Study area and origin of scions

Scions from six nine-year-old trees with average diameter and height of 20.61 cm and 16.25 m, respectively, were used. The trees belonged to a forest plantation that was established in August, 2011 on the Casa Redonda property of the Palo Bendito ejido, Huayacocotla municipality, in the state of Veracruz, Mexico, between 20°27'22” N and 98°29'00” W, at an altitude of 2 460 m.

The selection of the trees was based, essentially, on the phenological state of the shoots, that the trees were free of pests and apparent diseases, and that they showed good phenotypic characteristics such as self-pruning, straightness of the stem and absence of bifurcations, in addition to outstanding diameter (sociological dominance). The shoots collected on January 14^th, 2021, were still in the dormant stage, with an average length of 6 cm.

The shoots were collected from the outer middle part of the crown, where the recent vegetative growth appears; they were collected and grouped to later store them in bags labeled with the corresponding tree number; they were stored in a cooler with cans with frozen water, covered with a damp cloth to keep them at a low temperature and avoid stress and loss of turgor. The shoots were grafted 24 h after being cut.

Origin and preparation of rootstocks

Work was handled with nine-month-old P. patula plant rootstocks, 7.1 mm in basal diameter and 56.3 cm in height, from the Pueblo Nuevo Forest Nursery, located in Peñuelas Pueblo Nuevo community, Chignahuapan municipality, in the state of Puebla (19°57'36” N and 98°06'26.20” W, at 2 680 masl).

The seed came from a seed orchard located in the vicinity of the nursery itself. The plants were produced in polystyrene trays with 77 cavities 170 cm³ in capacity under 50 % shade mesh conditions. The specimens were grown in a mixture of sawdust, peat moss, perlite (Multiperl^®) and vermiculite (Agrolita^®) substrate in a 70:10:10:10 (vol:vol) ratio. The nutritional management consisted of the use of Osmocote^® 18-06-12, in doses of 7 g L^-1, in addition to Micromax^® at a rate of 200 g m^-3.

One month before grafting the plants, they were transplanted into two types of containers: 1.9 L pots and 310 cm³ tubes. The substrate consisted of a homogeneous mixture of peat moss, PRO-MIX FLX^®, perlite and vermiculite in a 6:3:1 ratio, respectively. Half of the standard plants of each type of container (30 plants) were added 2 g L^-1 of Osmocote Scotts^® 14-14-14, with 8 months release (light fertilization) and the other half (30 plants), 7 g L^-1 of the same material. To this second group, a nutrient solution from ^{Landis et al. (1989)} every 15 days (heavy fertilization) until the end of the experimental period, to ensure a sufficient nutritional supply (full fertilization). It was added the aforementioned nutrient solution contains all the macro and micronutrients, and was formed by using regionally available soluble fertilizer materials (Table 1). Irrigation was applied every third day; a group of 60 plants was irrigated at 100 % saturation and another only at 40% saturation.

Grafting procedure

Grafting was carried out during the second week of January 2021 at the facilities of the Montecillo Campus Forest Nursery, Colegio de Postgraduados, Texcoco, State of Mexico (19°27'46.8" N, 98°54'23" W; 2 240 masl). The grafting technique implemented in the experiment was the terminal cleft technique described by ^{Muñoz et al. (2011} and ²⁰¹³⁾, which has produced better results than other techniques in P. patula (^{González-Jiménez et al., 2022}).

Prior to grafting, the scions were washed for three times: 1) immersion in water with a prepared solution of commercial liquid soap and 1 % diluted chlorine and were manually shaken for 5 minutes to ensure the removal of dust, resin and organisms that could be on its surface, 2) rinsed with running water and submerged again in a second solution prepared with Captan^® 1.5 g L^-1 for 15 minutes to prevent and avoid subsequent phytosanitary problems, and 3) rinsed again with running water and deposited in a cooler to keep the shoots cool and turgid before grafting scion by scion.

