Feeding is a crucial point in the development of livestock and in each stage of its growth or production. An adequate balance among the amount of nutrients will result in high levels of production without affecting the animal's body condition. Therefore, the formulation of balanced foods becomes a critical point to avoid metabolic disorders, growth delays or deterioration in the reproductive part. When formulating, the nutritional value of the inputs must be taken into account, since the nutritional content of the ration will be given by the composition of each of the ingredients that make it up (Fernández Curi, 2013FERNÁNDEZ, C.E.: Guía Técnica - Formulación de alimentos balanceados y mejoramiento genético en ganado lechero, Per de 2013.).
Meat production in Cuba is below its historical levels and it is insufficient to satisfy the needs of the population, as stated by the National Statistics Office of Cuba (ONEI). As a strategy, the country has begun to raise buffalo cattle since 2012. These are distributed throughout the island and have specific characteristics in terms of feeding and development. This complies with the provisions of Guideline 167 of the Sixth Congress of the Communist Party of Cuba (2017)PCC- CUBA: “Actualización de los Lineamientos de la Política Económica y Social del Partido y la Revolución para el período 2016-2021”, En: Actas del VII Congreso del Partido Comunista de Cuba. La Habana: Oficina Central del Estado, Ed. Editora Política, La Habana, Cuba, p. 30, 2017., since it emphasizes the use of pastures and fodder for feeding cattle.
The Institute of Animal Science (ICA) has carried out various works related to the current situation of the country in the livestock system; and with the aim of achieving better results, experiments have been conducted where the food balance of the animals is analyzed (Borroto, 2015BORROTO, G.O.: “La fisiología digestiva del rumiante, objeto de investigación en el Instituto de Ciencia Animal durante cincuenta años”, Cuban Journal of Agricultural Science, 49(2): 179-188, 2015, ISSN: 2079-3480.; Delgado et al., 2011DELGADO, D.; FRANZOLIN, R.; MAZZA, P.H.: “Fermentación y cinética ruminal en búfalos alimentados con heno de bermuda cruzada (Cynodon dactylon (L.) Pears) y concentrado con soya integral o extrusada”, Revista Cubana de Ciencia Agrícola, 45(1): 39-43, 2011, ISSN: 0034-7485.). The Ruminants Department of the research institute has carried out experiments related to the nutritional development of the buffalo species. It is under these premises and in the current context, that the objective of the present work is the development of a computerized system to support decision-making, based on historical data, research and mathematical models that provide the most beneficial diet combinations for the development of the buffalo species.
In order to carry out a balance of rations, it must be known that a ration is the amount of food given to an animal. Rations must be well balanced to cover the requirements of the species according to its state of development. The balance of rations is done through calculations and mathematical procedures applied to achieve a balance between the nutritional requirements of the animals and the nutrient contributions of the food that make up the ration. Before balancing a ration, the following elements should be considered (Instituto Nacional Tecnológico, 2016):
The amount of feed available.
The composition of the feed.
The requirements of the animals for which the ration is formulated.
The method by which the ration is to be offered.
There are several methods available for the formulation of the rations according to Montes et al. (2017)MONTES, A.A.; CASTELLANOS, G.M.; MARTÍNEZ, G.G.; CHÁVEZ, E.T.: “Uso de la programación matemática para la formulación de raciones de animales”, En: Congreso Interdisciplinario de Ingeniería, vol. Revisión bibliográfica, p. 154, 2017., among them are the Score method, the Trial and Error method, the Pearson Square method, the Pearson Double Square model and the Dual Simplex method, the latter being a linear programming technique. Each of them was analyzed and, after observing results obtained in several publications such as those made by Gámez (2016)GÁMEZ, M.N.A.: Modelo De Optimización Para La Gestión De La Información De Producción Porcina En Pie De La Finca La Esperanza De La Vereda Tierra Negra Fusagasugá Cundinamarca, 2016., Ibáñez and Muñoz (2016)IBÁÑEZ, A.A.; MUÑOZ, J.S.: Desarrollo de un Programa Informático para la Formulación de Piensos Compuestos, 2016., Alvarado (2018)ALVARADO, M.P.A.: Elaboración de raciones como suplemento alimenticio del ganado bovino, empleando residuos de cosecha de maíz, maní y arroz, JIPIJAPA-UNESUM, 2018., it was decided to implement the Dual Simplex method and the Double Square Method.
