Gathering of cuttings

The cuttings were gathered on December 16, 2016, at the facilities of the Universidad Autónoma Chapingo (Chapingo Autonomous University), from branches of the middle and lower parts of the crown ―according to their accessibility― of healthy adult Acer negundo trees. Only the material of the first three knots from the tip to the base of each branch was used. The cuttings had an average length of 15 cm, and were collected in a plastic ice box; they were kept moist and shaded until they were placed in the substrate.

Establishment of the experiment

The experiment was installed in Tepetlaoxtoc, Mexico State; black polyethylene trays with 54 cavities and a volume of 200 mL per cavity were utilized. The substrate was a mixture of peat moss, agrolite, and medium texture vermiculite, in a proportion of 40, 20 and 40 %, respectively. Controlled release fertilizer (Osmocote ^®) was added in a dose of 1 kg m^-3 of substrate. The seedlings were irrigated with running water.

The factors taken into account were: cutting type, application of a root starter, and fertilization level. The cuttings were divided into two levels, the first of which is known as “soft cutting” and corresponds to the material gathered from the buds in the branches, with diameters of 4.5 to 8.5 mm, a slightly flexible consistency, and a light green color. The second level corresponded to “hard cuttings” from the base of the branch, with diameters ranging between 9 and 16 mm, with a rigid consistency, and a dark green to brown color.

The levels of root starter (Radix 10 000^®, with 1 % indolebutyric acid (IBA) and 99 % inert material) were: with and without application. The utilized dose of the soluble fertilizer, with the formula 20-20-20 (20 % N, 20 % P₂O₅, 20 % K₂O, the rest of inert material), was rated as high (150-60-120) or low (100-50-80) (N, P, K, ppm).

480 cuttings were placed in the trays, one per cavity, with a total of 60 cuttings for each of the eight treatments. At the moment of planting, the application of root starter was assigned at random, the cuttings that required it were moistened at the base and impregnated with Radix 10 000^®, the excess of which was removed before they were planted in the substrate. In order to seal the exposed incision, white acrylic paint was used, mixed with Captán ^® fungicide, in a dose of 1 g L^-1 of paint.

After the transplant, the trays were placed in a tunnel with a white polyethylene cover for greenhouses measuring 1.5 m in width, 3 m in length, and 1.5 m in height, with both ends covered with shade mesh, and a first irrigation was applied, adding 1 g L^-1 of Captán.

After eight weeks of the establishment of the experiment, at the start of the emission of adventitious roots, Peters Professional ^® soluble fertilizer for general use, convenient for the rapid growth stage in the nursery (20 - 20 - 20), began to be administered in the doses mentioned before, with irrigation every other day in order to maintain the moisture of the substrate.

After fourth months, the rooting percentage (in regard to the total number of cuttings) was registered, and the cuttings were considered to be rooted when side roots were observed. A sample of 40 cuttings was then obtained, in which the shoot diameter, the length of the shoot from the base to the bud, the length of the shoot from the base to the leaf tip, the root length, the (average) height from the base at which the roots emerged, the fresh weight of the leaves, the root, the aerial part, and the underground part, and the dry weight of the leaves, root, aerial part, and underground part were measured, using a plastic ruler calibrated in centimeters, a Truper^® digital vernier caliper, and a OHAUS Scout ^® Pro digital scale.

The leaves, roots, aerial and subterranean pats were dried inside craft paper bags, in a (Ríos Rocha H-41) drying oven at 80 °C until they attained a constant weight (generally after 72 h).

Furthermore, the shoot/root ratio was estimated based on the relationship between the dry weights of each of one of the parts and Dickson’s quality index, which is estimated dividing the total dry weight of the plant (g) by the sum of the slenderness coefficient (the ratio of the height of the plant, in cm, to its diameter, in mm) and the shoot/root ratio (^{Birchler et al., 1998}).

In this study, neither the dry or fresh weights nor the quality indicators included the weight of the cuttings.

Experimental design and statistical analysis

The experimental design was completely random and constituted a 2 × 3 factorial experiment. For the statistical analysis of the rooting, the percentage of rooted cuttings was transformed with the arc sine function of the square root of this percentage expressed as per-unit. The PROC ANOVA procedure of the SAS^® software (Statistical Analysis System), V. 9.4, was utilized, and the means were compared using the Tukey’s test.

