RESEACH WORK

Initial evaluation of integrated systems for food and energy production in Cuba

F. R. Funes-Monzote¹, G. J. Mart¹, J. Suárez¹, D. Blanco¹, F. Reyes¹, L. Cepero¹, J. L. Rivero², E. Rodríguez³, Valentina Savran⁴, Yadiris del Valle⁵, Marlenis Cala⁵, María del C. Vigil⁵, J. A. Sotolongo⁵, S. Boillat⁶ y J. E. Sánchez⁷
1 Estación Experimental de Pastos y Forrajes «Indio Hatuey», Central España Republicana CP 44280, Matanzas, Cuba.
E-mail: mgahonam@enet.cu
2 Estación Experimental de Pastos Las Tunas, Instituto de Investigaciones de Pastos y Forrajes, Cuba
3 Estación Experimental de Pastos Sancti Spíritus, Instituto de Investigaciones de Pastos y Forrajes, Cuba
4 Dirección Provincial de Planificación Física, Sancti Spíritus, Cuba
5 Centro de Aplicaciones Tecnológicas para el Desarrollo Sostenible (CATEDES), Cuba
6 Agencia de Cooperación Suiza para la Cooperación y el Desarrollo (COSUDE), Cuba
7 Instituto de Cibernética, Matemática y Física (ICIMAF), Cuba

ABSTRACT

The objective of the study was to define a typology of integrated food and energy production systems with agroecological approach, in Cuba. The results are based on a preliminary evaluation of the performance of integrated farms belonging to the international project Biomas-Cuba. A methodology is described which can be helpful as basis for future evaluations of the diversity-efficiency-energy-productivity relationship, in the search for typologies that can characterize, as accurately as possible, such integrated systems. The methodology allowed quantifying the performance of diversity, energy efficiency and productivity of 25 production systems in different conversion stages and distributed in the western, central and eastern regions of Cuba. The evaluated indicators (previously validated in Cuba for this type of study) and the use of empirical calculation and multivariate statistical methods allowed identifying and characterizing three main types of integrated systems, which were termed BIOMAS-1A, BIOMAS-1B and BIOMAS-1C. The results of the indicators, as well as the analysis of similarities and differences among the different types constitute elements to be taken into consideration for further studies on integrated food and energy systems with a higher number of farms.

Key words: Bioenergy, diversification, food production.

INTRODUCTION

The increase and instability of prices of fossil fuels, together with the need to reduce greenhouse gas emissions, has motivated in recent years the search for energy alternatives for agricultural development (Pimentel and Pimentel, 2008; Bogdanski et al., 2010). Especially, the use of renewable energy from biomass, including biofuels, attempts to solve environmental and socioeconomic problems of current food systems. These alternatives aim at utilizing the natural potential, insufficiently explored, of agroecosystems in the uptake and transformation of ecological energy sources by sustainable ways. For that purpose it would be necessary, in the future, to establish new approaches for the use of renewable energy sources from biomass, which allow increasing food production, preserving the environment and promoting social inclusion. In this scenario, the design of multifunctional agricultural systems, which are more resilient and promote energy and technology sovereignty to attain food sovereignty, is a priority; this combination of factors, according to Altieri (2009), makes up the three sovereignties of agroecology.

Plants, as photoautotroph organisms, can make use of only 1% of the solar energy that has incidence on the Earth surface (Pimentel and Pîmentel, 2008). A special case is C4 plants, such as corn (Zea mays), sugarcane (Saccharum officinarum) and sorghum (Sorghum bicolor), among others, which have higher photosynthetic efficiency. They include about 7 600 species (around 3% of the total known species). Particularly, the Poaceae family encloses 61% of C4 species (Zhu et al., 2008). They are capable of capturing up to 5% solar energy and fixing higher quantity of CO2 and turning it into longer-carbonated-chains organic compounds. Thus, they have potential to produce large energy quantities per cultivated area unit in a given time. Likewise, other C3 plants such as Jatropha curcas, Moringa oleifera, soybean (Glycine max) and sunflower (Helianthus annus), among oil plants, are capable of producing high-energy products, such as food or feed and/or fuel.

