Servicios personalizados
texto en 
Inglés (pdf)
Articulo en XML
Referencias del artículo
Enviar articulo por email
Citado por SciELO 
Similares en
SciELO 
Permalink
Milk, due to its nutritional balance, is a very complete essential food for human nutrition (FAO 2018 and FEPALE 2018). Dairy industry focuses its attention on the valorization and innovation of its products, because of the properties of some dairy components for therapeutic purposes (Bauman et al. 2006).
Livestock is responsible for most of the global usage of lands. Grasslands and croplands dedicated to production of cattle feed represent approximately 80% of agricultural land in the world (Friedrich 2014). Forage crops are sown on one third of the cultivated land, while the total area of pastures is equivalent to 26% of the ice-free land area (FAO 2015).
Bovine milk production, based on pastures and forages, may be profitable in tropical areas, if pastures are intensively used and an adequate management of the system is ensured. For this, pastures and forages that adapt the best to the edaphoclimatic conditions must be selected, as well as animals with the best adaptability to the challenges imposed by the tropics and climate change (Davis and Matamoros 2016). The objective of this study was to analyze the main factors that influence on dairy productivity of two dairy farms of Dos Ríos enterprise (UEB, initials in Spanish), of Manguaco agricultural enterprise.
Materials and Methods
This research was carried out in the UEB Dos Ríos, of Managuaco agricultural enterprise, located at 210 56’ 01.2” N and 790 20’ 43.8” W, in Sancti Spíritus municipality, Cuba. Livestock control data (white papers), corresponding to 2015 and 2016, from dairy farms 16 and 17, were collected. Siboney de Cuba (dairy 17) and tropical Holstein (dairy 16) were the main evaluated breeds. DeLaval mechanical milking was used in both farms.
The information was collected each month, for two years, and the following aspects were analyzed:
Total and milking cows
Milk production per milking group, monthly and seasons
Area of each pasture and forage species established at the beginning
Botanical composition was determined at the beginning of the study, according to dry weight range method (Mannetje and Haydock 1963)
Availability was monthly registered, at the entry and exit of the animals, according to the method described by Haydock and Shaw (1975). Cutting height was 10 cm from the ground. Between 80 and 100 observations hectare-1 were taken
Pasture intake was monthly estimated from the difference between the entry and exit availabilities of animals
Food intake in feeders, as well as supplements and complements, was registered every 15 days, weighing food offer without rejection
Number and size of paddocks
The control of reproductive status included the number of fresh cow, inseminated, pregnant and open cows, according to reproduction cards
Lost and affected quarters by mastitis were analyzed every 15 days by sampling the California animals in milking
Quality of milk in cooling tanks was daily sampled, as well as density, presence of mastitis and acidity. Percentage of fat and reductase were analyzed twice a month, according to sampling
The predominant soil in the farms under study is carbonate soft brown (Hernández et al. 2015).
For the analysis, estimates of the potential production curve of each farm were conducted by evaluating the efficiency of milk production, according to days of real lactation per months.
Likewise, forage balances were carried out by year, season and month, according to estimated intakes for each species of grass and forage and used supplements.
For the statistical analysis, the model of principal components, proposed by Torres et al. (2008), was applied.
The database, organized as a matrix, included the information on 21 variables, presented in columns and scenarios or studied months (48 in total, two units for two years), which corresponded to rows. Thus, the premise that the number of variables is lower than scenarios was fulfilled.
For the selection of principal components, the one that was higher than the unit was taken as eigenvalue. Each component was labeled with a name and the variables that best explained its performance were selected, considering preponderance values superior to 0.60.
With the selected principal components, factorial points were calculated, which can be used as an absolute measure of impact or performance (positive or negative) of the variables of the greatest importance in each farm. This allowed to classify the units using cluster analysis, which enables the selection of groups of similar units by using the dissimilarity coefficient.
The statistical package IBM SPSS (2013) was used for processing information.
Results and Discussion
Analyzes of feeding base, in both units (table 1), indicate that 85.6% of the areas dedicated to livestock activity in farm 16, are used for grazing. Out of these, 25.8% is established with Cuba CT-115 clone (Cenchrus purpureus cv. Cuba CT-115) as a biomass bank strategy. This percentage is very similar to that proposed by Martínez et al. (2012) for this technology, which is 30%. Meanwhile, farm 17 only dedicates 60.2% to grazing its livestock areas. Out of them, the areas with Cuba CT-115 clone represent 17.2% of the pasture.
