Location of the experiment

The experiment was conducted in the Sahuatoba forest nursery, belonging to the Ministry of Natural Resources and Environment of the Government of the state of Durango, located between 24°01'34.4" N and 104°41'34.3" W, at an altitude of 1 890 m, annexed to Sahuatoba Park in the city of Durango, Mexico.

The spread was carried out in a forest plant acclimatization area, covered by shade netting with 50 % light filtration. The cuttings were placed in black polyethylene containers, 12.7 cm diameter×20 cm long and 2.5 L capacity, which were filled with the substrates to be evaluated. 20 cuttings were placed in each container to generate one experimental unit. The containers were then placed on the floor on a white plastic base.

Collection and preparation of cuttings and origin of parent plants

Cuttings were collected in mid-April from 50 eight-year-old parent plants, propagated in containers of 38.2 cm diameter and 60.0 cm height, with 68.7 L capacity. These plants were produced in the “Praxedis Guerrero” forest nursery, which belongs to the Ministry of Natural Resources and Environment of the Government of the state of Durango, located at 23°57'01.0" N and 104°34'08.6" W, at an altitude of 1 883 m, at km 12 of the Durango-El Mezquital highway, Durango, Mexico.

The cuttings were collected the middle part of the canopy, in new (tender cuttings) and one-year-old shoots (lignified cuttings); once cut, in order to keep them hydrated and prevent fungal damage, they were placed and transported in bulk in 20 L buckets with water and Captán^® fungicide (2 gL^-1) and were put to root 20 hours after collection. Vigorous parent plants with abundant foliage and a symmetrical and uniform crown were selected.

In order to prepare the vegetative material, a transverse cut was made at the bottom of each cutting with pruning shears, exposing the cambium at the base. During rooting, three factors were evaluated: a) four rooting agents and a control, b) two conditions of maturity of the cuttings, and c) two types of substrates. A total of 20 treatments (5×2×2) with four replications were assessed (Table 1). Each experimental unit consisted of 20 cuttings.

Three powdered rooting agents with different concentrations of indole-3-butyric acid (IBA) were evaluated: Radix^® 1 500 (0.15 % IBA), Radix^® 10 000 (1 % IBA), and Raizone^® Plus (0.06 % IBA). Also tested were a liquid rooting agent (Raizplant^® 500) containing a combination of phytohormonal compounds with a 0.05 % concentration and a control (without rooting agent). The rooting agents were impregnated in the first 5 cm of the tip of each cutting. The maturity conditions considered were: tender cuttings (obtained from the last growth) and lignified cuttings (collected from the previous year's growth). The substrate used was 100 % perlite and a mix base (composed of 55 % sphagnum peat moss, 24 % vermiculite, and 21 % perlite).

Irrigation was performed every third day manually with a 7 L watering can; 2 L of water per irrigation were applied to each experimental unit. The temperature and relative humidity were measured with a maximum and minimum thermohygrometer (HOBO^®) and the following values were recorded: maximum temperature, 31.2 °C, minimum temperature, 22.8 °C, average temperature, 26.9 °C, 76 % maximum relative humidity, 58 % minimum relative humidity, and 69 % average relative humidity. During the four months of evaluation, three daily measurements of temperature and relative humidity were taken.

The survival rate and the tap root length of the rooted cuttings were recorded four months after the cuttings were established. The survival of each stake was assessed based on whether or not it had live tissue. In order to measure the tap root, all rooted cuttings were extracted from the container (Figure 1).

Figure 1 Extraction of cuttings in order to evaluate tap root growth. 

Statistical analysis

The results of survival and root growth were analyzed with the Shapiro-Wilk and Kolmogorov-Smirnov tests to determine the type of statistics to be used. Since a normal distribution of the survival and root growth data was not detected, the Kruskal Wallis nonparametric test was performed to determine possible significant differences between the factors evaluated (type of substrate, type of rooting agent, and condition of the development of the cuttings); paired medians were compared by applying Dunn's post hoc test according to the Bonferroni method.

The effect of the interaction between the evaluated factors on survival and root growth was analyzed with Dunnett's median test. These analyses were performed on the R^® statistical software platform, version 3.5.3 (^{R Core Team, 2020}).

Survival and root growth by factor

Statistical differences (p=5.3x10^-5) were observed in survival from the effect of the type of rooting agent and the maturity condition of the cuttings; however, there were none linked to the type of substrate utilized (p=0.92). Significant differences in taproot growth also occurred as an effect of the maturity condition of the cuttings (p=2.8x10^-9), the substrate type (p=5x10^-4), and the rooting agent type (p=0.008).

Bonferroni's comparison of medians showed significant differences in the survival of cuttings in the following comparisons: Raizone^® Plus vs Radix^® 1 500, Radix^® 1 500 vs Raizplant^® 500, and Radix^® 1 500 vs control. With regard to root growth, significant differences were observed in the following comparisons: Raizone^® Plus vs control, Raizplant^® 500 vs control, and Radix^® 10 000 vs control (Table 2).

