elocation-id: e3855
The sugarcane agribusiness is important for food sovereignty in Mexico, but climate change affects its production and yield. This study analyzed how climatic variables and prices impact sucrose yield in fresh cane at the Tres Valles mill, Veracruz, from 1995 to 2022, using a model of two simultaneous equations. The objective was to analyze to what extent climatic variables and respective relevant prices affect the yield of sucrose in cane, expressed in its equivalent of fresh cane, in order to quantify the magnitude of this impact. The results show that thermal oscillation has the greatest positive impact (elasticity of 1.01), whereas solar radiation has the greatest negative impact (elasticity of -2.06). Carbon dioxide emissions have a moderate positive effect (elasticity of 0.78) and precipitation has a smaller positive impact (elasticity of 0.17). The index of prices of cane sugar shows a positive elasticity of 0.96, while the wholesale price of standard sugar has a negative elasticity of -0.2. It was concluded that this type of study provides valuable information on the sensitivity of sucrose yield to climatic and economic factors, useful for mitigating climate change.
climatological normals, distorted prices, Gauss-Markov theorem, sucrose in fresh cane.
Climate change is a global phenomenon with primary manifestations in variations in meteorological normals (Arnell et al., 2019). Its consequences include ecological, social, and economic impacts at the local and regional levels (Byg and Salick, 2009). Research shows that agricultural yields will decrease if climatological variables exceed certain thresholds (Habib-ur-Rahman et al., 2022; Bibi and Rahman, 2023).
In Mexico, sugarcane is an important crop, economically and in terms of job creation. In 2023, it ranked third in agricultural production value with 5.8% of the total. Nevertheless, the sector faces challenges such as high sugar prices and increased imports, partly due to droughts. Models such as the DSSAT simulate crop growth under different scenarios, supporting decision-making (Everinghama et al., 2002). Nonetheless, these models focus on biophysical processes, making it difficult to incorporate complex economic analyses (Stern et al., 2022). Although some incorporate basic economic analysis (Morris et al., 2019), there is a gap between agroclimatic and economic models, as the former do not incorporate econometric structural analysis (McNunn et al., 2019).
This research adopts a conventional econometric approach, modeling yield based on relevant variables, with the aim of estimating a simultaneous equation model to quantify the effects of climatic variables and prices on sucrose yield in cane.
The study area of the research is the one delimited by the geographical coordinates of the four municipalities in which the supply area of the Tres Valles sugar mill is located. That is, between 18° 45’ 00” and 18° 08’ 00” north latitude and -96° 38’ 0” and -95° 46’ 00” west longitude. The study is carried out with data from the Tres Valles sugar mill, Veracruz, for the period from 1995 to 2022. The time series of temperature, precipitation, and solar radiation were obtained from NASA (2024). Data for sugarcane variables were obtained from the National Union of Sugarcane Growers of the National Confederation of Rural Owners for several years (UNC-CNPR, 2004; UNC-CNPR, 2014; UNC-CNPR, 2024).
Table 1 shows the variables that were used for the final run of the estimated empirical model.
[i] For Tosc, Rai, SRad, the source is NASA (2024); for Sacapa and WPri, it was the UNC-CNPR (2014, 2004 and 2024), the latter was deflated with the national producer price index (INPP, for its initialism in Spanish), thus obtaining the real price of standard sugar in the main supply centers (RWPri), which is the one finally used in the empirical model. The base year of the INPP was (2014-2016= 100) and the source of this index was Banco de México (2024). CO2 was obtained from NOAA (2024), and D1 from the SAS Institute (2024). Finally, the cane sugar price index (CSPI) was that from FAO (2024).
The empirical model to be estimated is the following system of two simultaneous equations.
2). Where the β's and α's are the parameters of the respective equations. The parameters β 0 and α 0 correspond to the ordinate to the origin (intercept) of equations (1) and (2). The respective expected signs are the following: β 1> 0 , β 2≶ 0 , β 3> 0 , β 4≶ 0 , β 5> 0 , β 6> 0 , β 7≶ 0 and α 1> 0. In addition, ε 1 and ε 2 are the statistical errors of the equation in question, with the property that they are normally distributed with mean μ and finite variance σ 2 .
