Introduction
The world will have a population of more than nine billion people by the year 2050, which will require 70% more food in the form of meat and milk (Wadhwa and Bakshi, 2013). The greater milk intake in developed countries is a result of increased per capita income, urbanization growth, and changes in eating habits (Guyomard et al., 2013). Thus, it is estimated that animal production must increase to meet the consumer demands, which implies an increase in the price of grains since its growth in the global market cannot meet the increasing demand for animal-derived foods.
A third of all feedstuffs annually produced (1.3 billion tons) is wasted by consumers and retailers or is lost due to the lack of adequate technology during and after harvesting. Of this total, around 300 million tons per year are foods available for consumption, and according to UNEP (2013), it is enough to feed part of the global population. In Northeast Brazil, the high production of mango associated with inappropriate management during production and after harvesting results in 28% average loss index (Choudhury, 1995), which end up as residues accumulated in inappropriate places, contaminating water and soil resources.
The most adequate use of discarded feeds in animal feeding is not well known due to variations in chemical composition associated with a large number of cultivars, edaphoclimatic conditions, and industrial processing. In general, mango fruit is rich in carbohydrates and organic acids, which are important energy and flavor providers, with volatile compounds as the factors responsible for the flavor (Liu et al., 2013).
Aragão et al. (2012) evaluated corn replacement with mango meal in sheep diets and found no effects on nutrient intake, even with 100% replacement, demonstrating its acceptability. On the other hand, Cavalcante et al. (2006) reported that inclusion of dehydrated mango above 36.1% DM caused a decrease in nutrient intake of sheep. Negative effects on intake and nutrient digestibility were observed by Anigbogu et al. (2006) in sheep fed 45 and 60% mango almonds. The authors justified the results due to the toxic properties of phenolic compounds present in the fruit. Huber et al. (2012) reported that mango has 4.89 and 6.84% of total phenols in the skin and kernel, respectively. Depending upon the ratio of mango peels and seed kernels fed to ruminants, decreased intake and digestibility can occur due to the astringent effect of phenolic compounds, such as tannins, which interact with proteins, carbohydrates and minerals, reducing the nutritional value of the diet (Shahidi and Naczk, 1995).
Thus, we hypothesized that mango meal, which is high in carbohydrates, could replace corn used as an alternative feed in goat production systems. Therefore, the purpose of this study was to evaluate the effects of corn replacement by whole mango meal (WMM) on milk yield and composition, and rumen kinetics of lactating goats.
Materials and methods
Ethical considerations
Management and care of the animals was performed in accordance with the guidelines and recommendations of the Committee of Ethics on Animal Use (CEUA) at the UFRPE, under the license number 143/2014.
Location
The experiment was conducted in the Goat Sector of the Animal Science Department of Universidade Federal Rural de Pernambuco (UFRPE), Recife, Brazil.
Experimental diets
Diets were formulated to meet dairy goat requirements in early lactation according to the National Research Council (NRC, 2007). The forage:concentrate ratio was 60:40 on a dry matter (DM) basis with Tifton-85 (Cynodon spp.) hay as the forage. Diets consisted of four replacement levels of corn by whole mango meal: 0, 330, 660, and 1000 g/Kg on DM basis. Chemical composition of ingredients is shown in Table 1. Proportions and chemical composition of experimental diets are shown in Table 2.
Mango meal was elaborated with the entire fresh fruit of Mangifera indica, “Rosa” variety. Fruit pulp, peel, and seeds were ground in a forage machine, dehydrated under the sun for 48 hours, and then ground passing through a 10 mm screen.
Animals, management, and sample collection
To determine nutrient intake and milk parameters, eight crossbred Saanen goats with 48.7 ± 1.99 Kg average initial body weight (BW) and 48.0 ± 2.0 days in lactation were distributed across a 4×4 double Latin square design, with four treatments and four animals. Feed was offered in two daily meals at 7:00 and 16:00 h in a sufficient quantity to obtain about 10% orts. The trial lasted 76 days and each of the four experimental periods lasted 19 days, divided as follows: 14 days for adaptation to the diets and 5 days to collect supplied feed, and ort samples.
1WMM: whole mango meal.
2Nutrients/Kg of product: Vitamin A - 135,000 IU; Vitamin D3 - 68,000 IU; Vitamin E - 450 IU; Calcium - 240 g; Phosphorus - 71 g; Potassium - 28.2 g; Sulfur - 20 g; Magnesium - 20 g; Copper - 400 mg; Cobalt - 30 mg; Chrome - 10 mg; Iron - 250 mg; Iodine - 40 mg; Manganese - 1,350 mg; Selenium - 15 mg; Zinc - 1,700 mg; Fluorine - 710 mg.
Orts were removed each day before the morning meal and weighed, sampled, placed in plastic bags, and frozen at -18 ºC. The feed was sampled three times per week d also placed inside plastic bags and frozen at -18 ºC. At the end of each experimental period, feed samples and orts were thawed and subjected to pre-drying at 60 ºC for 72 h, then ground in a Wiley- TE-625 knife mill (Tecnal, São Paulo, SP, Brazil) passing through a 1 mm mesh. Composite samples from each animal were saved on a dry-weight basis for each time period.
