Introduction

The carbon dioxide (CO2), methane (CH4) and nitrous oxide (N2O), greenhouse gases have increased in the atmosphere since pre-industrial times due to anthropogenic factors and this increase contributes to climate change; CO2 specific, increased by 40%; went from 278 ppm in 1 750 to 390.5 ppm in 2011 (^{IPCC, 2013}). The causes of the increase in the concentration of carbon gases in the atmosphere are well documented in the case of CO2 are industrial and domestic use of carbon-containing fuels (oil, coal, natural gas and wood), deforestation which causes the organic-matter decomposition and burning of plant biomass (^{Jaramillo, 2004}). According to ^{Maass (2009)}, the increase of CO2 in the atmosphere causes an increase in the retention of the radiation emitted by the earth's surface, altering the thermal regime and inducing global warming through the greenhouse effect.

The greenhouse effect is due to atmospheric gases just called greenhouse or greenhouse, the most important are water vapor (H2O) and CO2 (^{Garduño, 2004}). The importance of water vapor and CO2 in the atmosphere is such that without their presence the current average temperature of the planet would be about 33 °C colder and therefore the planet would be frozen (^{Schlesinger, 1997}). The CO2 is a natural component of air and has always existed in the atmosphere, what man has done is increase it and this is what leads to increased greenhouse effect (^{Garduño, 2004}).

To determine the biomass of trees and their allometric equations generate direct or destructive methods involving destruction of trees evaluated are used, since as indicated Fehse, et al. (²⁰⁰²) direct methods are based on the demolition and weighing vegetation. Therefore, because it is not always possible to tear down and destroy trees to evaluate their allometric equations to generate most of the work on the subject ^{Acosta et al. (2002)}; ^{Díaz et al. (2007)}; Juárez (2008); ^{Avendaño et al. (2009)}; have used destructive to estimate biomass and have focused primarily assess species of timber interest methods, it was considered important to estimate biomass and carbon using a nondestructive method for a kind of scarce commercial importance, but of great environmental importance, as they are species of riparian vegetation, because although it is assumed that this vegetation plays a crucial role in the carbon cycle is little information available on the stores of this element in these ecosystems.

The objective of this research was to generate allometric equations and biomass carbon by a nondestructive method for tree species (Alnus acuminata K) a transect gallery forest river Texcoco.

Materials and methods

Fieldwork was conducted in a forest transect gallery Rio Texcoco, located specifically in the section of the river that runs through the community of San Pablo Ixayoc, municipality of Texcoco, Estado de Mexico, Mexico. The transect was selected under the criterion best condition and abundance of riparian vegetation, measured 960 m in length, began in the Zacasolco bridge and ended up in the Aclamaxa bridge. The geographical coordinates of both points are 19° 28' 44.9" and 19° 28' 56.6" north latitude and 98° 47' 55.7" and 98° 48' 43.8" W, located at 2 512 and 2 463 m respectively.

The climate of the study area is semicold with an average annual temperature between 5 and 12 °C. The average temperature of the coldest month ranges from -3 to 18 °C and the hottest month between 6.5 and 22 °C being the wettest semicold sub-humid climates with summer rains month. The precipitation of the driest month is less than 40 mm and the percentage of winter rain less than 5 mm. In the area the total annual rainfall varies along the altitudinal gradient, registering rainfall of 800 mm at the bottom and up to 1 200 mm at the top where the pine forest and alpine meadow (^{INEGI, 2001}).

The arboreal vegetation of the study area also varies depending on the altitudinal gradient, at the height of San Pablo Ixayoc forests of oak, dominated by different species of the genus Quercus (oak), these forests are as a transition between the present coniferous forests and jungles can reach from 4 to 30 m in height; They can be open or very dense and develop in different ecological conditions (^{Martínez, 2010}).

The work area is located on the Sierra Nevada mountain range, the Feozem soils occur in the top and bottom Cambisol in the top two of medium texture (^{INEGI, 2001}).

A quantitative floristic sweep transect was performed for this sample plots of 600 m2 (12 m wide by 50 m long) were defined, these were established longitudinal and zigzagging manner according to the river and covered both the current water, which at that time of year (November- October, 2012) measured between 1 and 1.5 m wide, and two strips of gallery forest. In each sample plot all tree species present with a diameter at breast height (DAP), measured at 1.3 m, greater than 5 cm and value indices of relative importance (IVIR) were calculated using the following equation counted taken from ^{Stiling (1999)}.

