Production conditions

The study was conducted in the nursery of the Facultad de Ciencias Forestales de la Universidad Juárez del Estado de Durango (School of Forest Sciences of the Juarez University of the State of Durango), located in Durango city, Durango State, at 24°00'49'' N and 104°40'58'' W, at 1 860 masl. The experiment lasted five months and the first four were under greenhouse conditions (covered with 720-gauge polyethylene plastic treated against ultraviolet rays, a 60 % shadow mesh was placed on the polyethylene plastic); during the fifth month the plant was out in the open. The average temperature recorded under greenhouse conditions was 27.9 °C and in the open air of 29.0 °C.

The plant was produced in polystyrene trays of 77 cavities, 35 cm wide, 60 cm long, 15 cm high and 170 mL per cavity. As a substrate, a mixture composed of 50 % peat (peat moss) and 50 % pine bark was used. Prior to sowing, the seed was subjected to a pregerminative treatment with immersion, for 60 seconds in water at 94 °C; to prevent the development of fungi, 2.5 g L^-1 of Benomyl was applied during sowing.

Assessed treatments

Nine treatments composed of two doses of controlled-release fertilizer and three types of water-soluble fertilizers were evaluated, as well as the control in which only water was applied (Table 1). The controlled release fertilizer (8-9 months of release) Multicote^® 18N-6P₂O₅-12K₂O + 2MgO + micronutrients (Haifa Chemicals Ltd.) was incorporated into the substrate in doses of 3 and 6 g L^-1. The water-soluble fertilizers were applied every 48 h at a constant dose of 100 mg L^-1. The fertilization process began 20 days after sowing (DDS) and lasted three months, with July 10, 2015 as deadline.

Evaluation

At 116 DSD, eight plants were evaluated per experimental unit. The response variables considered were: the height of the aerial part (cm), which was registered with a graduated rule of 30 cm (Trupper^® 14387) and the measurement was taken with approximation up to tenths of a centimeter; the diameter of the stem (mm), taken at the height of the root collar with a digital vernier with precision of hundredths of a millimeter (SURTEK^™ 122204).

For the dry biomass of the aerial part and the root system (g) a FELISA^™ FE-291D drying oven was used at 70 °C for 72 h, then weighed on a Ohaus^™ PA214 analytical balance. Each section was previously packed in paper bags with its respective registration data (treatment, replication and plant number). With the above variables, the robustness index was determined, which is an indicator of the quality of the plant.

The cost of the fertilizer used during the experiment was calculated based on the applied proportion of each type of water-soluble fertilizer per treatment. Additionally, the cost of controlled release fertilizer was included based on the defined treatments, with average prices of 2017.

Experimental design and statistical analysis

The treatments were distributed in a completely randomized experimental design with a factorial arrangement, with four replicates per treatment. As the data did not fulfill the assumption of normality, the Kruskal-Wallis non-parametric statistical test (^{Kruskal and Wallis, 1952}) was used for all the variables evaluated, in addition to the Bonferroni-Dunn means separation test (p <0.05) (^{Pohlert, 2014}). The statistical analysis was performed with the statistical software R 3.2.3 (^{R Core Team, 2015}). The statistical model used was the following:

Where:

Y _ij = Response variable

μ = Mean general effect

A _i = Effect attributed to the _i -^eth of the water-soluble fertilizer factor level

B _j = Effect attributed to the _j -^eth of the controlled release fertilizer factor level

(AB) _ij = Effect attributed to the interaction between the _i -^eth level of the A factor and the _j -^eth level of the B factor

e _ijk = Random error

Controlled release fertilizers

The application of controlled release fertilizer Multicote^™ had significant effects on the variables evaluated (p <0.001); in the height of the aerial part and the root collar diameter, the dose of 3 g showed statistical equality with respect to the specimens that received the dose of 6 g; however, in the total dry biomass the highest dose was located in the upper statistical group (Table 2). It is evident that the lack of supply of the controlled-release option caused a lower growth in the plants.

These results coincide with those of ^{Bustos et al. (2008)}, who evaluated the growth of three tree species (Nothofagus dombeyi (Mirb.) Oerst., Nothofagus nervosa (Mirb.) Oerst. and Eucryphia cordifolia Cav.) by applying three doses of the controlled release fertilizer Osmocote^™ (2.5 to 7.5 g L^-1); they concluded that the highest doses were the most favorable.

On the other hand, ^{Aguilera et al. (2016)} also observed that in Pinus montezumae Lamb seedlings. produced with three doses (4, 6 and 8 g L^-1) of Basacote^™ Plus, Multicote^™ and Osmocote^™ Plus on two substrates, the doses of 6 and 8 g L^-1 further stimulated their growth.

Water-soluble fertilizer

The use of water-soluble fertilizers during irrigation only recorded differences in the response variables (p <0.001) with respect to the control; that is, where it was watered without water-soluble fertilizer. In the treatments in which the different water-soluble fertilizers were applied in water, the results of the variables evaluated (p <0.001) were similar to each other (Table 2).