It is important to note that, in the case of the parameter β 2 , which weights the thermal oscillation (Tosc), it is not possible to establish a priori the type of relationship it will have with the yield of sucrose in sugarcane (Sacapa) because, although in some studies such as that by Cardozo and Sentelhas (2013), it is mentioned that there is a positive correlation during the cane ripening phase, this is not always the case, as this relationship can vary depending on other factors, such as the variety of sugarcane, the age of the crop and fundamentally, the environmental conditions .
In the case of solar radiation (SRad), although a positive relationship with sucrose in cane would also be expected, under stress conditions; for example, drought, an excess of radiation could even have negative effects (Inman et al., 2005) . Regarding the parameters of the average temperature (β 1 ) and precipitation (β 3 ), they are expected to be positive in advance; however, if these are not in fact linear, the sign can be negative . It is important to note that when these variables are outside the optimal limits due to climatic variations, it can cause a decrease and even loss of sucrose in the cane (Inman et al., 2005; Cardozo and Sentelhas, 2013).
In the case of the dummy variable (D1), a negative relationship would be expected since both ‘El Niño’ and ‘La Niña’ cause stress in sugarcane, especially in the phenological stage of ripening (de Souza et al., 2015). In the case of the parameter α 1 , a positive relationship between the mean temperature (Taver) and carbon dioxide (CO2) is expected. Lacis et al. (2010) have established the causal relationship between temperature increase and carbon dioxide emissions.
In the first equation of the system, the FAO cane sugar price index (CSPI) is introduced as a proxy economic variable that explains the sucrose yield in cane expressed in tonnes of fresh cane. The CSPI is a global indicator that reflects trends in the international sugar market, which can have an impact on production decisions and management in sugarcane crops at the local level and influence producers’ management practices, potentially affecting sucrose yield (Cheavegatti et al., 2011).
Finally, the introduction of the real average price of standard sugar in the main supply centers of Mexico (RWPri) is due to the fact that it is used by CONADESUCA (2010) to determine the reference price that serves to determine the liquidation price for sugarcane producers. It is assumed that a higher reference price incentivizes producers to increase the planted area and production of sugarcane, but more importantly, the reference price in Mexico is based partly on sucrose content because this incentivizes sugarcane producers to focus not only on volume, but also on the quality of the cane in terms of its sugar content.
For the identification of the empirical model and following Gujarati and Porter (2009); Wooldridge (2016), the order and rank conditions were applied and for the estimation of the structural and reduced models, the SAS/ETS 9.4 system was used . Lastly, it is necessary to mention that the calculation of the elasticities evaluated at one point was carried out as indicated in Tomek and Kaiser (2014).
The empirical model that was used consists of two simultaneous equations that are composed of seven exogenous or predetermined variables and two endogenous variables. It was found that the first equation of the system (Sacapa) is exactly identified, whereas the second equation (Taver) is over-identified; therefore, the model as a whole is identified and it is possible to estimate the structural parameters . Table 2 shows the descriptive statistics of the variables used in the model.
Table 3 shows the structural parameters of equation (1) of the estimated model.
As observed in column six, the value of 11.7 of the F statistic with 7 degrees of freedom indicates that, statistically, the model as a whole is highly significant; there is strong evidence to reject the null hypothesis that all coefficients of the independent variables are simultaneously equal to zero. Additionally, given that the adjusted coefficient of determination (R2adjusted) is 0.74, this model explains approximately 74% of the variability in the dependent variable (Sacapa).
Additionally, the Durbin-Watson (DW) statistic has a value of 1.97 (very close to 2), so it can be stated that there is no significant first-order autocorrelation in the model residuals, which suggests that the model is well specified in terms of its temporal structure and that important variables that could cause dependence on errors have not been omitted (Gujarati and Porter, 2009). Finally, in the case of the variance inflation factor (VIF), it indicates that there are no problems of multicollinearity. Although in the case of the cane sugar price index, it has a VIF of 2.6, it can be considered as a mild to moderate multicollinearity that does not cause problems (Hair et al., 2019).
