Introduction

The plant cell wall is made up mainly of cellulose (35-50 %), hemicelluloses (10-35 %) and lignin (15-40 %); these compounds are therefore known as lignocellulosic materials (^{Chen, 2014}). Due to the biological origin of the plants and to their chemical composition, they are considered as a potential source for obtaining clean energies and various chemical products. The lignocellulosic composition of wood is about 90 to 99 % of dry weight (^{Fengel and Wegener, 2003}); which it varies among species, as well as among individuals, depending on the genetic combinations, age, and growth conditions of each species (^{Kilpeläinen et al., 2003}; ^{Yang et al., 2012}).

Besides the main components of wood, there is a small percentage of extractives (4-12 %) made of fats, waxes, phenols, terpenes, resin acids and gums, among others (^{Rowell et al., 2005}), as well as ashes, varying between 0.08 and 2 %. Extractives affect the properties of wood, such as natural durability, basic density, color and hygroscopicity, and also, certain processes like pulping and paper manufacture, drying and adhesion (^{Umezawa, 2001}).

Cellulose is one of the most abundant biopolymers in nature; from which textiles, food or pharmaceutical products, pulp and paper, and bioethanol can be obtained. Xylitol, ferulic and lactic acids can be obtained from hemicelluloses (^{Saha, 2003}), while lignosulfonates, polymeric resins and additives for adhesives, aromatic compounds like vanillin-which is used in pharmaceutical and cleaning products, perfumes, ice cream and chocolate-can be obtained from lignin (^{MacKay et al., 2009}). The products that can be derived from wood depend on its inherent chemical composition and on the knowledge of it; therefore, it is important to research its basic composition in order to know the amounts of the main chemical compounds that allow the assessment of the technical and economic feasibility to obtain products (^{Sadhukhan et al., 2014}).

There are in Mexico 46 species, 3 subspecies and 22 varieties of pine trees (^{Sánchez, 2008}), which represent 42 % of the pine species in the world. Twenty of them have been recorded in the state of Durango (García and González, 1998), of which 18 are located in the Pueblo Nuevo region (^{Gutiérrez et al., 2008}). Because of the available volume, the most exploited species in this region are Pinus cooperi C. E. Blanco, P. durangensis Martínez, P. teocote Schiede ex Schltdl. & Cham., P. leiophylla Schlecht. ex Cham, P. ayacahuite Ehrenb. ex Schltdl, P. engelmannii Carr., P. herrerae Martínez, P. pseudostrobus Lindley, P. douglasiana Martínez, P. michoacana Martínez, P. oocarpa Schiede and P. maximinoi H. E. Moore (^{Meraz, 2001}).

Despite the economic importance of pine wood in the state of Durango, there are no studies on the wood chemical composition of any local species, and in general, the studies of wood chemistry in Mexico are limited to a few species such as Pinus ayacahuite, P. cembroides Zucc., P. johannis M.-F. Robert, P. leiophylla, P. maximartinezii Rzedowski, P. michoacana var. cornuta Martínez, P. montezumae Lamb., P. oocarpa, P. patula Schltdl. et Cham, P. pinceana Gordon, P. pringlei Shaw, P. pseudostrobus, P. rudis Endl., P. strobus L. and P. teocote (^{Islas, 1992}; ^{Ávila, 2011}; ^{Bernabé et al., 2013}; ^{Lima, 2013}). Their results show that the wood of these species is made up of celulose (45.84 ± 12.26 %), hemicelluloses (24.5 ± 8.5 %) and insoluble lignin (27.9 ± 2.3 %), as well as extracts, which may vary according to the type of the solvent used, reaching values of 6.9 ± 6.6 % for hot water, of 8.7 ± 6.4 % for ethanol-toluene and of 20.2 ± 6.3 % for sodium hydroxide (NaOH) at 1 %.

Thus, the objective of this study was to determine the lignocellulosic composition, the content of extracts in ethanol- benzene and ethanol, as well as the amount of ashes in wood from three of the most exploited pine species in Pueblo Nuevo, Durango (Pinus ayacahuite, P. leiophylla and P. herrerae) as a contribution to the technical knowledge that describe them.

