Minimizing the ruminal fermentation of starch is the major challenge in formulating dietary supplements that will reduce the acid content in the rumen. The fermentation pattern is indicative of the potential nutritional value of food in promoting better performance (Van Soest, Reference Van Soest1994). Thus, acidification of the rumen environment, demonstrated by the decline in ruminal pH, occurs mainly after eating foods with a rapid rate of fermentation (Ørskov, Reference Ørskov1986). Ruminants consuming forage-based diets show the pH values between 6.2 and 6.8, while those who consume concentrates can vary from 5.8 to 6.6 (Church, Reference Church1979).
The use of alternative energy sources, such as dehydrated and pelleted orange peel (in Brazil called citrus pulp), have a good chance of minimizing the negative effects arising from increasing intake of concentrate-based diets. Orange peel is particularly rich in pectin, which has a high potential for rumen degradation but does not acidify the rumen environment to any great extent due to degradation of the final product being acetic acid (Van Soest, Reference Van Soest1994).
Fresh orange peel has shown potential for producing good quality silage (Ítavo et al., Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000a), especially for its high content of soluble carbohydrates and pectin. These are responsible for the improvement in digestibility and degradability of the silages, linear increase in the turnover rate of solids, reduced concentrations of N-ammonia in the rumen and efficiency and flow of bacterial protein nitrogen (Lykos et al., Reference Lykos, Vargas and Casper1997). The Brazilian orange juice industry produces fresh orange peel as a by-product comprising approximately 500 g/kg of the total fruit. According to Ítavo et al. (Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000a, Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000b), this by-product in silage form has a high nutritional quality for ruminants with non-fibrous carbohydrates apparent digestibility of 892.0 g/kg and 840 g/kg for dry matter apparent digestibility.
According to Ishler and Vargas (Reference Ishler and Vargas2016), carbohydrates are the major source of energy for rumen microorganisms and the single largest component (60 to 70%) of a dairy cow's diet. They represent the major component of net energy for support of maintenance and milk production. Carbohydrate nutrition influences the composition of milk as precursors for lactose, fat, and protein. Achieving an optimum balance between structural and non-structural carbohydrates in dairy cattle rations is a challenge faced by nutritionists. The changes in milk fat percentage, and ruminal pH, or chewing activity of dairy cows are caused by altering dietary NDF or NFC contents (NRC, 2001).
The concentrations of volatile fatty acids (VFA) rumen are highly variable and dependent on feeding frequency, time after feeding, and also the diet composition (Bergman, Reference Bergman1990). Ítavo et al. (Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000c) studying orange peel silage for sheep observed no differences among treatments (silage additives) for the pH (6.97), ammoniac-N (6.78 mg/100 ml of ruminal liquid), volatile fatty acids (acetic 45.99 mM/ml, propionic 11.16 mM/ml, and butyric 5.50 mM/ml). They concluded additives did not alter the nutritional value of the diet when evaluated through rumen fermentation characteristics. However, for VFA production, there was a better energy conversion efficiency (kcal of VFA/kcal of glucose), 729.2 g/kg for the control treatment (without additive).
As for milk production and composition, most research has reported no significant differences when cows are fed citrus pulp and comparative diets (Gehman et al., Reference Gehman, Bertrand, Jenkins and Pinkerton2006; Williams et al., Reference Williams, Chaves, Deighton, Jacobs, Hannah, Ribaux, Morris, Wales and Moate2018). However, according to Alnaimy et al. (Reference Alnaimy, Gad, Mustafa, Atta and Basuony2017), citrus by-products are utilized as a low-cost nutritional supplement to the diets of cattle but should not be used at high levels for lactating cows as milk production tends to decrease.
Our hypothesis was increasing levels of replacement of whole plant corn silage (fibre and starch-rich) by orange peel silage, (pectin-rich), could improve the rumen environment and increase nutrient intake and milk yield without affecting the milk composition.
Materials and methods
The experiment was carried out at the Iguatemi Experimental Farm of the Maringa State University (Maringá, PR, Brazil) and was approved by the Local Ethical Committee for Use of Animals in Experimentation at the State University of Maringa (Protocol no. 027/2011). Full details of the design are given in the online Supplementary File.
