| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Puyallup Research and Extension Center, Washington State University, Puyallup 98371
2 Department of Animal Sciences, University of California, Davis 95616
3 Department of Dairy and Animal Science, The Pennsylvania State University, University Park 16802
4 Department of Animal Sciences, Ohio Agriculture Research and Development Center, The Ohio State University, Wooster 44691
5 Department of Animal Sciences, The Ohio State University, Columbus 43210
6 Department of Animal Sciences, Washington State University, Pullman 99163
7 Church & Dwight Co. Inc., Princeton, NJ 08543
Corresponding author: Joseph Harrison; e-mail: jhharrison{at}wsu.edu.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: manure • nitrogen • phosphorus • potassium
Abbreviation key: ASAE = American Society of Agricultural Engineers, DMD = DM digestibility, DME = manure DM excretion, KE = K excretion, ME = manure excretion, MF = milk fat percentage, MILK = milk yield, MilkP = P in milk, MTP = milk true protein, NE = N excretion, PE = P excretion.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Estimates for manure and nutrient excretion by dairy cattle are found in the American Society of Agricultural Engineers (ASAE) Standard D384.1 (ASAE, 2001). These estimates are limited in their utility and accuracy as they are based on data from the late 1960s and early 1970s, and were taken from a data set that was of limited known origin. In addition, the ASAE standard was revised in 1988 to merge a dairy heifer column with a dairy cow column for one single column for all dairy cattle categories.
Recent reports of excretion data from dairy cows were compiled from lactating Holstein dairy cows producing an average of 20.3 and 29 kg of milk/d (Tomlinson et al., 1996; Wilkerson et al., 1997). Today, many dairy cattle are producing milk at twice those levels. Contemporary manure and nutrient excretion estimates are needed to more precisely predict excretion from higher producing cows. Most important, equations need to be developed to reflect the relationship between milk production and manure or nutrient excretion.
Previous evaluations of manure excretion from dairy cattle indicated that the ASAE (2001) manure excretion estimates underestimated excretion from high-producing cows (Tomlinson et al., 1996) and that using only BW is not an accurate method of predicting manure or nutrient excretion (James et al., 1999). As production and management of dairy cattle have changed in recent decades, changes may also have occurred in manure and nutrient excretion. In addition, excretion estimates for calves and heifers have not been published in recent years. Feed consumption (quality and quantity) differs from young calves to springing heifers. It is reasonable to assume that BW alone is not a good predictor of manure and nutrient excretion for all heifer categories. As an example, dairies with high culling rates and approximately equal number of replacement heifers as lactating cows need to have precise estimates for manure and nutrient excretion to adequately establish a nutrient plan. Furthermore, heifer data are essential for replacement heifer operations. Technical assistance providers, dairy operators, and staff from regulatory agencies are seeking site-specific information on manure volume and nutrient content to more precisely develop nutrient management plans and design adequate manure storage systems. Updated information is critical to owners of animals residing in environmentally sensitive areas.
In 2001, a committee was developed by the ASAE Structures and Environment Committee 412 and members from the Federation of Animal Science Societies to revise the ASAE manure excretion values using data from contemporary diets and levels of productivity (ASAE Standard Tables D384.1). The committee structure consisted of animal species subgroups, which included dairy cattle. The outcomes of the dairy subgroup’s efforts are reported in this paper. The overall goal of the dairy subgroup’s effort was to evaluate data collected from studies with contemporary diets that represented a broad geographical context and included data from cows milking >40 kg of milk per day. The primary objectives of the dairy subgroup were to develop regression equations to predict manure, DM, and nutrient excretion of calves, heifers, and nonlactating and lactating Holstein dairy cows, and to document nutritional parameters associated with manure and nutrient excretion.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
). The overall data set included records from a wide variety of animal ages, ranging from calves to multiparous lactating cows. Data were categorized into 4 groups: lactating cows (LACT), dry cows (DRY), heifers (HEIFER), and calves (CALF). The LACT data set (554 cows or cow-periods from Latin square experiments) included multiparous lactating Holstein cows. The DRY data set (18 cow-periods) animals were defined as multiparous, nonlactating cows that were pregnant. The HEIFER data set (60 animal-periods) included female, nonlactating animals that had not calved and included animals of various ages weighing >250 kg. The CALF data set (46 animal-periods) included animals weighing <250 kg (Table 2).