Materials and Methods

Six Pinus ayacahuite trees were selected and felled in the community known as “Peña” (23°34’38.24” N, 105°22’22.44” W), six P. leiophylla trees in the community “Mesa del Venado” (23°26’45” N, 105°21’56” W) and 5 P. herrerae trees in “Colorada” (23°24’35” N, 105°23’4” W), in the Pueblo Nuevo ejido, Durango. Samples of leaves and cones were taken from each felled tree in order to identify the species. The botanical samples were identified by the curator of the “Luciano Vela Gálvez” National Forest Herbarium(INIF) of the Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias. The selected trees were healthy, with a straight stem, and free of external visible damage. The dimensions were: 32.5 to 42.5 cm for the normal diameter (ND) and 17 to 23.5 m for the height of Pinus ayacahuite, 32 to 39 cm for ND and 15 to 22 m for the height of P. leiophylla and 32 to 40 cm for the ND and 17 to 25 m for the height of P. herrerae.

Two 3 m long logs were taken from the lower part of each tree and sawn to produce a 7 cm wide section of the central part of each log and, subsequently, 5 cm x 5 cm x 3 m pieces. Two pieces and a 20 cm long portion-free of defects such as knots or resin pockets-were then selected at random from each log (Figure 1). Because of the manner in which the samples were selected, it was not possible to separate the sapwood from the heartwood.

Figure 1 Illustration of the sample collection. 

The 5 cm x 5 cm x 20 cm pieces were planed with a Mizutti^® double-blade edge power planer in order to obtain the shavings, which were dried at room temperature (20 °C). The shavings of each piece were mixed, ground in a Thomas Wiley mill and sieved through No. 40 (0.425 mm) and No. 60 (0.250 mm) meshes. The ground material retained on the No. 60 mesh was used for various assessments; five repetitions were carried out per assessment per tree, adding up to a total of 25 to 30 measurements per species.

The samples were prepared for chemical analysis according to TAPPI T264 cm-97 standard (^{TAPPI, 2007c}); the extracts in ethanol-benzene and in ethanol were quantified according to the procedures of the TAPPI T 204 cm-97 standard (^{TAPPI, 2007a}), and the lignin, using the technique of TAPPI T 222 om-02 (^{TAPPI, 2007b}). The procedure described by Rowel et al. (2005) was utilized to determine the quantity of holocellulose, and ASTM D1103-77 (^{ASTM, 1977}) was used for cellulose, while hemicelluloses were estimated based on weight differences.

The data obtained from each assessment were subjected to Shapiro-Wilk test for normality (^{Razali and Wah, 2011}) and to Bartlett’s test for homogeneity of variances (^{Milliken and Johnson, 2009}). In order to determine the differences among species, a multiple comparison of means was carried out with Bonferroni’s method (^{Milliken and Johnson, 2009}). These analyses used a significance level of α = 0.05, and the UNIVARIATE, GLM and MIXED procedures, respectively, of the SAS software (^{SAS, 2000}). The GLM procedure was also used to determine the variation among species and within each species.

Results and Discussion

Table 1 shows the results for the normality tests of the observations and the homogeneity of variances for each of the assessments that were carried out, in which the data follow a normal tendency. However, the variances of lignin, holocellulose, cellulose and ash contents are not homogeneous; therefore, the mean comparison was carried out with the mixed linear model, which maximizes the restricted likelihood of the data by the generalized least squares fitting, unlike the general linear model, which uses the ordinary least square fitting (^{SAS, 2010}). With the exception of the ash contents, the results showed significant differences between species (p < 0.05) for the average values.

Table 1 Summary of the data normality and homogeneity of variances tests. 

The ash content is an approximate measure of the mineral salts and other inorganic materials in the wood (^{Rowell et al., 2005}). In this study, the ash content showed no significant differences between the wood of the three species (Table 2). The obtained values for P. ayacahuite wood are higher than those reported (0.15 %) by ^{Islas (1992)}. In P. leiophylla wood, the values are lower than those obtained by ^{Bernabé et al. (2013)} and ^{Lima (2013)} with 0.30 and 0.31 %, respectively. There are no values for ash in P. herrerae. In general, this content is within the values of 0.26 to 0.30 % recorded for P. michoacana, P. montezumae, P. oocarpa, P. patula, P. strobus and P. teocote (^{Islas, 1992}; ^{Bernabé et al., 2013}; ^{Lima, 2013}), but is lower than the 0.34 to 0.52 % for P. caribea, P. cembroides, P. johannis, P. maximartinezii, P. pinceana, P. pringlei, P. pseudostrobus and Pinus rudis.