Animals, as heterotrophic organisms, depend on plants to survive; thus, animal production systems are intrinsically less efficient than crop production ones, in obtaining energy for human feeding (Pimentel and Pimentel, 2008). Nevertheless, animals play a key role in the sustainable management of human resources by closing ecological cycles which guarantee a better use of the nutrients and energy circulating in the system (Schiere et al., 2002). The enhancement of livestock production/agriculture integration mechanisms can provide valuable opportunities to facilitate adaptation to the climate change, the increase of productivity and the reduction of energy costs of food production, among other socioeconomic benefits (Bogdanski et al., 2010).

The objective of this study was to define a typology of integrated food and energy production systems with agroecological approach, in Cuba, and the previous studies conducted by Funes-Monzote et al. (2009) about the relation among diversity, productivity and energy efficiency of agroecological production were taken as analytical basis. The results are based on a preliminary evaluation of the performance of integrated farms belonging to the International project Biomas-Cuba. This project has been executed by the Experimental Station «Indio Hatuey» and a group of Cuban institutions between 2008 and 2011. Its objective has been to prove and communicate, through pilot experiences, what local technological alternatives for energy generation from biomass are economically, socially and environmentally effective to improve the living conditions of women and men in rural zones of the country.

Still in its initial stage, the study documented in this paper comprises an analysis of the results of 2009.

MATERIALS AND METHODS

A group of 25 productive systems (farms) were monitored among the 53 that participate in the project Biomas-Cuba and about which there is enough information for their characterization and analysis. The farms are located in the provinces: Matanzas (7), Sancti Spiritus (7), Las Tunas (6) and Guantánamo (5) (table 1). They vary regarding the form of cooperative organization, size and design.

Heterogeneity (among farms) and the different diversity levels of crop, animal and forestry species (in the farms) characterized the sampled productive systems. Each farm represents a special case which is not comparable to the others for its production purposes, market, relations, management characteristics, etc. In such sense, the limits and surface (area) of the system, the subsystems, main interactions, as well as the inputs and outputs were described in detail, according to the system analysis and evaluation methods proposed by Spedding (1988) and Checkland (1999).

The farmers (and their farms) participating in the study were chosen as they are good innovators in agroecological practices and are sensitized with the search for more sustainable alternatives for livestock production. Permanent communication among them, the researchers and technicians who participate in the project allowed the constant interaction, trust and joint work, which guaranteed the design and participatory implementation of the technological innovations and the permanent information exchange during the monitoring. The data collection for a one-year period (2009) was in charge of the farmers and technicians of the local operational committees (LOC) of the project Biomas-Cuba and the members of the general management, in systematic visits to the sites.

The farms were characterized in detail to learn as much as possible about their structure and functioning in this initial stage of the study, which will comprise six years (2009-2014), until the end of the second stage of the project Biomas-Cuba.

For the information collection different elements were used of participatory research approaches: fast rural diagnosis, functional and interactive investigation methods and participatory rural diagnosis, among others (McCracken et al., 1988; Bellon, 2001). Afterwards, in order to simplify the information and allow global appreciation of the available natural and physical resources in each farm, bioresource and infrastructure diagrams were elaborated, adapted from Lightfoot et al. (1998). They served as reference for the analysis of critical points at farm level (McCracken et al., 1988). The diagrams, created together with the farmers, covered the system, subsystem levels and their biophysical components. There was information about field size, infrastructure of the agricultural system and its limits, agrodiversity components and production levels. All the information compiled in the diagrams attempted to improve communication among the researchers and the other shareholders involved in the study. The characterization of the agricultural systems was based on the information obtained during the participatory diagnosis, including workshops, field days, scenario construction and bioresource and infrastructure diagrams of the farm. Agroecological, economic and social aspects were also taken into consideration, in order to achieve an adequate system analysis (Altieri, 1995; Checkland, 1999).

Evaluation of the indicators

The evaluated indicators were: 1) species richness (Eq. 1); 2) production diversity (Eq. 2); 3) quantity of persons fed by the system in energy (Eq. 3); 4) quantity of persons fed by the system in protein (Eq. 4); 5) land use index (Eq. 5); 6) energy balance (Eq. 6) and 7) energy cost of protein production (Eq. 7), according to the methodology proposed by Funes-Monzote (2009).

Species richness (MI). The richness of cultivated species of the agroecosystem was evaluated through the Margalef Index (MI) (Magurran, 1988). For the calculation of this indicator crop, tree and domestic animal species were included.