|
|
The independent variables included in the LACT data set were BW, DIM, DMI, DM digestibility (DMD), milk yield (MILK), percentage milk fat (MF), and percentage milk protein (Table 3). Milk CP values in the data set were converted to milk true protein (MTP) values using a conversion factor of 0.9345 (Mackle et al., 1999). Dietary ingredients and characteristics were used as additional prediction factors for equations, including dietary concentrations of CP, NDF, P, and K. Dependent variables included in the LACT data set were (Table 3) manure excretion (ME), DM excretion (DME), N excretion (NE), P excretion (PE), and K excretion (KE). Dry matter excretion included both fecal and urinary DM and was determined by adding actual fecal DM and 4.5% of urinary excretion.
|
Data on P and K intake and excretion for lactating animals were only available for a subset of animals in the LACT data set. The MINERAL data set (85 cow-periods) included cows for which excretion of feces and urine were known. Intakes of minerals were determined through analyses of both feed and orts. One study of early lactation cows (Johnson et al., 1998, experiment 2; 15 cow-periods ranging from 16 to 61 DIM) was not included in the MINERAL data set due to negative P and K balances for the early lactation animals.
Diets fed during the metabolism trials included a wide variety of protein supplements and forage types. Forages included corn silage, grass silage, alfalfa silage, and grass hay. The remainder of the diets included various grains, by-product feeds, and mineral supplements. Cows in these trials were fed ad libitum.
Equations given for each parameter include residual standard error (SE) and interstudy SE. Equations with lower SE are expected to provide a more precise estimation of excretion and should be used when values for the input variables are available.
Sample Collection and Analyses
Contemporary data available for use in our study were predominantly from lactating cows. Few scientists have completed total collection studies on dry cows, replacement heifers, or young calves. Total collection metabolism studies (16 studies) conducted at Washington State University included both lactating cows (399 cow-periods) and dry cows (7 cow-periods). Feeding, sample collection, and sample analyses were conducted by methods outlined by Johnson et al. (1998) and Timmermans et al. (2000). Metabolism studies from the University of California, Davis included 3 calf studies, 3 heifer studies, 2 dry cow studies, and 4 lactating cow studies. Feeding and sample collections were described by James et al. (1999) and Meyer et al. (2000). Methods used for collection during the metabolism studies (6 studies; 139 cows or cow-periods) with lactating multiparous Holstein cows at The Ohio State University were reviewed by Weiss and Wyatt (2004). Studies from The Pennsylvania State University that included 32 observations from weaned calves and 32 observations from heifers are summarized by Gabler and Heinrichs (2003a, b). Minerals in feces were analyzed by the University of Nebraska Soil and Plant Analytical Laboratory (Lincoln, NE) and urine minerals were analyzed by Dairyland Laboratories (Arcadia, WI).
Mineral analyses were not conducted on milk samples. Therefore, milk mineral contents were assumed equivalent to values outlined in the 2001 Dairy NRC. Milk P and K were estimated at 0.9 and 1.5 g/kg of milk, respectively (NRC, 2001).
Statistical Analyses
Regression analyses were performed using PROC MIXED of SAS (SAS Institute, 1999) with the discrete effect of study included as a random variable (St-Pierre, 2001). Equations were developed by running multiple iterations in MIXED and removing the least significant effect at each iteration. For data sets equal to or greater than 200 observations, variables were kept if P < 0.10. For data sets with less than 200 observations, variables were kept if P < 0.25. Adjusted observations were calculated for graphing purposes by adding the residual from each individual observation to the predicted value of the study regression (St-Pierre, 2001).
