of salt tolerance in terms of mandarin fruit yield conferred by a segregating population of rootstocks derived from a medium salt-tolerant C. volkameriana and a salt-sensitive P. trifoliata cultivar. As possible tools for selection in future rootstock breeding programs, trait correlations and the association of candidate genes (primarily those involved in Na+ and Cl− homeostasis) with traits involved in salt tolerance were also investigated.
Materials and Methods
Plant materials and growth conditions.
A seedling population was developed by open pollination from an intergenic hybrid between ‘Volkamer’ lemon [V (female parent)] and ‘Rubidoux’ trifoliate orange [P (male parent)] obtained by J.B. Forner [Instituto Valenciano de Investigaciones Agrarias (IVIA), Valencia, Spain]. Twenty-five random seedlings derived from each individual V × P fruit-yielding tree were genotyped for five isozymatic loci (García et al., 1999) to infer their genetic origin (apomictic or sexual; sexual by cross-pollination or self-pollination). Seedlings derived from cross-pollination were discarded. The seedling population, and 10 seedlings each from two control genotypes (‘Volkamer’ lemon and ‘Flying Dragon’ trifoliate orange), were grafted with the ‘Hashimoto’ Satsuma mandarin in 1996.
Eighteen genotypes (13 derived from apomictic V × P hybrids) were grown as control treatment (no NaCl added), and 32 genotypes (20 derived from apomictic V × P hybrids) were grown under salinity conditions. Thirteen genotypes derived from V × P apomictic hybrids (i.e., those yielding seedlings that are genetically identical to the mother hybrid) were present in both treatments. The number of repetitions (nucellar seedlings) from apomictic hybrids varied between two and 25. Eight-year-old trees (163 total) were grown individually in 17-L pots with a sterilized substrate mix (50% peat, 30% coconut fiber, 15% sand, and 5% perlite) and placed in a screenhouse at IVIA according to a completely randomized design. Salt treatments were applied for 5 years (2004−08) between June and September with three irrigations per week of 500 mL per plant. The NaCl concentration in the irrigant was gradually increased during 2 weeks to 25 mm [4 dS·m−1; pH 7.85; conductivity meter (HI9033 multirange; Hanna Instruments, Smithfield, RI)]. Plants in control conditions received the same irrigation regime using tap water (1.4 dS·m−1, pH 7.43; 106 mg·L−1 Cl–, 158 mg·L−1 SO42–, 250 mg·L−1 CO3H–, 115 mg·L−1 NO3–, 138 mg·L−1 Ca2+, 43 mg·L−1 Mg2+, 62 mg·L−1 Na+, 2 mg·L−1 K+). Both control and salinity-treated plants received the same pest, disease, and weed control, fertilization, and pruning work. Fertilizer was proportioned automatically by mixing in a 1:100 proportion the stock solution A (200 mm NH4H2PO4) and solution B. The stock solution B contained 12 M Ca(NO3)2, 2.6 M KNO3, 36 g·L−1 iron chelate (Sequestrene 138 Fe; Syngenta, Madrid, Spain), and 5% microelement solution. This solution contained 0.3 mm copper SO4·5 H2O, 3.1 mm zinc SO4·7 H2O, 109 mm manganese Cl2·4 H2O, 92 mm BO3H3, 2 mm NH4MoO4, and 0.4 mm V2O5. A record of weather data variables including temperature and relative humidity measured under the net was obtained from the Moncada meteorological station located at IVIA (Supplemental Fig. 1).
Trait evaluation.
Four yield-related traits and 25 vegetative and physiological traits were evaluated for each plant. Eighteen of these traits were measured during the 5 experimental years under control and salinity conditions. The following fruit-yield traits were evaluated annually between October and November [‘Hashimoto’ Satsuma mandarin collection period (Soler, 1999)]: fruit number per plant (FN); mean fruit weight per plant (FW), measured as the average weight (grams) of 10 randomly sampled fruit; and total yield per plant (TFW), measured as the weight (kilograms) of all fruit produced per plant. The total accumulated production per plant (AcTFW), measured as the total fruit weight (kilograms) produced during the 5 experimental years (2004–08), was also measured.
The 10 randomly sampled fruit per tree also were evaluated every year for the following internal and external fruit-quality traits: fruit diameter [FD (millimeters)] was measured in the transversal section; rind thickness [RT (millimeters)], including flavedo and albedo, was measured in the transversal section; juice volume (milliliters) was quantified as the volume of juice without pulp; acidity (pH) was measured with a digital pH meter (PH25; Crinson, Barcelona, Spain); and total soluble solids (percent) was measured with a digital refractometer (Pallete PR-101; Atago, Tokyo, Japan).
