Study area

The work was carried out in two plots located in the aforementioned Mixteca Alta region: in Magdalena Zahuatlán locality and in Tlacotepec Plumas.

The Tlacotepec Plumas lot is located between 17°52’11.02’’ N and 97°26’16.97’’ W at 2 143 masl. There, the climate is temperate semi-dry, with an average annual temperature of 16 °C, average rainfall of 420.8 to 650 mm; soils have 7.66 pH, 2.80 % organic matter content, 16.50 mg kg^-1 phosphorus and 0.16 % nitrogen (^{Sedesol-CIBCEC, 2003}).

The Magdalena Zahuatlán lot is located at 17°24’14.97’’ N and 97°12’32.22’’ W, with 2 166 meters above sea level. The climate in the place is temperate semi-dry, with an annual average temperature of 15 °C, average rainfall of 650; soils have 8.12 pH, 2.68 % organic matter content, 13.1 mg kg^-1-phosphorus and 0.08 % nitrogen (^{Valencia et al., 2006}; ^{INEGI, 2010}, ^{Velasco et al., 2012}).

Experimental material

With the plan of the plantation provided by the INIFAP-Valles Centrales de Oaxaca Experimental Field, the distribution of the provenances within the plot was specified. The plantations were established in 1997 with 13 provenances of P. greggii, of which six grow naturally in northern Mexico (P. greggii var. greggii Engelm.) and seven in central Mexico (P. greggii var. australis Donahue & López) (Table 1).

The plantation was carried out in a common strain system of 40 × 40 × 40 cm in a staggered arrangement at a distance of 3 m, with an experimental design of complete random blocks with 12 blocks (replications) and nine plants per experimental unit (3 × 3), which gave a total of 1 404 trees per location.

Measurement of growth parameters

The study plantations are 17 years old. The three central trees of the experimental unit were selected, considering that they showed lower competition effects with trees from other sources. The height of each one was measured with a Haga altimeter, the normal diameter at 1.30 m from the ground and the basal diameter at 10 cm from the ground with a Forestry Suppliers Model 283D/5M diameter tape, and the crown diameter with two measurements, one in the north-south direction and another in an east-west direction with a Spencer Logger’s Model 975DC tape. A total number of 36 trees were measured by provenance in each location, with a total of 468 trees per location. A dead tree or a stump without a tree was recorded, with which survival was calculated. From the crown diameter, the crown area was calculated for each individual.

Data analysis

An analysis of variance was performed by using a randomized complete block model to compare the means of height, normal diameter, basal diameter, crown diameter and crown area between provenances and localities. When significant differences were defined, the Tukey test was applied with a confidence level of 95 % for the comparison of means. The R program was used with the agricolae library version 1.2-3 (^{Mendiburu, 2015}). The survival data were transformed with the arcsen function (p '= arcsin√p) for homoscedasticity and normality in which p are the proportions from 0 to 1 (^{Zar, 1999}).

Where:

y _ijk = Value of the observation in the i ^th locality, in the j ^th provenance of the k ^th block

µ = General mean effect

L _i = Effect of the i ^th locality

P _j = Effect of the j ^th provenance

B _k = Effect of the k ^th block

LP _ij = Effect of the interaction between the i ^th locality and the j ^th provenances

LB _ij = Effect of the interaction between the i ^th locality and the j ^th block

E _ijk = Effect of the interaction between the i ^th locality of the j ^th provenance and of the k ^th block (experimental error)

Differences between localities

The statistical analysis showed the existence of significant differences between localities (p <0.001) in all the assessed variables (Table 2). In Magdalena Zahuatlán, the provenances recorded an average height of 7.18±2.85 m, 14.07±6.03 cm in basal diameter, 10.55±5.54 cm in normal diameter, 3.48±1.24 m in crown diameter and 10.74±7.55 m² tree^-1 in the crown area. In Tlacotepec Plumas, the average growth was higher, with an average height of 10.50±2.20 m, 18.28±5.48 cm in basal diameter, 14.33±4.64 cm in normal diameter, 4.22±1.13 m in crown diameter and 15.03±8.03 m² tree^-1 in the crown area.

The higher growths of Tlacotepec Plumas suggest better quality of the site for the growth of provenances (Table 3), which coincides with ^{López et al. (2017)} as there is a greater accumulation of litter in soils of Tlacotepec Plumas than in Magdalena Zahuatlán. Growth was favored by 50 % more nitrogen and 42 % more potassium concentration in soils of Tlacotepec Plumas (^{Velasco et al., 2012}), which favor growth in height and diameter (^{Vázquez et al., 2018}).