Where: IVIR= index value relative importance; Dr= relative density; Pr= relative predominance; and Fr= relative frequency.

Where: D= density of each species; D1= density of all species.

Where: P= basal area of each species; P1= basal area of all species.

Where: F= number of plots in each species occurs; F1= number of sampled plots.

The branches were classified into four diametric categories: thin, medium, thick and extra thick.An inventory of branches by the glass was made, the number of branches present in these are recorded, their respective diameter class was recorded and pruned in each cup a branch of each existing diameter class using a hacksaw forest.

In the tree 1 only one branch pruned as this thin branches presented only in 2 two branches were pruned, this appeared thin and medium branches. Tree 3 to 8 per copy three branches were cut, these trees had thin, medium and thick branches. Trees 9 and 10 presented the four categories of branches, but the first three categories were part of the fourth, for it was considered that these trees only had the kind of extra thick branches, this category only cut one branch per tree.

The 23 total branches pruned: 8 thin, 7 medium, 6 thick and 2 extra thick, all were separated on the wood components of the branches (mr) and leaves (h). The components of each branch were weighed with a scale clock 15 kg capacity and their values were pooled and averaged by type of branch to generate fresh weight (PF) for each component according to diameter class. With these values and the inventory of tree branches the contribution was calculated in fresh weight of each component biomass of each cup.

Field sample leaves 500 g per tree and branches 23 a basal slice of wood of about 5 cm thick was extracted was collected. In the laboratory the PF was obtained from slices of wood with a scale mark Volke SF- 400 with capacity of 5 kg and all plant material was placed in a drying oven at 90° C until it reached a constant weight or dry weight (PS), this happened at 24 h in leaf samples at 48 h in slices with diameters between 5 and 20 cm and 96 h in slices with diameters larger than 20 cm.

The values PF and PS per component were used to determine moisture content in percentage (CH%). Moisture content resulting component were averaged and the results are assumed as their respective average moisture content percentage. With the average percentages of moisture and branches stock biomass content was determined by dry weight of each component by the glass (Table 1) by using the following formula:

Table 1 Characteristics and dasometric biomass and carbon content 10 alder trees (Alnus acuminata K). 

DAP= diámetro a la altura de pecho (1.3 m), At= altura total, Af= altura de fuste limpio, BMR= biomasa de la madera de las ramas, Bf= biomasa del follaje, Bc= biomasa de la copa, BF= biomasa del fuste, BTa= biomasa total por árbol.

Where: B= biomass dry weight; Pf= Fresh weight; CH%= percentage moisture content.

Adding biomass dry weight of the wood of the branches (Bmr) and leaves (Bh) total biomass was obtained dry weight of the respective cup (BC) (Table 1).

Of the 23 slices of wood were selected at random per tree was determined volume using sand (ground commercial mixture of stone Company Cantocreto Palsa SA de CV) and a plastic container so neiloide (basal diameter: 16 cm, upper diameter: 13 cm, height: 11.5 cm) which it determines the volume (1 905.77 cm3) using the formula:

Where: V= volume (cm3); π= 3.1416; h= height; R= radius greater; r= smallest radius.

The procedure was to sift through a mesh No. 60 an amount of sand twice the volume of the container to give 3 500 g of sifted sand were deposited in a plastic bucket. After the neiloide container 3 500 g of sieved sand was poured and the excess material formed a mound, without pressing, was abraded with a metal ruler to flush the container, all the abraded material was collected and discarded. Then the container was weighed along with the sand that was inside after roughing (2 842.1 g) and this material was returned to the plastic bucket. Subsequently, for each volume determination in neiloide container is poured directly from the pail a sand layer about 5 cm thick and without pressing was placed on this slice of wood to be evaluated, then it poured into the container all the remaining sand in the bucket.

