Plantation site

The experiment was established in the ejido Aquiles Serdán, municipality of Durango, Durango, Mexico. The ejido is located at the geographic coordinates 23° 53' 39.24'' N and 104° 33' 43.94'' W, at an altitude of 1 898 m. The average annual temperature is 16.59 °C; the coldest month is December with an average temperature of 2.58 °C and the warmest month is June with an average temperature of 28.91 °C. The average annual rainfall is 715.8 mm (^{Instituto Nacional de Investigaciones Forestales Agrícolas y Pecuarias [INIFAP], 2017}). The land has slight waves and a slope lowerthan 3 %. The soil has loamy texture, neutral pH (6.9) and low organic matter (1.05 %). These and other soil physicochemical properties were determined based on the Norma Oficial Mexicana NOM-021-RECNAT-2000 (^{SEMARNAT, 2002}); the information obtained is shown in Appendix 1.

Establishment, treatments and experimental design

The seed was collected from 20 trees located in Alberca, Coahuila (23° 32´ N and 100° 52´ W, altitude 2 400 m). The plants were grown in a nursery using polystyrene containers with 77 cavities, with a capacity of 170 mL per cavity. The plantation was established on July 24, 2014 using 10 month-old seedlings with average height and diameter of 34 cm and 3 mm, respectively. Soil preparation was made with agricultural tractor (subsoiling at 40 cm). The plants were deposited in strains of 25 to 30 cm deep and their distribution was in real frame with a spacing of 2 m between plants and lines.

The experiment had a completely random design; seven fertilizers were evaluated in doses of 7 and 14 g and a control (unfertilized). Table 1 shows the nutrient composition of the fertilizers evaluated. Each treatment had four replicates and each experimental unit consisted of 12 plants. Prior to planting, fertilizers were incorporated into the bottom of the strains, to avoid direct contact with the roots; then about 40 g of soil was added and the plants were placed. Plants were not irrigated during the evaluation period.

Plant evaluation

The test was carried out on July 30, 2015. Survival was quantified by a binary scale, assigning values of 0 to dead plants and 1 to living plants. Height increments (cm) and base diameter (mm) were obtained from the differences between the initial and final measurement of each plant. The stem volume was estimated with the following equation (^{Balám-Che, Gómez-Guerrero, Vargas-Hernández, Aldrete, & Obrador-Olán, 2015}):

V = 0.0333 0.7854 * DBT2* H

On the other hand, the foliar concentration of N, P and K was determined in needles from the last flow of growth of three seedlings per treatment; also, dry weight of 100 dehydrated needles was obtained using a drying oven at 70 °C for 72 h. With these data, vector nomograms were made (^{Haase & Rose, 1995}; ^{Timmer, 1991}) to know the relationship between nutrient concentration and biomass. N was estimated with the Kjeldahl method, P by means of yellow vanadomolybdate complex colorimetry and K by means of atomic emission (^{SEMARNAT, 2002}).

Statistical analysis

Analyzes were carried out using the statistical software R 3.2.3 (^{R Core Team, 2015}). To know the variability of growth explained by the addition of nutrients, coefficients of determination (R²) were obtained from a linear regression between the contents of N, P₂O₅ and K₂O in each fertilizer and height and diameter increment variables. Seedlings survival values and foliar N-P-K concentration were transformed with the arcsine and square root function, since these values were expressed in percentage (^{Steel & Torrie, 1988}). The Kruskal-Wallis non-parametric statistical test (^{Kruskal & Wallis, 1952}) was used for the variables evaluated, because the data did not meet the assumption of normality; subsequently, a mean separation test of Bonferroni-Dunn was carried out (P < 0.05) (^{Pohlert, 2014}). The statistical model used was as follows:

Yi=µ + αi+ ei

Height, diameter and volume increments of Pinus greggii var. greggii

The application of K favored the growth of seedlings (Table 2), even though this element was not deficient in the site (0.99 cmol_c·kg^-1, Appendix 1). The NOM-021-RECNAT-2000 indicates average values of K from 0.3 to 0.6 cmol_c·kg^-1 of soil (^{SEMARNAT, 2002}). In addition to fertilizer, factors such as water availability, type of soil and temperature influence the growth of plants, which together are associated with the capacity to assimilate nutrients (^{Núñez, 2013}).