Table 1 Structure of feeding base of dairy farms under study for two years
| Areas | Unit | Dairy farm 16 | Dairy farm 17 |
|---|---|---|---|
| Total livestock area | Hectares | 76.9 | 62.4 |
| Total grazing area | Hectares | 65.8 | 37.9 |
| Bahia grass ( |
Hectares | 48.8 | 31.4 |
| CT 115 ( |
Hectares | 17.0 | 6.5 |
| Percentage of grazing area as biomass bank | % | 25.8 | 17.2 |
| Total paddocks | Number | 27 | 41 |
| Average paddock area | Hectares | 2.44 | 0.92 |
| Total forages | Hectares | 11.1 | 24.5 |
| Sugar cane ( |
Hectares | 6.1 | 15.2 |
| King grass ( |
Hectares | 5 | 9.3 |
| Average cows | Number | 84 | 107 |
| Total stocking rate | Cows/ha | 1.09 | 1.71 |
| Grazing stocking rate | Cows/ha | 1.28 | 2.82 |
The rest of the areas corresponding to farms are dedicated to permanent forages, such as sugar cane (Saccharum officinarum) and king grass (Cenchrus purpureus). Stocking rate and grazing areas of dairy 17 are relatively high for the current use conditions of the farm, while, in 16, they are slightly superior for these conditions (1.71 and 2.82 vs 1.09 and 1.28 LAU ha-1, for units 17 and 16 in global stocking rate and per grazing area, respectively).
Total stocking rate of dairy farm 17 leads to having 39.8% of its livestock areas as permanent forages, which determines higher expenses on dairy production due to the dependence on workers for the cultivation, cutting, carrying and processing of food in the trough. This result differs from that obtained by Martínez et al. (2012) in an experiment, in which the animal obtained food directly from pasture.
Figure 1 shows the results of forage balances per seasons and years in both dairy farms. As a consequence of the above described in the feeding base (table 1), only positive DM balances are presented in rainy season of both years, which were between 6.8 and 11.7 t DM, for dairies 16 and 17, respectively. Meanwhile, in dry season of both years, a deficit between 6.5 and 28.7 t DM was indicated, for dairy farms 16 and 17, respectively.
These aspects are features of milk production systems based on pastures and forages under tropical conditions (Davis and Matamoros 2016). They indicate the need to establish more productive pastures and forages, and better adapted to exploitation conditions, and to strategically use complementary foods in dry season, such as those that are conserved in the form of silage or hay, and the agro-industrial by-products of the locality (Salado 2012 and Calderón et al. 2017).
Results of estimating the relative minimum potential production in both farms (figure 2) demonstrate that farm 16 presented greater potential in 7.61% with respect to dairy 17 (3,154 and 2,931 liters of milk per lactation of 305 days, for 16 and 17, respectively). This may be due to superior feeding conditions (figure 1) and to the genotype used in dairy 16, which is tropical Holstein. This breed shows more dairy potential than mestizo Siboney breed, although it requires greater demands regarding feeding and management (Roca et al. 2013, Coffey et al. 2016 and Vite et al. 2017).
The estimate of potential vs. real production efficiency, according to real lactation days per months, seasons and years, in the analyzed periods did not exceed 85%. This shows the existence of feeding and managing problems in these two herds in the analyzed periods, resulting in difficulties in reproductive aspects (Meikle et al. 2013 and Rojas et al. 2019) and in the compositional quality of milk (Hernández and Ponce 2006). Therefore, the general efficiency of the production system decreases (Senra 2011).
Results of the principal component analysis (PCA) (table 2) indicated that, out of the 21 initially studied variables, 10 were the ones with the highest factor loading in the study. These were grouped into four PCs, labeled according to the factorial loading of the variables that comprise them: PC1) productivity, with variance of 30.75%, PC2) herd, with 21.94%, PC3) reproduction, with 16.39% and PC4) milk quality, with 10.52%. Together, the four PCs explained 79.6% of total variability of the system.
