The study area

The study sites were located in La Peñita community, Santiago municipality Nuevo León and Santa Rita ejido, Arteaga municipality, Coahuila, both embedded in the Sierra Madre Oriental. The first one, in the 362409 and 2804175 UTM, coordinates, at 2 300 masl, a climate of the C type [C (W’)] subhumid temperate, with a mean temperature between - 3 and 18 °C and annual precipitation between 600 and 900 mm (^{García, 1973}). The second one is between the 353988 and 2791690 UTM coordinates, at 2 500 masl, a C (w₁) (i)w’ climate of the subhumid temperate and subhumid with a temperature of -8 to 30 °C and an annual precipitation between 146 to 632 mm. The vegetation of each one is listed in Table 1.

Table 1 Average characteristic of the coniferous forest and the Quercus in the study area. 

DBH = Diameter at breast height (cm); Proportion = Percentage that each species amounts.

Douglas fir (Pseudotsuga menziesii) is found in isolated populations of the Sierra Madre Oriental, Sierra Madre Occidental, Eje Neovolcánico and central part of the country (^{Domínguez et al., 2004}) and in small stands of Oaxaca state (^{Del Castillo et al., 2004}). Pinus teocote has a broad natural distribution in México and Guatemala and grows from 1 000 to 3 000 masl, in loam and clay loam soils. It is adapted to subhumid to humid climates, with annual precipitation from 1 000 to 1 500 mm (^{Dvorak and Donahue, 1993}).

Field sampling

The collection of growth nuclei (cores) was carried out in La Peñita, NL, in June 2015, with 52 trees of Pinus teocote (Pt) (26 healthy and 26 infested) and in April 2016 in Santa Rita, Coah. 72 from Pseudotsuga menziesii (Pm) (36 healthy and 36 infested). From each tree two cores were extracted (one from north-south and one from east-west) using a Pressler drill (Haglof FOI0308) and size at 1.3 m height (^{Mast and Veblen, 1994}; ^{Camarero and Martín, 2002}) The healthy trees (As) had a very strong vigor and no sawdust or galleries of any type of pest; the infested trees (Ai) had reddish needles and dead branches, bark with lumps of sawdust, galleries or exudation; both (As and Ai), should cover the same diameter categories and be free of competition (distance > 3 m to the nearest tree).

The mensuration characteristics were measured in plots of 50 x 50 m. The Ai were stripped a section of the shaft (30 x 20 cm) at the breast height to do a sampling of barking insects which were taken to the forest department of the UAAAN´s Laboratory for later identification.

Dendrocronological analysis

The cores were placed in a wooden base and were sandered to highlight the annual growth structures (^{Swetnam et al., 1985}). The growth rings were counted and dated through standard dendrochronological techniques (^{Stokes and Smiley, 1968}; Swetnam et al., 1985); they were individually measured through a Velmex with a 0.001 mm accuracy measuring system of sliding phase (^{Cerano et al., 2011}) and each sample from the same tree and the same species was compared through crossdating (^{Stokes and Smiley, 1968}8; ^{Camarero and Martín., 2002}). The dating and the exact measurements of each ring was checked with the COFECHA program (^{Holmes, 1983}; ^{Grissino-Mayer, 2001}).

With the ARTSAN program the biological (competition, suppression, release) and geometric (increase of the stump area with age) tendencies not related to climate and forest plagues were removed (^{Swetnam et al., 1985}; ^{Cook and Holmes, 1996}). ARTSAN generated the standardized and residual ring width index (IAA) (^{Swetnam et al., 1985}).

Analysis of growth and radial increase of Pinus teocote and Pseudotsuga menziesii

In order to corroborate that the average diameter and age between As and Ai were equal between species, a Student t test was applied through the Statistical Analysis System (SAS) software (^{SAS, 2014}). To identify differences between As and Ai, the ring width (AA) and IAA data were divided into two groups: 1) total chronology and 2) last 10 years. In order to evaluate the radial increase of each species before and after the pest outbreak (2011), a cumulative AA was performed, five years before and five years later, in which a Student t test was applied to both groups data.