In agreement with Table 3, volume and height and diameter increases showed significant differences (P < 0.001) due to the effect of fertilization treatments. Slow-release fertilizers had increments similar to those caused by agricultural fertilizers in the three variables evaluated; slow-release fertilizer 12-25-12 stood out, although, statistically, the effect was the same as the rest of the treatments. In comparison with the control, only the slow-release fertilizers 12-25-12, 09-23-14 and 18-06-12 of N-P-K promoted height and diameter increments, as well as wood volumes, significantly higher.

On the other hand, according to Table 4, the doses of fertilizers (7 and 14 g) caused significant differences in height (P = 0.002), diameter (P < 0.001) increases and in volume (P = 0.001) only with respect to the control (without fertilizer). ^{Oliet et al. (2009)} tested two slow-release fertilizers (9-13-18 and 17-10-10 NPK) in doses of 3, 5 and 7 g, and an unfertilized control, and determined that the formulation 9-13-18 of NPK in dose of 7 g generated the greatest increments in diameter and height of Pinus halepensis Mill.

In general, agricultural fertilizers had lower efficiency in the nutrient contribution. This agrees with the results of ^{Reyes-Millalón et al. (2012)}, who found a better response in the growth of P. radiata D. Don with slow-release fertilizers than with water-soluble fertilizers and the control (without fertilization); ^{Everett, Hawkins, and Kiiskila (2007)} also observed a similar trend with Pseudotsuga menziesii var. glauca (Beissn.) Franco. ^{Gotore, Murepa, and Gapare (2014)} found a particular response in the growth of Pinus patula Schiede ex Schltdl. & Cham.; these authors suggest that the addition of nutrients is not necessary in the initial stage of planting. However, the present study demonstrates that nutrient management is important in the plantation of P. greggii var. greggii, since it generated a significant response in growth; this is indicated by ^{Román, Vargas, Baca, Trinidad, and Alarcón (2001)}, who demonstrated that the nutrient management of P. greggii var. australis with nutrient solutions increased height, diameter and biomass.

The variables height, diameter and volume had significant differences between treatments; however, it is convenient to make future evaluations of the plantation to determine if the same differences remain or if others occur, which will probably happen during crown closure, which is the stage of greatest nutrient demand.

Survival of Pinus greggii var. greggii

Survival did not have significant differences due to the effect of fertilization (P = 0.799) and the doses tested (P = 0.241). It is possible that the reduced percentage of survival of control trees (52.1 %), agricultural fertilizers 16-16-16 (59.4 %) and 18-46-00 (64.4 %), as well as slow-release fertilizers 09 -23-14 (64.6 %), 12-25-12 (67.7 %), 14-14-14 (63.5 %), 17-17-17 (59.4 %) and 18-06-12 (67.7 %), owes more to the conditions of low moisture availability in the soil, than to plant needs for nutrients (^{Haase & Jacobs, 2013}; ^{Reyes-Millalón et al., 2012}). ^{Perez, Valeri, Cruz, and Vasconcelos (2016)} indicate that an adequate or even high level of K promotes the most efficient use of water in plants. In this study, probably, K exerted this effect and, therefore, influenced survival.

It is worth mentioning that the values obtained by any of the treatments evaluated exceed the national figure of survival after one year of planting in the reforestations: 36 % according to ^{Wallace et al. (2015)} and <50 % according to ^{Burney et al. (2015)}. In addition, taking into account that P. greggii var. greggii is an exotic species in Durango, survival values are acceptable, even though CONAFOR demands 80 % survival in the first year.

Foliar concentration of nutrients

Table 5 shows the results of the determination of foliar concentration of N, P and K in P. greggii var. greggii. The fertilizers used did not generate significant differences in foliar concentrations of N (P = 0.536) and K (P = 0.603), but there were differences in those of P (P = 0.030). The existence of similar concentrations of N may be due to its condition of deficiency in the site, which allows to suppose that the nutrient was absorbed by the plants and diluted within the biomass (^{López-López & Estañol-Botello, 2007}).