Pearson's Double Square model allows calculating a diet taking into account a mixture from a food composition, of which one or half of the mixture must be energetic and the rest protein (Wagner and Stanton, 2012WAGNER, J.; STANTON, T.L.: “Formulating rations with the Pearson square”, Fact sheet (Colorado State University. Extension). Livestock series ; no. 1.618, 2012.). Two to n foods can be used, which can be grasses, forages or supplements, fulfilling the characteristics of this method, that one of the foods contains more proportion than the rest. In the case of assigning more feeds to the mix, half of them must have a higher proportion of nutrient than the rest, with the aim of balancing the feeds. The essence of this method is to calculate the proportion of foods that will achieve the desired content of the nutrient to be balanced. According to Zalapa (2010)ZALAPA, A.: Realidades del Cuadrado de Pearson simple, compuesto y el agregado, Sitio argentino de Producción Animal, 2010., the procedure of this model consists in making a box where in the upper left side, the name of the product to be balanced and its desired nutrient content is marked. In the lower end, the name of another desired product to be combined and its respective nutrient value are put. In the middle of the box, the desired value of the nutrient is shown, which can be dry matter, metabolizing energy, crude protein, crude fiber, different minerals or vitamins. The rule or restriction is that there must be a higher and a lower value. In the following example, two foods are analyzed in terms of crude protein.
Step 1: The selected foods are added with their corresponding nutritional contributions and the value for which the mixture is to be made. One of the foods must have a higher proportion of nutrient than the other food.
Step 2: The value of the nutrient in the ingredient is subtracted and marked as positive.
Step 3: Then the sum is considered as part of 100% and both fractions by logic, the part that they represent of that 100%.
In this example, pollinaza contains 25% of raw protein, ground corn and stubble contains 7.8% of raw protein and for calf fattening11.1% of raw protein is required. By adding the values in cross form, 13.9 is obtained, therefore, no longer correspond to the pollinaza, but to its opposite that is corncob with stubble. With 3.3 parts (kg, pounds or the unit used) of pollinaza and 13.9 parts of corncob, the ration will be 11.1% of raw protein. Generally, the values are expressed in percentage. In this case 3.3 parts plus 13.9 parts add up to 17.2 parts, this would be 100%, obtaining 19.186% of pollinaza, 80.814% of ground corncob and stubble.
On the other hand, the Dual Simplex method is used to give solutions to linear programming problems since it is used to find an adequate solution in maximization and minimization problems. Among the advantages offered by this model is that it is useful for solving large and complex problems, serves to maximize profits and decrease production costs (STEEL, 2017). This method forms the basis of linear programming as it facilitates decision-making in complex cases, as it solves situations where the number of variables exceeds the number of equations. It is used to solve problems where it is necessary to optimize resources in the best possible way and it is used to solve linear problems in which three or more variables are involved. This method performs an interactive process. In a general way, this model performs the following steps: determining the objective function, establishing the pivot column and pivot row and performing the Gauss-Jordan methodology. The pivot column are the dependent values of the inequalities and the value that is selected is the most negative. The pivot row is obtained from the division of the value of the objective function that corresponds with the value of the row; the pivot value is the less negative of the previous results (Cormen et al., 2009CORMEN, T.H.; LEISERSON, C.E.; RIVEST, R.L.; STEIN, C.: Introduction to algorithms, Ed. MIT press, 2009, ISBN: 0-262-53305-7.)