Table 2 Matrix of monthly rotated components
| Principal components | ||||
|---|---|---|---|---|
| Productivity | Herd | Reproduction | Quality | |
| Accumulated total cows (head) | 0.099 | 0.284 | 0.193 | |
| Accumulated milking cows (head) | 0.456 | 0.096 | 0.052 | |
| Birtds (number) | 0.058 | 0.099 | 0.123 | |
| Total production (kg milk) | 0.331 | 0.198 | 0.216 | |
| Lactation (days) | 0.051 | -0.013 | -0.204 | |
| Efficiency percentage (%) | -0.105 | 0.188 | -0.035 | |
| Fresh cow percentage | 0.089 | 0.023 | 0.131 | |
| Pregnant cow percentage | -0.253 | -0.276 | -0.289 | |
| Density (g liter-1) | -0.011 | 0.022 | 0.033 | |
| Forage intake (kg cow-1 day-1) | -0.096 | 0.245 | 0.351 | |
| Total | 3.075 | 2.194 | 1.639 | 1.052 |
| Eigen variance, % | 30.747 | 21.940 | 16.391 | 10.521 |
| Accumulated variance, % | 30.747 | 52.687 | 69.078 | 79.599 |
In the case of productivity (PC1), variables total milk production and explotation efficiency potential presented high positive values of preponderance. Meanwhile, the intake of bulky food showed high negative values. This is due to the fact that the highest forage intake occurred during dry season, because there is low availability of grass. This makes the quality of the base diet to be low, which brings about a decrease of dairy productivity, due to a low nutrient intake (Reyes et al. 2012 and Pineda et al. 2016).
Regarding herd (PC2), all the variables with high values of positive preponderance indicate that with the increment of the number of total animals, there will be more milking cows and also the percentage of pregnant cows increases, which favors milk production.
In the analysis of reproduction (PC3), the two variables with the greatest weight in the preponderance are positive, and are related to each other. As births increase, reproductive indicators improve and percentage of fresh cows also increases, which can directly influence on milk production increases.
Regarding milk quality (PC4), the variable days of lactation presented a negative preponderance value, as it was negatively correlated with milk density. At the beginning of lactation, milk density decreases, as a consequence of the greater volume of milk produced (Hernández and Ponce 2006 and Castillo et al. 2019).
In order to identify the variables with the greatest impact on dairy productivity, these PC analyzes have been used in Cuba with satisfactory results in several livestock enterprises in Mayabeque province (Torres et al. 2008 and Rodríguez et al. 2013) and Villa Clara (Martínez et al. 2012). They have also been applied to the characterization of the factors that influence milk production on cooperative farms (CCS, initials in Spanish) in Ciego de Ávila province (Martínez et al. 2013).
When applying cluster analysis to achieve the grouping of 48 scenarios by similarity, groups I, II, III, IV were formed, composed by 14, 10, 11 and 13 individuals, respectively (figure 3). These four groups respond, in a general sense, to each dairy (16 and 17) in each season of the year (rainy and dry).
These results correspond to milk production systems based on pastures and forages under tropical conditions (Roca-Fernández et al. 2013 and Davis and Matamoros 2016). In this type of system, the influence of climatic conditions, fundamentally of precipitations, in both seasons, allows a marked difference in the availability and quality of basic food, in favor of rainy season. This brings about greater intake of animals (Reyes et al. 2012), greater fermentative activity at ruminal level and superior total VFA production (López et al. 2016 and Restrepo et al. 2016), which leads to an increase of milk production (Roncallo et al. 2012 and Davis and Matamoros 2016).
In the typifications of the four groups (table 3), composed by the ten studied variables, which had the highest preponderance in the analyzed systems, it is demonstrated that milk production (kg month-1) and efficiency with which the dairy potential of each farm (%) was exploited (%), as well as forage intake of cows (kg cow-1 day-1), were the most significant variables among the groups. The rest showed a very similar typification in each group.
Table 3 Typification of variables in the four groups
| Variables | Measures | Dairy farm 16 Dry period | Dairy farm 16 Rainy period | Dairy farm 17 Rainy period | Dairy farm 17 Dry period | ||||
|---|---|---|---|---|---|---|---|---|---|
| Group I (14 individuals) | Group II (10 individuals) | Group III (11 individuals) | Group IV (13 individuals) | ||||||
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | ||
| Total cows | Heads | 2587.8 | 285.0 | 2644.1 | 288.2 | 3163.9 | 224.3 | 3268.5 | 117.6 |
| Milking cows | Heads | 1306.4 | 176.6 | 1467.5 | 232.1 | 1848.2 | 92.6 | 1683.5 | 161.1 |
| Births | Heads | 7.1 | 3.4 | 6.0 | 2.9 | 7.1 | 2.7 | 8.5 | 3.2 |
| Milk production | kg month-1 | 10704.1 | 327.1 | 10663.1 | 116.0 | 16581.0 | 200.3 | 11527.1 | 230.1 |
| Lactation days | days | 134.5 | 18.3 | 165.0 | 11.1 | 137.8 | 12.7 | 141.1 | 15.4 |
| Use efficiency of the potential | % | 71.6 | 16.9 | 71.7 | 9.0 | 83.7 | 7.8 | 65.9 | 10.8 |
| Fresh cows | % | 9.6 | 3.2 | 8.0 | 4.1 | 9.4 | 2.1 | 10.2 | 2.5 |
| Pregnant cows | % | 32.1 | 4.4 | 38.6 | 4.1 | 36.0 | 3.4 | 42.6 | 3.7 |
| Density | g liter-1 | 1032 | 13.5 | 1031 | 12.8 | 1032 | 14.2 | 1032 | 13.6 |
| Forage intake | kg cow-1 day-1 | 21.7 | 5.4 | 17.6 | 5.0 | 15.7 | 5.5 | 29.2 | 5.1 |
Monthly milk production, as an average per season, showed no variation in dairy 16, only 0.38% during dry season. However, in dairy farm 17, it decreased 30.48% in dry season compared to the rainy period. The performance of use efficiency of dairy potential was similar in dairy 16, with similar values in both seasons (71.6 and 71.7% for rainy and dry seasons, respectively). However, dairy 17 showed a marked difference of 17.8% between periods (83.7 and 65.9% for rainy and dry periods, respectively).
