The cost of feeding makes the greatest contribution (exceeding 70%) to the total cost of milk production (Alqaisi et al., Reference Alqaisi, Ndambi and Hemme2011; Zambom et al., Reference Zambom, Fernandes, Schmidt, Gonçalves, Pozza, Javorski, Souza and Tinini2015). In this context, the use of by-products from agroindustry in livestock feed is a promising strategy to reduce costs, if there is availability of residues in the region (Correddu et al., Reference Correddu, Lunesu, Buffa, Atzori, Nudda, Battacone and Pulina2021; Quintero-Herrera et al., Reference Quintero-Herrera, Zwolinski, Evrard, J and Rivas-García2023). By-products from industrialization of cassava (Manihot esculenta Crantz) stand out among these residues with potential for use in livestock feeding (Serpa-Fajardo et al., Reference Serpa-Fajardo, Hernández-Ramos, Fernández-Lambert, Sandoval-Herezo and Andrade-Pizarro2022) due to cassava's high residual starch level and high energy value. After starch extraction, the residual fibrous material (cassava starch waste (REAM) or cassava pulp) is generated constituting about 30% of the original materials. There are only a few studies on the addition of REAM to the diet of dairy cows. However, this work is essential in order to evaluate the effects of these residues on feed intake and satisfactory performance for milk production of cows (Johansen et al., Reference Johansen, Lund and Weisbjerg2018).
The structure of the food as well as its nutrient content can affect digestive and metabolic utilization, as well as animal performance (Celi et al., Reference Celi, Cowieson, Fru-Nji, Steinert, Kluenter and V2017). Similarly, the milk components may undergo changes resulting from the diet composition through the nutrients that reach the mammary gland (Wanderley et al., Reference Wanderley, Itavo, Santos, Ccbf, Dias, Difante, Gurgel, Cunha, Mateus, Cardoso and Oliveira2022), while some blood parameters may indicate the metabolic condition of the animals. The purpose of this study was to evaluate the effect of different levels of REAM in the cow's diet on the intake and digestibility of dry matter and nutrients, blood parameters and milk production and composition of lactating Holstein cows
Material and methods
The experiment was carried out at the dairy cattle raising sector of Experimental Station of the Universidade Estadual do Oeste do Paraná (UNIOESTE), in Marechal Cândido Rondon, Paraná, Brazil. The experimental protocol was approved by the Ethics Committee on the Use of Animals of UNIOESTE (number 06511).
Design and management
Five multiparous Holstein cows with 150 ± 21 days of lactation and a mean milk production of 30.65 ± 1.17 kg/day and mean body weight of 593.6 ± 61.1 kg were used. The cows were housed in tie-stall barns in individual stalls with automatic water dispensers and concrete flooring. After feeding the cows were released into a resting area (30 m2 per cow) with natural shade for 2 h (at midday), and during the night. Milking was performed mechanically twice a day at 06:00 and 16:30 h.
Cows were distributed in a Latin square design with five treatments and five periods. Each period lasted 21 days, composed of 14 days for adaptation of cows to the diet and seven days for data and sample collection, totaling 105 days of experiment. Weighing of cows was performed at the beginning of each experimental period, after milking and before morning feeding. The treatments consisted of five levels (0, 8, 16, 24 and 32% of dry matter, DM) of REAM in the total diet offered to the cows. Feeds had a 50:50 (w:w) concentrate-to-roughage ratio, with corn silage being used as a roughage source and corn, REAM, soybean meal and minerals as concentrate sources. Feeds were balanced to meet the nutritional requirements for breeding animals (Table 1), according to recommendations of the National Research Council (National Research Council – NRC, 2001).
Composition of the experimental diets
1 Composition (quantities per kg of product): calcium – 260 g; phosphorus – 40 g; cobalt – 13 mg; magnesium – 15 g; sulphur – 12 g; manganese – 1100 mg; zinc – 3000 mg; selenium – 20 mg; iodine – 40 mg; fluorine – 170 mg and iron – 400 mg.