Where: S = total number of species; N = total number of individuals of all species (including animal, crop, fruit and forestry species).

Production diversity (HS). Production diversity was also calculated through the Shannon Index (Magurran, 1988), which includes the total production of each agricultural or livestock product and total production of the system.

Where: S = number of products; pi = production of each product; P = total production.

Quantity of persons fed by the system in energy (Pe) and protein (Pp).Indicators related to the system productivity were also evaluated, such as the quantity of energy (GJ/ha/year) and protein (kg/ha/year) produced and, in correspondence, the quantity of persons the system could support according to the average demand of such nutrients, of one person per year. The energy and protein contents of animal and plant products for the calculations were taken from Gebhardt et al. (2007). The energy equivalences used to calculate the expenses in direct and indirect inputs were those reported by García-Trujillo (1996). The energy and protein intake values per day recommended for the Cuban population were the ones described by Porrata et al. (1996).

Quantity of persons fed by the system (energy):

Where: S = number of products; mi = production of each product (kg); ri = percentage of edible product weight; ei = energy content of each product (MJ); A = area of the farm (ha); Re = requirement of one person (MJ/year).

Quantity of persons fed by the system (protein):

Where S = number of products; mi = production of each product (kg); ri = percentage of edible product weight; pi = protein content of each product (g/100 g); A = area of the farm (ha); Rp = requirement of one person (kg/year).

Land use index (LUI). The land use index (LUI) was also evaluated, combined with the analysis of the polycrops used in the farm, using the following calculation method:

Where S = number of products; Pi = crop (i) yield in polycrop; Mi = crop (i) yield in monocrop.

Energy balance (EB). An annual energy balance was made, taking into consideration the energy cost implied by producing the food energy.

Where S = number of products; m = production of each product (kg); e = energy content of each product (MJ/kg); T = number of productive inputs; I = quantity of productive inputs (kg); f = energy required for the input production (MJ/kg).

Energy cost of protein production (ECP). The energy cost of the protein production (ECP) in the system was evaluated through the following formula:

Where T = number of productive inputs; I = quantity of productive inputs (kg); f = energy required for the input production (MJ/kg); S = number of products; m = production of each product (kg); Pi = protein content of each product (%).

Analysis of the results

The farm was the experimental unit for the analysis. The best performance obtained for each indicator among all farms was pondered. The calculated values were transformed to a 1-10 scale to obtain a more normal distribution. If the indicator is intended to be maximized (e.g. Pp), the indicator value was expressed as percentage of the maximum value (% = Value/Max x 100). If the indicator is intended to be minimized (e.g. ECP), the indicator value was expressed as the inverse of the minimum value percentage

(% = 1(Value/Min) x 100). The indicators of biodiversity (MI+H), productivity (Pe+Pp) and efficiency (LUI+EB+ECP) were added to obtain the respective values of Biodiversity Index (DIV), Productivity Index (PROD) and Energy Efficiency Index (EE), respectively. Then, the best performance of each index among all farms was pondered. The values of the three indexes obtained for each farm were added, transformed in a 1-100 scale and the resulting value was divided by the sum of the best value of each of the three indexes to obtain the value of the Diversity-Productivity-Efficiency Index (DPE). A ranking was elaborated of the farms, regarding the value of the DPE index, which allowed empirically defining the farm types with regards to their performance.

The validity of the results was tested through the statistical analysis of principal components (PCA), in order to identify how the selected indicators accounted for the performance of the evaluated farms. The null hypothesis (Ho) corresponded to not finding any correlation significantly different from 0 among the evaluated variables, and the alternative hypothesis (Ha) meant that at least one of the correlations among the variables was significantly different from 0. A discriminant analysis (DA) was made for the formation of groups that represented the farm types with different characteristics. The statistical pack XLSTAT version 2008.6.07 (XLSTAT, 2008) was used. The PCA was made from the matrix of correlations. For the graphic presentation of the results the biplot was used. The individuals' points were obtained from applying the transformation proposed by the principal component analysis to the original values of the observations. A supplementary datum built with the best values reached for every indicator was added to the results of the PCA.

