Figure 1. Relationship between milk urea nitrogen (MUN) and plasma urea nitrogen
(PUN); MUN = 3.62 + 0.65 PUN; R2 = 0.73; n = 283.
E. T. Lyatuu
M. L. Eastridge 1
The Ohio State University
Department of Animal Sciences
1 For more information, contact at: The Ohio State University, 221B Animal Science Building, 2029 Fyffe Road, Columbus, OH 43210-1095; 614-688-3059: fax: 614-292-1515; e-mail: eastridge.1@osu.edu
Twelve Holstein dairy herds in Ohio were visited once, and 25% of the lactating cows in each herd were randomly selected for data collection. Blood and milk were sampled approximately four hours postfeeding. Many of the feeds and feed mixtures used on the farms were sampled and analyzed for chemical composition.
Nutritional factors were found to be more important than genetics in accounting for variations in milk production, milk composition, milk urea nitrogen, and plasma urea nitrogen values. Nutrient intake was observed to provide better correlation estimates on these variables than nutrient content in diets. Milk protein yield was more responsive to dietary differences than milk protein percentage. The interaction between intakes of neutral detergent fiber (NDF) and nonfiber carbohydrates (NFC) was the most important factor accounting for the variation in milk protein yield, explaining 58% of the variation.
The mean plasma urea nitrogen (PUN) concentration (13.4 ± 0.2 mg/dl) was higher than mean milk urea nitrogen (MUN) concentration (12.4 ± 0.2 mg/dl). The regression of milk urea nitrogen on plasma urea nitrogen resulted in the following relationship: MUN = 3.62 + (0.65 x PUN) (R2 = 0.73).
Further regression analysis showed that both plasma and milk urea nitrogen concentrations were affected by ration composition and intakes of various dietary components. Dietary intake of crude protein (CP) was found to explain 16% of the variation observed in milk urea nitrogen concentration, while the squared term for fatty acid (FA) intake explained 11% of the variation.
The concentrations of urea nitrogen in milk and plasma can be used to monitor feeding programs on dairy farms and thus used to improve management of the nutrition program. This provides another tool to dairy farmers for minimizing feed costs while maximizing production.
Following an increased importance of the protein concentration on milk pricing, a number of research projects have been conducted to study various factors that can affect milk protein content. Improvement through genetics and nutritional manipulation are the main strategies used to increase milk protein concentration. Factors such as feed composition, digestibility, and intake are some of the nutritional factors that have been identified as important in influencing the process of milk protein synthesis. Other factors include body weight, level of production, stage of lactation, and environmental conditions. Most of these impart their effects indirectly by affecting DM intake.
In recent years, studies on mechanisms behind urea formation from dietary protein found that the concentration of urea nitrogen in blood is directly related to ruminal ammonia absorption. However, since MUN is closely correlated with PUN (Baker et al., 1995; Roseler et al., 1993), and the fact that MUN is more representative over time of ruminal ammonia levels, less invasive to the animal, and less variable, most researchers would rather analyze for MUN concentrations. Optimum MUN concentration for individual cows ranges from 8 to 25 mg/dl, while optimum MUN concentration for a herd ranges from 12 to 18 mg/dl (Roseler et al., 1993). A number of factors are known to influence PUN and MUN including nutrient intake, sampling time, days in milk, and method of analysis (Butler et al., 1995; Gustafsson and Palmquist, 1993). Therefore, objectives of this study were to investigate the relationship between plasma and milk urea nitrogen and the effects of dietary protein and energy fractions on plasma and milk urea nitrogen; and to examine the possible relationships among dietary protein and energy fractions and milk protein concentration and yield.
Twelve Holstein farms in Ohio were selected primarily based on having at least 50 cows in the herd, participating in the Ohio DHI program, a rolling herd average (RHA) for milk greater than or equal to 18,000 lb, and a RHA for milk protein greater than or equal to 3.4% or a RHA for milk protein percentage being higher than or equal to the RHA for milk fat percentage. These criteria were set so that some farms with abnormal fat/protein ratio and (or) high milk protein percentage would be included in the study.
A one-time visit to each farm was undertaken whereby information on feeding systems was gathered. The visits were arranged to collect blood samples four hours after feeding. Data were collected from 25% of the lactating cows selected at random from each of the production groups. Genetic information (predicted transmitting ability for milk yield and milk components) for selected cows was taken from DHI records.
Blood samples were collected directly from the tail vein. During the visit, milk samples from 10 of the farms were hand-stripped from all quarters of the udder directly into a vial containing a preservative (2-bromo-2 nitropropane-1,3 diol). For the remaining two farms, milk was sampled at both the morning and afternoon milkings. Data for milk fat and milk protein percentages for each cow in 11 of the farms were taken from the most recent DHI report. For the other farm, samples of morning and afternoon milkings for each cow selected were sent to the DHI laboratory (DHI Cooperative, Inc., Powell, Ohio) for milk fat and protein analyses. The weighted average milk fat and milk protein percentages for each cow were calculated using the proportion of milk yield from each milking.