Milk
Goats were hand milked twice daily at 6:00 and 15:00 h and milk yield was recorded during the five days of data collection. Milk samples were collected and pooled by proportion according to milk yield at each milking. Milk samples were analyzed for fat, protein, and lactose concentration by infrared analysis (Bentley-2000, Bentley instrument, Inc. Minnesota, USA). Total solids were determined using the oven method according to AOAC (1990).
Chemical analysis
Dry matter (DM), organic matter (OM), crude protein (CP), and ether extract (EE) analyses were performed according to the AOAC (1990), method number 934.01 for DM, 930.05 for OM, 981.10 for CP, and 920.39 for EE. The EE was analyzed by Soxhlet extraction with petroleum ether. The concentration of neutral detergent fiber (NDF) was assayed with a heat-stable amylase and corrected for ash and nitrogenous compounds using the techniques described by Mertens (2002), with corrections for protein according to Licitra et al. (1996) and adding thermostable alpha-amylase. Lignin was extracted with sulfuric acid 720 mL/L (Van Soest and Wine, 1967). Non-fibrous carbohydrates (NFC) were calculated as follows, according to Hall (2000):
Where:
CP = crude protein.
NDFap = neutral detergent fiber corrected for ash
and protein.
EE = ether extract.
Concentration of condensed tannins (CT) and total In vitro semi-automatic technique of gas production was used to evaluate rumen fermentation kinetics, according to Theodorou et al. (1994). The rumen fluid was obtained from a rumen-cannulated goat. Pressure from the accumulated gas was measured by a pressure transducer (type T443A; Bailey & Mackey, Birmingham, England).
To assess the fatty acid profile, a 200 mL composite milk sample was used to separate the fat with the methodology proposed by Murphy et al. (1995). Triglycerides were submitted to transmethylation for methyl esters by using the 5509 methods (ISO, 1978). Methyl esters of fatty acids were analyzed by gas chromatography using a Varian 431-GC equipment and a mass spectrophotometer Varian 220-MS in a capillary column Zebron ZB-5MS Phenomenex (30 m × 0.25 mm × 0.25 µm). Fatty acids were quantified by area normalization of the methyl esters in percentage per area (%).
Crude income (economic analysis) was estimated considering only the feeding costs, assuming that all other components, fixed and variable, would be common for all treatments. The cost of whole mango meal was calculated based on the cost of labor to prepare it. Net income was calculated by the difference between crude income (milk production × price of a milk L) and the daily cost of feeding.
Statistical analysis
Parameters for cumulative gas production kinetics were analyzed by the nonlinear regression procedure (NLIN) with SAS software (2003, SAS 9.1 Institute Inc., Cary, NC, USA).
Intake and milk performance were submitted to variance and regression analysis using the R software (R Development Core Team version 2.15.3, 2013), according to the following model:
Where: phenols (TP) in whole mango meal were analyzed according to Wolfe et al. (2008).Yijk = dependent variable measured in animal j that was subjected to the i treatment in period k.
µ = general mean.
Ti = fixed effect of treatment i. Aj = random effect of animal j. Pk = random effect of period k. eijk = random error.
All statistical procedures were conducted with
0.05 as the critical probability level for a type I error.
Results
Intake and milk performance
No effects were found for WMM level on the intake of DM (1890 g/d), CP (278 g/d) and NDF (959 g/d; p>0.05; Table 3). However, intake of NFC and total digestible nutrients (TDN) decreased linearly around 27.9 g/d and 0.082 Kg/d, respectively, as dietary WMM increased (p<0.05).
Milk production decreased linearly 0.04 Kg/d as WMM increased (p<0.05; Table 3). The 4% fat- corrected milk yield showed no difference between treatments (p>0.05; Table 3). WMM levels had no effect (p>0.05) on milk composition variables, with average values of 4.03% lactose, 2.40% protein, and 10.0% total solids, respectively. Milk fat content increased linearly 0.09% for every 1% increase of WMM levels (p<0.05).
Milk fatty acid profile showed no difference between treatments (p>0.05), except for myristoleic fatty acid C14:1 cis-9, which decreased linearly 0.15 g/100g of fatty acids for every 1% increase in WMM levels (p<0.05; Table 4). Saturated fatty acids (SFA), polyunsaturated fatty acids (PUFA) and desired FA content showed no difference, with mean values of 82.9, 17.1 and
25.2 g/100 g, respectively (p>0.05).
In vitro fermentation kinetics
Gas produced from total carbohydrates (VTC) and fiber carbohydrates (VfFC) decreased linearly (106 and 137 mL/g of DM, respectively) as WMM increased (p<0.05; Table 5). A quadratic effect was observed for gas volume produced from non-fiber carbohydrates (Vf NFC), with a value of 58 mL/g of DM estimated with 32.7% corn replacement with WMM (p<0.05). A similar effect was found for non-fiber carbohydrate fractions and fiber carbohydrate (KdFC) rates, with values of 0.053 and 0.011 %/h estimated with 46.4 and 53.0% of replacement, respectively (p<0.05).