Equation evaluation was done by regressing residuals (predicted values subtracted from observed values) on the predicted values (St-Pierre, 2003). Predicted values were centered by subtracting the mean of all predicted values from each prediction. This makes the slope and intercept estimates in the regression orthogonal and, thus, independent. Mean biases were assessed using the intercepts of the regression equations, and the slopes of the regression equations were used to determine the presence of linear biases.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The historic ASAE (2001) table listed excretion values on a basis of 1000 kg of BW. The newly adopted table values (Table 4), based on prediction equations described in the following narrative, provide excretion estimates on a per-animal basis and include descriptions of animals and dietary assumptions used to develop the values. Table values, common in historic ASAE standards, were generated with corresponding nutritional parameters to provide an average value for predicting manure excretion. In addition, prediction equations were developed for each of the excretion parameters when enough data were present. The goal of including the prediction equations was to provide excretion estimates that are adaptable to particular operations in lieu of a general table value that is used by all dairy operations, regardless of the production level, breed, or size of the animals.
|
 | ([1]) |
The simple equation with MILK provides flexibility not previously given in the ASAE standards. The ASAE (2001) value for ME was listed as 86 kg/d per 1000 kg of BW (54.2 kg/d for a 630-kg cow). Using equation [1], ME would average 58.5 and 70.8 kg/d for cows producing 15 and 40 kg/d, respectively. The lower ME estimates given in the 2001 ASAE tables are a result of lower milk production and DMI of cows used to generate those values.
Body weight was a predictor of ME (P < 0.01), and similar findings were reported in equations developed by Wilkerson et al. (1997). When comparisons were made between equations including either MILK or BW, equations with MILK improved the prediction across studies by 18% when predicting ME. The less accurate relationship of BW to ME in the prediction equations indicated that basing ME on BW, as was done in 2001 ASAE, is not the most accurate method of predicting ME from lactating cows.
Inclusion of DMI in equation [2] provided a more precise estimation of ME than equations that did not include DMI. The best single independent variable for predicting ME in the LACT data set was DMI [2].
 | ([2]) |
Figure 1 shows the relationship between DMI and ME. Including DMI as an independent variable improved the precision of the estimation by reducing the residual SE by 28.9% as compared with using MILK. Inclusion of DMI in the equations has become a more realistic option as producers have improved record-keeping skills and increased the use of scales.
|
Most of the nonlinear models evaluated resulted in prediction equations that were less accurate predictors of ME than linear equations. The best nonlinear equation for describing the LACT data set included several of the same independent variables given by Wilkerson et al. (1997) (BW, DMI, DIM, dietary CP, and dietary NDF) and included the interaction of DMI and dietary CP as well as squared terms of DIM and dietary NDF. The nonlinear equation improved the residual SE of equation [2] by only 1%. Because the nonlinear equations provided only a very slight improvement over the linear equations, we suggest the use of the linear equations for predicting ME.
In addition to the development of new prediction equations using the LACT data set, previously published equations for predicting ME were evaluated. Predictions using the equation of Wilkerson et al. (1997) [ME = (0.0286 x BW) + (0.0378 x DIM) + (1.0689 x MILK) + (9.67 x Dietary CP, g/g of DM) + (61.4 x Dietary NDF, g/g of DM) – 21.94] resulted in mean and linear biases (P < 0.01) of 5.6 kg/d and – 0.25, respectively. Although ME in the LACT data set was greater than that predicted using the Wilkerson et al. (1997) equation, the difference was less than the standard deviation of the residuals. The comparison of previous equations with the LACT data set provides an evaluation of those equations using this data set. An independent data set would be required to compare the accuracy of previous and new equations.
DM excretion.
The average fecal DM excretion was 7.3 kg/d and DME was 8.5 kg/d (Table 5). Tomlinson et al. (1996) reported fecal DM excretion ranging from 6.2 to 7.4 kg/d, values similar to those in the LACT data set. The best predictor of DME was DMI [3].
|
 | ([3]) |
Equation [3] indicated a direct relationship between DMI and DME. In 1994, Van Horn et al. reported that DME could be determined by multiplying DMI by 0.33 and adding the urine DM.