Representative samples of feeds fed to the animals were collected for analysis. The composition of the feeds not sampled (such as baled hay) was estimated using values from NRC (1989).
The DMI for individual cows was estimated (eDMI) based on an assumed BW for each animal (parity 1 = 1,250 lb, parity 2 = 1,300 lb, and parity 3 = 1,350 lb), 4% fat-corrected milk (FCM) from DHI records, and DIM using the following equation (Edwards, 1991):
eDMI = (0.011BW) + [{2 x ((0.08 x BW0.75) +
(0.74 x 4%FCM))} / 4.4] - 3.056 +
(0.0364DIM) - {(7.0772 x 10-5) x (DIM)2}
where BW is in kilograms, and FCM is in kilograms per day.
Because this equation does not provide realistic estimation of DMI when DIM exceeds 305, DIM greater than 305 were adjusted back to 305.
The general linear model and stepwise regression procedures of SAS® (1988) were used to analyze the data.
A total of 295 animals were sampled from the 12 farms. The dietary CP ranged from 13.6 to 21.5% of DM, NDF ranged from 23.1 to 45.3% of DM, NFC ranged from 27.4 to 52.8% of DM, and FA ranged from 2.57 to 5.80% of DM.
The mean eDMI among farms ranged from 37 to 48 lb/cow/day, milk yield from 47 to 90 lb/cow/day, MUN from 8.0 to 18.5 mg/dl, and PUN from 8.5 to 22.5 mg/dl. Both farms with the lowest and the highest PUN had the corresponding lowest and highest MUN, respectively.
Considering all animals, the mean milk production was 63 lb/cow/day (Table 1). Milk yield was correlated with DMI ( r = 0.70), and intakes of FA ( r = 0.56), CP ( r = 0.71), NDF ( r = 0.45), and NFC ( r = 0.49) (Table 2). Intake variables were more highly correlated with milk yield compared to their dietary concentration. These results suggest that nutrient intake variables are more important than nutrient composition of a diet when analyzing nutritional status for a farm or an animal.
| Table 2. Correlation Coefficients for the Variables Examined Among the 12 Farms. | ||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Lact. Number | DIM (day) | Milk Yield (lb/day) | Milk Protein (%) | Milk Protein (lb/day) | DMI (lb/day) | CP (% DM) | CP Intake (lb/day) | NDF (%DM) | NDF Intake (lb/day) | FA (% DM) | FA Intake (lb/day) | NFC (% DM) | NFC Intake (lb/day) | MUN (mg/dl) | PUN (mg/dl) | |
| Lact. Number | 1.00 | |||||||||||||||
| DIM (day) | -0.02 | 1.00 | ||||||||||||||
| Milk Yield (lb/day) | 0.14* | -1.44** | 1.00 | |||||||||||||
| Milk Protein (%) | -0.09 | 0.47** | -0.47** | 1.00 | ||||||||||||
| Milk Protein (lb/day) | 0.10 | -0.32** | 0.94** | -0.17* | 1.00 | |||||||||||
| DMI (lb/day) | 0.21 | -0.47** | 0.70** | -0.28** | 0.69** | 1.00 | ||||||||||
| CP (% of DM) | -0.07 | -0.16* | 0.26** | -0.32** | 0.17* | 0.22** | 1.00 | |||||||||
| CP intake (lb/day) | 0.10 | -0.17* | 0.71** | -0.36** | 0.66** | 0.68** | 0.74** | 1.00 | ||||||||
| NDF (% of DM) | 0.09 | 0.07 | 0.13* | 0.28** | -0.05 | -0.07 | -0.83** | -0.53** | 1.00 | |||||||
| NDF intake (lb/day) | 0.21** | -0.01 | 0.45** | 0.05 | 0.52** | 0.51** | -0.51** | 0.14* | 0.73** | 1.00 | ||||||
| FA (% of DM) | 0.10 | -0.02 | 0.26** | -0.21** | 0.19** | 0.16* | 0.22** | 0.29** | -0.14* | 0.06 | 1.00 | |||||
| FA intake (lb/day) | 0.17* | -0.06 | 0.56** | -0.27** | 0.51** | 0.46** | 0.23** | 0.57** | -0.11* | 0.34** | 0.90** | 1.00 | ||||
| NFC (% of DM) | -0.09 | -0.06 | -0.03 | -0.21** | -0.10 | -0.05 | 0.56** | 0.26** | -0.87** | -0.72** | -0.11 | -0.14* | 1.00 | |||
| NFC intake (lb/day) | 0.05 | -0.11 | 0.49** | -0.31** | 0.45** | 0.47** | 0.57** | 0.72** | -0.74** | -0.17* | 0.05 | 0.29** | 0.76** | 1.00 | ||
| MUN (mg/dl) | -0.13* | -0.09 | 0.05 | 0.06 | 0.08 | 0.10 | 0.24** | 0.16* | -0.16* | -0.09 | -0.38** | -0.31** | 0.10 | 0.10 | 1.00 | |