The terminal grafting technique consisted of making a transversal cut in the rootstock, at a height of 15 cm, after eliminating the needles or fascicles present three cm below the height of the cut; subsequently, a longitudinal cut of 3 to 5 cm was made right in the center of the stem of the rootstock, so that the scion could be inserted into this slit, to which two diagonal cuts were previously made on opposite sides of its base (the size pattern slit), leaving the scion in a “v” shape.

The scion was inserted into the slit of the rootstock, in such a way that the cambium of both coincided. Once the union was made, it was fixed with Parafilm^® tape throughout the wound, to avoid phytosanitary problems in the graft area, or dehydration of the tissues at the union point. Finally, a transparent plastic bag (30-40 days) was placed covering the entire scion, including the Parafilm^® tying, to create an atmosphere of adequate humidity and temperature for the graft. The bag was gradually withdrawn until the graft was "successful" and adapted to greenhouse conditions. Once the grafting process was finished, the grafted plants were accommodated in a space inside the greenhouse in a completely random manner and the complementary irrigation and fertilization activities were implemented according to the corresponding treatments.

Irrigation and soluble fertilization

Irrigation was carried out every third day from the month prior to grafting (December 11, 2020) until the conclusion of the experimental phase (April 20, 2021) and at all times the pH was adjusted to 5.5. On each irrigation date, four grafted plants were randomly selected for each type of container (1.9 L pots and 310 cm³ tubes) and the amount of water needed (in grams) to saturate their substrates was determined by weighing them. The average amount necessary (of the four containers) to saturate the substrate was incorporated into all containers with a 100 % saturation irrigation regime. Only 40 % of that quantity was applied to the containers whose irrigation regime 40 % irrigation.

For the treatments that received soluble fertilization, the nutrient solution suggested by ^{Landis et al. (1989)} (Table 1), which was applied every 15 days as part of the complementary fertilization to the group of plants that received 7 g L^-1 of slow-release fertilizer, was added so that, as a whole, it constituted a complete fertilization.

In addition, in the week that no nutrient solution was applied, a Captan^® 1.5 g L^-1 solution was prepared to irrigate all the plants with the proportion of water that corresponded to them according to the irrigation regime, in order to prevent phytosanitary problems. that could influence the success of the graft.

Assessed variables

The graft success (PI) was evaluated 70 days after grafting was carried out, on the live scions only. Graft basal diameter (DBI) and graft length (LI) were also measured at 70 and 100 days after grafting. From these measurements, the robustness index (IR) and the increments in basal graft diameter (IDBI) and graft length (ILI) were derived.

The DBI measurements were obtained with a CD67-S6”PM Mitutoyo digital vernier caliper, and the LI was measured with a conventional 30 cm ruler. The DBI was measured immediately above the grafting point, while the LI was measured from this point to the apical bud of the graft. The IR was obtained by dividing the DBI/LI; the IDBI based on the basal diameters corresponding to the first and second evaluation, that is:

In a similar way, the ILI was calculated as the difference in length growth between the second and the first evaluation:

Experimental design and statistical analysis

A complete randomized experimental design was used with a 2×2×2 factorial arrangement: two types of container (310 mL tube and 1.9 L pot), two irrigation levels (100 % and 40 % irrigation of saturation) and two fertilization levels (2 g L^-1 of Osmocote^® 14-14-14 only and 7 g L^-1 Osmocote^® 14-14-14 + nutrient solution from ^{Landis et al. [1989]} ). The combination of the levels of the three factors tested resulted in a total of eight treatments, which were replicated 15 times, and the experimental unit was represented by a grafted plant.

An analysis of variance (ANOVA) was performed to determine the effect of the factors and the interactions on the DBI, LI, IR, IDBI and ILI with the data from the evaluation at 100 days.