Other variables evaluated for predicting DME were BW, DIM, MILK, MF, and MTP, but these variables were not significant (P > 0.25) when included in the equation with DMI. Conversely, dietary NDF concentration was a significant variable (P < 0.01) when included in an equation with DMI, but resulted in a less precise prediction equation across studies than when DMI was the only independent variable for predicting DME.
Two equations were developed for predicting DME in the absence of DMI:
 | ([4]) |
 | ([5]) |
Equation [4] provides a prediction of DME based solely on MILK, and equation [5] includes BW, MTP, and MILK. Predictions of DME using equations [4] and [5] are expected to be less accurate than predictions using equation [3], but provide estimates in instances where DMI is not known.
Nitrogen excretion.
The simple linear equation, using MILK as the only independent variable [6], indicated a positive relationship (P < 0.01) between NE and MILK. When MILK was used as the only prediction variable, it resulted in a less precise prediction than subsequent equations evaluated, but using MILK as the only variable resulted in a 2.6% improvement in accuracy compared with use of BW to predict NE.
 | ([6]) |
Equations were also developed for situations where intake of CP is known. When evaluated in a simple equation, CP intake was positively related (P < 0.01) to NE [7].
 | ([7]) |
As expected, an increase in CP consumption resulted in greater NE. The most precise equation developed for predicting NE included CP intake as an independent variable [7]. Equation [7] improved the residual SE by 27.5% compared with equation [6]. The direct relationship between N intake and NE indicates that future improvements in balancing diets to better meet the specific amino acid needs of the animal while decreasing dietary CP concentrations may be an important step in decreasing NE (Harrison et al., 2002).
Nitrogen intake was directly related to NE in previous experiments (Tomlinson et al., 1996; James et al., 1999; Krober et al., 2000; Frank et al., 2002). Excess intake N is mostly excreted via urinary excretion. Tomlinson et al. (1996) indicated that NE was closely related to N intake and DMI and somewhat related to BW, whereas Van Horn et al. (1994) stated that NE could be estimated by subtracting the N in milk from N intake.
Quadratic models were evaluated to determine if the predictions were improved with the addition of squared terms and interactions in the model. When quadratic models were evaluated, the resulting equations did not reduce the residual SE or interstudy SE compared with the linear models. Conversely, Wilkerson et al. (1997) reported that development of quadratic models led to a statistical improvement over the linear models developed for predicting NE. These authors, however, did not account for the imbalance of the predictor variables across studies (the random study effect in the model) and thus, might have induced the apparent nonlinearity of the prediction.
Previously published equations for predicting NE (Wilkerson et al., 1997) were evaluated using the LACT data set. Evaluation of the linear equation published by Wilkerson et al. in 1997 [NE = (0.000232 x BW) + (0.000342 x DIM) + (0.00649 x MILK) + (1.83 x Dietary CP, g/g of DM) + (0.280 x Dietary NDF, g/g of DM) –0.440] resulted in a mean bias of 37.4 g/d (P < 0.01) of excreted N and a linear bias of –0.264 (P < 0.01). The standard deviation of the residuals of the LACT data set was 0.0784 kg/d, indicating that the mean bias was less than the variation expected between studies.
Phosphorus excretion.
Dietary P concentrations in the MINERAL data set averaged 0.0044 g/g of DM. Many of the studies in the MINERAL data set were conducted before reduced P feeding was emphasized in dairy diets, which resulted in dietary P concentrations greater than needed to meet animal requirements. Because cows were fed diets with P concentrations greater than their requirements, equations developed with the MINERAL data set may not accurately account for diets with P supply at or below animal requirements.
The 2001 ASAE estimate of PE for a 630-kg cow was 0.0592 kg/d, 0.0147 kg/d less than the average PE for cows in the MINERAL data set (Table 5
). In contrast, an average PE of 0.057 kg/d was reported by Weiss and Wyatt (2004), which was similar to the 2001 ASAE value. However, the P intake in the MINERAL data set averaged 0.013 kg/d more than the cows in the Weiss and Wyatt (2004) study, and would account for most of the difference in PE.