2 Sulphur source (24% N and 24% S).
3 Estimated values by the following equation: TDN (%) = DCP +(DEE x 2.25) + TDC.
The cassava starch waste (REAM) used was collected in a cassava processing factory localized in the Marechal Cândido Rondon/PR. The collected material was stored over a cement floor and covered with black plastic sheeting. The REAM was stored for 21 days, corresponding to an experimental period, and was used within this period. Feed was provided twice a day, at 07:00 and 18:00 h, in the 60 and 40% proportion, respectively, of the total dry matter offered. Water was provided ad libitum. To avoid feed restriction and monitor feed intake, the uneaten feeds were weighed daily maintaining a proportion of leftover feed from 10 to 20% of the total DM offered daily. All feeds were weighed at the time of providing to the cows.
Sampling and analysis
During the seven days reserved for data collection in each experimental period, samples of corn silage, REAM, concentrate and the leftover feed of each animal were collected. The samples were then dried at 55 ºC in a forced-air oven for 72 h to determine DM and then ground to 1.0 mm in a Wiley mill. For determination of dry matter and nutrient digestibility, from the 15th to 20th day, faeces samples were collected directly in the rectal ampulla of the cows every two hours from 08:00 until 18:00 h each day. The estimate of faecal excretion was measured using as indicator the indigestible neutral detergent fiber (INDF), as proposed by Cochran et al. (Reference Cochran, Adams and Wallace1986). The sample of feeds, leftover feed and faeces were estimated according to Association of Official Analytical Chemistry – AOAC (1995) methods for dry matter (DM), crude protein (CP), ether extract (EE), lignin (L) and ash. Organic matter was calculated by the difference between the total DM and the ash content. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) were determined as previously described by Van Soest et al. (Reference Van Soest, Robertson and Lewis1991). Insoluble crude protein concentrations in NDF (NDICP) were analyzed according to procedures described by Licitra et al. (Reference Licitra, Hernandez and Van Soest1996). The non fiber carbohydrate concentration (NFC) was calculated according to the following equation described by Van Soest et al. (Reference Van Soest, Robertson and Lewis1991): NFC, % = 100 – (CP% + (NDF% – NDICP%) + EE% + ash%), in which: NFC, Non-fiber carbohydrates; CP, crude protein; NICP, Neutral detergent insoluble crude protein. Total carbohydrate (TC) was calculated according to the following equation: TC (%) = 100 – (CP% + EE% + ash%). Total digestible nutrient concentration (TDN) was calculated according to the following equation described by Sniffen et al. (Reference Sniffen, O'Connor, Van Soest, Fox and Russell1992): TDN (%) = DCP + (DEEx2.25) + TDC, in which: DCP, digestible crude protein; DEE, digestible ether extract; and TDC, total digestible carbohydrates.
Milk yield was measured, and samples were collected during morning and afternoon milking, with samples pooled to yield one sample per day. Fat concentration, protein, lactose, total nonfat solids, total solids and minerals were measured using a Milkoscope® automatic analyzer. The calculation of the milk production corrected for 3.5% fat (MPc) was done using the equation described by Sklan et al. (Reference Sklan, Ashkennazi, Braun, Devorin and Tabori1992). Feed efficiency was calculated for each cow, dividing mean milk production by the mean DM ingestion of each data-collection period.
Blood sample collection was carried our soon after milking and before morning feeding on the 21st day of each experimental period using 10 mL test tubes with sodium fluoride through venepuncture of the coccygeal vein. The concentrations of glucose, blood urea nitrogen and creatinine were measured in blood plasma using a commercial colorimetric kit in accordance with the manufacturer's instructions.
Statistical analysis
Data were analyzed by ANOVA and regression analysis, and significant equations with the highest coefficients of determination were adjusted, with significance declared at P < 0.05. All analyses were performed using Sisvar software for Windows (version 5.1, Statistical Analysis Software, UFLA, Lavras, MG).
Results
Intakes of DM, organic matter, CP, TC, NDF and ADF were not altered by the addition levels of REAM in the diet of dairy cows (Table 2). There was a significant reduction in EE intake such that for every 1% addition of REAM in the cow's diet there was a reduction of 0.004 g/kg in the EEI. TDN intake also showed a significant linear reduction, and in this case, there was a reduction of 0.047 g/kg for each 1% addition of REAM. In relation to dry matter and nutrient digestibility, a significant negative linear effect of REAM addition was observed for the majority of variables (Table 3), but not for CP digestibility. There were reductions of 0.210 and 0.199 g/kg for each 1% addition of REAM for digestibilities of DM and OM, respectively. For EE, TC, NDF, ADF and TDN digestibilities the reductions observed were 0.168, 0.247, 0.261, 0.218 and 0.271 g/kg, respectively, for each 1% addition of REAM in the diet.