Three fully developed leaves per plant were sampled from vegetative spring shoots to measure the following leaf characteristics after each yearly saline treatment period (June to September): final leaf fresh weight [LFWf (grams)]; final leaf dry weight [LDWf (grams)] measured in samples dried at 100 °C for 3 d; accumulated dry matter percentage (%DMA) calculated as the difference between leaf dry weight at the beginning and at the end of salt treatment [%DMA = (LDWf − LDWi) × (100/LDWi)]; final leaf water content [LWCf (grams)] calculated as the difference between LFWf and LDWf; leaf area [LA (square centimeter)] measured with a leaf area quantifier (LI-3100C area meter; LI-COR, Lincoln, NE); leaf color parameters LCL*, LCa*, and LCb* defined by Hunter coordinates (L*, a*, b*) arranged in a Cartesian system, where a* (x-axis) varies in a green/red scale (−60 to +60), b* (y-axis) varies in a blue/yellow scale (−60 to +60), and L* represents brightness in a black/white range (0 to 100) (CR-200 colorimeter; Konica Minolta, Basildon, U.K.); and leaf color parameters Cr* and Hue* defined by the cylindrical coordinate system (L*, C*, h), where Cr* (chroma) represents color intensity (0 to 60) {Cr* = √[(a*)2 + (b*)2]} and Hue* (hue angle) represents leaf color (0° to 360°) [hue* = tan−1 (b*/a*)] (HunterLab, 1996). The trait increment (for LFW, LDW, and LWC) at the end of the treatment (dTrait) also was calculated each year.
The foliar concentration (milligrams per liter) of Cl− was evaluated (Model 926 chloride analyzer; Sherwood Scientific, Cambridge, U.K.) at the end of each saline treatment period (June to September) using a random sample of three leaves per plant from the vegetative spring shoots (Gilliam, 1971). The foliar concentrations (percentage) of Na+, Ca2+, Mg2+, K+, and phosphorus were evaluated by inductively coupled plasma analysis at G.E. Cota.2 (Valencia, Spain) only for 2005.
At the end of the experiment in 2008, all plants (12-year-old trees) were pulled up, cleaned, and dissected to evaluate the following plant growth traits: trunk diameter [TrunkD (millimeters)] measured 2 cm above the graft line (using a digital caliper); total plant weight (kilograms); foliage weight (kilograms) measured above the graft line; rootstock weight (kilograms) the weight of the roots and trunk below the graft line; and the root weight [RootW (kilograms)]. Four leaf traits were measured only at the end of the experiment: leaflet fresh weight [LFWhj (grams)]; rachis fresh weight [LFWrq (grams)]; leaflet dry weight [LDWhj (grams)]; and rachis dry weight [LDWrq (grams)].
Rootstock genotyping.
Genomic DNA was extracted from the bark tissue below the graft union in each plant according to the protocol of Ruiz et al. (2000) with the following minor modifications. One piece of bark (1 × 0.5 cm) was homogenized in liquid N2, and an additional ethanol wash was performed before resuspending the DNA pellet. Polymerase chain reaction (PCR) conditions were specific for each marker, and the resulting product was analyzed by electrophoresis in sequencing-type 10% polyacrylamide gels and revealed by silver staining as described in Ruiz et al. (2000). A set of five primer pairs was used to detect sequence characterized amplified regions targeting the following salinity tolerance candidate genes, all involved in ion Na+ or Cl− homeostasis: SOS1 and SOS2 [salt overly sensitive primers were derived from those used in Villalta et al. (2008)]; NHX1 [Na+/H+ exchanger primers designed from citrus cDNA sequence AY028416 at the National Center for Biotechnology Information (Porat et al., 2002)]; CCC [cation chloride cotransporter primers designed from the citrus chloride cotransporter DNA sequence kindly provided by J.M. Colmenero-Flores (Colmenero-Flores et al., 2007)]; and Ethrec (ethylene receptor), whose primers were obtained from the cDNA sequence AF092088 (C. sinensis putative ethylene receptor mRNA). In the case of CCC, four segregating alleles were revealed after AluI digestion of PCR products, obtaining a cleavage amplified polymorphism marker. These salt tolerance candidate genes were located at the following linkage groups: 10 + 5b (SOS1 and NHX1), 8 + 6 (SOS2), 7 (CCC), and 3 (Ethrec), as reported by Bernet et al. (2010) and Raga et al. (2012). The identity of the polymorphic amplification products was previously checked by sequence analysis. A citrus primer list is available for scientific purposes only at IVIA Genetic Laboratory (Asins et al., 2009).
Statistical and genetic analyses.
The fixed effects for rootstock G, E, year (Y), and all their possible interactions [i.e., genotype per treatment (G × E), genotype per year (G × Y), treatment per year (E × Y), and genotype per treatment per year (G × E × Y)] were analyzed by a repeated measures approach using trees (within each genotype and treatment) as a random subject factor and first-order autoregressive covariance structure between measurements taken from the same tree over the years. Pearson’s correlation analyses between fruit yield parameters and all evaluated traits were studied under control and salinity conditions using the 13 apomictic hybrids included in both treatments in all years (InfoStat, 2004). Considering rootstocks as a random effects factor, broad-sense heritability (H2) was estimated for all traits evaluated in 2008 only for nucellar rootstocks (repetitions) derived from apomictic V × P hybrids under control (H2c) or salinity (H2s) conditions based on the genotypic (VG) and environmental (VE) variance estimators calculated by minimum variant quadratic unbiased estimator, as previously reported (Villalta et al., 2007).
Associations between parental alleles at candidate genes and means per treatment for traits were studied using the Kruskal-Wallis procedure in MapQTL 6 (van Ooijen, 2009). All V × P plants were considered taking into account their genotypes and nucellar or self-pollination origins. The significance level for candidate genes was fixed at P ≤ 0.05.
Results
A significant effect (P ≤ 0.05) of the rootstock genotype and treatment was found for most fruit yield and quality traits (Table 1). Salinity decreased yield and juice volume, but improved two quality traits by ≈6% (TSS and RT). Year effects were highly significant in most cases.