The simple equation [8] developed using MILK as the only predictor of PE indicated a positive relationship between MILK and PE.
 | ([8]) |
The positive relationship between MILK and PE is most likely a result of greater intakes of high-producing cows. Although MILK may be used for predicting PE, accuracy of predictions increased when P intake was included in the equations. Development of a simple equation with P intake as the only predictor reduced the residual SE by 16% [9].
 | ([9]) |
In this data set, P intake was the best single independent variable for predicting PE (Figure 2). Similarly, Beede and Davidson (1999) and Weiss and Wyatt (2004) found that P intake was the most important single factor in determining PE. When an additional equation that included MILK and MTP was developed for predicting PE, there was a reduction in the residual SE from 9.7 to 9.3. However, the precision of future predictions across studies was not improved with the addition of MILK and MTP to the equation.
|
 | ([10]) |
Subtracting MilkP from P intake to estimate PE assumes there is no tissue mobilization or retention. In the MINERAL data set, evaluation of [10] resulted in a linear bias (–0.409, P < 0.01), but no mean bias (P > 0.22). In 2004, Weiss and Wyatt proposed using [11] to estimate PE.
 | ([11]) |
Evaluation of equation [11] indicated a mean bias (0.0155 kg/d, P < 0.02) but no linear bias (P > 0.26) when evaluated using the MINERAL data set. The standard deviation of the residuals for the MINERAL data set was 0.0132 kg/d.
Potassium excretion.
Potassium excretion occurs mainly in urine, with some unabsorbed K excreted in feces (NRC, 2001). Total KE in the MINERAL data set averaged 0.200 kg/d with urinary K excretion accounting for approximately 75% of KE. The 2001 ASAE standards estimated KE to be 0.01827 kg/d.
Potassium excretion was directly related to both MILK [12] and dietary K concentration [13]. When MILK was evaluated as the only factor to predict KE, a positive relationship between MILK and KE was found.
 | ([12]) |
Milk production has been reported to have a curvilinear relationship to K intake, with the peak milk yield occurring at a dietary K concentration of 0.015 g/g of DM (NRC, 2001). Conversely, inclusion of squared terms and interactions did not improve the models for predicting KE in the MINERAL data set. The lack of a curvilinear relationship in our data set was most likely due to the low dietary K concentrations in the MINERAL data set (0.0129 g/g).
Future prediction of KE is expected to be more accurate if dietary K concentrations or K intakes are used to predict excretion. The best equation for prediction of KE included DMI and dietary K [13].
 | ([13]) |
When DMI and dietary K concentration were included in the equation, the SE between the studies in the data set was very small compared with equations that included other variables for predicting KE.
Nutrient excretion in early lactation cows.
Cows (15 cow-periods) from an early lactation study (average of 38 DIM) were evaluated separately due to the negative balances of N, P, and K for these cows. The average N balance for the early lactation cows was –0.437 kg/d. When the early lactation cows were evaluated using [7], the cows were excreting an average of 0.203 kg/d more N then predicted by equation [7]. However, evaluation of these cows with equation [6], in which MILK as the only independent variable, only underestimated N excretion for the early lactation cows by an average of 0.040 kg/d.
The early lactation cows were not included in the MINERAL data set because of the greater excretion of P and K for the early lactation animals compared with cows in later lactation. For early lactation cows, P intake was not a significant factor (P > 0.25) to predict PE, and PE was not related to MILK (P > 0.25), DMI (P > 0.25), or Ca intake (P > 0.16). Phosphorus excretion of these cows averaged over 0.023 kg/d more than would be expected based on P intake and MILK. The greater PE for these early lactation animals is most likely a result of greater endogenous fecal P losses, possibly related to bone mobilization.
Potassium excretion of cows in the early lactation data set was greater than KE of cows in the MINERAL data set. On average, KE for early lactation cows was 0.143 kg/d greater than cows in the MINERAL data set, even though K intakes were only 0.008 kg/d greater (Table 2). Due to the greater KE and the greater secretion of K in milk, early lactation cows were in negative K balance. Silanikove et al. (1997) found that cows in early lactation are often in negative K balance and suggested that increased amounts of K in the diet may be beneficial to milk production.