Differences between provenances

Growth between provenances showed significant differences (P <0.001) in all the variables evaluated (Table 2). The analysis of the localities as a whole indicated greater growth of the provenances of P. greggii var. australis. The provenance with the highest average growth in height was El Madroño, Qro. (10.99±3.07 m) and in basal diameter, normal diameter, crown diameter, and crown area, El Piñón Hgo., with values of 20.33±7.27 cm, 15.88±6.32 cm, 4.52±1.50 m and 17.81±10.75 m² tree^-1, respectively. The provenances of P. greggii var. greggii recorded lower average height; the highest growth was reached by the specimens from Puerto San Juan, Coah. with 8.86±3.15 m in height, 16.21±7.47 cm in basimetric diameter, 12.63±6.57 cm in normal diameter, 3.79±1.44 m in crown diameter and 12.89±10.06 m² tree^-1 in crown area (Tables 4 and 5).

In the provenances of P. greggii var. greggii at higher altitudes, a lower growth was observed, while those of P. greggii var. australis present at longitudes, latitudes, altitudes and lower soil pH, the data were higher. Being distributed near the area of the establishment of the plantations, the provenances of P. greggii var. australis showed better adaptation in the eroded areas of the Mixteca Alta de Oaxaca with a larger crown area, which protects the ground and provides woody residues and mulch, reduces the dragging of soil particles and decreases erosion, suitable conditions for reforestations in arid zones.

Pinus greggii shows a good response to erosion in the Mixteca Alta de Oaxaca, which coincides with the results of ^{Gómez et al. (2012)} who concluded that in degraded areas, P. greggii increases growth in height and basal diameter as the years have passed since its planting. This has a positive effect by providing more needles and carbon to the soil than P. oaxacana, a native species of the region where the present work was carried out (^{López et al., 2017}). However, the provenances of Pinus greggii var. australis grew faster than P. greggii var. greggii, as described by ^{Donahue and López (1996)} and ^{Velasco et al. (2012)}.

The difference in growth between provenances and localities in the study area began to be verified 2.5 years after the plantations were established, according to the work carried out by ^{Velasco et al. (2012)}, who recorded the highest growth in Magdalena Zahuatlán, in particular with the same outstanding provenances in this work. The variation in growth between varieties may be due to the absence of nutrients in the soil (^{Velasco et al., 2012}), which had repercussions in both provenances as reported by ^{Sánchez et al. (2017)} in P. greggii var. australis and ^{Vázquez et al. (2018)} in P. greggii var. greggii, a variation that recommends the application of nitrogen and phosphate fertilizers in plantations.

Outstanding growth results are shown by provenances of P. greggii var. greggii in this study are opposite to those of ^{Rodríguez et al. (2013)} obtained in Galeana, Nuevo León; the provenances of P. greggii var. greggii with a distribution closer to the poles showed a slow growth, which can be explained by the influence of temperature and the precipitation regime (^{Chhin et al., 2018}; ^{Taïbi et al., 2018}). The highest growth of the of P. greggii var. australis provenances is related to its distribution at lower latitudes, where conditions are less restrictive for its development, compared to the of P. greggii var. greggii provenances (^{Valencia et al., 2017}); therefore, they are better adapted to warmer, more humid environments and with higher rainfall (^{Donahue and López, 1996}).

The P. greggii var. australis provenances have genotypes with sufficiently vigorous roots to reach the available moisture in the deep layers, as well as a greater accumulation of biomass, characteristics that influence a greater protection of the soil from erosion (^{Asbjornsen et al., 2004}; ^{Villegas et al., 2013}); consequently, the nearness of the distribution environment of the provenances of P greggii var. australis with the study area favored its adaptation and growth by resisting temperature and drought Some essays established in Brazil, Colombia and South Africa with the same species and the same provenances (^{White et al., 2007}; ^{Aitken and Bemmels, 2016}). ^{Dvorak et al., (1996)} produced similar results in terms of higher growth of this species.

The results of the present work suggest the existence of a strong relationship between growth and the descriptive variables of the places of natural distribution of each provenance, which agrees with the findings of ^{Salazar et al. (1999)} in Puebla, that at a younger age there is a positive correlation between height and altitude and, at an older age, the correlation becomes negative. In addition, with time, provenances of lower altitude adapt better by showing more noticeable growth in height, basal diameter and crown diameter.