Probability values of rootstock effects from the three-way repeated analyses using nucellar rootstocks from 13 apomictic hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange.z
Salinity greatly reduced tree growth and accumulated fruit yield. Leaf concentrations of Cl− and Na+ increased strongly (215% and 149%, respectively). No significant rootstock effect on leaf Cl− concentration was detected in contrast to that for leaf Na+ concentration. A significant G × E interaction was found for this trait, suggesting that the rootstock effect depends on the presence of salinity (i.e., differences among rootstocks for Na+ translocation to the leaf were not predicted from control conditions). Few traits showed a significant G × E interaction (differential rootstock behavior) that did not depend on the year (no significant G × E × Y); these included FW, TFW, and LWC. There were two additional traits for which a G × E interaction was significant (AcTFW and leaf [Na+]), but these were evaluated in only 1 year. Estimates for trait heritability ranged from low to very low (Table 1). In general, plants under saline treatment had smaller estimates than those under control conditions. Exceptions to this included pH, LDW, LDWrq, LFWrq, and TrunkD.
Considering accumulative fruit yield under salinity as the most relevant criteria for selection of rootstocks conferring salt tolerance, only 5% of V × P hybrids conferred higher yield than the ‘Volkamer’ lemon parent when used as rootstock (Fig. 1B) Noteworthy, differences between control genotypes were larger under control than under salinity condition (Figs. 1A and 1B). If only the last 3 years of evaluation are considered (2006−08), and P > 0.02 is chosen as significant, then TFW is strongly correlated with the FN component (Table 2), which is quite consistently related to TSS, particularly under salinity. The FW component is related to FD, JV, and RT, and indirectly to TSS and FN, under both control and salinity conditions. Correlation coefficients for traits evaluated under control conditions were usually higher than the same traits under salinity conditions. A significant correlation between both conditions with respect to fruit traits was found only for AcTFW (0.79) and FN (0.73) in 2007, despite the FN distribution differences between both treatments (Figs. 1C and 1D).
Frequency distributions of (A–B) total accumulated fruit yield (AcTFW) and (C–F) fruit number (FN) of ‘Hashimoto’ Satsuma mandarin grafted on nucellar seedlings from hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange for the two last yielding years under control (A, C, and E) and salinity (B, D, and F) conditions. The phenotypic means of the population (M) and the control genotypes [‘Volkamer’ lemon (V) and ‘Flying Dragon’ trifoliate orange (P)] are indicated.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Frequency distributions of (A–B) total accumulated fruit yield (AcTFW) and (C–F) fruit number (FN) of ‘Hashimoto’ Satsuma mandarin grafted on nucellar seedlings from hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange for the two last yielding years under control (A, C, and E) and salinity (B, D, and F) conditions. The phenotypic means of the population (M) and the control genotypes [‘Volkamer’ lemon (V) and ‘Flying Dragon’ trifoliate orange (P)] are indicated.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Frequency distributions of (A–B) total accumulated fruit yield (AcTFW) and (C–F) fruit number (FN) of ‘Hashimoto’ Satsuma mandarin grafted on nucellar seedlings from hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange for the two last yielding years under control (A, C, and E) and salinity (B, D, and F) conditions. The phenotypic means of the population (M) and the control genotypes [‘Volkamer’ lemon (V) and ‘Flying Dragon’ trifoliate orange (P)] are indicated.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Pearson correlation coefficients and probability values of significant (P < 0.02) correlations between fruit yield traits (last 3 years) and all evaluated traits (Corr. trait) using nucellar rootstocks from 13 apomictic hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange under control and salinity (25 mm NaCl) conditions.z
Fruit yield traits did not show a consistent, significant correlation with any non-fruit (vegetative) trait. Leaf [Cl−] appeared related to FN and indirectly to FW under control and salinity conditions in 2007 and 2006, respectively, whereas leaf [Na+] was only related to TFW and FN in 2006. The leaf color hue* index also appeared related to fruit yield traits under control (2006 and 2007) and salinity (2007) conditions.
The poor consistency of correlations through all experimental years indicated the importance of the year effect. Thus, the distribution of FN (under both the control and salinity conditions) changed depending on the year, as shown in Figure 1C–E for the last 2 consecutive years (2007 and 2008).
Four genotypes, named ac, ad, bc, and bd, were observed for the hybrids derived from the cross between ‘Volkamer’ lemon (ab) and ‘Rubidoux’ trifoliate orange (cd) at each of the five candidate gene loci (Fig. 2). Similar to the results from correlation analyses, candidate genes were associated with some traits only in certain years, except for TFW, which was consistently associated with SOS1 (Fig. 2A) under control conditions (1.4 dS·m−1 tap water) for the 3 last experimental years (Table 3). NHX1 (Fig. 2B) was associated with fruit yield traits (FN and FW) only under control conditions. The same allele at NHX1 that correlated with higher FW7 (fruit weight in year 2007) and FW8 also correlated with higher LWC7. The same allele at SOS1 that correlated with higher FN6 and FN7 correlated with higher TFW in 2006, 2007, and 2008. Leaf [Cl−] was found associated with CCC (Fig. 2C) and Ethrec (Fig. 2B), but only under control conditions. For Ethrec, the same allele that correlated with a higher increment for leaf water content in control (dLWC8_C) was associated with a high leaf [Cl−]. CCC was associated with LWC (but not with leaf [Cl−]) under salinity conditions. Under salinity conditions, variations for LWC were associated with ‘Volkamer’ lemon alleles (at CCC and SOS1), whereas ‘Rubidoux’ trifoliate orange alleles in the rootstock at SOS2 were associated with significant variations in leaf [Na+] in Satsuma mandarin.
