Introduction

One of the current trends in Mexican citriculture is rootstock diversification, which arises as a result of the appearance of diseases such as the citrus tristeza virus and the need to improve the yield and productive efficiency of orchards. The diverse mechanisms of graft-rootstock interaction have been suggested in a general way and are currently being detailed, thereby increasing the knowledge for correct decision-making from the time of orchard planning.

In citrus, particularly in 'Tahiti' lime (Citrus x latifolia Tanaka ex Q. Jiménez), there are several studies related to the use of rootstocks, focused mainly on the development of vigor, yield, and fruit quality (^{Curti-Díaz, Hernández-Guerra, & Loredo-Salazar, 2012}; ^{Jiménez, Frometa, & García, 2009}; ^{Machado, Siqueira, Salomão, Cecon, & Da Silva, 2017}; ^{Mademba, Lemerre-Desprez, & Lebegin, 2012}; ^{Piña-Dumoulín et al., 2006}). In these studies, it has been found that 'Volkamer' lemon (Citrus volkameriana Tenn. & Pasq.) is a rootstock that induces high vigor in the 'Tahiti' lime graft and that increases with age; the plant can reach heights of 1.04 m from the first year to more than 5 m in adulthood with crown volume greater than 100 m³. By contrast, when the trifoliate orange (Poncirus trifoliata L. Raf.) was evaluated as a rootstock, the heights developed by the 'Tahiti' lime were low during the first years (0.85 to 2.70 m) but increased to 3 and 4.5 m with variable crown volumes, from 9 to 98 m³. For this reason, trifoliate orange is considered a rootstock with a dwarfing effect (^{Cantuarias-Avilés et al., 2012}; ^{Jiménez et al., 2009}; ^{Machado et al., 2017}; ^{Stenzel & Neves, 2004}). On the other hand, ^{Jiménez et al. (2009)} and ^{Espinoza-Núñez, Mourão, Sanches, Cantuarias-Avilés, and Dos Santos (2011)} highlighted that sour orange (C. aurantium L.) induces high vigor with heights of 3.2 m and up to 111 m³ of crown in adulthood.

Although there are studies focused on graft vigor characteristics, few analyze their relationship with the anatomical characteristics of the rootstock stem’s secondary xylem. In this regard, ^{Tadeo, Moya, Iglesias, Talón, and Primo-Millo (2003)} reported that the most vigorous rootstocks have greater physical capacity for water transport, since their stems have larger vessel elements. Citruses have secondary xylem of diffuse porosity, with vessel elements surrounded by parenchyma cells and fibers.

^{Saeed, Dodd, and Sohail (2010)} studied the stem xylem of seven rootstocks in three-month-old citrus seedlings, among them the sour orange and 'Flying Dragon'. The researchers observed that more vigorous rootstocks, such as sour orange, have fewer vessel elements and comparatively narrower phloem in stems and roots than less vigorous ones. They also explained that the difference in vessel number and size is due to the synthesis of auxins in the young leaves, where high levels induce the formation of numerous and small vessels, due to their rapid rate of differentiation. On the other hand, if the auxin concentration is low, the differentiation is slower and therefore there are fewer and larger vessels. With respect to 'Flying Dragon', the authors attributed the slow growth and reduced height of the rootstock to a high bark:xylem ratio, both in root and stem. The less vigorous growth in rootstocks with smaller vessel elements in the stem and root xylem suggests that low auxin/cytokinin levels favor phloem differentiation, and higher auxin/cytokinin levels favor xylogenesis (^{Saeed et al., 2010}).

