Detailed diagnosis of the biophysical environment

The parametric method of landscape classification (^{Duchart et al., 1989}) was used to obtain relatively homogeneous basic units of the territory (UP) that resulted from the superposition of the thematic maps of physiography, climate, land use and vegetation. To each basic unit was assigned a name for the use of soil and vegetation, followed by the type of climate and finally with the geoform of the land.

To design the physiographic map, terrain forms, slope and altitude were taken into account. The climate map was prepared based on the methodology of ^{Gómez et al. (2008)}, for which 23 meteorological stations with complete data of at least 20 years were selected (^{IMTA, 2013}). The temperature map was generated using the simple linear interpolation method to obtain specific monthly and annual average temperature data for the study area (^{Gómez et al., 2008}).

The mapping of the average annual rainfall was based on the graphic method, establishing analogies of the areas with meteorological stations with which they have no information. The selected weather stations were geographically located according to their coordinates in the ArcMap 10.1 Geographic Information System; its annual average precipitation value was included in a digital elevation model map with contour lines. The impact of the relief on the meteorological systems that generate the precipitation was analyzed and, expressed in the value of the annual average precipitation at each geographical point with information, the annual average isohyetal lines were drawn up at 100 mm intervals, starting from those that have more elements to delimit their location. Subsequently, the isohyetal lines plotted on a false color map denoting the vegetation cover were overlaid and to which the polygons were added with data on the types of land use and vegetation; isohyetal lines were adjusted based on the association that precipitation may have with vegetation cover (^{Gómez et al., 2008}).

The average monthly precipitation was estimated through an analysis of the distribution of regional precipitation from the annual average precipitation intervals (^{Gómez et al., 2008}).

The annual average temperature and precipitation maps were put together to define areas of climatic influence (AIC). In each AIC, the monthly and annual mean temperature and precipitation values were estimated. Subsequently, the type of climate was determined using the Köppen Classification System modified by García (^{García, 2004}).

To identify all land and vegetation uses of the study area, we worked with the INEGI Series VI polygons (^{Inegi, 2016}), in which the boundaries were redefined and detailed with greater precision by means of Google Earth satellite images (2018) at a scale of 1: 10 000.

Based on the satellite images, the tree density of the polygons of the secondary deciduous forest vegetation consigned by the INEGI Series referred to as high, medium and low was classified. Likewise, new polygons of roza-tomb-burning agriculture and mechanized agriculture were generated. For this, the results of previous work carried out in the study area were considered (^{Uribe et al., 2015}).

Agroforest diagnosis

In each UP a forest inventory was made to identify the species that best develop in each of them. Three 400 m² circular shape sampling plots were established in each UP, with a radius of 11.28 m. Trees and shrubs with a normal diameter> 2.5 cm were recorded throughout the sampling surface. For trees, the normal diameter was measured at 1.30 m. The diameter of the bushes was measured at 15 cm above the ground. The height of the trees and shrubs was determined with a 43890 Haga altimeter, and the diameters with a 283D Forestry Suppliers cloth diametric tape.

The Importance Value Index (IVI) was calculatd with the folwing formula (^{Mueller-Dumbois y Ellenberg, 1974}):

Where:

VI= Importance Value of the i species

Dominance = Basimetric area of the i species / ∑ basimetric area of all the species

Frecuence = Number of plots in which the i species is present / total number of plots

Abundance = Number of individuals of the i species/ total number of individuals of the whole sampled area

Based on what was observed in the field and in previous studies in the area (^{Uribe et al., 2015}; ^{Burgos et al., 2017}; ^{Yescas et al., 2017}), those UPs that had two or more limitations described in Table 2 were selected, to give a possible solution through the implementation of agroforest technologies, with special interest in increasing the low productivity of the system.

Social contribution

There were 32 semi-structured interviews that represented 80 % of the total producers in the study area. The interviews were prepared based on the Diagnostic and Design Methodology (^{Raintree, 1987}) which was adapted to the conditions of the region and the purpose of this research. The questions focused on the identification of other possible limitations in each UP, as well as in the main economic interest on the part of the producers; in a similar way, the preference of tree and shrub species of multiple uses for the design of agroforest technologies was asked too.

Design of agroforest technologies

The main limitations of each UP, the tree and shrub species of multiple uses indicated by the producers with special interest to the economic aspects and with representativeness in the landscape through the value of importance (VI) were considered. Aspects of physiography and climate were included in order to ensure adequate development in the field of these taxa.

Based on the literature review (^{Chacón and Harvey, 2006}; ^{Soto-Pintó et al., 2011}; ^{Bolaños et al., 2013}) the morphology of each species studied was determined, the density of trees per hectare was established, as well as the spatial arrangement of trees in each technology.