Silver-stained gel electrophoresis of markers developed for (A) salt overly sensitive genes SOS1 and SOS2; (B) Ethrec (ethylene receptor) and the Na+/H+ exchanger gene NHX1; and (C) CCC (cation chloride cotransporter) of V × P hybrids between between ‘Volkamer’ lemon (V) and ‘Rubidoux’ trifoliate orange (P). Four possible genotypes of the V × P progeny are indicated as ab, ac, bc, and bd, where a and b correspond to the alleles of ‘Volkamer’ lemon (Volk.) and c and d correspond to those of ‘Rubidoux’ trifoliate orange (Rubi.).
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Silver-stained gel electrophoresis of markers developed for (A) salt overly sensitive genes SOS1 and SOS2; (B) Ethrec (ethylene receptor) and the Na+/H+ exchanger gene NHX1; and (C) CCC (cation chloride cotransporter) of V × P hybrids between between ‘Volkamer’ lemon (V) and ‘Rubidoux’ trifoliate orange (P). Four possible genotypes of the V × P progeny are indicated as ab, ac, bc, and bd, where a and b correspond to the alleles of ‘Volkamer’ lemon (Volk.) and c and d correspond to those of ‘Rubidoux’ trifoliate orange (Rubi.).
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Silver-stained gel electrophoresis of markers developed for (A) salt overly sensitive genes SOS1 and SOS2; (B) Ethrec (ethylene receptor) and the Na+/H+ exchanger gene NHX1; and (C) CCC (cation chloride cotransporter) of V × P hybrids between between ‘Volkamer’ lemon (V) and ‘Rubidoux’ trifoliate orange (P). Four possible genotypes of the V × P progeny are indicated as ab, ac, bc, and bd, where a and b correspond to the alleles of ‘Volkamer’ lemon (Volk.) and c and d correspond to those of ‘Rubidoux’ trifoliate orange (Rubi.).
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
List of significant trait candidate gene associations and corresponding genotypic means using seedlings from 13 or 32 hybrids between ‘Volkamer’ lemon and ‘Rubidoux’ trifoliate orange as rootstocks under control and salinity (25 mm NaCl), respectively.z
Discussion
Many non-genetic factors can affect plant responses to salinity such as the composition and concentration of salts (Lauchli and Grattan, 2007), drought, irradiance, leaf temperature, atmospheric evaporative demand, soil type, and agronomic practices (Adnan, 2004; Syvertsen and Levy, 2005). We experimentally tried to minimize non-genetic effects and interactions by growing plants in large pots under a screenhouse and controlling salinity with irrigation. We approached the citriculture practice by evaluating the rootstocks in terms of the fruit yield of the scion during several years. For this purpose, a population of genetically related rootstocks was grafted with the same citrus cultivar. This experimental design allowed us to genetically analyze the rootstock effects on the grafted cultivar. The approach of Tozlu et al. (1999) used a non-grafted segregating progeny derived from Citrus grandis (a sexual species not used as rootstock) and P. trifoliata. Our approach intended a more realistic evaluation of the salt tolerance conferred by the rootstock. Nevertheless, our findings, using a semihydroponic system, might differ from those that would have been obtained under real field conditions.
Highly significant year effects were obtained for most traits (Table 1). Differences among years for relative humidity and temperature existed (Supplementary Fig. S1), which might explain the year effects, at least in part. Mandarins are perennial crops with alternate-bearing, and so year effects also can be attributed to this phenomenon and interfere with the salt tolerance response in terms of fruit yield. For these reasons, a multiyear approach was used.
A general reduction of tree vegetative growth by ≈30% was observed under salinity conditions (Table 1). Similar effects have been reported in grafted citrus crops with reductions in plant height, canopy volume, and trunk diameter (García-Sánchez et al., 2006; Grieve et al., 2007). Our results showed that the rootstock greatly influenced fruit traits of the grafted cultivar (in agreement with the citrus bibliography), but no single vegetative trait could fully predict fruit yield. Heritability estimates of rootstock-mediated effects on fruit yield under salinity (Table 1) were low, 0.18 at most, below those reported for tomato (Solanum lycopersicum) rootstocks (0.3 in Estañ et al., 2009). In general, trait heritabilities were higher under control than under salinity conditions, which suggested that the rootstock segregant population had larger genetic variability (i.e., more genes involved and/or larger effects on the traits) under control conditions than under salinity conditions. However, there were exceptions, because traits might be related to physiological mechanisms of salinity tolerance, including LDW, LDWrq, %DMA, LFWhj, LFWrq, LWC, and TrunkD. All these exceptional traits were related to vegetative growth and the osmotic stress response (Munns and Tester, 2008).
