Data collection

For the selection of producers to whom the evaluation survey was applied, simple random sampling was used, sample size (n) was defined by ^{Fernández (1996)} using Equation 1.

Where: N= total population; Z_α= 95% confidence level; p= expected proportion of 5%= 0.05. Based on the above, the sample size of 92 producers was determined, applying 56 questionnaires for FIMMC and 36 for NMMV.

A structured and administered study instrument was used, through a personal interview with the head of each household. The data collected included socio-economic characteristics of households, management of the production system, as well as the use of corn RC by producers in these two regions for the observation period of one year prior to the survey (2014), through the production data reminder method. The data analysis and the application of the ordered Probit model were performed in the statistical software package STATA version 12®.

Methodology for estimating opportunity cost

In order to estimate the opportunity cost of competitive uses, the production volume of corn RC was determined in both localities, using the weighted average yield of corn stubble by ^{Reyes-Muro et al. (2013)}, who indicate that the production obtained in one hectare is of 46.6% grain and 53.4% of stubble. The calculation of stubble production was obtained by equation 2.

In addition, according to the methodology ^{Gasque (2008)}, the consumption of ms per animal was determined by 3.2% of its live weight per day, for which the equivalences per animal unit (UA) were established, according to the ages established by the SAGARPA, was considered a UA, with average weight of 450 kg (^{INIFAP, 2011}). Therefore, the consumption of more day^-1 of an UA was of 14.4 kg, with this information was estimated the period that could maintain the existing MR to the herds farms in both localities. Also, taking as reference that the rental price of pasture per UA is $160.00 per month, the income that could be obtained by each producer was calculated, allocating the amounts of existing stubble as a source of fodder.

On the other hand, to determine the amount of N, P and K that can release 1 t of stubble in the soil, the percentages established by Nijhof (1987) of 0.9, 0.2 and 1.4%, respectively, were considered. To estimate the total N, P and K incorporated in the soil with the retention of the RM, these percentages were multiplied with the average amount of RC produced in NMMV and FIMMC. Subsequently, considering the chemical fertilizers existing in the local market: urea (46-00-00), triple calcium superphosphate (00-46-00) and potassium chloride (00-00-60), the kilograms of N, P and K that could be released in the soil, with the amounts of stubble existing in each common, multiplying by the prices of these fertilizers, obtained from ^{SNIIM (2015)} in the second half of August 2015.

Empirical model: probit ordered

The econometric model Probit Ordenado, proposed by ^{McKelvey and Zavoina (1975)}, was used to analyze the determinants that affect the use of RM for soil cover. The distribution of the proportion of maize residues for soil cover, reported by the sample producers, was concentrated in values of 0, 50 and 100%, which group the proportional use of the residues in three categories (0 if < 34%, 1 if the percentage of use is between 34 and 66%, and 2 if the percentage of use is > 66%), represented using the probit ordered model, by equation 3.

Where: Y^*= latent proportion that RC are used for soil cover; X= vector of explanatory variables for the use of RM for soil mulch; β= vector of the coefficients of the explanatory variables; and ε= random error following a standard normal distribution. The following set of explanatory variables is included for the equation: characteristics of the producer (age, educational level of head of household, family workforce in agricultural activities), size of plots where maize was produced, size of herd, amount of RC generated in the plots, area available for pastures and training services received.

The current proportion of maize residues used as soil mulch (Y_m,i) reported by farmers were grouped into three categories (0, 1 and 2) which is specified by equation 4.

Where: γ_m,1 and γ_m,2= unknown limits of the categories in the latent variable. The probability of a given producer being located in one of the possible combinations is given in equation 5.

Therefore, the probability that the observation of an individual (i) is located in option j is presented in equation 6.

Characteristics of inhabitants and production systems

The results indicate that in the common NMMV, the heads of households had an average age of 59.3 years and 3.9 years of schooling. On the other hand, the average area used for agricultural activities is 8.5 ha, of which 5.7 ha (67%) were used for the production of maize in the spring-summer 2014 cycle (SS-2014), obtaining a yield of 4.9 t ha^-1.

In FIMMC, on average, producers registered 61.4 years of age and 3.2 years of schooling. The area devoted to agricultural activities was 12.2 ha, of which only 20.5% was destined to maize, with 1.2 t ha^-1 of maize grain (Table 1). Comparing the results of both production systems, the agricultural area was statistically higher in the common FIMMC; however, the averages of surface area for maize and the yields obtained were statistically higher in the NMMV common.

Table 1 Characteristics of the inhabitants and production systems of the study communities. 

It was also detected that 100% of the interviewees in the two common cultivate maize, either alone or intercropped with another crop: 90.3% grow maize alone, 5.4% maize and beans, and 4.3% maize and squash in NMMV and 33% maize alone, 47% maize and beans and 19% maize and pumpkin in FIMMC, in the latter it was observed that 66% of the producers diversify their crops much more.