Cows, diets and design
We used eight multiparous Holstein lactating cows (587.5 ± 39.6 kg BW) fitted with large diameter (4 inch) rumen cannula, starting on 110 ± 22 d of lactation. We prepared experimental (OPS) and control (WPCS) silages: fifty tonnes in natural orange peel (Citrus sinensis L. Osbeck) from the COCAMAR-CITRUS SA in Paranavaí, PR, Brazil, were ensiled uncoated in vertical trench silos, which were then first opened after 120 d. Similarly, fifty tonnes of whole-plant corn (Zea mays L.) was ensiled for 120 d. The concentrate, based on soybean meal, corn meal and minerals, was formulated to meet the nutritional needs of dry cows (National Research Council, 2001). The total mixed diet (TMD) was intended to provide an isoprotein diet for groups of cows, at a forage : concentrate ratio of 750 : 250. Details are given in the online Supplementary Table S1.
The experiment was in the form of a duplicated (double) 4 × 4 Latin square design each with four cows and four replacement levels of corn silage by orange peel silage and four periods (21 d each) of the sampling, carried out two times. During the experimental periods, the cows were fed a lactating cow total mixed diet (TMD) until their assignment to treatment. All cows were fed the same 750 : 250 g/kg roughage : concentrate ratio TMD with treatment differences achieved by replacement of corn silage by orange peel silage at rates of 0 (control), 250, 500 and 750 g/kg DM into each cow's individually weighed allocation of TMR at feeding. The TMD was fed three time by day.
Measurements and analytical methods
The animals remained in the adaptation period of 14 d followed by 7 d of sample collection (food, leftovers, milk and rumen fluid). Both the offered diet and the leftovers were collected by composite sampling to determine the nutrient intake. The samples were dried and ground. The diets and the leftovers were analysed according to Association of Official Analytical Chemists (2000) methods to determine dry matter (DM); organic matter (OM); crude protein level from total nitrogen (CP), and ether extract (EE). The determination of neutral detergent fibre (NDF) was performed according to Mertens (Reference Mertens2002).
The samples of milk of morning and afternoon milkings from the last 7 d of each experimental period were analysed according to AOAC (2000) methods to determine dry matter (DM); organic matter (OM); crude protein level from total nitrogen (CP) × 6.34 factor. The Gerber method chemical test was used to determine the fat content of milk samples.
The pH and ammoniac-N of rumen fluid was determined immediately upon withdrawal of the liquid (Ítavo et al., Reference Ítavo, Valadares Filho, Silva, Valadares, Leão, Cecon, Ítavo, Moraes, Rennó and Paulino2002). Analyses of volatile fatty acids in rumen fluid were done according to Wilson (Reference Wilson1971).
Calculations
Orange peel and corn silage dried samples and ground through 5 mm mesh were kept in nylon bags and incubated in the rumen of the cows. The incubation times were 3, 6, 12, 24, 48 and 72 h, and the zero time was used with washing in water. The DM, CP and NDF degradation for each incubation time was calculated by disappearance after incubation (g/kg). Degradation of feeds was calculated using the equation d = a + b(1 − e −ct) of Ørskov and McDonald (Reference Ørskov and McDonald1979), where d is the degradation rate at time t; a is an intercept representing the portion of feed solubilized at time 0; b is the rumen potentially degradable fraction; c is the degradation ratio of b fraction, and t is the incubation time. Non-linear parameters (a, b and c) were estimated by minimal square iterative procedures by Gauss-Newton procedure. The DM, CP and NDF effective degradability (ED) was calculated using the equation of Ørskov and McDonald (Reference Ørskov and McDonald1979). Effective degradability of feed was estimated according to ARC (1984).
To convert synthesis and production of fatty volatile acids, we used the factors 0.864, 1.0 and 0.618 for acetate, propionate and butyrate, respectively. Relations between effective production and VFA concentration in the rumen was described by the regression equations cited by Ítavo et al. (Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000c).
Statistical analysis
Data of milk production, milk composition and nutrients intake were analysed as a 4 × 4 Latin square with diet, period and cow as factors using the general linear model procedures of Statistical Analysis System – SAS (2000). The ruminal parameters data experimental design was randomized blocks, split plots (plots = four levels, five times subplot = sampling of rumen fluid), with four replicates per treatment. Replacement level, time of sampling of rumen fluid and period were used as factors in the analysis, performed using the general linear model procedures of SAS (2000).