The experimental design model was:

Where:

Yijkl = Value of the response variable of level i of A, level j of B, level k of C

μ = General mean

Ai = Effect of the i level of the A factor (container)

Bj = Effect of j level of the B factor (irrigation)

Ck = Effect of the k level of the C factor (fertilization)

ABij = A×B interaction corresponding to the i level of A and the j level of B

ACik = A×C interaction corresponding to the i level of A and the k level of C

BCjk = B×C interaction corresponding the j level of B and the k level of C

ABCijk = A×B×C interaction corresponding to the i level of A, the j level of B and the k level of C

εijkl = Experimental error corresponding to the l replication of the i level of A, the j level of B and the k level of C

In order to analyze graft success (PI), due to the fact that the assumptions of normality and homogeneity of variance were not met, even testing several transformations of the variable, the non-parametric Mann-Whitney test was applied to detect the effects of the factors studied. A comparison of means was performed by Tukey test, with a confidence level of α=0.05. Statistical analyzes were performed with version 9.0 of the statistical software Statistical Analysis System (^{SAS, 2002}).

Grafting success

After 70 days from the day of grafting, the average grafting success(PI) was 72.5 %; which is relatively high, compared to that obtained by ^{Barrera-Ramírez et al. (2021)}, who registered a PI of 41 % in Pinus pseudostrobus Lindl., but close to that achieved by ^{Muñoz et al. (2013)} in that same species (P. pseudostrobus; 82.8 %). Despite this percentage, the Mann-Whitney statistical test indicated that the tested factors did not significantly affect the process (Table 2).

^{Hildebrant (2017)} mentioned that the rootstock container should be 5-10 cm in diameter by 10-15 cm deep to provide the plant with enough space for root growth. However, the present study shows that success is similar with a container volume of 310 mL than with one of 1.9 L. The absence of the effect of container volume is probably due to the fact that none of the container sizes limited root growth, provided the relatively small size of the rootstocks.

In regard to fertilization, ^{Barrera-Ramírez et al. (2021)} also did not recognize significant effects of rootstock fertilization on graft success in Pinus pseudostrobus Lind. var. oaxacana (Mirov) S. G. Harrison; instead, they determined that the succes is affected by factors associated with genetic affinity (origin of the material), phenology (grafting time) and type of graft. Studies by ^{Barnett and Miller (1994)}, ^{Yin et al. (2012)} and ^{Goldschmidt (2014)} confirm the effects of the genetic affinity of the rootstock and the scion on graft success. These authors point out that a greater genetic compatibility between the components of the graft implies greater anatomical, morphological, physiological and biochemical affinity and a greater probability of successful grafting.

In the study described here, the amount of water supply before and after the grafting operation did not affect graft success (Pr<0.3057, Table 2). It is highly probable that this lack of effect is due to the decrease in the transpiration rate of the scion, achieved by placing a transparent polyethylene bag during grafting. ^{Dabirian and Miles (2017)} demonstrated that the reduction of the transpiration rate in the scion, in this case through the application of antiperspirant substances, effectively improves the success rates in watermelon (Citrullus lanatus [Thunb.] Matsum. & Nakai). In the grafting of forest species, the use of transparent bags to cover the scions and reduce transpiration is common (^{López, 2020}; ^{Pérez-Luna et al., 2020}; ^{González-Jiménez et al., 2022}). In this work, the transpiration rates of the scions were not assessed, but it is possible to deduce that they were reduced, in such a way that the two irrigation levels tested could be sufficient to adequately supply water to the buds.

Morphological variables

Graft basal diameter (DBI) and graft length (LI) showed significant differences (p≤0.05) between container types, fertilization levels, and irrigation levels, but not for interactions (Table 3), so that the proposed null hypothesis is rejected. This suggests that the individual effects of one factor on these variables are independent of the level taken by the other factors, and that the effects of such factors are additive (^{Berrington and Cox, 2007}).

The importance of preparing the rootstocks before subjecting them to grafting lies mainly in the fact that they are responsible for providing the root system that in turn will provide water and nutrients to the graft (^{Ranjith and Ilango, 2017}; ^{Kita et al., 2018}). The development of the root system depends on the space available to grow, and according to ^{Landis (1990)}, the type of container (size and color) directly affects the temperature of the substrate and this in turn, affects the development of the root. In addition to determining the growth space for the root, the volume of the container constitutes the size of the water and nutrient reservoirs for the plant (^{Pershey, 2014}). Consequently, the type of container affects not only the root system of the composite plant, but also the growth in diameter and height of the plants (^{Castro-Garibay et al., 2016}).