| PUN (mg/dl) | -0.13* | -0.14* | 0.09 | -0.01 | 0.11 | 0.13* | 0.33** | 0.24** | -0.18* | -0.08 | -0.30** | -0.23** | 0.09 | 0.10 | 0.86** | 1.00 |
| * P < 0.05. ** P < 0.01. | ||||||||||||||||
|
Table 1. Simple Statistics for the Variables Observed Among the 12 Farms1. | |||
|---|---|---|---|
| Variable | N | Mean | Standard Deviation |
| Milk yield (lb/day) | 294 | 62.9 | 21.3 |
| Milk fat (%) | 294 | 3.48 | 0.80 |
| Milk protein (%) | 294 | 3.39 | 0.40 |
| Milk protein yield (lb/day) | 294 | 2.09 | 0.64 |
| DMI (lb/day) | 295 | 43.3 | 7.5 |
| CP (% DM) | 295 | 17.0 | 1.98 |
| CP intake (lb/day) | 294 | 7.63 | 1.45 |
| NDF (% DM) | 295 | 35.4 | 5.42 |
| NDF intake (lb/day) | 294 | 15.8 | 3.2 |
| FA (% DM) | 295 | 3.59 | 0.86 |
| FA intake (lb/day) | 294 | 1.63 | 0.51 |
| NFC (% DM) | 295 | 36.5 | 6.07 |
| NFC intake (lb/day) | 294 | 16.3 | 3.26 |
| MUN (mg/dl) | 294 | 12.38 | 3.10 |
| PUN (mg/dl) | 294 | 13.38 | 4.04 |
| 1 Average performance for an individual cow; FA = fatty acids, NFC = nonfiber carbohydrates, MUN = milk urea nitrogen, and PUN = plasma urea nitrogen. | |||
While the 1996 average milk protein percentage for the state of Ohio was 3.24% (average for 878 official DHI herds), the average among the 12 farms visited in this research was 3.39% (Table 1). Both dietary composition and intake of CP, FA, and NFC were negatively correlated to milk protein percentage. Considering the fact that these nutritional variables had a positive correlation with milk yield (as discussed earlier), these decreases in milk protein percentage can be explained, in part, by a dilution effect of increased milk yield (Firkins and Eastridge, 1992).
Apart from the dilution effect, other reasons may explain the observed relationship between milk protein percentage and individual nutritional components. The negative relationship between CP and milk protein percentage can also be an indicator of excess rumen undegradable protein in diets. Excessive rumen undegradable protein has been associated with a decrease in microbial protein synthesis and amino acid flow to the small intestine.
The FA intake had a negative effect on milk protein percentage which may have resulted from changes in mammary amino acid utilization as suggested by Cant et al. (1993) or dilution effect resulting from increased milk yield. The interaction between CP and NFC intakes explained 13% of the variation in milk protein percentage.
Regressing nutritional variables on milk protein yield gave a higher coefficient of determination (R2 = 0.76) compared to when the same variables were used to estimate milk protein percentage (R2 = 0.22). These results support a conclusion made by Schingoethe (1996) that diet can influence milk protein yield more than it can influence milk protein content.
The FA percentage and FA percentage squared accounted for 10% of the variation in milk protein yield. Interaction of NDF and NFC percentages was also found to be important, explaining 5% of the variation. The interaction between dietary NDF and NFC intakes explained 58% of the variation in milk protein yield, and protein intake explained 6% of the variation.
The closeness by which the relationship of various variables to milk protein yield follows that of the same variables to milk yield suggests that the total amount of milk produced was more related to the milk protein yield (r = 0.94) than to milk protein percentage (r = - 0.47), results which agree with Wu and Huber (1994).
The MUN concentration among cows ranged from 5.8 to 27.6 mg/dl, with a mean value of 12.4 ± 3.1 mg/dl per cow (n = 294), and the PUN concentration ranged from 5.9 to 33.5 mg/dl with a mean of 13.4 ± 4.0 mg/dl (n = 284). The correlation between MUN and PUN values was 0.86, close to the correlation of 0.88 observed by Roseler et al (1993).