Economic evaluation
The mango used in the study was discarded fruit (mango wastes) from municipal markets, and the cost involved in its acquisition and transportation was R$0.33/Kg DM. The lowest total cost of feeding (Table 6) was found at a replacement level of 1000 g/Kg (BRL, R$1.17; where BRL = code for Brazilian currency in “Real” = R$). For each R$1.00 invested, R$2.52 was returned.
BRL - code for Brazilian currency, real (R$); USD$1.00 = BRL, R$2.24.
1. Ingredient costs (R$/Kg DM): Whole mango meal: 0.33; Ground corn: 0.80; Soybean meal: 0.95; Urea: 1.80; Dicalcium phosphate: 1.35; Mineral mix: 1.50; Tifton hay: 0.59.
2. Price of a liter of milk in the region of the study: R$1.50.
Discussion
The absence of effects on milk yield (4% fat- corrected) and total milk solids suggests that goats have adapted to variations in diet quality, characterized by increasing levels of indigestible nutrients such as lignin and phenolic compounds. This result is in agreement with Hofmann (1989) that reported high selectivity in goats, and ability to adapt to seasonal variations in the quality of fodder.
Seventy-five percent of the milk fatty acid (FA) profile of goats is composed of capric (C10:0), myristic (C14:0), palmitic (C16:0), stearic (C18:0), and oleic (C18:1) acids, according to Park et al. (2007). In the present study, these FAs totalized 89.1% of the total FA in milk fat (Table 4), which probably was influenced by the source of the tested feed, as the kernel of mango seed is basically composed of C16:0, C18:0, and C18:1, which together represent 91% of the total FA (Rukmini and Vijayaraghavan, 1984). An average concentration of stearic acid C18:0 of 8.04 g/100 g FA was observed in this study, which is within the range suggested (from 6 to 11%) by Chilliard et al. (2001). This is a positive result, since stearic acid is not involved in the increase of cholesterol levels. Thus, as soon as stearic acid is ingested, it is metabolized to oleic acid C18:1.
The reduction in concentration of myristoleic acid C14:1 cis-9 per 100 g FA could be related to the presence of tannins in mango meal. A similar result was recorded by Toral et al. (2011), when tannins with sunflower oil were added to the diets of Assaf sheep, observing a reduction in C14:1 cis-9 in comparison to a diet without tannins. Literature results referring to the impact of tannin intake on milk fatty acid profile is still inconsistent. Reductions in 18-carbon FA contents can occur in the presence of tannins for grazing cows or sheep, or it may increase due to biohydrogenation of dietary PUFA, which depends on tannin type or dose (Cabiddu et al., 2009).
The lowest gas production obtained from total and fiber carbohydrates can be attributed to the greatest percentage of lignin in the fiber fraction and the presence of tannins in the WMM (Table 1). Digestion (gas production) is assessed by the ratio between soluble and insoluble fractions. Thus the higher concentration of the fraction for low digestibility observed in the diets where corn was totally replaced (1000 g/Kg) with WMM (Table 5) contributed to the lower gas production.
Considering the principle that rate of digestion is essentially a physicochemical function of the diet (Van Soest, 1994), we can say that lignin content, physicochemical interaction between NDF and the other dietary compounds, and the presence of tannins affected the degradation dynamics. The presence of seeds in the WMM contributed to the increased contents of lignin (Table 2), which explains the reduction in rumen fermentation due to inhibition of microbial adhesions and, consequently, the degradation process.
The results related to rumen degradation kinetics were correlated with those of ingestion behavior. When the goats were fed 1000 g/Kg WMM, there was a daily rumination time (554 minutes) close to the physiologic time (600 min/d) cited by Welch (1982), a fact that may be attributed to a greater proportion of fractions of low digestibility carbohydrates and phenolic compounds that block the access of cellulolytic enzymes. As a result, there was an increase in the time required to process the forage into small particles, which reduces rumination efficiency.
Knowing that the mango origin was from discarded food at municipal markets, the cost involved in its acquisition and transportation represented only R$0.33/Kg of DM (Table 6). The low cost of discarded mango resulted in 23.5% economy of the diet when 1000 g/Kg of corn was replaced by WMM. Despite the decrease of crude income with the replacement of corn by WMM, the control diet (R$3.06/d) and the diet with 330 g/Kg replacement (R$3.05/d) were similar. This decrease can be explained by the fact that crude income takes into account only the gain from milk produced. However, an opposite behavior was observed for net income, since diets with higher WMM were more profitable due to a lower cost of the feed. The costs reduction of the diet represents a 17% increase in net income for the diet with 1000 g/Kg replacement when compared to the control diet. This suggests a better benefit:cost ratio and indicates a higher economic return. Besides being economically viable, replacement of corn with WMM becomes interesting from the standpoint of environmental preservation, since discarded mango fruit -unacceptable for human consumption- can be used in a sustainable way as animal feed instead of being improperly discarded.
We recommend replacing corn with whole mango meal up to 330 g/Kg DM of diet, because of the better fermentation of fiber carbohydrates, it does not compromise milk yield of lactating goats, and it also reduces feeding costs.