Evaluation of the economic viability of agroforest technologies

The total cost per hectare of the establishment of the proposed agroforest technologies was calculated through the charge for labor, charge for machinery and equipment, and charge for consumption. These charges were obtained from the concept and performance catalog of the National Forestry Commission (^{Conafor, 2014}), in which the current cost per wage and per tool to be used was modified. It is worth mentioning that the cost estimate was for April 2019, when the average exchange rate was $ 19.00 Mexican pesos per US dollar.

If the main economic interest of the producers is the sale of firewood, the economic benefits were estimated only for such activity. It was considered that the planting of the species proposed in each technology will be taken care of by the producers in the early stages and based on what is defined by the National Forestry Commission, the total volume of the tree was estimated at the age of 20, to obtain the number of wood loads per tree, under the assumption that there will be a 10 % mortality of trees planted in each technology (Conafor, 2013).

The volume was multiplied by the local cost of firewood load. The load of firewood is equivalent to 0.17 m³, so 1 m³ of firewood contains 5.8 loads. According to ^{Vázquez et al. (2017)} its cost depends on the type of species, and varies between $ 85.00 Mexican pesos and $ 100.00 Mexican pesos, which is equivalent to US $ 4.47-5.26.

Detailed diagnosis of the biophysical medium

It is recognized that 42 % of the surface of the study area has moderately sloping slopes with slopes between 2 and 8 %, at altitudes of 1 200-1 400 m (Table 3) (Figure 1).

Figure 1 Spatial distribution of the delimited geoforms in Ajuchitlán, Tlaquiltenango, Morelos. 

14 areas of climatic influence were delimited and three different climates were determined in the area evaluated (Figure 2). The warm climate, the driest of the subhumid ones, with a summer rainfall regime [Aw₀(w)(i')g] has a greater predominance in the area, and is presented in 53.47 % of the total area studied (1 436 ha); it is located in the river valley and the surrounding areas of moderately sloping slopes of the western part.

Figure 2 Spatial distribution of the climatic types found in Ajuchitlán, Tlaquiltenango, Morelos. 

On the sloping slopes, there is a gradient of greater precipitation, where the weather is warm, the normal of the sub-humid ones and with summer rainfall [Aw₁(w)(i')], the surface with this climate is of 609 ha (22.65 %). At the top of the slopes, the climate is characterized as semi-warm of the warm group, the normal of the subhumid and summer rainfall regime [A(C)w₁(w)(i')g], with an area of 641 ha (23.88 %).

Regarding the use of soil and vegetation, 11 types were delimited in the study area. The secondary vegetation of low deciduous forest (SBC) of medium density (42.30 %) predominates. The area dedicated to agriculture and grassland accounts for 19.30 % of the total (Figure 3).

SBC = Low Deciduous Forest; BE = Oak forest.

Figure 3 Spatial distribution of the climatic types found in Ajuchitlán, Tlaquiltenango, Morelos. 

By integrating the three thematic maps of physiography, climate and land use and vegetation, 14 UP were obtained, of which, the secondary vegetation of SBC in its different densities and climatic conditions, as well as physiographic, represent the use with greater surface area in the region, both for livestock and for the community (Figure 4). This is because during the rainy season, natural vegetation is the only source of food for livestock, which decreases the expenses of the producers (^{Uribe et al., 2015}).

1. Slash and burn agriculture in an Aw₀(w)(i´)g climate on a moderately sloping slope; 2. Mechanized annual temporary agriculture in an Aw₀(w)(i´)g climate in an almost flat river valley; 3. Grassland induced with medium density BE in a climate A(C)w₁(w)(i´)g on a hillside peak; 4. Grassland induced and introduced in an Aw₀(w)(i´)g climate on a moderately sloping slope; 5. Grassland induced with medium density SBC in an Aw₀(w)(i´)g climate on a moderately sloping slope; 6. Grassland induced with low density SBC in a climate A(C)w₁(w)(i´)g on a hilltop; 7. Secondary vegetation of low-density SBC in an Aw₀(w)(i´)g climate in an almost flat river valley; 8. Secondary vegetation of high density SBC in an Aw₁(w)(i´)g climate on a sloping slope; 9. Secondary vegetation of medium-density SBC in an Aw₀ (w) (i´) g climate on a moderately sloping slope; 10. Secondary vegetation of medium-density SBC in a climate A(C)w₁(w)(i´)g on a hilltop; 11. Secondary vegetation of medium-density SBC in an Aw₀(w)(i´)g climate on a sloping slope; 12. Secondary vegetation of high density SBC in an Aw₁(w)(i´)g climate on a hillside summit; 13. Secondary vegetation of high density SBC in an Aw₁(w)(i´)g climate on a sloping slope; 14. Secondary vegetation of high density SBC in an Aw₁(w)(i´)g climate on a moderately sloping slope; 15. Urban Zone; 16. Body of water.