Table 2 Lignocellulosic composition, extracts and ashes of the wood of three pine species. 

^† = Values with the same letter are not significantly different (p < 0.05); * = Values based on oven-dry weight of extract-free samples.

In the extraction of wood with ethanol-benzene, waxes, fats, resins, photosterols, non-volatile hydrocarbons, low molecular weight carbohydrates, salts and other water-soluble substances are eliminated (^{TAPPI, 2007a}). The content of extracts of the wood in ethanol-benzene was significantly higher (p < 0.05) in P. ayacahuite than in P. leiophylla and P. herrerae (Table 2). The obtained values are within the range of extracts in ethanol- benzene (0.5 - 12 %) reported for different pine species in other countries (^{Shimizu, 1991}; ^{Shupe et al., 1997}; ^{Fengel and Wegener, 2003}; ^{Hafızoğlu and Usta, 2005}; ^{Rowell et al., 2005}). There are no records for this type of extracts in Mexico. The high content of extracts in the wood may cause problems for the pulping, gluing and finishing processes (^{Umezawa, 2001}).

During the extraction of wood with ethanol, certain compounds including waxes, fats, resins, tannins, gums, sugars, starches and pigments are removed (^{ASTM, 2012}). Ethanol extracts showed significant differences (p < 0.05) between P. herrerae wood (1.28 %) and P. ayacahuite (1.02 %) and P. leiophylla wood (1.06 %) (Table 2). The values obtained are lower than those determined (3.01 %) by ^{Lemos et al. (2005)} for Pinus oocarpa wood.

The lignin content was significantly higher (p < 0.05) in P. ayacahuite, with 28.6 %, followed by P. leiophylla (27.50 %) and P. herrerae (26.3 %) (Table 2). The amount of lignin obtained in P. ayacahuite is less than 29.9 %, according to the findings of ^{Islas (1992)}, while for P. leiophylla the value is below 28.5 % according to ^{Bernabé et al. (2013)} and below 29.4 % according ^{Lima (2013)}, respectively, although these authors applied a quantification technique for lignin which differs from the one used in this study. Compared to other pine species, the amount of lignin is greater than the values of 24-25.3 % recorded for Pinus michoacana, P. montezumae, P. oocarpa, P. pringlei and P. teocote (^{Ávila, 2011}; ^{Bernabé et al. 2013}), similar to the values of 26.1 to 28.5 % estimated for Pinus caribea, P. montezumae, P. patula, P. pseudostrobus, P. rudis and P. strobus (^{Islas, 1992}; ^{Lima, 2013}), and lower than the values estimated (28.9-30.4 %) for P. cembroides, P. johannis, P. maximartinezii and P. pinceana. The obtained values are within the range of extracts in ethanol- benzene (20.2 - 30.3 %) reported for different pine species in other countries (^{Shimizu, 1991}; ^{Shupe et al., 1997}; ^{Fengel and Wegener, 2003}; ^{Hafızoğlu and Usta, 2005}; ^{Rowell et al., 2005}).

Significant differences (p < 0.05) were found in the holocellulose contents of the wood of three species, which are higher in P. herrerae (70.0 %) and lower in P. leiophylla (66.7 %) (Table 2). The obtained value of holocellulose for P. ayacahuite is lower (69.1 %) than the value determined by ^{Islas (1992)}, while for P. leiophylla the value is lower than 60.7 %, found by ^{Lima (2013)}, and 69.2 %, reported by ^{Bernabé et al. (2013)}. The amount of holocellulose of the three species is higher than that reported (63.6-65.9 %) for Pinus cembroides, P. montezumae, P. pringlei and P. pseudostrobus (^{Ávila, 2011}; ^{Lima, 2013}), but below that (71.5-74.7 %) found for P. caribea, P. michoacana, P. oocarpa, P. patula, P. rudis, P. strobus and P. teocote (^{Islas, 1992}; ^{Bernabé et al., 2013}). The amount of holocellulose of the wood of the studied pine trees is within the range of 62.9 to 79.8 % reported for pines in other countries (^{Shimizu, 1991}; ^{Shupe et al., 1997}; ^{Fengel and Wegener, 2003}; ^{Hafizoğlu and Usta, 2005}; ^{Rowell et al., 2005}).