The shafts were subdivided upwardly in cylindrical sections as possible and ensuring that the dimensional change between the basal and apical diameter of each section was at least possible dramatic. The ends of each section were scored on the shafts with horizontal lines of approximately 15 cm made forest hacksaw. The basal and apical diameter of each section and the average of these measured was taken as the overall diameter of the respective section while the existing length between both diameters was regarded as its height. With these data, the volume of each section is calculated as if they would have been perfect cylindrical shape, using the following formula:

Where: V= volume; π= 3.1416; r= radius; h= height

The sum of the individual lengths of the sections in which the shaft is subdivided assumed as the height of clear bole (Af l), and the sum of the individual volumes of the sections was assumed as their actual or total volume. With these data the form factor of the shafts which is a reduction factor which indicates the taper of a tree and is expressed as the actual volume divided by the volume of a cylindrical solid reference tree determined dimensions corresponding to DAP height and clear bole of the tree representing (^{Romahn et al., 1994}).

The biomass of stems (BF) (Table 1) was obtained with the following formula proposed by Fehse et al. (²⁰⁰²):

Where: BF= stem biomass (Mg); DM= density of the wood species (Mg m-3); DAP= diameter at breast height (m); Afl= clear bole height (m); Ffp= average shape factor of the species.

Adding biomass drink (BC) to the biomass of the respective shaft (BF) was obtained the total biomass per tree (BTA).

Additionally, the percentage of carbon from the wood of the branches and leaves, for that, of the 10 trees evaluated quantified they chose five covering equally the range of DAP and analyzed by each tree a representative sample of the wood of the branches and one of the sheets. From each sample, 100 g of plant material is milled to pulverize using a mill Thomas Wiley Mill Model ED-5; in the case of wood slices, these first had to be converted into thin slivers carpenter using a chisel and mallet. The 10 samples were analyzed in automatic carbon autodeterminador brand Model Shimadzu TOC 5050A SSM according to the procedures of the Laboratory of Soil Fertility Postgraduate College (^{Etchevers, 1992}).

The carbon content of each sample was used to obtain the average percentage of carbon component. Because no trees are felled, the average carbon from the wood of the branches was used as the average percentage of carbon shafts. With carbon values per component and data of biomass dry weight of the cup and shanks the carbon content of these structures it was determined by the following expression taken from ^{Díaz et al. (2007)}:

Where: CCC= carbon content of the component; BTC= total biomass of component (kg); (%)C= carbon percentage component. By adding the carbon content of the cup to the carbon content of the respective stem the total carbon content per tree (CCA) (Table 1) was obtained. Allometric equations to generate biomass and carbon two models were used: linear and exponential, as indicated Rügnitz et al. (2008); ^{Gayoso (2002)} and expressed respectively as follows:

Where: Y= biomass (Mg); X= DAP (cm) and B0 and Bi are the model parameters.

Where: Y= biomass (kg); X= DAP (cm); and b and k are the function parameters.

The fit of the models to generate proposals allometric equations was performed using the SAS statistical analysis program (2002).

Results and discussion

The 762 trees were inventoried belonging to 9 species with the following IVIR, alder (Alnus acuminata K.); 36.81%, ash (Flaxinus uhdei W): 14.73%, oak (Quercus rugosa N): 14.12%, weeping willow (Salix spp): 11.92%, tepozan (Buddleia cordata K): 7.62%, pepper tree (Schinus molle L): 6.37%, cedar (Cupressus lusitanica M): 4.63%, hawthorn (Crataegus spp): 3.33% and eucalyptus (Eucaliptus spp): 0.46%.

With regard to the biomass of the tops, branches of diameter category: thin, medium, thick and extra thick showed basal diameters measured between 5 and 10, 10 and 15, 15 and 20, and between 20 and 25 cm respectively, with an average wood 03.05, 15.2, 49.5 and 174.6 kg and leaves 2.5, 7.5, 23.5 and 46.6 kg fresh weight respectively. While their average values for wood dry weight of the branches and the leaves were 1.79, 7.75, 25.25, and 89.05 kg and 0.95, 2.85, 8.93 and 17.71 kg respectively. The wood of the branches and leaves had humidity which ranged from 43-57% and 55-68% with averages of 49 and 62% respectively. Wood of branches provided a greater biomass compared to the cup with the leaves, on average wood of branches provided to each tree biomass 0.42 Mg and 0.1 Mg leaves.