As for the rooting agent utilized, the highest survival was obtained in cuttings rooted with Radix^® 1 500 (32.8 %); the rest of the treatments were placed in the following statistical groups: Radix^® 10 000 (17.2 %), Raizone^® 500 (12.8 %), Raizplant^® 500 (8.7 %), and control (5.9 %) (Figure 2a). The highest tap root growth was observed in the unrooted cuttings (control), with 10.75 cm average length. The intermediate group was composed of cuttings rooted with Radix^® 1 500 and Raizone^® 500, exhibiting a root growth of 8.61 and 7.89 cm, respectively. The lowest root growth, of 7.84 and 6.73 cm, respectively, was observed in cuttings rooted with Radix^® 10 000 and Raizplant^® 500 (Figure 2b).

Rd-10 000 = Radix^® 10 000; Rz = Raizone^® Plus; Rp = Raizplant^® 500; Rd-1 500 = Radix^® 1 500; C = Control. Different lowercase letters indicate significant differences between treatments (p<0.05) according to Dunn's test for median difference.

Figure 2 Survival and primary root growth of Cupressus macrocarpa Hartw. ex Gord. var. lutea Topacio cuttings, by the effect of three factors: a) and b) by type of rooting agent; c) and d) by the maturity condition of the cutting, and e) and f) by substrate type. 

As a result of the maturity of the cuttings, the survival rate of the lignified cuttings was 21 %, while that of the tender cuttings was 10 % (Figure 2c). According to ^{Shiembo et al. (1997)}, tender cuttings exhibit greater natural auxin generation, and when the concentration of this hormone is increased by the application of IBA, the toxicity increases considerably, which causes greater mortality of cuttings. The root length in lignified cuttings was 9.1 cm, while in tender cuttings it was 6.2 cm (Figure 2d).

Regarding the type of substrate, there was no statistical difference in survival between cuttings rooted in a mix base (16 %) and those rooted in perlite (15 %) (Figure 2e). On the other hand, root growth was higher in cuttings rooted in perlite (9.07 cm), compared to those rooted in the mix base (7.63 cm) (Figure 2f).

In a rooting study of cuttings of Cupressus cashmeriana Royle ex Carrière in which tender and lignified cuttings were compared, survival was higher in the lignified ones (52.8 % vs 31.9 %) (^{Raup and Taylor, 2012}). In contrast, ^{Lahiri (1975)} cited for the same species a survival rate of less than 5% in the rooting of lignified cuttings.

In the rooting of cuttings of woody species, it has been pointed out that the more mature the tissues are, the more auxin is produced after the cuts (wounds) are made in the cutting; auxin is a phytohormone that promotes plant growth and is important for generating roots (^{Scalabrelli and Couvillon, 1986}).

Auxin production during rooting promotes the generation of callus, which favors the survival of the cuttings (^{Mackenzie et al., 1986}) as well as root proliferation (^{Cabrera et al., 2022}).

In a rooting study of Cupressus macrocarpa Hartw. ex Gord. var. goldcrest, ^{Mamani and Fernández (2017)} achieved 27.8 % survival when the cuttings were propagated in a medium containing equal parts sphagnum peat moss, charcoal and silt. The survival rate attributed to the substrate in the present study was considerably lower, with values of 16 % for both perlite and the mix base. This may be due to the fact that perlite has greater porosity and aeration, which prevents excess humidity, reduces the presence of fungi, and favors the formation of roots (^{Rivera-Rodríguez et al., 2016}).

^{Rosier et al. (2004a)} evaluated the survival in rooted cuttings of Pinus virginiana Mill. when applying 10 and 60 mMol of IBA, representing 0.2 and 1.2 % IBA per solution liter, respectively, and obtained a survival rate of 85 and 55 %. In the rooting of cuttings of Abies fraseri (Pursh) Poir., a survival of 100, 97 and 77.5 % is documented when using concentrations of 10 mMol (0.2 % IBA), 1 mMol (0.02 % IBA), and 60 mMol (1.2 % IBA), respectively (^{Rosier et al., 2004b}).

The higher survival obtained in the present work, with an intermediate dose of IBA (0.05 %) indicates that the hormone concentration should be high enough, but without excesses that may negatively affect survival, such concentration may increase as a consequence of the toxicity generated by the rooting agent on the cuttings. This effect was observed in the rooting of cuttings of Pinus patula Schiede ex Schltdl. et Cham. by ^{Bautista-Ojeda et al. (2022)}, who indicate a survival of 56.7 % when using a rooting agent with a 0.5 % concentration of IBA. ^{Saini (2001)} also observed toxicity caused by a high concentration of IBA, he determined that the survival rate of rooted cuttings of Cupressus cashmeriana was less than 50 % when he applied a rooting agent with more than 0.8 % of IBA, while when using concentrations of IBA below 0.5 %, the survival rate was higher than 70 %.