The wood of Pinus herrerae showed the highest content of cellulose (45.6 %), with respect to the wood of P. ayacahuite (43.9 %) and P. leiophylla (43.3 %) (Table 2). ^{Islas (1992)} estimated the content of cellulose of P. ayacahuite in 58 %, while ^{Lima (2013)} estimated a value of 60.7 % for P. leiophylla. These values are above those determined in this study for the same species. The above mentioned authors estimated cellulose values of 56.7 to 64.8 % for P. caribea, P. montezumae, P. patula, P. pseudostrobus, P. rudis and P. strobus wood. However, lower cellulose contents were identified in Pinus cembroides (36 %), P. johannis (35.5 %), P. maximartinezii (37.6 %) and P. pinceana (35.6 %) (Revilla, 201, cited by ^{Bernabé et al., 2013}). The cellulose contents in the studied species also ranges between 30 and 61.6 % for the wood of pine species of other countries (^{Shimizu, 1991}; ^{Shupe et al., 1997}; ^{Fengel and Wegener, 2003}; ^{Hafızoğlu and Usta, 2005}; ^{Rowell et al., 2005}).

The highest concentrations of hemicelluloses were obtained in Pinus ayacahuite and P. herrerae wood, compared to P. leiophylla wood (Table 2), but are lower than those obtained for Pinus cembroides (29.9 %), P. johannis (30.9 %), P. maximartinezii (30.6 %) and P. pinceana (31.1 %). The values for the wood of foreign pines range between 18 and 41.2 % (^{Shimizu, 1991}; ^{Shupe et al., 1997}; ^{Fengel and Wegener, 2003}; ^{Hafızoğlu and Usta, 2005}; ^{Rowell et al., 2005}); therefore, the contents of these compounds (23.4-24.4 %) are within the range for foreign pine species.

The results of the SAS GLM procedure show that the variation in the chemical composition within each of the species was lower than that observed among species (Table 3). The highest occurred in Pinus ayacahuite and in P. leiophylla for the content of extracts in ethanol, while P. herrerae had the lowest variation in the content of ethanol-benzene extracts (Table 3). Among species, the highest variation was found in the content of ashes and extracts, both in ethanol-benzene and in ethanol. Variation within the species is important for genetic improvement, while variation among species is essential for pulping to paper making and for chemical conversion in chemical products derived from wood (^{Yang et al., 2012}).

Table 3 Variation of the chemical composition (%) within and among the studied pine species. 

In general, wood with low contents of lignin and extracts can be easier to delignifying, which favors the pulping process and bioconversion for the production of liquid biofuels (^{Sable et al., 2012}; ^{Zeng et al., 2014}). P. herrerae wood showed lower contents of lignin and extracts than other species, and therefore, it could be more suitable for producing pulp for paper and for bioconversion, due to its high content of cellulose.

The differences and variations in the wood chemical composition of the studied species may be due to the genetic traits of each species, to the age of the trees and to the growth conditions (^{Campbell et al., 1990}; ^{Berrocal et al., 2004}; ^{Yang et al., 2012}). However, more specific studies addressing these variables are needed.

Conclusions

Wood of Pinus ayacahuite, P. herrerae and P. leiophylla shows statistically significant differences (p < 0.05) in their lignocellulosic components and their contents of extracts, with a larger variation among species than within species. The first has a high content of total extracts (13 %) and a higher content of lignin. In contrast, P. herrerae wood has lower contents of total extracts (5.6 %) and of lignin, as well as higher contents of holocellulose and cellulose. P. leiophylla wood contains less holocellulose, cellulose, hemicelluloses and ashes than the other studied species.

In general, the wood chemical composition and extract contents of the studied species may be considered as a feedstock for obtaining pulp for paper and liquid biofuels; however, these processes may be more difficult for P. ayacahuite wood due to its high extracts content.