Regarding stem biomass, the minimum and maximum number of sections in which they were defined ranged from 4 to 7, all basal diameter was greater than apical, no two sections that measured the same or within a tree or all trees within the shortest section measured 30 cm long and the longest 3.5 m. The higher was the height of the highest bole was also the number of sections in which they were defined, the height of clear bole varied from 3.93 to 13.10 m and the form factor (Ffp) of 0.5 to 0.75 with an average value of 0.66. The results showed, with the exception of shafts 7 and 8, the volume of clean shafts was not proportional to length if not DAP, the greater was the largest volume was also DAP. The volumes ranged from 0.08 to 4.38 cubic meters for trees with lower and higher respectively DAP and the average volume was 1.66 m3. Biomass ranged from 0.05 to 3.05 Mg for trees with the lowest and highest DAP respectively.

The percentage of biomass in trees was distributed as follows: 64.92, 27.06 and 8.02 in the shaft, the wood of the branches and leaves respectively. The carbon content varied from 48.2 to 50.3% and 47.1% to 46.5 in the wood of the branches and foliage with averages of 48.9% and 47.1% respectively. These results differ to those reported in similar jobs where the foliage has a higher carbon content (^{Díaz et al., 2007}; Juárez, 2008; ^{Gómez et al., 2011}). This could be due to the high proportion (50%) of small branches in samples analyzed foliage. By multiplying the values of biomass dry weight of the wood component of the branches and foliage by their respective percentages average carbon content carbon content of these components per tree (CCMR and CCf) was obtained and by adding these defined the carbon content of each cup (CCc) (Table 1).

Table 1 Carbon content in 10 alder trees (Alnus acuminata K). 

The carbon content of the shanks (CCF) was obtained by multiplying the value of their biomass by the average percentage of carbon content in the wood of the branches. This percentage was used since not been felled trees not had samples of stem wood and overall carbon content in the shafts is always more similar to the branches compared to the foliage (^{Gómez et al., 2001}; ^{Díaz et al., 2007}; Juárez, 2008). Lastly, the sum of the carbon content of the cup (CCc) and the shaft (CCF) generated the value of the carbon content in each tree (CCa) (Table 1).

The allometric equations generated with DAP values of the 7 trees and determined by the linear model showed a highly satisfactory coefficient of determination (R2= 0.99) and were expressed as:

Where: B and CCA are biomass content (Mg) and tree carbon (Mg) respectively and DAP is the diameter at breast height (1.3 m). However, both equations generated negative values of biomass and carbon to use them in DAP less than 23 cm, unlike the equations generated by the exponential model although ratios lower determination (R2= 0.95) these did not generate negative values to use them DAP less than 23 cm. Allometric equation to estimate biomass in alder (Alnus acuminata K.) using the exponential model was expressed as:

Where: B is the content of biomass per tree (Mg) and DAP is the diameter at breast height.

Exponential model parameters and the scattering points biomass values observed are presented in Figure 1.

Figure 1 Dispersion values of Alnus acuminata K. and the regression line generated with the content data using the exponential model biomass. 

By adjusting the equation to determine the biomass according to the DAP was presented a determination coefficient of 0.95.

The allometric equation to estimate the carbon content alder (Alnus acuminata K.) with the exponential model was expressed as:

Where: CCA is the carbon content per tree (Mg) and DAP is the diameter at breast height.

Exponential model parameters and the scattering points observed values carbon presented in Figure 2.

Figure 2 Dispersion values of Alnus acuminata K. and the regression line generated with carbon content data using the exponential model. 

By adjusting the equation to determine the carbon content depending on the DAP was presented a determination coefficient of 0.95.

The high IVIR of alder (36.81%) was influenced by their percentages density (27.82%) and relative frequency (26.32%), but fundamentally was determined by the percentage of predominance (56.30%), this indicated an outstanding development of their shanks compared with the other eight species. These values coincided with that observed in the field where his DAP were from 5 to 105 cm, and the height of a large number of its elements was greater than 20 m, while ash (Flaxinus uhdei W), the species the second largest IVIR (14.73%), the DAP of the species measured only 5 to 74.80 cm and its height was generally lower than in the alders. The marked development of the alders was attributed to its height captured more solar radiation and could photosynthesize more besides found growing in a favorable environment for growth, since as noted ^{Rzedowski (1978)}, as in many other parts of the world, forests of Alnus (alder) in Mexico are two main ecological affinities: live along streams and small rivers, or are successional communities arising as a result of the destruction of other forest types.