Citrus salt tolerance is related to the capacity to restrict Cl− transport from root to leaves (Levy and Syvertsen, 2004; Maas, 1993; Ream and Furr, 1976). Heritability estimates of the leaf concentration for both toxic ions clearly decay under salinity, primarily as a result of non-genetic components. It is deduced from Table 1 that genetic variability (H2) to improve scion leaf [Na+] is available at the segregating rootstock population, but apparently not for [Cl−]. Therefore, is scion leaf [Cl−] a good criterion to select V × P rootstocks for salt tolerance? The [Cl−] is directly related to FN (0.69) and indirectly related to FW (−0.65) under control conditions (Table 2). Under salinity conditions, correlation coefficients decreased to 0.50 and −0.57 for FN and FW, respectively. Therefore, the value of leaf [Cl−] to predict salt tolerance in the V × P population is low. For candidate gene analyses, the CCC of the citrus rootstock is only associated with leaf [Cl−] under control conditions (Table 3); it does not appear useful to predict leaf [Cl−] under salinity in the V × P progeny. Chloride is an essential plant micronutrient that participates in osmoregulation, cell elongation, vacuolar turgor regulation, membrane potential stabilization, and pH regulation (White and Broadley, 2001). The double role (nutritious/toxic) of Cl− in citrus plants might explain why different genes control leaf [Cl−] depending on the absence/presence of salinity. However, the lack of a significant association of CCC genotypes with leaf [Cl−] variations under salinity does not prove that CCC is not involved in salinity responses. The large non-genetic variability and/or a decay of its putative individual contribution under salinity conditions might explain the failure to detect it under salinity conditions.
Considering the scion leaf [Na+] as a good criterion to early select rootstocks for salt tolerance, P. trifoliata was reported as an efficient Na+ excluder at low salinities (Sykes, 2011; Walker, 1986), similar to some of its hybrids (Sykes, 1992). This is in agreement with the association of leaf [Na+] and AcTFW with NHX1 under control conditions (Table 3), in which high yield and low leaf [Na+] are related to the same ‘Rubidoux’ allele, suggesting a putative contribution of the citrus rootstock NHX1 to the variation of both scion traits. There is increasing evidence to show that NHX antiporters regulate the homeostasis of K+ and pH in intracellular membranes under normal and saline conditions (Barragán et al., 2012; Bassil et al., 2011; Leidi et al., 2010; Venema et al., 2003; Yamaguchi et al., 2013). The tonoplast vacuole NHX1 antiporter participates in Arabidopsis thaliana salt tolerance by compartmentalization of Na+ (Leidi et al., 2010; Munns and Tester, 2008). However, under salinity conditions, the association between AcTFW with NHX1 was not detected, and leaf [Na+] was significantly associated with ‘Rubidoux’ trifoliate orange alleles at SOS2 (Table 3).
The ‘Rubidoux’ alleles of the salt tolerance candidate gene SOS1 were consistently associated with fruit yield under control conditions (Table 3). As it has been recently reviewed by Yamaguchi et al. (2013), SOS1 appears to have several other functions that are not directly related to Na+ homeostasis but to pH homeostasis, both in the cytosol and the vacuole of root cells. Noteworthy, SOS1 is also associated with root and rootstock weights (Table 3). The ‘Volkamer’ lemon alleles at both SOS1 and CCC were associated with LWC under salinity conditions. Therefore, maintaining the scion water content might be an important role of the citrus rootstock derived from ‘Volkamer’ lemon to manage excess Cl− and Na+ under moderate salinity, genetically connecting the physiological mechanisms of salt tolerance defined by Munns and Tester (2008). Among the exceptional traits whose heritability estimates increased under salinity conditions, LWC showed statistical evidence of differential rootstock behavior depending on the presence of salinity [significant G × E (Table 1)]. This result supports the hypothesis that salt tolerance in citrus depends to a great extent on water use (Moya et al., 2003; Syvertsen et al., 2010), although we found no significant genetic correlation between LWC and leaf [Cl–] under salinity nor control conditions. This salt tolerance strategy based on the rootstock’s ability to maintain the scion water content was also found in tomato (Asins et al., 2010). Therefore, exploiting this tolerance mechanism might be useful for rootstock breeding in general.
Our results did not show a significant effect of salinity on mandarin juice pH (Table 1). Other citrus studies (reviewed by Navarro et al., 2010), with different experimental designs (one or two unrelated rootstocks) and different grafted cultivars, showed different salinity effects on juice pH, either increasing (Navarro et al., 2010) or decreasing (García-Sánchez et al., 2003) juice acidity.
Significant effects of salinity on other fruit quality traits were observed. Some traits such as JV and FD were reduced, whereas others were improved such as TSS and RT (Table 1). Higher TSS under salinity stress was previously observed for other grafted citrus cultivars (Boman, 2005; Dasberg et al., 1991; García-Sánchez et al., 2003, 2006; Navarro et al., 2010). Under salinity conditions, JV was directly correlated with FW, whereas TSS was directly correlated with FN and indirectly correlated with FW. Therefore, the TSS increase could be related to the reduction in JV (and FW) as a result of reduced water availability, as found previously for a Clementine mandarin (Navarro et al., 2010). The main components of variance influencing TSS were the rootstock genotype and the year (P < 0.0001; Table 1) and no significant interaction (G × E, G × Y, E × Y, and G × E × Y) was found suggesting that all genotypes increase TSS similarly under salinity. Because this is not the case for JV, other than reduction of JV could contribute to the increment of TSS under salinity. The enhancement of sucrose hydrolysis, increasing the concentrations of hexose sugars, might be involved such as was proposed under moderate water deficit stress in Satsuma mandarin (Yakushiji et al., 1996; Yukushiji et al., 1998) and ‘Valencia’ sweet orange (Barry et al., 2004).
