Introduction

Forests and tropical rainforest provide important environmental goods and services, which can contribute to the improvement of the local economy (^{Méndez et al., 2018}). Temperate forests are ecosystems with great diversity in the American continent; they extend from the United States of America to Honduras (^{Mora-Donjuán y Alanís-Rodríguez, 2016}).

Knowledge of the structure and composition of plant communities is fundamental, as it allows the creation of strategies that promote the growth and development of forest stands, ensuring rational use without compromising the original scenarios (^{Aguirre-Calderón, 2015}; ^{Manzanilla et al., 2020}). In addition, it is necessary to know the richness, composition and degree of similarity of the communities (^{Domínguez et al., 2018}), as well as the components that influence diversity and richness, the most prominent of which are the environmental factors and the physical land characteristics (^{Saldaña, 2013}).

Species diversity may decrease towards higher latitudes and altitudes (^{Malizia et al., 2020}). ^{García-Aguilar et al. (2017)} confirm that the development capacity of forests is closely linked to the physical conditions of the terrain. Likewise, ^{McIntire et al. (2016)} confirm that exposure is a limiting factor for the establishment of shade-intolerant species.

At present, several studies have been carried out on the floristic structure and composition in forest ecosystems, which focus on changes in the altitudinal gradient (^{Alves et al., 2010}; ^{Rascón et al., 2018}). However, the influence of geographic exposure has not been considered, resulting in insufficient information about Mexico.

Therefore, the objective of this research was to determine differences in the structure, richness, composition and degree of similarity of plant communities, with regard to exposure and altitude range in a temperate forest in the state of Durango. The following hypotheses are analyzed: (i) the northern exposure will record higher values of basal area; (ii) as altitude increases, the richness and diversity of species will decrease, and (iii) the degree of similarity in the species composition will be defined by exposure.

Materials and Methods

The study area was located in the mountain system called Western Sierra Madre in the state of Durango, between the geographic coordinates 26°26'25.7" N, and 106°03'57.5" W, and 23°10'29.5" N and 105°22'09.2" W (Figure 1).

Figure 1 Location of the sampling sites (green dots) in the study area. 

According to the Köppen classification modified by ^{García (1988)}, the predominant climate types in the area are temperate sub-humid, with four subdivisions: (A)c(w₂), (A)c(w₁), C(E)(w₂), and C(E)(w₂)(x'). Temperature variation ranges from 12 to 18 °C, with the exception of some areas with values of 10 to 12 °C; rains usually occur in summer and droughts in winter (^{Quiñones et al., 2012}). The soil types form associations between Regosol, Lithosol, Cambisol, and Phaeozem (^{INEGI, 2007}). The existing vegetation types are pine forest, oak forest, pine-oak forest, oak-pine forest, and shrub secondary vegetation in all plant communities (^{González et al., 2012}).

Results

Forty-eight species were registered (Table 2), distributed into eight families; of these, Fagaceae had the highest number of taxa and comprised 39.58 % of the total, followed by Pinaceae, with 31.25 %. Convolvulaceae and Betulaceae had the lowest values.

Abundance (density). Areas 1 and 2 exhibited higher values. On the southern exposure, the species with the highest abundance were: P. arizonica Engelm., with 600 ind ha^-1, and J. deppeana Steud., with 66 ind ha^-1. On the northern exposure, P. arizonica had 344 ind ha^-1, and Q. sideroxyla Bonpl., 111 ind ha^-1. Areas 3 and 5 showed similarity in the total number of individuals per hectare. The most abundant taxa in area 3 were Q. laeta Liebm., with 122 ind ha^-1, and P. leiophylla Schiede ex Schltdl. et Cham., with 68 ind ha^-1. In area 5, P. durangensis and P. oocarpa Schiede ex Schltdl. had a greater presence (Table 3). Areas 4 and 6 recorded a difference compared to the higher altitude intervals (p<0.05). In area 4, the species with the highest density were Q. crassifolia Humb. & Bonpl., with 121 ind ha^-1 and P. leiophylla, with 83 ind ha^-1. In area 6, the most abundant taxa were P. durangensis, with 77 ind ha^-1, and Q. resinosa Liebm., with 47 ind ha^-1.

Basimetric area. The southern exposure exhibited the largest basal area in the three altitude intervals. In interval 1 (2 700-3 000 m), the average was 56.42 m² ha^-1; P. arizonica stood out with 21.26 m² ha^-1 in area 1 (Table 3). The ANOVA test showed no significant differences for areas 4 and 6; however, the average was lower, of 15.70 m² ha^-1 and 15.94 m² ha^-1, respectively. The species with the largest basimetric area were P. leiophylla, with 2.94 m, and P. durangensis, with 3.36 m² ha^-1 (Figure 2B).

Figure 2 Means and standard error of (A) Abundance (ind ha^-1), (B) Basimetric area (m² ha^-1), (C) Richness of species (S), (D) Shannon True Diversity Index (¹ D), (E), Margalef Index (D _mg ). Means with different letters (a, b) indicate differences (p<0.05). 

Ecological importance value index (EVI). The Pinus and Quercus genera exhibited the highest percentages of IVI; P. arizonica reached the highest percentage in the interval 1, north and south; Q. laeta Liebm., on the southern exposure, and Q. crassifolia, on the northern exposure, dominated the mean altitude interval. In the lowest height interval, the highest IVI was obtained by P. durangensis; Table 3 shows the three most important species by area.