Data were evaluated by analysis of variance and regression. Means at 0.05 significance level by Tukey test were compared, and the regression equations followed by the matrix model: Y = Xβ + ɛ
Where:
Y = Vector of observed values;
X = matrix corresponding to the values of independent variables;
β = Vector of unknown parameters;
ɛ = vector of random errors (N × 1); where ɛ ~ N (ϕ, Iσ 2);
Results
There was a significant effect of substitution of corn silage by orange peel silage on the DM, EE, NDF, and crude energy intakes (Table 1), but crude protein intake was not affected. Intakes of DM (kg/d), NDF (kg/d), and energy (Mcal/kg DM) declined linearly with increasing percentage of orange peel silage in the diet, while the consumption of EE showed a significant linear increase.
Table 1. Nutrient intake, milk production and composition averages and regression equations adjusted in function of the replacement level (RL) of orange peel silage in replacement of corn silage
SEM, Standard error of means; RL, DM Replacement level of corn silage by orange peel silage (g/kg DM).
a Y DMI = 9.3471 − 0.0158*RL (r 2 = 0.93).
b Y CPI = 1.4577.
c Y EEI = 0.0952 − 0.0029*RL (r 2 = 0.99).
d Y aNDFI = 4.0919 − 0.0182*RL (r 2 = 0.99).
e Y ADFI = 1.9563 − 0.0112*RL (r 2 = 0.99).
f Y NFCI = 38 625 + 0.0690*RL (r 2 = 0.99).
g Y DEI = 37.6333 − 0.0673*RL (r 2 = 0.93).
h Y Milk production = 19.3687 − 0.0024*RL (r 2 = 0.88).
i Y 4% Fat corrected milk = 16.6178 + 0.005658*RL−0.00000545*RL2 (r 2 = 0.99).
j Y Protein = 2.816 + 0.0009852*RL (r 2 = 0.99).
k Y Fat = 3.500 + 0.000825*RL (r 2 = 0.93).
There was a significant effect of substitution of corn silage by orange peel silage on milk production, 4% fat corrected milk (kg/d), protein and fat content on milk (Table 1). Milk production presented a negative linear effect, whilst 4% fat corrected milk showed quadratic effects of substitution of corn silage by orange peel silage. The regression equation predicted a maximum effect on fat corrected milk production at a level of 519.1 g/kg of replacement of corn silage by orange peel silage. Protein and fat content in milk showed a positive linear effect of substitution of corn silage by orange peel silage.
Ruminal degradation parameters are presented in Table 2. The soluble fraction (a) of CP of silages presented high values for both silages, above 500 g/kg, suggesting these silages have a lot of soluble CP. DM presented similarly high values, above 400 g/kg of soluble fraction whilst NDF a-fraction showed values of 120 g/kg, all with no difference between silages. The rumen potentially degradable fraction (b) of NDF of orange peel silage was higher (900 g/kg) than that of corn silage (680 g/kg: Table 2). Effective degradability of nutrients also showed elevated values, suggesting that orange peel silage has rapid degradability in the rumen environment. The effective degradability values of DM of corn silage were 740, 640 and 580 g/kg for 20, 50 and 80 g/kg hour-1 of passage rates in the rumen (ARC, 1984).
Table 2. Averages of dry matter (DM), crude protein (CP) and neutral detergent fibre (aNDF) ruminal degradation parameters (a, b and c) and effective degradability (EfD)
a,b Values in a same row with a different superscript are significantly different by t-student test (P < 0.05).
There was significant differences (P < 0.05) in rumen pH values both for replacement levels and also according to the collection times of rumen fluid (Table 3). The results show that there was intense fermentation of the diets. Replacing corn silage by orange peel silage (0, 250, 500 and 750 g/kg, in dry matter basis) with supplementation of concentrate for confined lactating Holstein cows showed the lower pH values (6.18), one hour after feeding. Initially, the values of ammoniac-N had risen due to the rapid degradation of rumen orange peel silage; such behaviour was accompanied by a consequent lowering of pH (Table 3).