Even though it is an adequate nutrient in soil (Appendix 1), K favored more the growth of seedlings (Table 2), because it probably promoted a better use of water (^{Perez et al., 2016}). If this effect had occurred, it is likely that foliar K has tended to be diluted in the biomass, generating only small and insignificant changes in the foliar concentrations of the nutrient (Table 5) (^{López-López & Alvarado-López, 2010}). In contrast, the foliar concentrations of P were modified as the nutrient was added in the fertilizing formula; apparently the immediate release of nutrients by the agricultural fertilizer 16-16-16 allowed the increase of foliar P, while, the slow-release fertilizer, due to its durable condition, only released small portions of nutrients for a longer time. This implies that P was deficient in the site before applying the treatments (3.0 mg·kg^-1, Appendix 1), since the NOM-021-RECNAT-2000 indicates as mean values 5.5 to 11 mg·kg^-1 (^{SEMARNAT, 2002}).

Regarding fertilizer doses, there were no significant differences in foliar concentration of N (P = 0.079), P (P = 0.535) and K (P = 0.592). In the comparison of nutrient foliar concentrations, the absence of differences between treatments is frequent, due to the presence of dilution effects due to seedling growth (^{López-López & Alvarado-López, 2010}).

The vector analysis, shown in Figure 1a, indicates that the slow-release fertilizers 09-23-14, 14-14-14, 18-06-12 and the agricultural fertilizer 18-46-00 N-P-K, generated an increase in the weight of needles and the reduction in the concentration of foliar N, nutrient that was deficient before applying it; such an interpretation agrees with the soil analysis (0.05 % is considered low [^{SEMARNAT, 2002}]) and the behavior of the foliar concentration of N (Table 5) (^{López-López & Alvarado-López, 2010}). This same effect is observed in P with the slow-release fertilizers 09-23-14, 14-14-14 and 18-06-12, and the agricultural fertilizer 18-46-00 (Figure 1b), and in the case of K with the slow-release fertilizer 06-18-12 (Figure 1c). This indicates that these nutrients were insufficient to increase the biomass of needles, which is often directly related to the total biomass of seedlings (^{Timmer & Stone, 1978}).

Figure 1 Vector nomograms of fertilization treatments with different formulations in Pinus greggii var. greggii. AF: Agricultural fertilizer and SRF: Slow-release fertilizer. 

The nomogram of N also shows that the agricultural fertilizer 16-16-16 and that of slow-release 12-25-12 have a vector that indicates a decrease in the weight of needles, and an increase in the foliar concentration of N (Figure 1a). Apparently it is an effect of N concentration, due to the limitation of growth by some factor probably linked to fertilizers. Likewise, the vector of the slow-release fertilizer 17-17-17 reflects weight reduction of needles and foliar concentration, this may be due to an antagonistic effect with another nutrient (^{López-López & Alvarado-López, 2010}) that decreased the availability of N. That same tendency is observed in P with the slow-release fertilizer 17-17-17 and 12-25-12 of N-P-K, while in K that was only observed with the slow-release fertilizer 12-25-12 of N-P-K.

On the other hand, in the nomogram of K it is observed that the vectors of slow-release fertilizers 09-23-14, 14-14-14 and that of agricultural use 18-46-00 indicate increase in weight of needles and increase in concentration (Figure 1c), this suggests that K was deficient before treatment; K insufficiency was corrected and a high consumption occurred (^{López-López & Alvarado-López, 2010}). According to soil analysis, K is sufficient and it seems that this nutrient improved the water status of the seedlings and their growth (^{Román et al., 2001}), despite the fact that water is a limiting factor in the area.

In general, slow-release fertilizer 17-17-17 induced reduction of foliar biomass (Figure 1). Apparently, P was the deficient nutrient in the plots that received slow-release fertilizer 17-17-17 from N-P-K (Figure 1b), in which case, the foliar concentration decreased up to 0.11 %. P is a nutrient highly prone to pH conditions (<5 precipitation, > 5 adsorption, occlusion and reversion) and it is likely that it has restricted absorption (^{Núñez, 2013}) in fertilizer 17-17-17, decreasing the growth, including that of the roots; the same effect may also have reduced the absorption of N and, to some extent, that of K.