^{Copes (1980)} mentioned that the vigor of the rootstocks is determinant for the subsequent growth of the graft; so that when the rootstock is small or not very vigorous, the graft frequently presents a plagiotropic growth, in contrast to the orthotropic growth of grafts on vigorous and well-rooted rootstocks. The vigor of the rootstock is determined, among other factors, by the size of the container (^{Landis, 1990}), which possibly explains the greater growth of the graft in the 1.5 L container, compared to the 310 mL container in the present study (Table 4).

Once the graft has got successful, we have a complete (composite) plant whose growth is affected by the same factors that define the growth of a non-grafted plant, in addition to others associated with the composite plant itself, such as hormonal aspects and permanence or not of the physical compatibility between rootstock and graft (^{Verdugo-Vásquez et al., 2021}). However, in the present study a firm union between rootstock and scion was observed without callous deformations until the time of evaluation of the experiment. The use of slow release mineral fertilizers increases the longitudinal growth of the plant (^{Cañellas et al., 1999}), depending on the amount and type of nutrients it contains. For this reason, to produce rootstocks, ^{Pérez-Luna et al. (2021)} used, with good graft survival results, 10 g L^-1 of Multicote^® 6: 18-6-12 (NPK) + micronutrients, while ^{Barrera-Ramírez et al. (2021)} applied three different fertilization doses: 3, 5 and 8 kg of controlled release fertilizer Multicote^® 15-7-15 (NPK) + 2 MgO + micronutrients per cubic meter of substrate, and achieved better results with the highest fertilizer dose.

In the present study, the level of fertilization influenced the growth in basal graft diameter (DBI) since it increased when complete fertilization was applied (7 g L^-1 of Osmocote 14-14-14 + nutritive solution from ^{Landis et al. (1989)} (Table 4). However, the same did not occur with graft length (LI) since there were no statistically significant differences in shoot elongation, nor did it occur with IR. On the other hand, the increases (between the first and second evaluation) in diameter and elongation of the shoot aumented with complete fertilization, in coincidence with what was reported in the study conducted by ^{Mutabaruka et al. (2015)} in Castanea sativa Mill.

The irrigation rate received by the grafts marked significant differences in DBI (5.6705 vs 6.0222 mm, with 40 and 100 % saturation, respectively) and LI (21.173 vs 27.236 cm, with 40 and 100 % saturation, respectively) with grafts growing the most when the substrate was saturated to 100 %. Despite these results, ^{Hibbert-Frey et al. (2011)} concluded that water availability did not affect scion growth when grafting Abies fraseri (Pursh) Poir. shoots on Abies bornmuelleriana Mattf. rootstocks, although in this case, the interspecificity of the graft may have influenced the result.

The graft robustness index (IR) showed a significant difference (p≤0.05) between irrigation rates (Table 3). The IR is an indicator of the resistance of the plant to desiccation by wind, its survival and growth in dry locations, furthermore, it is a criterion that must be considered in the production of quality plants. ^{Saenz et al. (2014)} mentioned that to improve the IR it is necessary to produce plants in large containers to favor root development. In grafting projects, this variable has not been included, however, it will be necessary to establish the grafts in the field and determine their behavior in relation to the index.

The increase in basal graft diameter (IDBI) was only significantly affected (p≤0.05) by the fertilization regime while the increase in graft length (ILI), by the type of container, fertilization, and the double interaction type of container × fertilization regime, which indicates that the effect of the container on the length of the graft is affected by the fertilization regime and vice versa.

When performing the comparison of Tukey means (α=0.05; Table 4), statistical differences were found between levels of the factors studied. Growth in both diameter and graft length is higher when using rootstock in larger container, full fertilization, and heavy irrigation (additive effects of factors tested) (^{Berrington and Cox, 2007}). This should be considered if increased growth rates of the composite (grafted) plants, especially in diameter and length is to be pursued.