Introduction

Distributed in the Sierra Madre Occidental (SMO) are 24 Pinus taxa (46 % of the national total), 54 of Quercus (34 %) and seven of Arbutus (100 %), which are the main physiognomic components of the forests in this region. Pine forests cover 30 197 km², which represent 12 % of the surface area of the SMO, at an altitude between 1 600 and 3 320 masl, while pine-oak forests cover 76 265 km², i.e. 30 % (^{González et al., 2012}). Both types of vegetation grow in temperate semi-cold and temperate semi-dry climates. They have a great biodiversity and provide considerable ecological and economic benefits (^{Semarnat, 2012}).

The state of Chihuahua has an area of 24.7 million hectares, of which 7.08 million (28.66 %) are temperate forests. The pine and pine-oak forests are the main ecosystems under exploitation and provide a great diversity of environmental goods and services to rural and urban communities (^{Gobierno del Estado de Chihuahua et al., 2016}). Forest Management Unit (Umafor) No. 0808, in the Guadalupe y Calvo municipality, Chihuahua, is one of the most relevant for its contribution to the timber production of the state. Thus, as indicated by ^{Delgado et al. (2016)}, it is important to know the structure, composition and diversity of tree species of these forests in order to plan their sustainable management and determine their conservation status.

^{Jiménez et al. (2001)} indicate that the Importance Value Index, among other parameters, determines the horizontal structure of plant communities. On the other hand, ^{Del Río et al. (2003)} suggest that a description of the vertical structure of its components is necessary to characterize a multicohortal and uneven-aged community.

Forest use modifies the forest in terms of its structure, composition and growth dynamics (^{Del Río et al., 2003}; ^{Folke et al., 2004}). At the stand level, irregular management practices focus more on forming a vertical structure that creates heterogeneous masses. It is important to note that multicohort stands are properly distributed across the entire landscape. A balanced vertical structure is achieved by regenerating fractions of the stand at different times at the level of individuals, groups or strips (^{Aguirre, 1997}).

For the planning, management and conservation of arboreal communities, the study of floristic structure and diversity is important (^{Villavicencio et al., 2006}). On the other hand, the assessment of diversity contributes to know the effect of forest exploitation on tree species and determine the practices that allow achieving sustainable forest management (^{Aguirre et al., 2008}).

Therefore, the purpose of this study was to determine if the vertical structure and diversity of the irregular pine-oak forest is different between forests with and without forest management.

Materials and Methods

Study Area

The work was carried out in the Forest Management Unit (Umafor) No. 0808, located in the Guadalupe y Calvo municipality, in the south of the state of Chihuahua (Figure 1). This Umafor has an area of 904 780 ha, with four main climates: warm, semi-dry, semi-warm and temperate, of which the last two predominate (^{Asociación Regional de Silvicultores de Guadalupe y Calvo A.C.- Conafor-ProÁrbol- Secretaría de Desarrollo Rural del Gobierno del Estado de Chihuahua-Semarnat, 2009}). The main soil types were classified according to the definition of soil units proposed by FAO / UNESCO are: Acrisol, Pheozem, Regosol, Planosol, Litosol and Cambisol.

Figure 1 Location of the study area 

The type of vegetation analyzed was pine-oak forest, which provides the industry with a wide variety of raw materials of great economic importance.

Sample analysis

Data from the National Forest and Soil Inventory 2004-2009 (Scale 1: 250 000) were used, extracting those corresponding to pine-oak forest. There were 95 clusters under management and 27 without management (Figure 1), which make up a network of points distributed every 5 km; each one consists of four sites of 400 m² for the inventory of trees, one central and the remaining three arranged in the form of an inverted "Y" with respect to the north. Additionally, In order to evaluate repopulation, a concentric site of 12.56 m² was plotted within each site (Figure 2).

Source: ^{Conafor, 2012}.

Figure 2 Scheme of a forest sampling conglomerate. 

For the analysis of the information, the following data were used: treatment (with or without handling), gender, species, normal diameter and total height. Within the sample plots of 400 m², all trees with a diameter, at 1.30 m in height, equal to or larger than 7.5 cm, were inventoried.

In the repopulation inventory, frequency was recorded by genus and species present, based on height ranges. In order for an individual to be considered as part of the repopulation within the sub-plot of 12.56 m², it should have height equal to or greater than 25 cm and normal diameter of less than 7.5 cm.

Results and Discussion

The height layers were defined according to the maximum height of the forest under management and without management, which were 42.0 and 30.5 m, respectively (Table 1).