Biomass values observed in the cups and derivatives supply of wood from branches and leaves maintained reported in other work pattern (^{Acosta et al., 2002}; ^{Díaz et al., 2007}; ^{Avendaño et al., 2009}) where these increases directly with increasing the values of trees dasometric form. However, it is considered that for greater accuracy in determining the biomass of the glass by the method employed is desirable to prune more branches by the glass instead of just trimming one by diameter class or failing to perform a more efficient classification branchesas can becombined with the variable length diameter branches.

The biomass contained in the shaft increased by increasing DAP than its height and this pattern was retained in most of the trees except 7 and 8. The shaft of the shaft 7, despite having 10 cm over the DAP shaft 6, presented a biomass content (1.17 Mg) slightly lower than the shaft 6 (1.27 Mg). The shaft 8 despite its DAP of 0.72 m was severely affected by the low length of its stem (3.93 m) which affected so that this presented a marked decline in the biomass of this structure, because although its DAP expect his shaft submit higher and consequently about 1.5 Mg biomass actually presented only 0.65 Mg; i.e. half expected.

This might be because many trees despite having a pronounced DAP their stems at some point tend to bifurcate or at some point in their development could be pruned by the residents of the community. However, the pattern reported in work related to the topic (^{Acosta et al., 2002}; ^{Díaz et al., 2007}; ^{Avendaño et al., 2009}) remained where it is reported that the contribution of the stem to the tree biomass increases directly by increasing the values of trees dasometric form. Regarding the form factor obtained (0.66) was considered to be less than 1 was calculated as satisfactorily as noted Romahn et al. (¹⁹⁹⁴), in the case of shafts of morphic coefficient value or form factor it is always less than unity.

The total carbon content per tree tended to increase with increasing DAP, total tree height (At) and the height of clear bole (Af l), these values ranged from 0.03 to 2.12 Mg for trees with the largest and smallest DAP respectively. The shaft contributed the most carbon to trees compared to the cup, on average bole contributed 0.54 and cup 0.29 Mg per tree (1), these ratios are consistent with those reported in work relating to the subject (^{Díaz et al., 2007}; Juárez, 2008; ^{Gómez et al., 2011}). So it was considered that there was no adverse when using the carbon percentage branches to estimate the percentage of this element in the shanks effect.

Compared to the average carbon content in the wood of the branches and leaves the values obtained (48.9 and 47.1%) varied slightly to those reported in similar jobs where the leaves have higher carbon content (^{Díaz et al., 2007}; Juárez, 2008; ^{Gómez et al., 2011}). This could be due to the high proportion of small branches, about 50%, in leaf samples analyzed in this paper. However, the pattern observed in relation to the partial contribution of carbon by the stem, branches and leaves to the total content of this element in the tree was preserved.

Removing the shafts 1, 7 and 8 of analysis, allometric equations generated with higher coefficients of determination in both models. The coefficients of determination obtained greater than 0.9, indicated that in determining biomass and carbon in Alnus acuminata K. is reliable take the DAP as explanatory variable. It was considered that the best model to generate allometric equations was exponential, coinciding with ^{Gayoso (2002)} who notes that there is a preference for this model because it expresses a proportionality of relative increases between two tree components, coupled with that is widely consistent for different forms of growth.

Conclusions

The alder (Alnus acuminata K.) was the tree species with the highest IVIR (36.81%) of gallery forest transect studied.

The method of container of known volume and sand employed to determine the volume of the wood samples was effective because the sand is effectively molded to any shape of the samples and the amount of sand displaced was always similar to the volume of sample assessed.

The non-destructive method used for calculating biomass was effective by permitting determine this variable without tearing down the trees and to the pattern observed in similar studies where the stem is the structure that gives the tree as many have remained biomass and carbon, followed by wood branches and leaves.

The obtained coefficients of determination (R2) greater than 0.9 indicate that certain equations to estimate biomass and carbon Alnus Acuminata K. are reliable for that species under such conditions and within range of DAP evaluated.

The exponential model was the best estimated the biomass and carbon content as it does not generate negative values when using it in trees with lower DAP to 23 cm. The average carbon content in the species was 48.84%.