The percentage of reduction of the IAA was obtained by means of its average in the total chronology of As and Ai in both species. The ring width data (mm) were transformed to basal area (cm²), from which the accumulated basal area (ABA) was derived; the Schumacher growth model was adjusted to these variables, with which the increase in basal area (IAB) in each species was determined. This model is expressed as follows:

Y=expa+b/x

Diameter and age of Pinus teocote and Pseudotsuga menziesii

The diameter and age in As and Ai of P. teocote was statistically the same (p = 0.6439 and p = 0.6922) (Table 2); the diameter was 39. 68 in As and 40.48 cm in Ai, with 84.81 and 82.91 years, respectively. In As and Ai of Pseudotsuga menziesii, the diameter was not statistically different (p = 0.1320), nor was age (p = 0.0867, Table 2); the diameter in As was 48.05 cm and 44.46 cm in Ai, with 79.64 and 86.28 years, respectively. ^{Hawkins et al. (2013)} have recommended that in order to assess this effect of increase by forest pests, the sample should be distributed over trees of the same categories; this study fulfills this premise.

Table 2 Student's t-test results in diameter and age of Pinus teocote Schiede. ex Schltdl. & Cham. in Nuevo León and Pseudotsuga menziesii Schiede. ex Schltdl. & Cham. in Coahuila, Mexico. 

Pt = Pinus teocote; Pm = Pseudotsuga menziesii; DBH = Diameter at breast height; n = Number of trees; D.S. = Standard deviation.

Growth and increment of tree ring of Pinus teocote and Pseudotsuga menziesii

In As of P. teocote a chronology of 140 years (1875 - 2014) and of 124 years (1891 - 2014) in Ai were obtained. This study stands for one of the first intents to demonstrate the dendrochronological potential of P. teocote. In Pseudotsuga menziesii the chronology was 118 años (1898 - 2015) and 135 years (1881 - 2015) in As and Ai, respectively. In this same species and specialty, ^{Zhang et al. (1999)} in Canadá, revealed chronologies near 600 years; in New Mexico, United States of America, ^{Swetnam and Lynch (1989)} studied trees of only 180 years.

The dendroclimatic resarch studies made in Mexico over Pseudotsuga menziesii have reported from 203 (^{Arreola et al., 2010}), 300 (^{Cerano et al., 2011}) up to 554 years (^{Villanueva et al., 2009}). The results here obtained prove that the forests of this species in this part of the country are in very contrasting developing stages, as differences from 68 to 436 years were recorded.

The Student t test (90 %) in the total chronology (G1) indicated that the ring width of P. teocote between As and Ai is statistically different (p = 0.0672) (Table 3, Figure 1A) with annual average growth of 1.73 mm in As and 1.92 mm in Ai. In the last ten years (G2), this species shows a different ring width between As and Ai (p = 0.0591; Table 3) with AA of 0.81 and 0.89 mm. With the total chronology (G1), in Pseudotsuga menziesii there was no significant difference in AA between As and Ai (p = 0.1128) (Table 3, Figure 1B), resulting in an annual average growth of 2.99 in As and 2.71 mm in Ai; however, in the analysis of the last 10 years (G2), data show that Pseudotsuga menziesii diameter growth was higher in infested trees (p = 0.0001; Table 3). What is found here differs from what has been recognized for other species, as has been recorded with Picea abies affected samples (Dendroctonus micans Kugelann, 1794), on which ^{Rolland and Lampériere (2004)} calculated a reduction of 50 % in ring width.

Table 3 Results of the Student t test of ring width in total chronology and the last 10 years in healthy and infested Pinus teocote Schiede. ex Schltdl. & Cham. and Pseudotsuga menziesii trees (Mirb.) Franco. 

Pt = Pinus teocote; Pm = Pseudotsuga menziesii; AA = Tree ring width (mm); N = Number of samples; D.E. = Standard deviation; G1 and G2 = Group one and two.

Figure 1 Ring width on total chronology of Pinus teocote Schiede. ex Schltdl. & Cham. (A) and Pseudotsuga menziesii (Mirb.) Franco (B) of two locations in Mexico. 