Given that mandarin juice volume (and FW) showed a significant G × E interaction, whereas TSS did not, it seems possible to select V × P rootstocks that would increase TSS and maintain JV under salinity conditions. As a consequence, these rootstocks would induce earlier fruit maturation for a given grafted cultivar under salinity conditions and expand the harvesting period of citrus cultivars, which is a major goal of the citrus industry worldwide.
In conclusion, our results on the inheritance of rootstock effects on salt tolerance in the progeny of ‘Vokamer’ lemon agrees the hypothesis that it depends to a great extent on water use (Moya et al., 2003; Syvertsen et al., 2010). Because only a 5% of V × P progeny induced higher fruit yield than the salt-tolerant parent, and the heritabilities of rootstock effects were low, selection assisted by associated markers might be useful to obtain rootstocks that confer salt tolerance to the grafted mandarin. Salt-tolerant candidate genes SOS1 and NHX1 were associated with fruit yield traits under control conditions (1.4 dS·m−1), and SOS1 and CCC were associated with LWC under salinity conditions (4 dS·m−1). Results from rootstock effects on fruit quality traits suggest that rootstocks inducing early fruit maturation can be selected within the V × P progeny to expand the harvesting calendar in citrus under moderate salinity conditions.
Literature Cited
Adnan, A. 2004 Influence of salinity on citrus. A review paper J. Central European Agr. 5 263 272
Apse, M.P. & Blumwald, E. 2007 Na+ transport in plants FEBS Lett. 581 2247 2254
Asins, M.J., Bolarín, M.C., Pérez-Alfocea, F., Estañ, M.T., Martínez-Andújar, C., Albacete, A., Villalta, I., Bernet, G.P., Dodd, I. & Carbonell, E.A. 2010 Genetic analysis of physiological components of salt tolerance conferred by Solanum rootstocks. What is the rootstock doing for the scion? Theor. Appl. Genet. 121 105 115
Asins, M.J., Ruiz, C. & Bernet, G.P. 2009 Citrus primer list. 12 May 2014. < http://www.ivia.es/∼mjasins/index.html>
Bañuls, M.J., Serna, M.D., Legaz, F., Talón, M. & Primo-Millo, E. 1997 Growth and gas exchange parameters of citrus plants stressed with different salts J. Plant Physiol. 150 194 199
Barragán, V., Leidi, E.O., Andrés, Z., Rubio, L., De Luca, A., Fernández, J.A., Cubero, B. & Pardo, J.M. 2012 Ion exchangers NHX1 and NHX2 mediate active potassium uptake into vacuoles to regulate cell turgor and stomatal function in arabidopsis Plant Cell 24 1127 1142
Barry, G.H., Castle, W.S. & Davis, F.S. 2004 Rootstock and plant water relations affect sugar accumulation of citrus fruit via osmotic adjustment J. Amer. Soc. Hort. Sci. 129 881 889
Bassil, E., Ohto, M.A., Esumi, T., Tajima, H., Zhu, Z., Cagnac, O., Belmonte, M., Peleg, Z., Yamaguchi, T. & Blumwald, E. 2011 The arabidopsis intracellular Na+/H+ antiporters NHX5 and NHX6 are endosome associated and necessary for plant growth and development Plant Cell 23 224 239
Bernet, G.P., Fernandez-Ribacoba, J., Carbonell, E.A. & Asins, M.J. 2010 Comparative genome-wide segregation analysis and map construction using a reciprocal cross design to facilitate citrus germplasm utilization Mol. Breed. 25 659 673
Bingham, E.T., Mahler, R.J., Parra, J. & Stolzy, L.H. 1974 Long-term effects of irrigation-salinity management on a Valencia orange orchard Soil Sci. 117 369 377
Boman, B.J. 2005 Salinity effects on Florida grapefruit in the Indian River region HortTechnology 15 89 95
Brumós, J., Colmenero-Flores, J.M., Conesa, A., Izquierdo, P., Sánchez, G., Iglesias, D.J., López-Climent, M.F., Gómez-Cadenas, A. & Talón, M. 2009 Membrane transporters and carbon metabolism implicated in chloride homeostasis differentiate salt stress responses in tolerant and sensitive citrus rootstocks Funct. Integr. Genomics 9 293 309
Castle, W.S. 2010 A career perspective on citrus rootstock, their development and commercialization HortScience 45 11 15
Cerdá, A., Nieves, M. & Guillén, M. 1990 Salt tolerance of lemon trees as affected by rootstocks Irrig. Sci. 11 245 249
Colmenero-Flores, J.M., Martinez, G., Gamba, G., Vazquez, N., Iglesias, D.J., Brumos, J. & Talón, M. 2007 Identification and functional characterization of cation-chloride cotransporters in plants Plant J. 50 278 292
Cooper, W.C. 1961 Toxicity and accumulation of salts in citrus trees on various rootstocks in Texas Proc. Florida State Hort. Soc. 74 95 104
Cooper, W.C., Olson, E.O., Maxwell, N. & Otey, G. 1956 Review of studies on adaptability of citrus varieties as rootstocks for grapefruit in Texas J. Río Grande Valley Hort. Soc. 10 6 19