Table 3. Values and regression equations adjusted for ruminal pH and Ammonia-N in function of collection time (t), for replacement levels (RL) of DM corn silage by orange peel silage
SEM, Standard error of means; RL, Replacement levels of DM corn silage by orange peel silage (g/kg DM).
a Y = 6.43468 + 0.00374832*RL − 0.130636*t + 0.0186622*t 2 (R 2 = 0.83).
b Y = 10.5902 + 0.0263425*RL − 0.87133*t (R 2 = 0.90).
The behaviour of the production of volatile fatty acids occurred cubically (Table 4 and online Supplementary Table S2). The daily yield (Y in mol/d) in the rumen of cows fed orange peel silage at replacement levels of the corn silage show that total VFA has modifications with orange peel inclusion in the diet. Diets with 500 and 750 g/kg DM of orange peel silage showed similar total VFA to the diet with corn silage only (0 g/kg DM).
Table 4. Average of Volatile fat acids production (Y in mol/d) and relationships between volatile fat acids results with total VFA, in ruminal fluid of cows fed with diets containing Replacement levels (RL) of DM corn silage by orange peel silage
SEM, Standard error of means; RL, Replacement levels of DM corn silage by orange peel silage (g/kg DM).
a Y = 4.310 + 0.362*RL − 0.15*RL2 (r 2 = 0.93).
b Y = 1.245 − 0.083*RL (r 2 = 0.90).
c Y = 0.305 − 0.036*RL (r 2 = 0.93).
d Y = 5.817 + 0.2825*RL − 0.001575*RL2 (r 2 = 0.92).
e Y = 0.7378 + 0.0282*RL − 0.0057*RL2 (r 2 = 0.98).
f Y = 0.2190 − 0.0291*RL + 0.0065*RL2 (r 2 = 0.98).
g Y = 0.0470 − 0.0028*RL (r 2 = 0.98).
Discussion
DM intake was reduced with the inclusion of orange peel silage in the TMR (Table 1), but this fact can be explained by a decrease in rumen fill, shown by the reduction in consumption of NDF and consequential increase in NFC consumption from the pectin present in orange peel silage. According to Van Soest (Reference Van Soest1994), orange peel is particularly rich in pectin, which has a high potential for rumen degradation without excessive ruminal acidification. However, total-VFA as well as acetic acid production was not greater than that of corn silage (Table 4).
The results for NDF intake suggest a metabolic regulation of consumption occurred. CP intake was constant, but energy consumption (MCal/d) decreased with increasing substitution level, leaving some doubt over this metabolic regulation which, according to Mertens (Reference Mertens1994) involves hunger and satiety signals that operate through different mechanisms to hormonal and neural controls of voluntary intake. The dominant role of physiological regulation of intake and physical limitations are modified by stimuli related to palatability, disease and food management so that the intake is affected by animal, food and feeding system characteristics. Overall, it appears that orange peel silage can be considered an acceptable alternative to corn silage at substitution levels up to 750 g/kg.
There were significant effects of substitution of corn silage by orange peel silage on milk production with an overall decrease of 0.0024 kg per replacement level (Table 1). Milk protein presented a positive linear effect of replacement with 0.009852 g/kg of replacement level. This increase could be associated with the higher proportion of propionic acid production in the highest replacement level (Table 4). Fat content of milk showed a positive linear effect with 0.00825 g/kg of replacement level. Some studies have shown a higher concentration of fat in the milk of cows fed citrus pulp compared to cows fed other diets (Van Horn et al., Reference van Horn, Marshall, Wilcox, Randel and Wing1975; Belibasakis and Tsirgogianni, Reference Belibasakis and Tsirgogianni1996). Leiva et al. (Reference Leiva, Hall and van Horn2000) worked with 20.5% citric pulp inclusions and observed greater fat production (+0.18%) compared to a diet with 19.5% cornmeal, as well as lower milk yield and lower milk protein.
The potentially degradable ruminal fraction (b) of the NDF of the orange peel silage was greater than 900 g/kg (Table 2), while the same fraction of corn silage was 680 g/kg. This result indicates the high potential of orange juice industry by-product in silage form as a fibre source for dairy cows.