In FIMMC 59% of the producers burned the remaining stubble, after grazing the cattle. Contrary to what happens in NMMV where only 3.2% burned them. This is due, on the one hand, to the fact that in the NMMV community a conservation tillage program was implemented in the region and, on the other, to the application of a state law enacted in the 1990s, which promotes the burning of waste not. However, this effort, although significant, has not been sufficient to assure the coverage of stubble required by AC, corresponding to 30% of the land cover (^{Erenstein and Cadena, 1997}), due to the small percentage of stubble left over plot after grazing (^{López et al., 2013}). This was evidenced by the study of Erenstein et al. (1998) in the Frailesca, since of 82% of the producers who had supposedly opted for non-burning, only 12% adopted the conservation of waste.

On the other hand, according to the characteristics of the plots used for maize (SS-2014) in the two localities, it was observed in NMMV that 69% of the producers considered that they have a soil with a medium fertility, besides the majority (72%) think that their soil is moderately deep and 75% indicate that their land has a topography. However, in FIMMC 16% mentioned that they have poorly fertile soil, almost half (45%) with shallow soil and 98% indicated that the slope of their plot is medium and abrupt.

Production and use of crop residues

According to estimates of stubble production, the average amount of RC generated in NMMV was 31.95 t ha^-1, much higher compared to FIMMC, with 3.79 t ha^-1, due to the average size of maize plot and to yield obtained; in addition, because of the soil characteristics, since the majority of the FIMMC farmers cultivate in soils with poor fertility and less depth, that do not allow to increase the biomass, with which could increase the volume of waste generated and made available for their use as food and soil mulch (^{Naudin et al., 2015}).

Also, it was detected that in NMMV the use of stubble for soil cover is more common, 33.3% of the producers use it as the only option and 47.2% combined with the feeding of cattle, so the positive effect that they have had AC programs in this region. Contrary to FIMMC, only 9% of farmers use RC as their sole use for soil mulch and 16% combine it with the use of forage for livestock (Table 2). On the other hand, the total of the interviewed producers used RC for at least one of the alternative uses (livestock feed, land cover, sale or burning).

Table 2 Uses of maize residues in two communities in the state of Chiapas.  

In the two localities the producers used RM as feed for cattle, 50 and 85% in NMMV and FIMMC, either as a sole purpose or shared with another alternative use (land cover, sale or burning), use it through grazing in situ. A similar result to FIMMC reported ^{López et al. (2012)} in a study of 21 producers of the Frailesca, where 80% of the interviewees introduced cattle to their plots.

Proportion of stubble used for feed vs. soil cover

As for the proportion of stubble used for food and mulch (Table 3), most producers use RC for livestock feed, for example, in NMMV 38.9% use more than 66% of the stubble available for their livestock. As a result, only 38% of the producers leave enough stubble (> 33% of the total) that can be considered as effective mulch for the AC, this proportion of producers coincides with the study carried out by ^{Erenstein and Cadena (1997)} in the municipality of Motozintla, Chiapas, who reported that 38% of the interviewed producers had retained stubble above the recommended threshold (2 t ha^-1).

Table 3 Proportion of stubble used for feeding vs soil cover in study communities. 

On the other hand, in FIMMC, the amount of stubble that was destined for cattle feeding was greater, since 51.8% of the respondents allocated >66% as feed for cattle, and 12.5% dedicated >66% of stubble for coverage of soil. Of the 92 producers surveyed, 43 of them used more than 66% of stubble for livestock; that is, 46.7% in both localities, and only 20.6% use more than 66% of RC for soil mulch. This clearly demonstrates that the use of maize residues was mainly used for cattle feeding, a condition that was similar to that established by ^{Hellin et al. (2013)}. Taking into account the different percentages of stubble left by producers as mulch and the amount of RC produced in both localities, it follows that it may be easier to introduce land cover practices where biomass production was high enough to satisfy competitive uses (^{Valbuena et al., 2012}). Also, authors such as ^{Valbuena et al. (2015)} concluded that in places where stubble production is reduced there are pressures on stubble, which give rise to opportunity costs.

It is important to point out that the producers, who averaged <33% of the waste as a mulch, did not manage to cover 30% of the soil surface, as a minimum requirement to promote the conservation of the upper soil layer, ^{Erenstein (2002)} concluded which with this percentage could reduce its erosion by 80%. It is necessary to keep in mind that 30% of the surface can be covered with a threshold of 2 t ha^-1 of RM (^{Erenstein, 1997}). Consequently, when the quantities of RC that are maintained as surface mulch are insufficient, minimum tillage alone can lead to lower yields compared to current agricultural practices, especially in soils that are prone to formation of crusts and compaction (^{Baudron et al., 2012}).