On the other hand, ^{Martínez-Alcántara et al. (2013)} studied the behavior of 'Valencia' orange grafted onto 'Flying Dragon' and 'Rubidoux' rootstocks and recorded that the latter induced 39.7 % more fresh weight and 27.3 % more leaf area than the former. They also observed that 'Flying Dragon' had the highest values in terms of hydraulic resistance, which limited the hydraulic conductance, which in turn decreased water potential, causing a reduction in stomatal conductance. These authors concluded that the main cause of dwarfism in the 'Flying Dragon' rootstock may be its low hydraulic conductivity. ^{Berdeja-Arbeu et al. (2013)} also evaluated six, six-year rootstocks, including 'Volkamer' lemon and sour orange. These authors found no significant difference in the height of 'Tahiti' lime trees but found negative correlations in the graft perimeter with the tangential diameter of the vessel and total vessel area, which indicates that the morphology and anatomy of the 'Tahiti' lime stem xylem was affected by the different rootstocks.

In this context, this research aimed to determine the relationship between the anatomical characteristics of the stem xylem of four rootstocks (two non-dwarfing and two dwarfing) and the initial growth characteristics of the 'Tahiti' lime graft in a nursery.

Materials and methods

The experiment was carried out in a glasshouse located at Chapingo Autonomous University’s Experimental Agricultural Field, situated at 19° 29’ NL and 98° 52’ WL, at an elevation of 2 240 m. Four types of rootstock plants, all six months of age, were obtained from the certified Maricruz nursery in Arroyo de Piedra, Tlapacoyan, Veracruz, Mexico. The plants were moved to Chapingo, Mexico; they were transplanted in 16-L individual pots and placed in the greenhouse for another nine months for their adaptation and development. The substrate was a mixture of sandy-silt soil + 10 % compost; the plants were irrigated every 72 h. The average temperature in this period was 29 °C.

The 'Tahiti' lime buds used for grafting were donated by the Maricruz nursery. The rootstocks used were 'Volkamer' lemon (C. volkameriana), sour orange (C. aurantium) (not certified), 'Flying Dragon' trifoliate orange (Poncirus trifoliata var. monstrosa) and 'Rubidoux' trifoliate (P. trifoliata). Grafting was carried out from August to September 2015. A budding-type graft was used.

The initial growth variables of the primary shoot were as follows: a) length obtained weekly with a flexometer from the base of the graft to the apex, until the measurement was constant, b) growth rate represented by the number of days required to reach constant length, c) diameter measured at 4 cm from the base of the graft with a Truper® digital Vernier caliper (model CALDI-6MP, Mexico), d) number of leaves; e) length of internodes obtained by dividing the total length by the number of primary shoot internodes, f) leaf area calculated by the sum of the area of ​​all the shoot’s leaves, through the Image Tool v3 program (^{University of Texas Health Science Center San Antonio [UTHSCSA], 1995}), and g) average leaf area obtained by dividing the total leaf area by the number of leaves.

The wood anatomical description work was carried out at Chapingo Autonomous University’s Wood Anatomy and Technology Laboratory. The anatomical variables of the rootstock stem were measured in cross-section preparations and in cell individualization. The preparations for the microscopic observation of the stem cross-sections were made according to the procedure described by ^{García, Guindeo, Peraza, and de Palacios (2003)}, with 24 h of safranin staining and coating with Entellán^® resin. Individualization of the xylem elements was done by following the maceration technique described by ^{Franklin (1945)}. The visible sediment was stained with safranin and a drop was placed on the slide; the excess water was removed with heat, it was covered with Entellan^® resin and the coverslip was placed. Both groups of preparations were observed on a Leica Microsystems DM3000 image analyzer (Germany). In the stem cross-sections, the density of the vessel elements, average area of ​​the vessels and the xylem radius were observed and measured. In the preparations with the maceration technique (individualized cells), the length and diameter of both vessels and fibers were measured; the sample size for the measurement of these variables was 27 observations. The percentage of area occupied by the vessels was calculated by multiplying the average vessel area by the density.

The data were analyzed in a completely randomized experimental design with six replicates per treatment. Analysis of variance and Tukey's range test (P ≤ 0.05) were performed for each variable; subsequently, the degree of association between the anatomical and growth variables was determined by Pearson correlation in the SAS v9 program (^{Statistical Analysis System, 1992}).