Effect of different levels of residue from the extraction on cassava starch (REAM) in the diet of lactating Holstein cows on the daily intake of dry matter and nutrients
Notes:
1 DMI: dry matter intake.
2 OMI: organic matter intake.
3 CPI: crude protein intake.
4 EEI: ether extract intake.
5 TCI: total carbohydrate intake.
6 NDFI: neutral detergent fiber intake.
7 ADFI: acid detergent fiber intake.
8 TDNI: total digestible nutrient intake. R: regression; P: level of confidence; R2: coefficient of determination; CV: coefficient of variation. ns: not significant (P > 0.05). L: linear equation. (L1) ŷ = 0.720 – 0.004x. (L2) ŷ = 13.718 – 0.047x.
Effect of different levels of residue from the extraction on cassava starch (REAM) in the diet of lactating Holstein cows on the dry matter and nutrients digestibility
Notes:
1 DMD: dry matter digestibility.
2 OMD: organic matter digestibility.
3 CPD: crude protein digestibility.
4 EED: ether extract digestibility.
5 TCD: total carbohydrate digestibility.
6 NDFD: neutral detergent fiber digestibility.
7 ADFD: acid detergent fiber digestibility.
8 TDN: total digestible nutrients. R: regression; P: level of confidence; R2: coefficient of determination; CV: coefficient of variation. ns: not significant (P > 0.05). L: linear equation. (L1) ŷ = 68.72 – 0.210x. (L2) ŷ = 70.26 – 0.199x. (L3) ŷ = 85.99 – 0.168x. (L4) ŷ = 69.45 – 0.247x. (L5) ŷ = 50.07 – 0.261x. (L6) ŷ = 43.72 – 0.218x. (L7) ŷ = 71.22 – 0.271x.
Milk production showed a negative linear effect with the addition levels of REAM in the diet (Table 4), specifically, there was a reduction of 0.140 kg/d milk for each 1% of addition level. The concentrations of protein, lactose, minerals and nonfat solids showed a positive linear response to the inclusion levels. For each 1% of REAM added to the diet, there was an increase of 0.004; 0.006; 0.001 and 0.011 g/kg, respectively. The concentrations of blood glucose, blood creatinine and blood urea nitrogen were not affected by inclusion of REAM.
Effect of different levels of residue from the extraction on cassava starch (REAM) in the diet of lactating Holstein cows on the milk production (MP), dry matter intake (DMI), milk production efficiency (MPE), other milk components and on the concentrations of plasma glucose, plasma creatinine, and plasma urea nitrogen (PUN)
Notes:
1 Non fat solids.
2 Total solids. R: regression. P: level of confidence. R2: coefficient of determination. CV: coefficient of variation. ns: not significant (P > 0.05). L: linear equation. (L1) ŷ = 31.10 – 0.140x. (L2) ŷ = 3.08 + 0.004x. (L3) ŷ = 4.58 + 0.006x. (L4) ŷ = 0.69 – 0.001x. (L5) ŷ = 8.34 – 0.011x.
Discussion
The diet was formulated assuming a DM intake of 3% of body weight and actual intake remained near or above this throughout. This confirms that there were no limiting factors for food intake in the diet, neither in physical terms or energetic terms (National Research Council – NRC, 2001). The lack of significance for the intakes of OM, CP, TC, NDF and ADF probably arise from the absence of significance for the DM intake since these components represent the greatest fraction of the diet (Table 1). Our results corroborate those observed by Abrahão et al. (Reference Abrahão, Prado, Perotto, Zeoula, JAC and SMB2006), who evaluated substitution of corn by REAM in finish feeding of young bulls and did not observe significant differences in DM intake in relation to body weight. Other authors (Zambom et al., Reference Zambom, Fernandes, Schmidt, Gonçalves, Pozza, Javorski, Souza and Tinini2015; Serpa-Fajardo et al., Reference Serpa-Fajardo, Hernández-Ramos, Fernández-Lambert, Sandoval-Herezo and Andrade-Pizarro2022) also did not observe a reduction in intake when they studied the addition of REAM to cattle diets.