Producers who own cattle and the use of corn stubble

The 41.6% of the producers have cattle in NMMV, with a mean of 14.8 UA, and 100% of them use RM as cattle feed in some proportion. These farmers have paddocks of 8.5 ha in average, the main pastures that they own are star (Cynodon plectostachyus Rich) and llanero (Andropogon gayanus Kunth). In FIMMC, 78.5% of the producers have cattle, on average with 16.6 UA, and the majority of the farmers (95.4%) allocate some percentage of stubble for their livestock. In addition, these have an average area of 11.1 ha for grazing, where predominate star grasses and zacatón (Sporobolus airoides Torr).

It was observed that there is a strong relationship of the producers that have cattle with the use of RC, since 97.7% allocate some percentage (<33, between 34-66 and >66%) of maize residues as fodder. On the other hand, 100% of the producers that own cattle in the two evaluated localities, graze their cattle from january to march, after harvesting in the plots where they planted maize: in both regions there is a serious shortage of forage during the (Erenstein et al., 1998), according to the UA registered in each region, an average consumption of 72 and 73% of the available stubble was estimated in NMMV and FIMMC, data similar waste consumption in the Frailesca region, 60-80% have been reported by ^{López et al. (2013)}, this was ratified by ^{Hellin et al. (2013)}, these authors determined that in Chiapas, 70% of the stubble produced is used for livestock and 20% for soil cover.

Opportunity cost of stubble use: food vs cover

This study focused exclusively on the use of dry stubble, after harvesting corn kernels. It was estimated that one cow consumed 5 256 kg ms year^-1. Assuming stubble as the sole source of maize for cattle feeding and that all maize biomass produced was used for feeding, the average stubble produced could maintain a cow for 2 218 days in NMMV and 263 days in FIMMC.

Considering that on average livestock ownership by producers is 14.8 and 16.6 AU in NMMV and FIMMC, the RM could maintain a producer&apos;s herd for 150 days (5 months) in NMMV, and 16 days in FIMMC. This shows that the stubble produced is much smaller than the annual requirement for livestock feed in both regions. On the other hand, taking the reference price of pasture rent per UA in the region, $ 160.00 month^-1, would yield an income of $11 840.00 and $1 328.00, respectively, using stubble produced as fodder. However, it is important to note that RM is not the only source of ms in the study regions, since most of the farmers own paddocks with pastures and use mineral salt as a nutritional supplement.

On the other hand, considering that the RM provides N, P and K based on ms, 1 t of stubble left in situ could incorporate to the soil 9, 2 and 14 kg of N, P and K, respectively. With this assumption, the RC generated at the producer level, 31.95 t ha^-1 on average, in NMMV could incorporate 287.55, 63.9 and 447.3 kg of N, P and K, respectively.

When considering the prices of the most common synthetic fertilizers in Chiapas, prices obtained from ^SNIIM/2015), and their percentage NPK content, savings of 625 kg of urea ($3 750.00), 138.9 kg of triple superphosphate of calcium ($1 361.00) and 745.5 kg of potassium chloride ($5 815.00) for NMMV. For FIMMC, the average stubble produced, 3.79 t, could incorporate 34.11, 7.58 and 53.06 kg of N, P and K, respectively, and generate a saving of 74.15 kg of urea ($445.00), 16.47 kg of triple superphosphate of calcium ($161.00) and 88.43 kg of potassium chloride ($670.00).

Table 4 presents the opportunity cost of using stubble for both cattle feed and soil cover. As can be seen, the opportunity cost represents 6.7 and 4% for NMMV and FIMMC, i.e. if producers used stubble for soil cover they would be “losing” $914.00 in NMMV and $52.00 in FIMMC, which is equivalent to $160.00 and $ 20.8 ha^-1, respectively. Although it is important to emphasize that the benefits of the use of stubble for livestock are short term and as mulch are reflected in the medium and long term.

Table 4 Opportunity cost of using corn stubble for cattle feed vs soil cover  

^{Beuchelt et al. (2015)} point out that producers hardly renounce current income from future profits because of scarce resources, limited liquidity and the risk-producing environment; in addition, high pressure on food resources, since they cannot afford to invest in biomass technologies for soil improvement, since they have strong priorities in ensuring the immediate feeding needs of their livestock (^{Valbuena et al., 2012}). Similarly, ^{Rusinamhodzi et al. (2015)} indicate that from an economic perspective, it is logical that farmers prioritize the livelihood of livestock with RC more than the management of soil fertility. However, in the study area, the potential of improving the availability of stubble as a mulch was detected, as well as increasing the availability of forage, since most of the producers who practice the maize-livestock system have areas of pastures that could be used to establish improved forages.