Results and discussion

The statistical analysis of variance indicated that the four rootstocks have an effect on all the initial growth variables of 'Tahiti' lime and that their anatomical characteristics are different from each other.

Table 1 shows the effect of the four rootstocks on the initial growth variables of 'Tahiti' lime. Comparative analysis of the statistical means showed that the sour orange and 'Volkamer' rootstocks had similar behavior in almost all the evaluated growth variables, with the exception of shoot length that was smaller in 'Volkamer' (21 % smaller). The 'Flying Dragon' and 'Rubidoux' dwarfing rootstocks coincided in all the measured growth variables, forming another statistical group. The effect of these rootstocks is reflected in the reduction of all the variables evaluated with respect to the non-dwarfing ones; for example, in the dwarfing rootstocks, the reduction in the shoot length, growth rate and shoot leaf area was 65 to 70 %, in the number of leaves 57 % and in the diameter of the primary shoot 42 %, while the area of ​​each leaf and internode length decreased by 26 and 32 %, respectively.

The ‘Volkamer’ lemon rootstock induces heights of up to 5 m and a crown volume greater than 100 m³ in adulthood, while the trifoliates of the genus Poncirus induce plant heights of less than 3 m (^{Curti-Díaz et al., 2012}; ^{Jiménez et al., 2009}; ^{Machado et al., 2017}; ^{Mademba et al., 2012}; ^{Piña-Dumoulín et al., 2006}; ^{Stenzel & Neves, 2004}). On the other hand, ^{Jiménez et al. (2009)} and ^{Espinoza-Nuñez et al. (2011)} have highlighted the great vigor induced by sour orange (C. aurantium) in 'Tahiti' lime; however, none of them has related any anatomical characteristics of the stem of the rootstocks with their results.

Table 2 shows the anatomical variables of the stem of the four evaluated rootstocks. The 'Volkamer' rootstock, in general, expressed the highest values ​​in each variable, with the exception of vessel density, as can be seen in Figure 1. In the case of fiber length and xylem radius, the 'Volkamer' and sour orange rootstocks were statistically similar (P > 0.05). By contrast, the 'Rubidoux' and 'Flying Dragon' trifoliates had the lowest values, except in vessel density; both rootstocks were only statistically different (P ≤ 0.05) in vessel element length and fiber length.

Figure 1 Size and density of secondary xylem vessels of the stem of two vigorous rootstocks (A: sour orange and B: 'Volkamer' lemon) and two dwarfing ones (C: 'Flying Dragon' trifoliate and D: 'Rubidoux' trifoliate) in 'Tahiti' lime grafts. Magnification 10x.  

Figure 2 shows that 'Volkamer' lemon had longer and larger-diameter vessel elements than sour orange, this in turn greater than 'Rubidoux' and 'Flying Dragon', which did not present difference in diameter, but did in length (Table 2).

Figure 2 Average vessel elements of the stem xylem of two vigorous rootstocks (A: sour orange and B: 'Volkamer' lemon) and two dwarfing ones (C: 'Flying Dragon' trifoliate and D: 'Rubidoux' trifoliate) in 'Tahiti' lime grafts. Magnification 10x. 

Figure 3 shows that the fibers of sour orange and 'Volkamer' lemon had similar length and were longer than those of 'Rubidoux' and 'Flying Dragon'; according to ^{Brenes-Angulo, Reyes-Cordero, and Moya-Roque (2012)}, all fibers smaller than 1 mm in length are classified as short fibers. Fiber diameter was higher in 'Volkamer' and similar in the other three rootstocks (Table 2).

Figure 3 Average fiber length of the stem xylem of two vigorous rootstocks (A: sour orange and B: 'Volkamer' lemon) and two dwarfing ones (C: 'Flying Dragon' trifoliate and D: 'Rubidoux' trifoliate) in 'Tahiti' lime grafts. Magnification 10x. 