Cuartero, J.M., Bolarín, C., Asins, M.J. & Moreno, V. 2006 Increasing salt tolerance in the tomato J. Expt. Bot. 57 1045 1058
Dasberg, S., Bielorai, H., Haimowitz, A. & Erner, Y. 1991 The effect of saline irrigation water on Shamouti orange trees Irrig. Sci. 12 205 211
Estañ, M.T., Villalta, I., Bolarín, M.C., Carbonell, E.A. & Asins, M.J. 2009 Identification of fruit yield loci controlling the salt tolerance conferred by Solanum rootstocks Theor. Appl. Genet. 118 305 312
Flowers, T.J. 2004 Improving crop salt tolerance J. Expt. Bot. 55 307 319
Food and Agriculture Organization of the United Nations 2008 FAO land and plant nutrition management service. 12 May 2014. < http://www.fao.org/ag/agl/agll/spush/>
García, M.R., Asins, M.J., Forner, J. & Carbonell, E.A. 1999 Genetic analysis of apomixis in Citrus and Poncirus by molecular markers Theor. Appl. Genet. 99 511 518
García, M.R., Bernet, G.P., Puchades, J., Gómez, I., Carbonell, E.A. & Asins, M.J. 2002 Reliable and easy screening technique for salt tolerance of citrus rootstocks under controlled environments Austral. J. Agr. Res. 53 653 662
García-Sánchez, F., Carvajal, M., Cerdá, A. & Martinez, V. 2003 Response of Star Ruby grapefruit on two rootstocks to NaCl salinity J. Hort. Sci. Biotechnol. 78 859 865
García-Sánchez, F., Pérez-Pérez, J.G., Botía, P. & Martínez, V. 2006 The response of young mandarin trees grown under saline conditions depends on the rootstock Eur. J. Agron. 24 129 139
Gilliam, J.W. 1971 Rapid measurement of chlorine in plants materials Proc. Soil Sci. Amer. Soc. 35 512 513
Grieve, A.M., Prior, L.D. & Bevington, K.B. 2007 Long-term effects of saline irrigation water on growth, yield, and fruit quality of Valencia orange trees Austral. J. Agr. Res. 58 342 348
Grieve, A.M. & Walker, R.R. 1983 Uptake and distribution of chloride, sodium, and potassium ions in salt-treated citrus plants Austral. J. Agr. Res. 34 133 143
HunterLab 1996 Hunter Lab color scale, insight on color 8 9 (Aug. 1–15). HunterLab, Reston, VA
InfoStat 2004 InfoStat software 2004 version. FCA. Universidad Nacional de Córdoba, Cordoba, Argentina
Koyama, M.L., Levesley, A., Koebner, R.M.D., Flowers, T.J. & Yeo, A.R. 2001 Quantitative trait loci for component physiological traits determining salt tolerance in rice Plant Physiol. 125 406 422
Lauchli, A. & Grattan, S.R. 2007 Plant growth and development under salinity stress, p. 1–32. In: Jenks, M.A., P.A. Hasegawa, and S.M. Jain (eds.). Advances in molecular-breeding toward drought and salt tolerant crops. Springer, Houten, The Netherlands
Leidi, E.O., Barragán, V., Rubio, L., El-Hamdaoui, A., Ruiz, M.T., Cubero, B., Fernández, J.A., Bressan, R.A., Hasegawa, P.M., Quintero, F.J. & Pardo, J.M. 2010 The AtNHX1 exchanger mediates potassium compartmentation in vacuoles of transgenic tomato Plant J. 61 495 506
Levy, Y. & Syvertsen, J. 2004 Irrigation water quality and salinity effects in citrus trees Hort. Rev. 30 37 82
Maas, E.V. 1993 Salinity and citriculture Tree Physiol. 12 195 216
Moya, J.L., Cadenas, A., Primo-Millo, E. & Talon, M. 2003 Chloride absorption in salt-sensitive Carrizo citrange and salt-tolerant Cleopatra mandarin citrus rootstocks is linked to water use J. Expt. Bot. 54 825 833
Munns, R. & Tester, M. 2008 Mechanisms of salinity tolerance Annu. Rev. Plant Biol. 59 651 681
Navarro, J.M., Gomez-Gomez, A., Perez-Perez, J.G. & Botia, P. 2010 Effect of saline conditions on the maturation process of Clementine fruits on two different rootstocks Span. J. Agr. Res. 8 S21 S29
Pardo, J.M. & Rubio, F. 2011 Na+ and K+ transporters in plant signalling, p. 79–90. In: Geisler, M. and K. Venema (eds.). Transporters and pumps in plant signaling. Springer-Verlag, Düsseldorf, Germany
Porat, R., Pavoncello, D., Ben-Hayyim, G. & Lurie, S. 2002 A heat treatment induced the expression of a Na+/H+ antiport gene (cNHX1) in citrus fruit Plant Sci. 162 957 963
Prior, L.D., Grieve, A.M., Bevington, K.B. & Slavich, P.G. 2007 Long-term effects of saline irrigation water on Valencia orange trees: Relationships between growth and yield, and salt levels in soil and leaves Austral. J. Agr. Res. 58 349 358
Raga, V., Bernet, G.P., Carbonell, E.A. & Asins, M.J. 2012 Segregation and linkage analyses in two complex populations derived from the salt tolerant citrus rootstock Cleopatra mandarin. Inheritance of seed reproductive traits Tree Genet. Genomes 8 1061 1071
Ream, C.L. & Furr, J.R. 1976 Salt tolerance of some Citrus species, relatives, and hybrids tested as rootstocks J. Amer. Soc. Hort. Sci. 101 265 267