Table 4 shows the results of the second equation of the empirical model.
| Variable | Estimated parameter | Standard error | t- value | Pr > | t | | Statistic | Value | VIF |
|---|---|---|---|---|---|---|---|
| Intercept | 18.1153 | 1.212693 | 14.94 | <0.0001 | F | 21.33 | 0 |
| CO2 | 0.014385 | 0.003115 | 4.62 | 0.0001 | R2adjusted | 0.46 | 1 |
| DW | 1.67 |
The interpretation of the statistics F, R2adjusted and DW is analogous to the previous case and therefore, the estimated parameters of the empirical model are unbiased, consistent, and efficient. Nevertheless, the analysis of the marginal effects of the covariates on the response variable can be confusing if it is carried out in terms of absolute values due to the physical units of the variables. To avoid this situation, in economics, it is more common to perform sensitivity analysis in relative terms, free of physical units, that is, through the analysis of elasticities evaluated at a point, which are shown in Table 5.
| Variable | Tosc | Rai | SRad | CO2 | CSPI | WPri | D1 |
|---|---|---|---|---|---|---|---|
| Sacapa | 1.01 | 0.17 | -2.06 | 0.78 | 0.96 | -0.2 | -0.02 |
| Taver | 0.17 |
In general, it can be assumed that the sensitivity analysis in relative terms of the results of the reduced model through elasticities considers taking a 1% change in the exogenous or predetermined variables, and the respective estimated parameter will be the percentage change that will occur in the endogenous or response variable. Thus, a 1% change in the thermal oscillation (Tosc) will increase the sucrose in cane by 1.01%, but expressed in terms of fresh cane (Sacapa), a 1% increase in precipitation (Rai) will increase the yield of the Sacapa by 0.17%; a 1% increase in solar radiation (SRad) will cause the yield of the Sacapa to decrease by 2.06%; in the event of a 1% increase in CO2 emissions, the Sacapa will increase by 0.78%, which is due to the so-called fertilization effect of this greenhouse gas on sugarcane.
The elasticity of the average temperature (Taver) with respect to the Sacapa, which is obtained from the structural model since both variables are endogenous, was 4.64. This implies that if the Taver increases by 1%, the Sacapa will increase by 4.64%. On the other hand, if the cane sugar price index (CSPI) increases by 1%, the Sacapa will increase by 0.96%. In the case of the real wholesale price of standard sugar (RWPri) as an exogenous variable of the Sacapa yield, it did not present the expected sign, since a 1% increase would cause a decrease in the Sacapa, which is, apparently, an economic irrationality on the part of the producer .
However, a possible explanation for this fact is that the sugarcane market is one of the most distorted, since; for example, local sugarcane leaders continuously promote social movements that can affect the prices of sugarcane as a fundamental input of the sugar industry; likewise, the price of sugarcane is a controlled price, as it is set by the government and not by the market (Aguilar et al., 2011) . Finally, in the case of the dummy variable D1, coded as 0 for the prevalence of ‘La Niña’ in the respective year and as 1 for the prevalence of ‘El Niño’ in that year, the corresponding parameter had a value of -0.02; its interpretation is as follows: on average, the Sacapa is 0.02% lower during the years of prevalence of ‘El Niño’ compared to the years of prevalence of ‘La Niña’, keeping all other variables in the model constant.
The research revealed complex relationships between climatic and economic variables and the yield of sucrose in cane (Sacapo). The mean temperature (Taver) showed a direct relationship with Sacapo, with a 1% increase in Taver increasing Sacapo by 4.64%. This contrasts with Chandio et al. (2020) findings in China, where there was a negative effect of temperature on agricultural production in the long term. Precipitation showed a positive impact, with a 1% increase causing Sacapo to grow by 0.17%. This result differs from what was reported in Chandio et al. (2020) in China, but is consistent with Habib et al. (2022), who found a positive impact of precipitation on agricultural production .