In this way, 'Volkamer' lemon and sour orange, which showed a lower density of vessels and vessel elements with larger dimensions, also produced higher values in the initial growth of the 'Tahiti' lime shoot. Conversely, 'Flying Dragon' and 'Rubidoux' had the smallest and most dense vessel elements, but the initial growth of the 'Tahiti' lime shoot was reduced. Some studies carried out with various fruit species such as avocado, peach, apple and citruses indicate that there is greater shoot growth when the stem or root vessels of the rootstock are larger, and less growth when there are small vessels, but in greater quantity (^{Bauerle, Centinari, & Bauerle, 2011}; ^{Berdeja et al., 2013}; ^{Fassio, Heath, Arpaia, & Castro, 2009}; ^{Martínez-Alcántara et al., 2013}; ^{Rodríguez, 2012}; ^{Tombesi, Johnson, Day, & DeJong, 2010}). In the same studies, the authors related the large xylem vessels with greater hydraulic conductance -whether measured or theoretical- and greater shoot growth; on the other hand, when the vessels were small, both the hydraulic conductance and growth were low.

In the present study, the diameter of the vessels and their density allowed us to distinguish, anatomically, the rootstocks that induce either high vigor or dwarfism. ^{Rodríguez (2012)} considers that the number of vessels and their size are factors that determine hydraulic conductance in plants, and this in turn influences the water relations, which are reflected in the vegetative growth. This suggests that rootstocks with larger vessel sizes would be inducing the formation of more vigorous 'Tahiti' lime plants, due to the availability of water, minerals and hormones in the solution, contrary to dwarfing rootstocks, which would have lower availability.

Table 3 shows the degree of correlation of the 'Tahiti' lime growth variables with the anatomical variables of the rootstock xylem. The correlation revealed a strong association of vessel size and density, fiber length and xylem radius with each of the initial growth variables of 'Tahiti' lime. The high negative correlation between vessel density and all the initial growth variables stands out, indicating that at a lower number of vessels per unit area there is a variable with a positive behavior and vice versa. ^{Zach et al. (2010)} found negative correlations of mean density of trunk vessels and twigs of forest trees with tree height, and a positive correlation between stem vessel diameter and tree height.

In this study, the correlation between the number of leaves and density is high and negative (-0.883), which is similar to the findings reported by ^{Saeed et al. (2010)}, who mention that vessel density is related to the number of leaves, induces and controls the formation of vascular tissue and, consequently, stimulates the high auxin levels near the leaves, which favors the formation of numerous small vessels due to their rapid rate of differentiation.

The xylem radius showed high positive correlations with all the initial growth variables. ^{Saeed et al. (2010)} suggest that rootstocks characterized by having little vigorous growth have smaller vessel elements caused by lower auxin levels that, in conjunction with the effect of cytokinins, favor differentiation of phloem, while higher auxin/cytokinin levels favor xylogenesis. In this study, vigorous rootstocks had a greater xylem radius than dwarfing rootstocks (Table 2).

Fiber length also showed a high correlation with the growth variables, but the diameter showed very low correlation. It is probable that these characteristics are not related to growth, since the main function of these cells is support (^{García et al., 2003}) and their length is only a consequence of cell differentiation; however, it is necessary to continue studying them.

Knowledge of the degrees of association of the anatomical characteristics of the rootstock stem and the growth variables is important, due to their probable use as indicators or indices to predict the behavior of a grafted variety.

Conclusions

The sour orange and 'Volkamer' lemon rootstocks showed vigorous growth of the 'Tahiti' lime primary shoot contrary to the 'Rubidoux' and 'Flying Dragon' trifoliate rootstocks that showed dwarfing effect. These effects on initial growth are strongly correlated with the characteristics of the vessel elements of the stem xylem of the rootstocks, indicating that the functions of this tissue influence the growth of the 'Tahiti' lime graft in the nursery. Therefore, the length, density and diameter of the stem xylem vessels of the rootstock can be considered predictive variables of the growth of the graft shoot. The obtained information represents an advance in the generation of knowledge to reinforce decision-making in the establishment of 'Tahiti' lime plantations.