Controlled-release and water-soluble fertilizers

The application of the controlled release fertilizer and the water-soluble fertilizers generated significant differences in the height of the aerial part, in the root collar diameter and in the total dry biomass (p <0.001); in addition, they showed a robustness index slightly higher than that indicated by ^{Prieto et al. (2012)} (< 8), but which can be satisfactory (Table 3).

The height of the aerial part was outstanding with treatment 7, which consisted of the highest dose of the controlled-release product and with the water-soluble fertilizer Triple 19. As for the root collar diameter, all the treatments fertilized, either with the controlled release, the water- soluble or combined, were in the upper statistical group and registereddifferences greater than 1.0 mm with respect to the control. The values of the most outstanding treatments varied from 3.4 to 3.6 mm, with a maximum difference of 0.2 mm between them. The values of total dry biomass had an average fluctuation between treatments of 0.48 and 2.74 g, with the best results in treatments 4 to 9 (Table 3).

The Mexican Standard NMX-AA-170-SCFI-2016 "Certification of forest nurseries", establishes that Prosopis laevigata must have a height interval between 25 and 30 cm and a root collar diameter of the minimum root of 4 mm at five months of age (^{SCFI, 2016}). In this study, heights were reached in approximately four months, except for treatments 1 and 2; with regard to the root collar diameter , the recommended minimum was not achieved, but the values were very close; therefore, all treatments where fertilized were effective. The root collar diameter is considered one of the most important variables to define the quality of the plant (^{Sáenz et al., 2010}; ^{Tsakaldimi et al., 2013}), and it is related to the height, as well as to the radical development of the plant (^{Jacobs et al., 2009}). On the other hand, ^{Prieto et al. (2009)} and ^{Sáenz et al. (2010)} argued that plants with large diameters support bending and resist more damages caused by insects and animals, so this criterion is considered important in the early performance of a plantation.

^{Prieto et al. (2013)} evaluated six mixtures of substrates in the production of Prosopis laevigata, and applied 7 kg m^-3 of controlled release fertilizer 15-07-15 of NPK, supplemented twice a week with the water soluble Peters^™ Professional (PP) fertilizer, growth (20-09-19 of NPK) in 100 ppm, and PP finalizer (4-25-35 of NPK), in a dose of 100 ppm. After 21 weeks of plant growth, they concluded that the average diameter of the plant fluctuated between 2.91 and 3.05 mm, while the height, from 19.7 to 25.7 cm. In the present study, a greater increase was observed in the two variables in less time (30 days less), which corroborates the influence of N P and K.

^{Prieto et al. (2012)} established that the robustness index for this species must be < 8, which defines a good balance between height and diameter, and allows the plant to survive under conditions of low moisture and desiccation from wind, due to the resistance of woody tissue, in addition to containing water reserves and photosynthates; a higher value describes a disproportionate plant and susceptible to damage by wind, drought and frost (^{Rodríguez, 2008}). Derived from the above, the treatments that received some nutrient source revealed adequate robustness indexes, which shows that the fertilization routines used are correct (Table 3).

Based on the results obtained in the present study, it can be seen that the factor that marked differences between treatments was the controlled release fertilizer applied in two doses, and even with best results in the highest dose (6 g L^-1); ^{Rose et al. (2004)} noted that the most outstanding feature of this type of fertilizer is its ability for nutrients to be supplied once and that they are released gradually and for a long time, which in turn prevents the loss of nutrients by leaching. On the other hand, the same authors indicate that it is convenient to evaluate mixtures of controlled-release fertilizers and water-soluble products, so that the optimal form of nutrition in plants can be determined.

Cost analysis of fertilizers in plant production

The cost from the application of controlled-release fertilizers and water-soluble fertilizers was higher in treatment 7 with MXN$ 0.077 per plant produced; and, in contrast, the lowest cost was that of treatment 2 with MXN$ 0.026 per plant, by investing only in the controlled release fertilizer in a dose of 3 g L^-1 of water (Table 3).

Thus, to produce one million plants using treatment 7, the cost due to fertilizer would be MXN$ 77 000.00 while with treatment 2 it would only require MXN$ 26 000.00 However, the response of the morphological variables that define plant quality show that treatment 5 is the best, given that in most of the variables evaluated there were no statistical differences with respect to treatment 7 and its cost due to fertilization, for the production of one million plants in this treatment amounts to MXN $50 000.00 which represents a saving of 35 %; that is, MXN$ 27 000.00 (Figure 1).

Costo = Cost; Tratamientos = Treatments

Figure 1 Fertilizer cost per treatment for the production of one million plants. 

^{Aguilera et al. (2016)} refer to the costs of the fertilizer, and determined that the controlled release fertilizer 18-6-12 of N-P-K is the most inexpensive to make plantations of Pinus montezumae for restoration purposes; this same product was tested in the present study, and in the same way a significant difference was found in the total dry weight. In addition, it is cheaper to fertilize with 3 g than with the higher dose (6 g), but the higher dose increases the morphological characteristics of the plant. Finally, it is feasible to obtain the minimum recommended sizes for the species with any of the two doses of fertilizer.