There was a significant reduction in EE intake in response to REAM inclusion. The increased amount of REAM required an increase in the amount of soybean meal to balance the nitrogen levels of the diets, and as both components (REAM and soybean meal) have EE contents less than the others, there naturally occurred a reduction in the EE concentration in the cow's diet. Lima et al. (Reference Lima, Veloso, Silva, Bonomo, Pinheiro, Dutra, Júnior I G and Veloso2008) evaluated the addition of REAM at percentages of 0, 5, 10 and 15% in DM of the diet of mixed Holstein x Zebu dairy cows and found a reduction in the EE intake with increased levels of REAM in the diet, and the authors also attributed this reduction to the EE content (0.60%) in the REAM. The observed reduction in TDN intake may be related to the increase of NDF and ADF concentrations and lignin in the diets with greatest levels of REAM (Table 1), since the effectiveness of fiber degradation reflects on the estimates of total dietary energy content (Arce-Cordero et al., Reference Arce-Cordero, Paula, Daniel, Silva, Broderick and Faciola2021).
The reductions in digestibility may be related to the increase of the fibrous fractions in the diet. According to Van Soest (Reference Van Soest1994), there is a negative correlation between the lignin concentrations and the digestibility of the feed. Dias et al. (Reference Dias, Silva, Veloso, V, Pires, Souza, F, Mendes and Nunes-Nascimento2008) evaluated the addition of cassava starch production waste in the diet of growing heifers and found that inclusion levels above 14% were responsible for reducing the nutrient digestibility and negatively affecting the heifer's performance. Lima et al. (Reference Lima, Veloso, Silva, Bonomo, Pinheiro, Dutra, Júnior I G and Veloso2008) recommended the addition of REAM up to the level of 15% in the total diet of mixed dairy cows, which could be done without altering the nutritional and physiological parameters of the animals. Abrahão et al. (Reference Abrahão, Prado, Perotto, Zeoula, JAC and SMB2006) also did not observe significance in the addition levels of REAM in the diet of finishing young bulls in respect to CP digestibility. The low expressiveness in reductions in DM digestibility of the diets is due to the characteristics of the starch present in REAM, which exhibits greater degradability and digestibility than corn (Zambom et al., Reference Zambom, Fernandes, Schmidt, Gonçalves, Pozza, Javorski, Souza and Tinini2015).
The results presented here are similar to those reported by Ramalho et al. (Reference Ramalho, Ferreira, A S C, Lima and Rocha2006), who assessed the replacement of corn with cassava scrapings in diets for lactating Holstein cows and found that milk production decreased progressively with inclusion (0, 25, 50, 75 and 100% in DM). Similarly, Santos et al. (Reference Santos, Ferreira, Santos, Damasceno, Oliveira, Agostino, Olivo, Soares and Silva2015) evaluating cassava peel as a replacement for corn in diets for lactating Holstein cows also found a reduction in milk production with increasing levels (0, 33, 66, and 100% DM). Based on these data and the results presented here, one can infer that the use of REAM reduces milk production raising questions as to the economic viability of using this residue in the diet of dairy cows.
The milk production efficiency was not significantly affected by the addition levels of REAM indicating that despite the reduction in milk production, the addition of REAM in the diet did not alter the efficiency of the cows in feed conversion into milk. The concentrations of protein, lactose, minerals and nonfat solids showed a positive linear response to the inclusion levels of REAM. These increases may be related to the reduction in the daily quantity of milk produced, which provides for an increase in the concentration of the milk components (Wang et al., Reference Wang, Nan, Zhao, Jiang, Wang, Hua, Zhang, Wang, Liu, Yao and Xiong2021). In a study carried out by Pires et al. (Reference Pires, Susin, Santos, Mendes, Oliveira, RC and Simas2008) using lactating Holstein cows, it was reported that the conversion efficiency of dietary protein into milk protein was less when cassava scrapings were used. However, in the present study, there was a linear increase in the levels of milk protein. Probably these data are directly related to the dietary manipulations to balance nitrogen lervels (increasing levels of soybean meal and, at 32% REAM inclusion, urea).