In Abies alba attacked by defoliating insects (Epinotia subsequana Haworth, 1811), ^{Camarero and Martín (2002}) showed that in a three year period (1996 - 1998) the trees showed a reduction of 1.4 to 1.5 mm in comparison with healthy individuals, which responds to the depletion of carbohydrate reserves that are stored in the needles; this has also been reported by ^{Muzika and Liebhold (1999)} in other genera such as Picea and Abies in which the behavior is different, as pests do not affect growth in diameter (^{Grigaliūnas and Zolubas, 2006}).

Before the barking beetle outbreak, the infested Pinus teocote trees did not show significant differences in regard to the healthy ones (p = 0.1169). For Pseudotsuga mensiezii, the degree of damage had already an effect upon ring width (p =0.0143). After colonization, significant differences were found between healthy and unhealthy individuals of both species, even though it was the other way around, since ring width was greater both in attacked Pinus teocote trees as well as in those of the second species free of these insects (Table 4, Figure 2).

Table 4 Student t test results of the ring width increase before and after the pest outbreak in healthy and infested trees of Pinus teocote Schiede. ex Schltdl. & Cham. and Pseudotsuga menziesii (Mirb.) Franco. 

Pt = Pinus teocote; Pm = Pseudotsuga menziesii; IA-AA = accumulated tree ring width (mm year^-1); N = number of samples; D.E. = standard deviation. .

Figure 2 Ring width increment before (A - B) and after (C - D) the appearance of plagues over Pinus teocote Schiede. ex Schltdl. & Cham. (Pt) y Pseudotsuga menziesii (Mirb.) Franco (Pm). 

The species studied here responded differently in this variable to the infestation of barking beetles. In the fir, the reduction was 22.6 % before the insects and 39.8 % after them. It is possible that additional factors (not analyzed here) such as drought, among others, have contributed to this reaction. Some authors indicate that the spruce studied here is sensitive to the attack of forest pests, which is manifested in reductions of up to 50 %, especially if they are long-lived trees (> 100 years) (^{Alfaro et al., 1985}; ^{Hawkins et al., 2013}).

Pinus teocote and Pseudotsuga menziesii ring width index

The average IAA of P. teocote in total chronology (G1) and in the last ten years (G2) in As (0.98; 0.85) and in Ai (0.99; 0.87) do not show statistical differences (p = 0.8912; P = 0.9360) (Table 5), in Pseudotsuga menziesii, the IAA in the total chronology and in the last ten years also did not differ statistically (p = 0.4728; 0.7061) between As (0.96; 0.87) and Ai (0.99; 0.81) (Table 5).

Table 5 Student's t test results of ring width index in the total chronology and in the last 10 years in healthy and infested Pinus teocote Schiede. ex Schltdl. & Cham. and Pseudotsuga menziesii (Mirb.) Franco trees. 

Pt = Pinus teocote; Pm = Pseudotsuga menziesii; IA-AA = accumulated tree ring width); N = number of samples; D.E. = standard deviation.

In a Student t test performed with the data of the accumulated IAA increase after the outbreak, statistical differences (p = 0.0113; p = 0.0193) were found for P. teocote and Pseudotsuga menziesii, the first with As and Ai Of 1.75 and 1.49 and the second with 1.67 and 1.93 in As and Ai, respectively. Compared with the results of the ring width accumulated after the outbreak, these results indicate that there is an impact of the bark insects on the radial growth of the trees.

Figure 3A shows a reduction of IAA in As and Ai of P. teocote of up to 30 % in the 1916-1918, 1920-1925 and 1950-1953, 1955-1958 periods, during which a reduction of the IAA in As and Ai of Pseuddotsuga. menziesii up to 45 % (Figure 3C) occurred; the behavior up to 50 % was observed in 2011 in both species (Figures 3B and 3D). In precipitation reconstruction studies for the southeast of Coahuila and north of the country (^{Cerano et al., 2011}; ^{Villanueva et al., 2009}), the following dry periods were obtained: 1890-1902, 1916-1933, 1951-1963, 1970-1985 and 1994 onwards, which coincides with the lapses of IAA reduction in As and Ai in the two taxa.

Figure 3 Percentage of reduction of the ring width index in the total chronology (A - B) and in the last ten years (C - D) in Pinus teocote Schiede. ex Schltdl. & Cham. (Pt) and Pseudotsuga menziesii (Mirb.) Franco (Pm) healthy and infested trees. 