The purpose of using this procedure in the formulation of rations is to determine the amount of feed that should be mixed to meet the nutritional requirements of the animal, taking into account the selected values in nutritional requirements. In order to carry out this mathematical model, the information stored in the database will be taken as dependent values of the equations; in this case, it would be the tables of the different types of feeds (grass, supplement and forage). The independent values will be obtained from the nutritional requirements of the animal from the selected nutrients.
For the design of the system, the Unified Software Process methodology was used (Martínez and Martínez, 2014) or Rational Unified Process also known by its acronym RUP, which uses the Unified Modeling Language (UML) as a notation language (Debrauwer and Van der Heyde, 2016DEBRAUWER, L.; VAN DER HEYDE, F.: UML 2.5: iniciación, ejemplos y ejercicios corregidos, Ed. Ediciones ENI, 2016, ISBN: 2-409-00372-9.). For the implementation of the web application, the Java programming language was used (Arnold et al., 2005), using NetBeans as the development environment (Gimeno and González, 2011). In addition, the Java Server Faces (JSF) framework was used, based on the Model-View-Controller architecture pattern (Müller, 2018). In addition, Prime Faces, a component library for JSF, was used, which has a set of enriched components that facilitate the creation of web applications. For the web design, the authors also worked with the Boostrap framewok that offers the possibility of creating responsive web sites (adaptive web design), through the integration of CSS and JQuery libraries (Spurlock, 2013SPURLOCK, J.: Bootstrap: Responsive Web Development, Ed. O’Reilly Media, Inc., 2013, ISBN: 1-4493-4460-7.).
On the other hand, the database management system PostgreSQL (Douglas and Douglas, 2003DOUGLAS, K.; DOUGLAS, S.: PostgreSQL: a comprehensive guide to building, programming, and administering PostgresSQL databases, Ed. SAMS publishing, 2003, ISBN: 0-7357-1257-3.) was used for the storage and persistence of the information. Spring Security was used to manage the security of the system, providing an authentication system through which users can authenticate themselves and access multiple functions through a single entry point (Mularien, 2010MULARIEN, P.: Spring Security 3, Ed. Packt Publishing Birmingham,, England, 2010, ISBN: 1-84719-974-7.).
Once all the analysis were done and the mathematical models to be used were defined, the computer system was developed. In the system, all the information regarding users, pastures, fodder, food supplements and nutritional requirements of the animal, among other information, is managed. In addition, the ration balance is calculated based on live weight and evolutionary state, the feed mixtures are elaborated and the yield of a pasture is estimated according to the hectares of land.
Figure 1 shows the initial interface, which has a main menu from which the different functionalities of the system can be accessed, as it is allowed.
To access the system each user must authenticate himself; and depending on his level of accessibility or role, he will be able to access the system's functionalities. The roles that make up the application are administrator, researcher and basic user. The system administrator will be in charge of the user management, the researcher is in charge of managing the nutritional requirements of the buffalo according to the live weight and its evolutionary state, the nutritional composition of the pastures, forages and supplements and the seasons of the year. The basic user can make the formulation of rations, as shown in Figure 2, and estimate the yield of a pasture according to the hectares of land.
Figure 3 shows the nutritional requirements of the animal (live weight, evolutionary state, vitamins, minerals, etc.) which forms the basis for all subsequent feed calculations.
For the elaboration of the mixture, the nutrient (Figure 4) and the percentage of it that is desired to be obtained are identified. The existing grasses, forages and supplements are selected (Figure 5) and the formulation is continued. The result is a pie chart with the recommended quantities for each of the foods (Figure 6).
From the study of bibliography consulted, it was possible to obtain a greater knowledge of the subject in question, identifying the main aspects involved in the feeding of the animal. In addition, it facilitated the understanding of the mathematical models Double Square of Pearson and Simplex Dual model, selected for the preparation of food rations. The system was implemented, thus favoring the formulation of rations, which will be a great help to specialists in this field. For the evaluation of the system, several tests were carried out using real data. These showed that the system is very useful and practical for specialists.