The following statistical model was utilized:

Where:

µ = Overall mean

α _i = Effect of the i ^th level of the cutting factor

β _j = Effect of the i ^th level of the root starter factor

γ _k = Effect of the k ^th level of the fertilization factor, the combinations of letters indicate the interactions between the levels of the involved factors

E _ijk = Experimental error

Response to the cutting type

According to the means comparison, the hard cuttings had superior results for shoot diameter, shoot length from the base to the bud, and shoot length from the base to the leaf tip (Table 3).

The hard cuttings were taken from the part closest to the stem; therefore, they may contain a larger amount of carbohydrates than soft cuttings. In this regard, ^{Veierskov (1988)} points out a rising gradient in the concentration of these reserve products toward the stem base. According to ^{Hartmann and Kester (1991)}, the thicker cuttings from the basal portion of the branches accumulate more reserve carbohydrates. This pattern is adjusted to, and accounts for the greater response in the variables mentioned above. This result is similar to that obtained by ^{Liao et al. (2012)}, who attempted the vegetative reproduction of Myrciaria dubia (Kunth) McVaugh and obtained the best average for the shooting of leaves with large diameter cuttings. Nevertheless, different responses are to be expected in different species. ^{dos Santos et al. (2011)} conclude that Cestrum laevigatum Schltdl. and Salix humboldtiana Willd. can be propagated with any cutting type; however, woody cuttings, which tend to be even thicker, work best for species like Ficus adathodigifolia L. and F. citrifolia Mill.

Response to the application of a root starter

The root starter (IBA) produced a tendency with a higher value in the variables stem length, both from the base to the bud and from the base to the leaf tips; it also influenced the fresh and dry leaf weight (Table 4). This agrees with the findings of ^{Boschini and Rodríguez (2002)}, who, after applying different doses of IBA to Morus alba cuttings, registered statistically significant differences in the overall shooting of buds and roots between the plants on which IBA was used. Likewise, ^{Maldonado et al. (2017)} achieved the rooting of Malphigia mexicana A. Juss. cuttings, and ^{Quintero et al. (2008)}, that of Symphoricarpos microphyllus H. B. K. stakes, using root starters.

^{Hartmann and Kester (1991)} state that the forming of adventitious roots may be due to certain inherent, non-translocable factors determined by the genotype of the individual tissue cells, as well as to the endogenous rooting factors present in each individual; furthermore, the lack of response to the application of synthetic auxin is due to the fact that several species have enough natural auxin to promote rooting. In the study by ^{dos Santos et al. (2011)}, conducted on cuttings of 20 Brazilian forest species, C. laevigatum and S. humboldtiana formed roots without a root starter, but 11 taxa ―e.g. Dendropanax cuneatus (DC). Decne. & Planch., Erythrina falcata Benth. and Casearia sylvestris Sw.― did not exhibit roots with any of their treatments.

Response to the level of fertilization

No significant differences were obtained in the mean comparison analysis for the levels of fertilization (Table 5), which may be due to the fact that “low fertilization” is sufficient to meet the nutritional demands of the plant, at least in terms of its morphological characteristics. The results were similar to those registered by ^{Bualó et al. (2006)} for Calibrachoa linearis (Hook.) Wijsman and P. kleinii L. B. Sm. & Downs., which were subjected to two levels of fertilization; furthermore, their growth and morphological characteristics did not differ significantly between treatments.

After applying a root starter (3 000 gL^-1 of IBA) during 5 s, ^{Caetano et al. (2010)} tried various fertilizers and a control for rooting semi-woody olive tree (Olea europaea L.) cuttings, and they point out that the largest number of roots was produced with a bio fertilizer containing 0.1 % K, 0.34 % Zn, 0.02 % Cu, 0.04 % Mn, 0.18 % B, 0.43 % Mo, 0.12 % S, 0.64 % Ca, and 0.27 % organic C.

Root starter-fertilizer interaction

The interaction between the root starter and the fertilizer had effects on the length of the shoot from its base to the bud and to the leaf tip (Table 2). The highest shoot length values were registered with the application of the low dose of fertilizer, and diminished with the high dose. When “high fertilization” was applied, the value for shoot length was similar, regardless of whether or not a root starter was used. Without the latter, the shoot length increased with a higher dose of fertilizer (figures 1 and 2).

Longitud del brote hasta la yema = Length of the shoot from the base to the bud; Fertilización = Fertilization; Alta = High; Baja = Low; Con enraizador = With root starter; Sin enraizador = Without root starter.