Lactose concentrations are of great importance in the milk synthesis process, being considered as the main osmotic factor in milk (Roche et al., Reference Roche, Friggens, Kay, Fisher, Stafford and Berry2009; Lin et al., Reference Lin, Sun, Hou, Qu, Gao and Q2016). There is the hypothesis that using REAM in the diet of animals, thereby increasing the supply of starch with high degradability, will result in a higher supply of propionate (the main precursor of glucose), thereby causing increased availability of glucose and, as a consequence, an increase in the production of lactose (Roche et al., Reference Roche, Friggens, Kay, Fisher, Stafford and Berry2009). However, the osmotic situation should then result in an increased milk volume, which is not what we observed. Mineral concentrations in milk remain relatively constant in healthy cows since, in addition to lactose, sodium and potassium are also quite stable components (Linn, Reference Linn1988). The increase in mineral concentrations we observed was due to the increase in the lactose concentration, since they are correlated characteristics. In addition, the increase in mineral concentrations may have been brought about by the increase in milk protein concentrations since some minerals like calcium and phosphorus are bound to the amino acids of the milk casein molecules (Linn, Reference Linn1988; Wanderley et al., Reference Wanderley, Itavo, Santos, Ccbf, Dias, Difante, Gurgel, Cunha, Mateus, Cardoso and Oliveira2022).
The increase in nonfat solids occurred due to the increase of protein, lactose and mineral concentrations. According to Johansen et al. (Reference Johansen, Lund and Weisbjerg2018) changes in solids concentration are explained by variations in their components. The component with the greatest capacity to cause this variation in the nonfat solids is protein. The metabolic state of the animal may affect its performance, leaving the animal more susceptible to diseases such as metabolic acidosis, which causes changes in the physiochemical characteristics of the milk (Jaramillo-López et al., Reference Jaramillo-López, Itza-Ortiz, Peraza-Mercado and M2017). Serum glucose contents typically show little variation, mainly due to efficient homeostatic mechanisms present in organism of the cows. Thus, diets in general have little effect on glycemia, with the exception of animals in chronic malnutrition conditions (Gomes et al., Reference Gomes, Oliveira, Pereira, Vieira, Henrique, Fernandes and Leonel2016). The serum glucose levels found in the present study are in agreement with those observed by Pires et al. (Reference Pires, Susin, Santos, Mendes, Oliveira, RC and Simas2008), where mean values of 56.56 mg dL–1 were reported, with no difference being verified in plasma glucose concentrations in animals where a diet with manioc residue as a starch source was supplied.
Serum creatinine concentrations did not show variation with the increase of REAM in the diet. According to Gregory et al. (Reference Gregory, Birgel, D´Angelino, Benesi and Birgel2004), serum creatinine is a non-protein nitrogen substance and is formed through muscular metabolism of creatinine and phosphocreatine. Individual factors and those related to the age of the animal may have an influence on the values of serum creatinine, with values from 0.90 to 1.50 mg dL–1 being found in young animals. Values near these are within a limit considered physiological. The formation of this metabolite is rarely affected, not by diet or by protein catabolism, and it is usually used as a supplemental tool in evaluation of renal function as it may identify possible nephropathies. The PUN serum concentrations also did not vary with the increase of REAM in the diet. These data demonstrates that the efficiency in protein use was not affected by the inclusion. According to Lima et al. (Reference Lima, Veloso, Silva, Bonomo, Pinheiro, Dutra, Júnior I G and Veloso2008) PUN concentrations from 7 to 23.5 mg dL–1 are within a physiologically normal range; this variation is acceptable for lactating cows.
In conclusion, feeding lactating Holstein cows diets containing different levels of residue from cassava starch extraction (REAM) resulted in similar dry matter intake. The inclusion of REAM at up to 32% of the diet reduced milk production by 15%, and so based on these results we recommend the inclusion of 16% of REAM in the diet of dairy cows. Inclusion did not change the efficiency of milk production or blood parameters, and increased the milk concentration of protein, lactose, minerals and total solids.
Acknowledgements
We would like to express our sincere gratitude to the National Council for Scientific and Technological Development (CNPq) for their generous financial support through the Universal Project, under grant No. 403.555/2021-3.