It should be emphasized that the higher levels of inclusion of orange peel in the diet promoted stabilization of rumen pH (Table 3), even though there were significant differences in the levels of substitution. This is explained by the fact that the final product of pectin degradation, acetic acid, does not acidify the rumen environment, but acts as a buffer (Van Soest, Reference Van Soest1994). Citrus by-products improve the utilization of other dietary NDF, possibly due to positive effects on rumen microflora. Because these by-products are high pectin energy sources, when included in diets for ruminants, they may increase the molar proportion of acetic acid and decrease the molar proportion of propionic acid, resulting in an increased acetate/propionate ratio (Alnaimy et al., Reference Alnaimy, Gad, Mustafa, Atta and Basuony2017). This event can explain the 20.6% of the increase in milk fat content of the higher orange peel silage level treatment compare to corn silage treatment (4.1 vs. 3.4%).
Ammoniac-N concentration showed a cubic behaviour. An initial increase due to rapid ruminal degradation of orange peel silage was followed by lower values as a consequence of a slight lowering of pH (Table 3). Then, about seven hours post-feeding, there was a partial elevation, probably due to the recycling of nitrogen through the saliva.
The behaviour of the production and concentration of volatile fatty acids (Table 4) occurred in cubic form, suggesting that degradation occurred in two phases. Other parameters accompanied such developments, reinforcing this hypothesis. Probably the degradation of the components of the slow-fraction of silages, cited by Sniffen et al. (Reference Sniffen, O'connor, Van Soest, Fox and Russell1992), started or maximized its degradation after the peak pH and ammoniac-N.
According to Bampidis and Robinson (Reference Bampidis and Robinson2006) citrus pulp contains both pectin and cellulose, with pectin comprising approximately 450 g/kg of cell-wall. Pectin is degraded very rapidly and extensively in the rumen is degraded but, unlike starch, yields little lactate, causing less decline in rumen pH (Table 3). As mentioned above, citrus by-products tend to increase the molar proportion of acetic acid and decrease the molar proportion of propionic acid, resulting in an increased acetate/propionate ratio (Bampidis and Robinson, Reference Bampidis and Robinson2006). We saw only partial evidence of this; the acetate/propionate ratios were 3.89, 4.10, 4.09 and 3.67, respectively, for diets with 0, 250, 500 and 750 g/kg DM of replacement levels of DM corn silage by orange peel silage.
The substitution of corn silage by orange peel silage can be interesting to animal production. When evaluating the total production in moles/d (Y-total VFA), it is observed that the diet with 250 g/kg orange peel silage had similar values of energy production (5.97 mol/d) to the control WCPS diet (5.87 mol/d), indicating that this low level of substitution would be the most suitable (Table 4). However, according to Ítavo et al. (Reference Ítavo, Santos, Jobim, Voltolini, Faria and Ferreira2000c), the values of efficiency of energy conversion (VFA kcal/kcal of glucose) reveal that total or individual VFA production values are not the most suitable to evaluate a food or diet because when studying the production of total or individual VFA in rumen fluid, you have no idea of the contribution of digestible energy that was being converted into metabolizable. After subtracting the energy losses as methane, nitrogen and urine, the individual fatty acids would be participating in 0.65 to 0.75 of total metabolizable energy (Bergman, Reference Bergman1990).
It can be speculated that differences in the physical and chemical composition of citrus pulp such as citrus variety, silage, skins proportion, dry matter content etc. may explain the variety of performance responses reported when citrus pulp is supplied to dairy cows (Williams et al., Reference Williams, Chaves, Deighton, Jacobs, Hannah, Ribaux, Morris, Wales and Moate2018).
In conclusion, high levels of inclusion of orange residues in the diets of lactating cows can have effects on the synthesis of milk and its components (proteins and fats), and on ruminal parameters and synthesis of VFA. The results suggest that orange peel silage may replace corn silage. Orange peel silage is well accepted, even with changes in cow production. The ruminal parameters remain within normal limits, with no evidence of any adverse changes in the rumen environment. Finally, orange peel silage as a substitute for corn silage for feeding dairy cows has shown promising results in increasing milk fat content.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022029920001028.
Acknowledgements
The authors would like to thank the National Council of Scientific and Technological Development for financing the research and granting of the scholarship. This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001.