Solar radiation (SRad) presented a negative effect, with a 1% increase decreasing Sacapo by 2.06%. This could be explained by the nonlinear relationship between solar radiation and sucrose accumulation in sugarcane, a C4 plant with high photosynthetic efficiency, as mentioned in Cardozo and Sentelhas (2013); Arnell et al. (2019) . The thermal oscillation (Tosc) showed a positive impact, with a 1% variation causing Sacapo to increase by 1.01%. This is in line with studies in Guatemala that associate greater thermal amplitudes with greater accumulation of sugar in sugarcane, as noted in Castro et al. (2010). A 1% increase in CO2 concentration increased Sacapo by 0.78%.
The FAO cane sugar price index (CSPI) was statistically significant (t = 2.14) as an explanatory variable for Sacapo. The real wholesale price of standard sugar (RWPri) was used as a proxy for the liquidation price of sugarcane (RLPri), as the latter was not statistically significant. This may be because the price of sugarcane is controlled by the government via the 2005 Sugarcane Law, which distorts the price and affects the allocation of resources and the producer’s decisions.
The study reveals that climate change negatively influences sugarcane production and yield, affecting Mexico’s food security. A simultaneous equation model was adopted to quantify the effect of climate variables and prices through elasticities. The main findings include the following: thermal oscillation has a significant positive effect on sucrose yield, solar radiation shows a negative impact on yield, and international sugar prices play a relevant role in cane yield. This approach integrates economic analysis with climatic variables, providing a more comprehensive assessment than previous studies focused primarily on biophysical aspects. The proposed model offers a more comprehensive understanding of the interrelationship between climatic and economic factors in sugarcane production.
Casellas, E.; Bergez, J. E.; Belhouchette, H.; Therond, O.; Adam, M.; Metay, A. and Wery, J. 2009. A methodology for the evaluation and improvement of a generic biophysical soil-plant-atmosphere crop model based on 'mini-application'. Ed. Integrated assessment of agriculture and sustainable development. Setting the Agenda for Science and Polic. Egmond and Zee, The Netherlands. 198-199 pp. https://hal.inrae.fr/hal-02755379v1/file/41721-20110225053142917-2.pdf.
Cheavegatti, G. A.; Couto-Abreu, H. M.; Arruda, P.; Bespalhok, F. J. C.; Lee, B. W.; Creste, S.; Ciero, L.; Aparecido, F. J. and Vargas, O. F. A. 2011. Sugarcane ( Saccharum X officinarum ): a reference study for the regulation of genetically modified cultivars in Brazil. Tropical Plant Biology. 4(1):62-89. Doi: 10.1007/s12042-011-9068-3.
CONADESUCA. 2010. Comité Nacional para el Desarrollo Sustentable de la Caña de Azúcar. Metodología del precio nacional de azúcar estándar al mayoreo. Definiciones y procedimiento de cálculo del precio de referencia del kilogramo de azúcar base estándar. 1-4 pp. http://conadesuca.gob.mx/politica%20comercial/metodolog%c3%8da%20del%20precio%20nacional%20de%20az%c3%9acar%20est%c3%81ndar%20al%20mayoreo.pdf.
Everinghama, Y. L.; Muchowb, R. C.; Stonec, R. C.; Inman-Bambera, N. G.; Singelsd, A. and Bezuidenhout, C. N. 2002. Enhanced risk management and decision-making capability across the sugarcane industry value chain based on seasonal climate forecasts. Agricultural Systems. 74(2) 459-477. Doi: 10.1016/s0308-521x(02)00050-1.
Habib-ur-Rahman, M.; Ahmad, A.; Raza, A.; Hasnain, M. U.; Alharby, H. F.; Alzahrani, Y. M.; Bamagoos, A. A.; Hakeem, K. R.; Ahmad, S.; Nasim, W.; Ali, S.; Mansour, F. and El Sabagh, A. 2022. Impact of climate change on agricultural production. Issues, challenges, and opportunities in Asia. Frontiers in Plant Science. 13(1):01-22. Doi: https://doi.org/10.3389/fpls.2022.925548.