RESULTS AND DISCUSSION

Table 2 shows the performance of the seven indicators in the 25 selected farms. In general, high heterogeneity was observed regarding the indicators of diversity, productivity and efficiency, which prevented making a lineal analysis of the results. The farms with high species richness (MI>8) or high production diversity (H>2) did not necessarily have high productivity values. Only farms 7 and 8 had high species richness and, in turn, high energy and protein productivity in terms of persons that were fed per hectare. Yet, the productivity levels of farm 8 were highly motivated by the high import of external inputs, which caused an unfavorable performance of the energy efficiency indicators. On the other hand, farm 12, with high energy efficiency and the lowest protein production costs, achieved low productivity.

This reinforces the notion that the diversification of livestock production systems in itself is not a factor that determines a productivity increase, but rather the design of functional biodiversity in terms of the use of such resources as nutrients, water and energy, to achieve a green agriculture (Koohafkan et al., 2011).

In spite of having similar characteristics in terms of diversified agroecological design, three farms of Matanzas province had contrasting values of the productivity and efficiency indicators (table 2). The medium-scale Cayo Piedra farm (diversified agricultural crop farm), achieved the highest productivity and efficiency values. Plácido (mainly destined to livestock production, but also diversified with crop, fruit trees and ornamental plants) showed moderate values. La Arboleda (diversified with livestock production, agriculture, but mostly dedicated to fruit production), with higher species richness, had lower productivity and efficiency levels.

A higher diversity did not necessarily result in higher productivity and efficiency, although it was an important component. La Arboleda had a lowest productivity in terms of energy and protein quantity per hectare, as it was mainly dedicated to the production of fruits, which are low in these nutrients and contribute little to the energy balance (table 2). The low energy efficiency of this farm is also ascribed to the high intensity of labor and to the fact that it is dedicated to other activities such as craftsmanship, which increases the family income. The diversity indicators are closely related not only to the number of individuals, but also to the equity among them, from the presence of the species and its relative abundance (Magurran, 1998). That is why La Arboleda, although having almost three times as much species richness as Cayo Piedra and Jesús María, reached production diversity similar to them (table 2).

La Caoba farm achieved the lowest energy cost of protein production (ECP) and the second highest value of energy efficiency (EB); however, its productivity was low, which would not be desirable in principle (table 2). The objective of agricultural biointensive systems, which try to maximize the use of renewable energy sources to attain a productivity increase, should achieve high productivity equivalent to high efficiency in the use of energy. The system integration level is an important factor to fulfill this objective (Funes-Monzote et al., 2009); nevertheless, moderate or low productivity can also be the result of low or moderate intensity management systems with conservationist objectives. Conversely, La Caoba farm, in particular, has an important forestry component which, logically, was not reflected in terms of productivity with regards to the quantity of persons fed by the system.

The least productive farms and the ones with less production diversity were the ones from Guantánamo, which were establishing the integrated systems of J. curcas with crops and two from Las Tunas province (farms 16 and 17), which corresponded to dairy units with low management level. These contrasting results, calculated through the productivity, diversity and efficiency indexes (table 3) required a more integrating analysis to arrive at conclusions about their later development. Empirically, no farm (table 2) reached the best values in more than one indicator (values highlighted in boldface), which led to identify the best average

performance of each, taking into consideration the grouping of the biodiversity, productivity and efficiency indexes. Thus, the farms that had a better performance of these indexes and a general index named DPE (diversity, productivity and efficiency) were preselected, which allowed grouping the farms into typologies, with the average performance of the group and a farm prototype that better defined the characteristics of each farm. For example, Cayo Piedra was the prototype farm that could best define the characteristics of the type BIOMAS-1A with the available data, and thus successively for the prototypes Plácido or La Quinta (table 3).

When calculating and ranking the farms according to their performance, the smallest ones (except Cayo Piedra) were found to achieve the best diversity, productivity and efficiency indexes and, thus, the highest DPE values, although this is not always true when the stage of the conversion process is also analyzed (table 3). This confirms the results obtained by Funes-Monzote et al. (2009), and in turn opens new questions about which would be the best strategies to implement highly diverse livestock production systems at larger scales.

Principal component analysis

The first principal component (F1) accounted for 49,5% of the variability, while the second, F2, 20,4%, resulting in 69,9%, which was considered sufficient to explain the performance of the evaluated variables (fig. 1). The PCA showed the high heterogeneity of the farms and their contrasting differences regarding the diversity, productivity and efficiency indicators. This result confirmed the need to group farm types in order to arrive at an approximation of the agroecological potentials and the satisfaction level of that potential in the studied sample. This also led to incorporate, as part of the analysis, a target agroecological system (No. 26), formed by the best value of each indicator reached among all the farms (fig. 1).