After the pest outbreak, in P. teocote the difference of the Ai reduction with respect to As was 9.3, 16.2, 11 and 5 % in the years 2011, 2012, and 2014, respectively (Figure 3B). In contrast, in Pseudotsuga menziesii, in which healthy trees recorded a lower reduction in IAA compared to the infested, with differences of 6.3, 11.6, 11.4, 15.7 and 24.2 in the years 2011, 2012, 2013, 2014 and 2015 respectively (Figure 3D), there is a positive response to the attack of barking insects for the pine and negative for the spruce.

Basal area and increment of basal area of Pinus teocote and Pseudotsuga menziesii

The Schumacher growth model adjusted to the basal area of As and Ai individually resulted in high statistical significance (p <0.0001) in the two forest species (Table 6). The AB of the Ai of Pseudotsuga menziesii is the one that showed the best fit with R²_aj = 0.8545; on the contrary, the Ai of Pinus teocote presented the lowest fit (R² _aj = 0.7015). The analysis of the above shows that the greatest variation in AB is observed in the Ai of the pine, while the smallest was calculated in the Ai of Pseudotsuga menziesii, with CV = 44.4829. The results show important variations within each species, as in P. teocote the As have a smaller variation (VC = 50.7667), opposite to the spruce in which this occurred in Ai (Table 6).

Table 6 Statistic for the fit of the growth model of basal area in Schiede. ex Schltdl. & Cham. and Pseudotsuga menziesii (Mirb.) Franco. 

a and b = Regression coefficient of Schumacher’s model; E.E.= Standard error of the regression coefficients; R² _aj = Adjusted determination coefficient; C.V. = Variation coefficient; Pr > F = F probability; CME = Mean squared error; Gl = Degrees of freedom of the total of the model.

The average cumulative basal area according to Schumacher's growth model, in P. teocote's As was 215.1 cm² and 229.5 in Ai. The highest increase in AB was in the first 25 years of age, in As with 190.1 cm² and in Ai with 196.6 cm² (Figures 4A and 4B). In Pseudotsuga menziesii, in As an average of 385.9 cm² and of 352.9 cm² was recorded in Ai, and the highest increase in AB was at 40 years of age with 431.5 cm² in As and 375.8 cm² in Ai (Figures 4C and 4D). The increase of AB in Ai of the latter species is lower compared to As; this is similar to the results of ^{Rolland and Lemperiere (2004)} in Picea abies, where this occurred in the first 35 years with 400 cm² in healthy trees and 350 cm² in trees affected by Dendroctonus micans.

Figure 4 Basal area observed and estimated in Pinus teocote Schiede. ex Schltdl. & Cham (Pt) and Pseudotsuga menziesii (Mirb.) Franco (Pm) healthy (A-C) and infested (B-D) trees. 

Figures 5A and 5B show the same pattern of increase in basal area in both species, but there are peaks below the average. In the pine, the As (14.8 cm² year^-1) and Ai (14.4 cm² year^-1) show these peaks in the 1952 - 1957, 1989, 1998 - 2002, 2006, 2008 - 2009, 2011 - 2012 periods, and although in the total chronology the average is similar, after the outbreaks of barking beetles (2011-2014), those figures vary more (9.5 cm² year^-1; 12.6 cm² year^-1) in As and Ai. In Pseudotsuga menziesii the peaks below the IAB average were at the same lapses in As and Ai, with values of 33.2 and 34.5 cm² year-¹ respectively; after the outbreak, they became smaller, up to 26.2 and 15.9 cm² year^-1, the latter of which means a difference of 10.3 cm² year^-1 between healthy and infested trees, thus affirming the negative impact on the radial increase of this forest species.

Figure 5 Increment of the basal area in regard to the age of the Pinus teocote Schiede. ex Schltdl. & Cham. (A) and Pseudotsuga menziesii (Mirb.) Franco. (B) healthy and infested trees. 

The IAB in the total chronology and after the outbreak on this latter species is rather low compared to the 54 cm² year^-1 recorded for dominant healthy trees of Pseudotsuga menziesii of the state of Chihuahua, where the age of the trees varies from 67 to 240 years (^{Castruita et al., 2016}), although dominant and codominant trees were selected for this study.