The dendrogram of plant communities by altitude intervals showed that there is 65 % similarity between areas 1 and 2, with 65 %, and 46 % between areas 4 and 6 (Figure 3). This indicates a similar floristic composition in these areas, where the dominant species tend to be the same as in other areas. Likewise, the southern exposure of area 5 exhibited lower similarity values compared to the rest of the areas; as a rule, the exposure is precisely what defines the degree of similarity between sites located at lower altitudes. The generalist species found in both exposures and within the three altitudinal intervals were P. durangensis, Q. sideroxyla Bonpl., J. deppeana, and A. xalapensis Kunth. At sites with lower altitudes, the species varied by type of exposure: Q. coccolobifolia Trel. occurred on the northern exposure, and A. pennatula Benth., L. acapulcense Benth., P. serotina Ehrh., and I. arborescens Humb. et Bonpl., on the southern exposure.

Figure 3 Dendrogram of similarity based on a Bray-Curtis analysis between areas with different altitudinal ranges. 

Richness. Differences (p<0.05) were found between areas 1 and 4, with a lesser number of species at sites with a higher altitude; the number of species within area 4 was observed to be around twice as high as in area 1. In the remaining areas, the average interval was five taxa per site (Figure 2C).

Diversity indexes. Shannon's true diversity showed no differences (p<0.05) between areas, with an average of 2.94 ± 1.58 on the southern exposure and 3.89 on the northern exposure (Figure 2D). Regarding the Margalef index, differences (p<0.05) were observed in the areas located at a higher altitude with respect to lower areas. The average values were 0.73 for both exposures in interval 1, and 0.98, in interval 3 (Figure 2E).

Discussion

The results obtained for density agree with those cited by ^{Delgado et al. (2016)}, who estimated 565 ind ha^-1 and 16 taxa in temperate forests of Durango, within an altitudinal range of 2 400 to 2 500 m. On the other hand, ^{López-Hernández et al. (2017)} registered lower values, 389 ind ha^-1 and 11 species in forests in the state of Puebla; this can be attributed to the fact that the evaluated forests exhibit more intensive harvesting activities than those of the present study.

The basimetric area was found to increase as altitude does, as confirmed by Muñoz et al. (2020); exposure and slope can also influence the productivity of different areas. Results of the study show that the basimetric area is larger at sites with higher altitudes, with values up to twice as high. Results are similar to those obtained by ^{Graciano-Ávila et al. (2017)}, who estimated a similar basimetric area in temperate forests of Durango, Mexico.

In most temperate forests of Mexico, Pinus and Quercus tend to be the most prevalent genera, as they include a greater number of species which have more dominant mensuration variables and, therefore, tend to have a higher value of ecological importance (^{Domínguez et al., 2018}). In this case, there were variations in the species with a higher IVI value; thus, areas located at both higher and lower altitudes registered higher EVIs in Pinus, a result that agrees with those obtained by ^{Hernández-Salas et al. (2013)} and ^{Graciano-Ávila et al. (2019)}, who report P. durangensis as the most ecologically important species in the forests of Chihuahua and Durango. ^{Paredes et al. (2019)} coincide with the results obtained in the present study and cite Q. crassifolia as the most ecologically important taxon in a temperate rainforest in Hidalgo.

Altitude, slope and exposure have a great influence on the floristic composition and species richness of plant communities. ^{Siles et al. (2017)} recorded 27 taxa over an altitudinal gradient of 1 300 to 1 500 m; these data are similar to those documented in the present study, in which 18 to 23 species were found to occur within the lower altitudinal range. Likewise, authors such as ^{Castellanos-Bolaños et al. (2019)} agree that the higher the altitude, the lower the number of taxa.

Regarding diversity, ^{Domínguez et al. (2018)} indicate similar records at similar altitudes for a temperate forest in the Ruiz Cortinez ejido, in the region of El Salto, Durango, where the values obtained for diversity are similar to those of the present study.

In order to assess the diversity and wealth of species among communities, it is necessary to know the biological characteristics and their proportional distribution (^{Moreno, 2001}). The value of the Margalef index was lower at higher altitudes, as shown by ^{Báez et al. (2015)}, who indicate that low temperatures and topographical features contribute to the decline of species. Likewise, the high number of taxa at lower altitudes is similar to that estimated by ^{Clark et al. (2015)}, who report a large wealth of species in plots located at low altitudes.

According to ^{Zarco et al. (2010)}, exposure influences the development of different species, particularly at medium and low altitudes; this information is in agreement with that recorded in the present study, where the number of taxa was found to be higher on the southern exposure.

The similarity between communities is linked to altitude and exposure, among other factors (^{Chust et al., 2006}) that determine which species can adapt to different plant communities. Accordingly, it is possible to observe both generalist and specialist species adapted to different areas with particular conditions.

Hernández et al. (2013) indicate that the grouping depends, to a large extent, on the degree of adaptation of the taxa, so that different species of conifers and oaks may occur in sites with different characteristics. The above confirms the data obtained for different species of Pinus and Quercus, which are present in all the areas studied. Likewise, we agree with ^{Delgado et al. (2016)}, who point out the occurrence of particular species established in very clearly defined microhabitats, given that in this study certain taxa were observed only in low altitude intervals and on the southern exposure.

Conclusions

The basal area does not exhibit significant differences between the different areas and altitudinal ranges. However, the density is higher at sites with a higher altitude, contradicting the hypothesis put forward at the beginning of the study.

Regarding the wealth and diversity of species, the hypothesis is fulfilled, since these are greater in areas with a southern exposure. Similarly, sites at lower altitudes have the highest number of species, and both generalist and specialist species are identified in specific areas.

The results indicate that, within the areas evaluated in the different altitudinal intervals, the structure, floristic composition, and wealth of species are strongly influenced by the topographic characteristics of the terrain. It is possible to identify that the degree of similarity of plant communities varies according to the altitudinal range and slope exposure, where temperature can support or hinder the development of species.