In the rooting of Pinus virginiana cuttings, the highest root growth (60 cm) was obtained when high concentrations of IBA were applied (60 mMol-1.2 %) (^{Rosier et al., 2004a)}. This result contrasts with those obtained in the present study, in which significant differences were recorded when comparing between the root growth of cuttings without a rooting agent (control) and cuttings with an IBA concentration of 0.8 % (Raizplant^® 500) and 0.05 % of auxins, cytokinins and gibberellins (Raizone^® Plus). According to ^{Benítez et al. (2002)}, rooting success is higher in semi-hardwood species; this would explain the greater growth in roots of P. virginiana, whose wood density varies from 0.61 g cm^-3 (^{McCormack et al., 2012}), compared to Cupressus macrocarpa, which has an average wood density of 0.436 g cm^-3 (^{Todoroki et al., 2015}).

Root growth was higher in cuttings with the 0.15 % IBA rooting agent (Radix^® 1 500), compared to that achieved with rooting agents with higher IBA concentrations (Raizone^® Plus with 0.8 % IBA and Radix^® 10 000 with 1 % IBA), which is the response to an apparent toxicity on the species under study in the presence of high concentrations of hormone induced by the rooting agent (^{Blythe et al., 2007}). The higher root growth of cuttings without a rooting agent (control) could be due to the fact that some species naturally generate enough phytohormones to achieve rooting (^{Valdés et al., 2003}).

Five months after the cuttings were made, ^{Mamani and Fernández (2017)} observed an average root growth of 7.19 cm when rooting the cuttings of Cupressus macrocarpa var. goldcrest in a sphagnum peat moss substrate. When using as a substrate a combination of clay, sphagnum peat moss, and sand in a proportion of 2:7:3 for the rooting of Chamaecyparis lawsoniana (A. Murray) Parl., a root growth of 6.28 cm was reported three months after the cuttings were made (^{Spanos et al., 1999}). Considering that the evaluation period for the base mixture was similar, the results obtained by these authors are similar to the average ones of the present work, presumably because this mix, which includes sphagnum peat moss, is a substrate with good porosity which provides support to the plant and has good levels of drainage and aeration, characteristics that are crucial to favor the propagation of forest species (^{Hartmann et al., 2010}).

Root survival and growth due to the effect of the interaction between factors

Significant statistical differences were recorded in the survival rate (p=0.008) and root growth (p=0.028) of the cuttings due to the effect of the triple interaction between the factors of the evaluated treatments (Table 3).

The best survival rate (45 %) was obtained with treatment 13: lignified cuttings-perlite-Radix^® 1 500 (0.15 % IBA), followed by treatment 4 with 37.5 %, whose common factor with the superior treatment was the rooting agent Radix^® 1 500, indicating that this was important for the rooting of the cuttings. The survival rates of treatments 1, 2, 3, 4, 7, 11, 15 and 19 coincide with each other in certain statistical groups, and even with treatment 13 when related to intermediate groups. 77.8 % of these treatments had in common the use of lignified cuttings. The lowest survival values (1.2 to 1.5 %) occurred in treatments 8, 10 and 20, in which the common factor was the tender cuttings; in two of them no rooting agent was applied, and in two of them the substrate was a base mix (Table 3).

In the rooting of semi-lignified cuttings of Pinus virginiana (^{Rosier et al., 2004a)} and lignified cuttings of Abies fraseri (^{Rosier et al., 2004b}) propagated in a substrate composed of 60 % perlite and 40 % sphagnum peat moss, the survival rate was higher than 90 % and 85 %, respectively, when a low IBA concentration was used (5 mMol=0.1 % IBA). These results coincide with those of the present study, at least in the interaction between the rooting agent with a low IBA concentration (0.15 %) and the lignified cuttings.

The root growth was statistically outstanding in treatments 10 (tender cuttings propagated in the base mix) and 19 (lignified cuttings propagated in perlite), both characterized by the absence of a rooting agent; the other treatments had a similar effect on root growth.

^{Rosier et al. (2004a}; ^2004b) evidenced a higher root growth in tender cuttings of Pinus virginiana when using a base mix and a low dose of IBA (6 mMol=0.12 % IBA). They also achieved good root development in tender Abies fraseri cuttings with a low dose (12 mMol=0.24 % IBA). Root growth in these two species was 1.94 and 7.13 cm four months after the cuttings were made. Both their results and those documented in the present work show that a low IBA concentration favors root growth in tender and lignified cuttings of conifer species, as pointed out by ^{Castro-Garibay et al. (2019)}, who indicate that high concentrations of IBA can inhibit root growth.

Because the interactions with the greatest taproot growth do not coincide with those with the highest survival rates, between the two alternatives it is desirable to ensure more live cuttings; in this sense, ^{Cabrera et al. (2022)} document that rooted vegetative material can be nourished and hormones can be applied to promote root development. On the other hand, the objective of rooting cuttings is the massive cloning of adult trees (^{Medina et al., 2014}). For this reason, in the rooting of lignified lemon cypress cuttings, the use of perlite supplemented with Radix^® 1 500 as a substrate will allow a better survival of the cuttings, which in turn will facilitate the obtainment of abundant vegetative material to produce ornamental plants (^{Cabrera et al., 2007}).