SBC = Low Deciduous Forest; BE = Oak Forest; * UP selected for the design of agroforest technologies.

Figure 4 Spatial distribution of Relatively Homogeneous Landscape Units (UP) in Ajuchitlán, Tlaquiltenango, Morelos. 

Agroforest diagnosis

64 tree and shrub species were identified. The species with the highest importance value per landscape unit are shown in Table 4. These results are similar to those of ^{Hernández et al. (2011)} in a study carried out in Ajuchitlán, where they determined that the most important species are ixtumeca (Euphorbia schlechten Boiss), tlahuitol (Lysiloma divaricata (Jacq.) McBride) and tecolhuixtle (Mimosa benthami JF Macbr) they represent 37.29 % of the total importance value index.

The UP selected for the design of agroforest technologies were: slash and burn agriculture (1), mechanized annual temporary agriculture (2), induced and introduced grassland (4), grassland with low-density SBC at the top of the slope (6) and secondary vegetation of low density SBC in the river valley (7).

It was observed that the main limitation in the first three UPs (1, 2 and 4) is the use of conventional agriculture and grassland with intensive use of nitrogen fertilizers. In semi-structured interviews and field trips, limitations were recognized in fencing with dead posts, due to its high cost of establishment and maintenance (^{Uribe et al., 2015}). This is due to the fact that the area is clearly cattle-raising, which generates pressure for tree species that are used as poles to make dead fences in small and fractional plots with an average area of 1 ha.

In UP 6 there is soil degradation caused mainly by water erosion. In UP 7, low tree density was observed, which can cause soil loss if proper management tasks are not applied.

Design of agroforest technologies

Live fences

In the UP with pasture and agriculture (1, 2 and 4) the establishment of living fences is recommended, which according to ^{Harvey et al. (2005)} are multi-purpose trees planted in rows, managed by farmers. They are used to divide pastures and protect crops from animals, or to mark the boundaries of plots, which helps maintain the landscape and connectivity of the agricultural-livestock region (^{Chacón and Harvey, 2006}). Because the properties are small and divided, no other type of agroforest technology is recommended.

For the establishment of living fences it is suggested to incorporate the following species: Gliricidia sepium (Jacq.) Kunth ex Walp. or kill rat; Acacia cochiacantha Willd. or brown cubata; Mimosa benthamii or tecolhuixtle and Guazuma ulmifolia Lam. or cuaulote.

These species are of multiple use and have a high value of importance. They develop properly in the almost flat river valley and on moderately sloping slopes with the Aw₀(w)(i’)g climate type. They can be used to sell firewood once they have reached the desired development, since they are appreciated by the villagers (^{Vázquez et al., 2017}).

For its establishment in the field, a spacing of 4 m between the plant is proposed, interspersing a dead pole every two meters, with a density of 100 ha^-1 plants. A smaller separation could reduce crop yields, due to competition for light and nutrients. The barbed wire should be placed when the plants are well rooted and have a thickness of more than 15 cm in normal diameter (^{Chacón and Harvey, 2006}).

According to ^{Yescas et al. (2017)} the establishment of living fences is one of the recommended technologies in the agricultural-livestock area of the low deciduous forest of the Sierra de Huautla, Morelos. This is because, in addition to serving as boundary delimitation, it can meet the needs of firewood for trees planted in such technology.

According to ^{Cortez et al. (2017)} the use of live fences may be an option to reduce the negative impacts of extensive livestock, loss of natural vegetation and soil degradation in the Sierra de Huautla. Likewise, ^{Soto et al. (2008)} mentioned that this technology has a beneficial effect on the soil by reducing erosion, management costs are low or zero and provide various products, among which the use for firewood and fodder.

Live barriers

In UP 6 a high degree of erosion was observed. For this reason, it is recommended to apply the practice of living barriers, which will be used to control the loss of soil due to water erosion, which reduces the speed of runoff water downhill and thus the removal of the soil; it also favors that the soil particles settle in the upper part of the living bars and form natural terraces (^{Quinton and Rodríguez, 1999}). With this strategy, the soil degradation process will stop and, over time, the productive potential of the land will increase.

Likewise, little diversity of species and an important presence of ixtumeca (Euphorbia schlechten) were identified, with an importance value of 27 %. This species is considered by the community as invasive, very aggressive and of little use. Therefore, it is recommended to use rustic species that serve to retain the soil and generate adequate conditions to restore long-term tree species that are useful. The carbon content on the site is expected to increase and gradually replace the ixtumeca.