Based on the relative Phytosociological Value (VFr), the predominant species in the lower forest layer under management were Pinus durangensis Martínez with 18.42 %, Quercus sideroxyla Humb. & Bonpl 10.05 %, Quercus fulva Liebm. 9.91 %, Pinus arizonica Engelm. 8.31 % and Arbutus xalapensis Kunth 8.02 %.

In the medium layer, the outstanding species were Pinus durangensis with 31.37 %; Pinus arizonica, with 13.36 %; Pinus lumholtzii Robins & Fern, with 8.53 %, and Pinus ayacahuite Ehrenb. ex Schltdl., with 6.98 %.

In the upper layer, Pinus arizonica was the most abundant, with 39.39 %, having presence in all the layers of the vertical structure. However, it was not the most dominant species in the lower and middle layers. The second place is occupied by Pinus durangensis with 27.27 %, followed by Pinus engelmannii, with 15.15 %, and by Pinus leiophylla Schl. & Cham (9.09 %) and Pinus ayacahuite (3.03 %). The genus that contributes the most is Pinus, followed by Quercus, with more than 80 % in each of the layers.

According to ^{Lamprecht (1990)}, there are species that exhibit continuous vertical distribution behavior, that is, they are present in all the layers. Also, ^{Acosta et al. (2006)} state that a species ensures its permanence in the composition and structure of the forest if it is present in all of its layers. In the present study the species with continuous vertical distribution are: Pinus durangensis, Pinus arizonica, Pinus leiophylla, Pinus ayacahuite, Pinus herrerae Martínez, Pinus engelmannii and Pinus chihuahuana Engelm. On the other hand, ^{Alanís-Rodríguez et al. (2010)} mentions that, in post-fire restored areas in Pinus-Quercus forests, not all species are found in all layers.

In the managed pine-oak forest, the lower layer had the highest percentage of individuals, with 84.22 %; the middle layer had 15.32 %, and the upper layer had 0.44 % (Figure 3). In this regard, ^{Acosta et al. (2006)} mention that when the number of individuals in the lower layer is greater than those in the middle layer, it corresponds to a forest in regeneration, as observed in the present study.

Figure 3 Percentage of individuals per height layer in the forest with and without management. 

In the unmanaged forest, the lower layer represents 67.67 % and the following species predominated: Pinus durangensis (19.13 %), Arbutus xalapensis (12.28 %), Quercus fulva (12.12 %), Quercus sideroxyla (8.93 %) and Pinus arizonica (6.85 %). All these species were important in the managed forests as well.

The medium layer comprises 28.61%, and Pinus durangensis was the most important species with 29.16 %, followed by Pinus herrerae with 11.93 %, Pinus arizonica 7.38 %, Pinus lumholtzii 5.49 % and Pinus teocote Schiede ex Schltdl. & Cham. with 5.49 %.

Finally, the upper layer comprises only 3.61 % and is made up of Pinus durangensis, with an important value of 64.17 %; Pinus teocote Schiede ex Schltdl, with 11.94 %; Pinus herrerae, with 8.95 %; Pinus lumholtzii and Pinus engelmannii, with 2.98 %, and Juniperus deppeana Steud and Pinus oocarpa Shiede, with 1.43 %. This implies that the Pinus genus is present in all the layers, predominating in the upper layer and therefore, according to ^{Human (2000)}, this genus belongs to this stratum. In coincidence with the area under management, Pinus and Quercus were structural dominants.

Most of the individuals were concentrated in the lower layer, which coincides with what ^{Castellanos-Bolaños et al. (2010)} reported for a temperate cold climate forest in Oaxaca, as did ^{Zarco-Espinosa et al. (2010)} in relation to the tree structure of an unmanaged tropical forest in southeastern Mexico. According to these authors, most tree species are represented by a few small, scattered individuals. In this regard, in the present study the species that had a continuous distribution in the unmanaged forest were Pinus durangensis, Pinus arizonica, Pinus herrerae, Juniperus deppeana, Pinus lumholtzii, Pinus teocote, Pinus engelmannii and Pinus oocarpa. As in the forest under management, the lower height layer recorded the highest percentage of individuals, of 67.67 % (Figure 2).

The analysis of the relative sociological position found that the most common species in the three layers of the vertical structure of the managed forest were Pinus durangensis (18.87 %), Quercus sideroxyla (9.91 %), Quercus fulva (9.59 %), Pinus arizonica (8.49 %) and Arbutus xalepensis (7.75 %), and those with the least presence were Quercus coccolobifolia Trel (2.08 %), Pinus engelmannii (1.95 %), and Alnus acuminata Kunth (1.77 %). In addition, there are 25 more species with a lower participation (Table 2). Alonso-Torrens et al. (2016) found that the most dominant species are those of the Pinaceae family.