Forage intake of cows in dairy 16 during dry season was 23.3%, higher than intake during rainy season. However, in dairy 17, this intake increased by 86.0%, compared to the same periods.
This could be caused by the high proportion of areas established with CT-115 as biomass bank in dairy 16, which favored its efficient use in dry season. Hence, the need for bulky food in animal feeders was less during this period, and stability in feeding was achieved (Martínez et al. 2012). Regarding dairy 17, the high global stocking rate that it showed did not allow the existence of a balance in the system in general, mainly in dry season (Lok et al. 2013).
When relating PC impacts with the formed groups (table 4), in the case of production (PC1), in both dairy farms, in dry season, its expression was negative, with a higher value in dairy 17 (-0.815). Meanwhile, in rainy period, impact values were positive, with a higher expression in dairy 17 (1.259). This performance of PC1 is typical of dairies with complex situations in base feeding, mainly during dry season (Motta et al. 2019).
Table 4 Mean impact factors in each group per principal component (PC)
| PC1 Production | PC2 Herd | PC3 Reproduction | PC4 Milk quality | |||||
|---|---|---|---|---|---|---|---|---|
| Mean | SD | Mean | SD | Mean | SD | Mean | SD | |
| Group 1 Dairy 16Dry season | -0.257 | 0.900 | -1.096 | 0.576 | 0.024 | 1.081 | 0.536 | 0.717 |
| Group 2 Dairy 16 Rainy season | 0.035 | 0.395 | -0.341 | 0.637 | -0.273 | 1.811 | -1.547 | 0.378 |
| Group 3 Dairy 17 Rainy season | 1.259 | 0.360 | 0.556 | 0.3951 | -0.256 | 0.633 | 0.587 | 0.416 |
| Group 4 Dairy 17 Dry season | -0.815 | 0.659 | 1.022 | 0.422 | 0.401 | 0.819 | 0.025 | 0.489 |
Regarding herd (PC2), obtained values were in accordance with the highest number of animals and, therefore, with the biggest global stocking rate and per grazing area, of dairy 17 in both seasons of the year. This makes it difficult, to a great extent, to be able to achieve stability of the productive system and its self-sufficiency (Senra 2011).
Regarding reproduction (PC3), there was a negative impact in both dairies with respect to fresh cows and births, which was very similar in rainy season. This is an undesirable aspect in the exploitation systems based on pastures and forages, since it indicates that, in the studied cases, calvings and the proportion of fresh cows are higher in dry season. At this stage, there is less availability and quality of pasture and forages, which makes it difficult to express the productive potential of dairy cattle at the beginning of lactation, and later affects the poor reproductive performance of the herd (Roja et al. 2019).
Milk quality (PC4) presented a negative value in group 2, which corresponds to rainy period of dairy 16. This aspect is directly related to problems that exist in the correct functioning of the milking equipment. This situation caused an increase of the presence of subclinical and clinical mastitis, affecting milk quality at dairy level. In addition, Holstein animals are more susceptible to these technological effects (Serraino and Giacometti 2014, Zumbado and Romero 2015 and Vite et al. 2017).
Results of this research allowed to identify the variables with preponderance in the performance of milk production system under the study conditions. The analysis of average impact factors per groups and PC made it possible to identify with better certainty which were the difficulties in order to be able to draw up strategies for minimizing, as much as possible, or eliminating them.