The species proposed for this activity are nopal (Opuntia lubrica Griffiths) and Chinese copal, which develop properly on the top of the slope and in the type of climate A(C)w₁(w)(i´)g. Although the cactus is not highly representative in this area (IVI 6 %), it can be a good option to be implemented in live barriers in areas with high erosion, because it is a rustic species and develops well in these conditions and It has a high survival (^{Bolaños et al., 2013}).

The Chinese copal (Bursera bipinnata DC. Engl.) is recommended from having a high value in the area (18 %) in addition to branching from ground level in its early stages. Likewise, it is a species highly prized by residents for its resin.

For the establishment in the field, the separation between cactus plants should be 20 cm from each other, along the contours with a distance between curves of 16.6 m (^{Bolaños et al., 2013}) and a planting density of 3 012 individuals ha^-1. The strips with prickly pearls and the strips with Chinese copal should be placed interleaved. Once established the plants will be pruned to give them structure.

In the nopal (Opuntia spp.) the main stems will be used to form the boards between plants that reduce or limit runoff. In the case of copal, the material removed in pruning will be accommodated between the plants to form boards. When the barriers are established, thinning will be carried out to allow the best development of the plants, taking care to keep the borders on the contour.

According to ^{Zúñiga et al. (1993)} the use of living barriers in some tropical regions of Mexico has been used successfully, which is why it is widely recommended in areas with soil degradation. It is expected that by using the living barriers in this UP, water erosion will be reduced, by means of the longitudinal decrease and the gradient of the slope; and soil moisture is increased to favor the good development of plants (^{Ruiz et al., 2001}).

Acahuales enrichment

In UP 7, species enrichment is recommended using the improved acahual technology proposed by ^{Soto-Pintó et al. (2011)}. The recommended species are tlahuitol, tepeguaje (Lysiloma acapulcensis (Kunth) Benth), tecolhuixtle and cuaulote. These species have a full development in the almost flat river valley with a kind of Aw₀ (w) (i´) g climate, which are the biophysical conditions where it is proposed to establish agroforest technology.

The species selected for this activity coincide with the results of ^{Moreno and Paradowska (2009)} who mention that the potential use of these species, apart from enriching the acahuales and generating environmental benefits, are preferred by the producers for the use of firewood and fodder in the low deciduous jungle of central Veracruz.

For the establishment of the plantation it is recommended to make gaps of 2 m wide in a direction perpendicular to the slope with a separation of 3 m between trees. Between line and line of trees 7 m of natural acahual are left, which forms a density of 476 ha^-1 trees. If trees larger than 20 cm in diameter are found, they should be left there (^{Soto-Pintó et al., 2011}).

According to ^{Soto-Pintó et al. (2011)} the enrichment of acahuales maintains the regeneration ability of natural vegetation and increases the potential of the soil by increasing coverage, biomass and productivity. This contributes to improve the environmental benefits and, therefore, the economic and social benefits (^{Canadell and Raupach, 2008}).

Economic viability of the proposed agroforest technologies

Table 5 shows the establishment costs per hectare of the technologies discussed. It is observed that the enrichment of acahuales turns out to be the cheapest to implement, mainly due to the fact that many inputs are not required, such as in live fences and barriers.

The costs per hectare obtained in this research are within the range calculated by ^{Sáenz et al. (2010)} ranging from $ 10 207 to $ 19 923 Mexican pesos, depending on agroforest technology. The implementation of agroforest technologies generates a considerable investment cost for the producers' economy, which varies depending on the fixed costs; as well as the current prices of inputs and labor. However, the long-term benefit, both economic and environmental, will make farmers recover their investment, which, although it can be paid for with their own resources, can also be financed totally or partially by government institutions.

As the main economic interest of the producers is the sale of firewood, the calculations were made only for this activity. It was determined that the economic benefit from the sale of firewood depends largely on the species. Because the volume is interrelated with the growth in diameter and height of the species. It is observed that the benefit is affected by the density of trees per hectare that is proposed, under the assumption that there will be 10 % mortality of trees planted in each technology (Table 6).

According to the results obtained, the sale of firewood turns out to have a high economic value. This is mainly due to the fact that firewood is considered by rural families as an indispensable natural resource for their way of life. Based upon the Food and Agriculture Organization of the United Nations (FAO) and the former Mexican Agriculture Department (Sagarpa) (^{FAO and Sagarpa, 2007}) an average family needs about 500 kg of firewood annually to meet the needs of energy in cooking food, which implies an annual consumption of 8 m³, equivalent to the extraction of 32 trees.

Because the extraction of firewood is practiced on a daily basis in the communities of Morelos, negatively affecting the forest resources of the deciduous forest (^{Yescas et al., 2017}), it is very important to generate and implement agroforest technologies with species that have quality dendroenergetics; It turns out to be a viable alternative that has both economic and environmental benefits. In this way, the pressure on those preferred species for firewood can be reduced and the negative impacts reduced.