Ruiz, C., Bretó, M.P. & Asins, M.J. 2000 An efficient methodology to identify sexual seedlings in citrus breeding programs using SSR markers Euphytica 112 89 94
Ruiz, D., Martinez, V. & Cerdá, A. 1999 Demarcating specific ion (NaCl, Cl, Na) and osmotic effects in the response of two rootstock to salinity Sci. Hort. 80 213 224
Shannon, M.C. & Grieve, C.M. 1999 Tolerance of vegetable crops to salinity Sci. Hort. 78 5 38
Soler, J. 1999 Reconocimiento de variedades de cítricos en campo. Consellería de Agricultura, Pesca y Alimentación, Generalitat Valenciana. Serie de Divulgación Técnica No. 43
Storey, R. & Walker, R.R. 1999 Citrus and salinity Sci. Hort. 78 39 81
Sykes, S.R. 1985 A glasshouse screening procedure for identifying citrus hybrids which restrict chloride accumulation in shoot tissues Austral. J. Agr. Res. 36 779 789
Sykes, S.R. 1992 The inheritance of salt exclusion in woody perennial fruit species Plant Soil 146 123 129
Sykes, S.R. 2011 Chloride and sodium excluding capacities of citrus rootstock germplasm introduced to Australia from the People’s Republic of China Sci. Hort. 128 443 449
Syvertsen, J. & Levy, Y. 2005 Salinity interactions with other abiotic and biotic stresses in citrus HortTechnology 15 100 103
Syvertsen, J.P., Melgar, J.C. & Garcia-Sánchez, F. 2010 Salinity tolerance and leaf water use efficiency in citrus J. Amer. Soc. Hort. Sci. 135 33 39
Tozlu, I., Guy, C.L. & Moore, G.A. 1999 QTL analysis of Na+ and Cl– accumulation related traits in an intergeneric BC1 progeny of Citrus and Poncirus under saline and non-saline environments Genome 42 692 705
van Ooijen, J.W. 2009 MapQTL 6 software for the mapping of quantitative trait loci in experimental populations of diploid species. Kyazma, Wageningen, The Netherlands
Venema, K., Belver, A., Marín Manzano, M.C., Rodriguez Rosales, M.P. & Donaire, J.P. 2003 A novel intracellular K+/H+ antiporter related to Na+/H+ antiporters is important for K+ ion homeostasis in plants J. Biol. Chem. 278 22453 22459
Villalta, I., Bernet, G.P., Carbonell, E.A. & Asins, M.J. 2007 Comparative QTL analysis of salinity tolerance in terms of fruit yield using two Solanum populations of F7 lines Theor. Appl. Genet. 114 1001 1017
Villalta, I., Reina-Sánchez, A., Bolarín, M.C., Cuartero, J., Belver, A., Venema, K., Carbonell, E.A. & Asins, M.J. 2008 Genetic analysis of Na+ and K+ concentrations in leaf and stem as physiological components of salt tolerance in tomato Theor. Appl. Genet. 116 869 880
Walker, R., Torokfalvy, E. & Downton, W. 1982 Photosynthetic responses of the citrus varieties Rangpur lime and Etrog citron to salt treatment Austral. J. Plant Physiol. 9 783 790
Walker, R.R. 1986 Sodium exclusion and potassium-sodium selectivity in salt-treated trifoliate orange (Poncirus trifoliata) and Cleopatra mandarin (Citrus reticulata) plants Austral. J. Plant Physiol. 13 293 303
White, P.J. & Broadley, M.R. 2001 Chloride in soils and its uptake and movement within the plant: A review Ann. Bot. (Lond.) 88 967 988
Witcombe, J.R., Hollington, P.A., Howarth, C.J., Reader, S. & Steele, K.A. 2008 Breeding for abiotic stresses for sustainable agriculture Philosophical Trans. Royal Soc. B 363 703 716
Yakushiji, H., Nonami, H., Fukuyuma, T., Ono, S., Takagi, N. & Hoshimoto, Y. 1996 Sugar accumulation enhanced by osmoregulation in Satsuma mandarin fruit J. Amer. Soc. Hort. Sci. 121 466 472
Yamaguchi, T., Hamamoto, S. & Uozumi, N. 2013 Sodium transport system in plant cell Front. Plant Sci. 4 410
Yukushiji, H., Morinaga, K. & Nonami, H. 1998 Sugar accumulation and partitioning in Satsuma mandarin tree tissues and fruit in response to drought stress J. Amer. Soc. Hort. Sci. 123 719 726
Undernet mean relative humidity (RH) and averaged temperature (T) recorded at Instituto Valenciano de Investigaciones Agrarias (Valencia, Spain) meteorological station for the 5 yielding years during the salt treatment period.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Undernet mean relative humidity (RH) and averaged temperature (T) recorded at Instituto Valenciano de Investigaciones Agrarias (Valencia, Spain) meteorological station for the 5 yielding years during the salt treatment period.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
Undernet mean relative humidity (RH) and averaged temperature (T) recorded at Instituto Valenciano de Investigaciones Agrarias (Valencia, Spain) meteorological station for the 5 yielding years during the salt treatment period.
Citation: Journal of the American Society for Horticultural Science J. Amer. Soc. Hort. Sci. 139, 5; 10.21273/JASHS.139.5.518