All the indicators (except ECP) explained the variability in the performance of the farms, which indicated that they were important in their differentiation. Such contrasting results of the calculated values for the ECP, mainly due to the technological heterogeneity (use of energy inputs) are seemingly the result of such

performance. Among the variables that accounted for the differences of the farms with regards to F1, LUI and HS stood out, which leads to state that, at least in the evaluated sample, there was a strong link among the diversity, productivity and efficiency indicators (fig. 1).

Discriminant analysis

The reclassification of the farms by using a discriminant analysis proved that they were all adequately classified empirically, with the exception of number 25 (Villa Josefa), which had been classified as BIOMAS-1B and instead it belonged to BIOMAS-1C, for 96% of effectiveness (fig. 2). A total of 92,6% of the variability was explained through F1, which was sufficient to express the variability of all the evaluated indicators. This indicated that the formed types responded to common characteristics among the groups according to the indicators and, besides, that it made sense to grant high relevance to the interaction among diversity, productivity and efficiency, known as DPE index.

Characterization of the identified farm types

BIOMAS-1A (strong food and energy integration). The farms comprised in this type had strong correlations among diversity of cultivated plant and/or animal species, high energy efficiency and high productivity, in terms of amount of food produced per area unit dedicated to crops or animal production. They are generally small-scale productive systems ( < 15 ha) and have vast traditional knowledge about animal rearing and local crops. They have high production stability, autonomy in the use of resources and are resilient before the effect of external factors. With little investment their energy food production potential could be increased, by incorporating new technologies for a more efficient use of the available biomass.

BIOMAS-1B (in the process of increasing food and energy integration). This type is characterized by having considerable advances in the diversification of the productive system. In many cases they achieve high energy efficiency, but with low productivity and viceversa. Although they have knowledge in the management of natural resources, they still require higher efforts in the integrated design of the productive system. For such reason the input availability cannot be conjugated to the established functional diversity and the increase of the efficiency and productivity indicators. With financial support in technology and some design changes, they can remarkably improve their performance and be considered as systems BIOMAS-1A.

BIOMAS-1C (initial stages of the food-energy integration). The farms included in this type are starting the integration process of the food system and show a strong energy unbalance. It may happen that a farm belonging to this group has considerable energy sources of industrial origin (diesel, machinery, irrigation, chemical products) or abundant energy of biological origin (manure, biomass, labor), but inefficient use is made of these resources. Those that are starting an integration process or in the establishment stage of oil crops or biodigestor installation, but in turn have low diversity and productivity, are included in this farm type. In general, to achieve integration in these systems a strong training component and higher financial support will be required, although their poor performance can also be given by the waste of wrongly used natural and financial resources. Moving up to type BIOMAS-1B will require conscious work, between two and three years.

CONCLUSIONS

A clear differentiation was found among the three types with regards to the DPE index which explains the interactions among the diversity, productivity and efficiency indicators. The implementation of technologies for the use of renewable energies from the biomass, such as biogas, biofuels and gasification, as well as other sources (windmills, hydraulic rams, solar panels, etc.) is expected to be evaluated at long term and to contribute to the increase of productive levels through a more efficient use of the available energy and the design of biodiverse farms. These results could serve as reference for further studies in the future. For the next years the following is recommended:

• Increasing synergies with other projects that allow increasing the sample size (including conventional specialized productive systems).

• Validating the evaluated typology through the use of multivariate statistical methods which incorporate the diversity, productivity and energy efficiency indicators, linked to biophysical, technological, management and socioeconomic factors.

• Conducting a temporary evaluation, during six years (2009-2014), of the food and energy production systems.

• Evaluating the integration potential with regards to the possible food and energy combinations to be applied to the prevailing agricultural systems in the country, aiming at making recommendations which could be extrapolated to systems at all scales (cooperative, municipality) and complexity levels.

ACKNOWLEDGEMENTS

The authors thank the Swiss Development Cooperation Agency (SDC), the Cuban institutions that participate in project Biomas-Cuba and the farmers involved in the project, for their valuable contributions to this study.