In the pine-oak forest without management, the results indicate that the species with the greatest presence in the three levels were: Pinus durangensis, with 20.76 %, Quercus fulva, 10.86 %, Arbutus xalapensis, 10.62 %, Quercus sideroxyla 8.30 %, and Pinus arizonica, with 6.93 %; while Pinus teocote, with 1.80%; Pinus leiophylla, with 1.64 %, and Quercus candicans Née, with 1.31 %, were the species with the least presence. In addition, other 19 species had a phytosociological value of less than 1%.

In the natural regeneration of the managed forest, the lower layer had the highest percentage of individuals, with 74.96 %; the medium layer had 15.66 %, and the upper layer, 9.38 %; while the unmanaged forest had a similar behavior, with values of 76.69 %, 18.16 % and 5.15 % respectively. This is considered reasonable because, as individuals grow, competition increases and therefore mortality does too.

On the other hand, the Pinus and Quercus genera predominate in the three regeneration levels of the forest with and without management, where Pinus durangensis stands out as the dominant species, except in the lower layer of the unmanaged forest, where Quercus rugosa dominates.

In regard to natural regeneration, of the total number of species present in the forest under management, the lowest category of regeneration has 93.93 % of the species present, and 75.60 % of them reach the arboreal level. The middle and upper categories of regeneration have 60.60 % of the said species present in the regeneration, and 41.78 % of them are consolidated as trees. It is also important to indicate that, of the seven species that have continuous vertical distribution in the tree stratum, only P. leiophylla and P. chihuahuana do not occur in the three levels of regeneration. This ensures the permanence of the five species found in all regeneration and woodland flats in the forests studied.

The lowest layer of the unmanaged forest has 91.30 % of the total species with regeneration, and only 51.21 % of them reach the arboreal layer; the middle layer has 65.21 % of the species with regeneration, and only 36.58 % become consolidated as trees. The lower layer of regeneration gathered 34.78 % of the species, of which only 19.51 % are grouped in the arboreal layer. In this case, the only species that has a continuous vertical distribution, both in regeneration and in the tree stratum, are Pinus durangensis and Pinus arizonica, which implies their competitive advantage over the rest of the species.

These results coincide with those of ^{Hernández (2007)}, who reveals a homogenization of the forest after the application of the treatments, in this case with and without management, and also reports that it benefits the species of economic importance slightly. This agrees with what was indicated by ^{Taylor (2016)}, who in a conifer forest in the Highlands of Chiapas, found an absence of many tree species, which is due to the forms of management utilized, i.e. through selective felling and promotion of the regeneration of fast-growing species with the highest economic value, such as pines.

The richness of the arboreal stratum with and without management was 41 and 34 species respectively. The values obtained for the Shannon-Wiener diversity index (H') were 2.773 and 2.723 for both forests, which are similar to those described by ^{Hernández-Salas et al. (2013)} for forests under management, with similar characteristics, in northwestern Mexico. Likewise, ^{Graciano et al. (2017)} determined a richness of 12 species and an H’ of 1.74 for a mixed pine-oak forest of Durango, and ^{Méndez et al. (2018)} reported a richness of five species and an H’ of 0.78 for a forest of the same type in the Sierra Madre del Sur.

On the other hand, with the Pielou index, values of 0.7468 and 0.7721 were obtained in forests with and without management, respectively. Based on this, and because the values of this index range from 0 to 1, we infer that in the present study the equitability of the abundance of their species is medium to high.

As for the statistical analysis carried out using the Hutchenson t test for the Shannon-Wiener diversity index, the difference between pine-oak forest with and without management was not significant (P> 0.05).

Conclusions

In the managed forest, five pine species with continuous vertical distribution are presented, both in the regeneration and in the tree stratum, while in the forest without management, only two of these pine species maintain this distribution.

Pinus durangensis is the predominant species considering the three levels of regeneration and trees, both in managed and unmanaged forests.

The vertical structure of the forests studied is similar in terms of the percentage of individuals per layer, with a greater number of individuals in the lower layer of regeneration and woodland, which means that they are considered regenerating forests.

The species with the best sociological position with and without management are Pinus durangensis, Quercus sideroxyla, Quercus fulva, Pinus arizonica and Arbutus xalapensis; they are also the most abundant.

The forests that are being exploited through the MMOBI maintain a statistically similar diversity to that of the unmanaged forest.