Figure 1 Effect of the root starter * fertilizer interaction on the length of the shoot from the base to the bud. 

Longitud brote hasta punta hoja = Length of the shoot from the base to the leaf tip Fertilización = Fertilization; Alta = High; Baja = Low; Con enraizador = With root starter; Sin enraizador = Without root starter.

Figure 2 Effect of the root starter * fertilizer interaction on the length of the shoot from the base to the leaf tip. 

Triple interaction

Although the use of cuttings with various levels of hardness is advisable for different groups of species, those of hard wood tend to contain more carbohydrates than those of soft wood (Hartman and Kester, 1991). Hence, regardless of the level of fertilization, IBA on hard cuttings favored the accumulation of the fresh root weight. This tendency was also present in soft cuttings with “low fertilization”, and was even more pronounced; however, the highest value was attained using “high fertilization” without a root starter (Figure 3).

Peso fresco de raíz = Fresh root weight; Fertilización alta= High fertilization; Fertilización baja= Low fertilization Low; Con enraizador = With root starter; Sin enraizador = Without root starter. A) Hard cuttings; B) Soft cuttings.

Figure 3 Effect of the interaction between the cutting type, the use of a root starter, and fertilization on the fresh root weight.  

The greatest response for fresh root weight in the hard wood cuttings with “high fertilization” may be partly related to K, as supplementation with this nutrient stimulates root growth, induced by the auxins. Likewise, K is very important for the emission of adventitious roots, since it influences the cellular expansion, the turgor pressure, the moisture content of the cell, and the stomatal action, and therefore modulates the effects of hydric stress in the cuttings due to their initial absence of roots. This was observed in hard wood cuttings of Junipers virginiana L., with a correlation between K and the percentage of rooting (^{Henry et al., 1992}).

High fresh weights occurred in soft cuttings with the application of a root starter and “low fertilization”, or without a root starter and “high fertilization”; this tendency can be associated with the concentration of nutrients in the cuttings when they are cut, which influences the redistribution of K and differs between cutting types (^{Henry et al., 1992}).

The concentration of nutrients may be expected to be lower in harder, more lignified cuttings; this caused the soft cuttings, which initially contain more K, to inhibit the accumulation of fresh weight when extra K was added and a root starter was applied. Similar interactions have been documented in other species. For example, in semi-hard Dalbergia sissoo Roxb. wood cuttings, fertilization (with NPK) favored rooting and shooting; furthermore, the indolebutyric acid-fertilizer interaction proved to be significant (^{Bakshi, 2008}).

Other plant quality indicators

Dry shoot weight / dry root weight ratio

With this indicator, a ratio equal to one means that the aerial biomass is equal to the underground mass. However, if the ratio is below one, then the underground biomass exceeds the aerial mass; the opposite is true when the value is above one. In the present study, all the values were above one; therefore, the root exhibited little development in relation to the aerial part, which impacts the plant’s resistance to hydric stress in the field (^{Rodríguez, 2008}; ^{Prieto et al., 2009}), although it was not statistically significant (Table 6). Similar values to those obtained when assessing the leaves and roots of Acer negundo (Table 7) have been cited for Tabebuia rosea (Bertol.) DC. and Swietenia humilis Zucc., both of which were rated as low-quality plants (^{Sáenz et al., 2014}). This indicator was developed for seed-produced individuals; therefore, it is advisable to carry out more tests on cuttings of other broadleaves and in plantations in an urban or rural environment.

Dickson’s index

Dickson’s index is considered as one of the best parameters to indicate the quality of the plant, as it expresses balance between the distribution of the mass and the robustness of the plant, preventing the selection of out of proportion plants and the rejection of shorter but more vigorous specimens (^{García, 2007}). In Mexico, the plant quality is generally categorized according to Dickson’s index. Plants with values of 0.20 or under are rated as low-quality; medium quality plants have values of 0.45 to 0.20, and high-quality plants are those with a value of 0.5, or above if they are conifers (^{Sáenz et al., 2014}); the higher the value, the higher the quality (^{Basave et al., 2017}).

The variance analysis evidenced statistical significance (P<0.05), influenced by the cutting factor (Table 6). Tukey’s test indicated the best results for this index in hard cuttings (Table 7), which may be directly related to a higher root production (^{Guzmán et al., 2012}).