In this study, when MUN was regressed against PUN, a linear relationship was determined, with an intercept of 3.62 mg/dl (intercept > 0; P < 0.05)( Figure 1). An intercept of 3.62 mg/dl for MUN implies that this amount of urea may originate from amino acid metabolism in the mammary secretory cells. However, high correlation between MUN and PUN found in other studies (Baker et al., 1995; Gustafsson and Palmquist, 1993; Oltner et al., 1985; Roseler et al., 1993) suggests that urea in the blood system is the major source of urea nitrogen in milk. Several factors should be considered when interpreting urea nitrogen data. These include such factors as diurnal variations and the lag time between the peak level of the BUN to that of MUN (Gustafsson and Palmquist, 1993), the different permeabilities of the mammary duct tissue (Linzell and Peaker, 1971), and the differences in specific gravity between various components of milk solids that could alter the relative concentration of urea in milk (Roseler et al., 1993).
Both MUN and PUN were affected by the dietary composition and intake of nutritional components. Both dietary FA content and intake were negatively correlated to MUN and PUN concentrations. The squared term for dietary FA content in the diet accounted for the greatest amount of the variation (partial R2 = 0.16) among nutrient composition variables for MUN concentration. The intake of FA accounted for 11% and 20% of the variation in MUN and PUN concentrations, respectively.
Dietary CP content and intake were found to affect both MUN and PUN; dietary concentration and intake of CP were positively correlated to both MUN and PUN. The CP content in the ration accounted for 11% of the variability observed in the MUN, and CP intake accounted for 16% of variation in MUN concentration.
The analysis for the effect of separate sampling time, i.e., morning versus afternoon ( n = 100), for MUN showed no significant difference in MUN levels despite morning milk yield being higher than that for afternoon milking. These results agree with Gustafsson and Palmquist's (1993) conclusion that the concentrations of MUN from samples collected at any time of the day may be reliable, provided the lag time between feeding and actual sampling is taken into consideration.
When evaluating the nutritional impact on various milk production characteristics, nutrient intake was observed to be more important than dietary composition. Milk protein yield was found to be more highly correlated to nutrient intake than to nutrient composition. These results showed that although it is difficult to improve milk protein percentage through feeding, milk protein yield proved to be quite responsive to dietary differences among the farms. With the current focus on the value of milk protein, it is important and more profitable for farmers to feed cows for higher milk production, rather than attempting to feed for higher milk protein percentage.
Our results support that both MUN and PUN can be used as indices for monitoring the feeding programs on dairy farms. Because of difficulties associated with collection of blood samples under field conditions compared to collection of milk samples, and the fact that MUN and PUN are highly correlated, MUN values can be used to access the utilization of protein and energy by ruminal microbes.
Baker, L. D., J. D. Ferguson, and W. Chalupa. 1995. Responses in urea and true protein of milk to different protein feeding schemes for dairy cows. J. Dairy Sci. 78:2424.
Butler, W. R., D. J. R. Cherney, and C. C. Elrod. 1995. Milk urea nitrogen (MUN) analysis: Field trial results on conception rates and dietary inputs. Page 89 in Proc. Cornell Nutrition Conference for Feed Manufacturers. Cornell Univ., Ithaca, NY.
Cant, J. P., E. J. DePeters, and R. L. Baldwin. 1993. Mammary amino acid utilization in dairy cows fed fats and its relationship to milk protein depression. J. Dairy Sci. 76:762.
Edwards, C. M. 1991. Characterizing milk production and estimating dry matter intake by calving season. M.S. Thesis. The Ohio State University, Columbus, OH.
Firkins, J. L., and M. L. Eastridge. 1992. Replacement of forage or concentrate with combination of soyhulls, sodium bicarbonate, or fat for lactating dairy cows. J. Dairy Sci. 75:2752.
Gustafsson, A. H. and D. L. Palmquist. 1993. Diurnal variation of rumen ammonia, serum urea, and milk urea in dairy cows at high and low yields. J. Dairy Sci. 76:475.
Linzell, J. L. and M. Peaker. 1971. The permeability of mammary ducts. J. Physiol. (Camb.) 216:701.
National Research Council. 1989. Nutrient requirements of dairy cattle. 6th Rev. Ed. Natl. Acad. Press, Washington, DC.
Oltner, R., M. Emanuelson, and H. Wiktorsson. 1985. Urea concentration in milk in relation to milk yield, live weight, lactation number, and amount and composition of feed given to dairy cows. Lives. Prod. Sci. 12:47.
Roseler, D. K., J. D. Ferguson, C. J. Sniffen, and J. Herrema. 1993. Dietary protein degradability effect on plasma and milk urea nitrogen and milk nonprotein nitrogen in Holstein cows. J. Dairy Sci. 76:525.
SAS/Stat® User's Guide. 1988. Release 6.03. SAS Inst., Inc., Cary, NC.
Schingoethe, D. J. 1996. Dietary influence on protein level in milk and milk yield in cows. Anim. Feed Sci. Technol. 60:181.
Wu, Z. and J. T. Huber. 1994. Relationship between dietary fat supplementation and milk protein concentration in lactating cows. A review. Livest. Prod. Sci. 39:141.