Dry Cows
The DRY data set was a small data set and consisted of 18 cows. Of these dry cows, 15 cows were fed diets specifically formulated for dry cows and 3 cows were fed diets formulated for lactating cows. Due to the limited data set, prediction equations were not developed for dry cows and only average intake and excretion values are reported.
Manure excretion from dry cows averaged 38.6 kg/d (Table 5), which was 12.3 kg/d more than reported by Wilkerson et al. (1997), but similar to the 36.3 kg/d estimate by Van Horn et al. (1994). The lower ME values from Wilkerson et al. (1997) are most likely a result of the restricted intakes for dry cows in their studies. The DME estimate of 4.5 kg/d for dry cows reported by Van Horn et al. (1994) was the same as the average value in the DRY data set.
Manure excretion for the DRY data set was 59% of the 2001 ASAE value for ME for a cow weighing 755 kg. However, the 2001 ASAE values do not differentiate between lactating and nonlactating animals.
Mean NE from dry cows was 0.228 kg/d, though the range in NE was large (Table 5), and was 0.049 kg/d greater than that reported by Wilkerson et al. (1997). Estimates of NE in the 2001 ASAE standards were 0.340 kg/d for dairy cattle and 0.257 kg/d for beef cattle. Clearly, estimates for dry cows are closer to the 2001 ASAE estimates for beef cattle than for dairy cattle. Addition of dry cows to the updated standards will be an improvement over ASAE 2001 values and will improve the flexibility and accuracy of the standards.
The DRY data set was used to evaluate linear equations for ME and NE published by Wilkerson et al. (1997). Dry cows were assumed 230 d pregnant because day of pregnancy was not available in the dry cow data set. When the Wilkerson et al. (1997) equation [ME = (0.00711 x BW) + (32.4 x Dietary CP, g/g of DM) + (25.9 x Dietary NDF, g/g of DM) + 8.05] was evaluated there were no mean or linear biases (P > 0.13) and the equation accurately described the cows in this small data set. Therefore, the Wilkerson et al. (1997) equation could be used for prediction of ME for dry cows. Conversely, evaluation of the NE equation [NE = (0.000107 x BW) + (1.11 x Dietary CP, g/g of DM) + (0.170 x Dietary NDF, g/g of DM) – 0.135] (Wilkerson et al., 1997) resulted in mean (0.2795 kg/d; P < 0.01) and linear biases (1.19; P < 0.01). The mean and linear biases indicate that updated regression equations need to be developed to adequately predict NE from dry cows. In the future, more research is needed on dry cows fed diets typical in the industry, to be able to develop regression equations.
Heifers
The data set for heifers included 60 observations ranging in BW from 274 to 613 kg (Table 2). The 2001 ASAE manure excretion estimates for dairy cattle were not categorized by animal age. Excretion estimates for heifers were not specifically available and would have to be approximated using either dairy cattle or beef cattle data. Determination of new prediction equations for growing dairy heifers was difficult due to a shortage of total collection metabolism trials recently conducted on this class of animals.
Manure excretion in the HEIFER data set was overestimated by ~54% for the average heifer (437 kg) using the 2001 ASAE dairy excretion estimates, but by only ~3% using the beef estimates. The most accurate equation for predicting ME for heifers included BW and DMI [14]. Manure excretion was dependent on BW of the animal alone [15], although the addition of DMI to the equation provided a more precise predictor of ME.
 | ([14]) |
 | ([15]) |
Nitrogen excretion from dairy heifers in the HEIFER data set averaged 0.1173 kg/d, which was 0.0804 and 0.0313 kg/d less than predicted by the 2001 ASAE dairy and beef values, respectively. An equation [16] developed for NE using the HEIFER data set increased NE when CP intake increased (Figure 3).
|
 | ([16]) |
Hoffman et al. (2001), James et al. (1999), and Wilkerson et al. (1997) also reported increased NE for heifers fed greater levels of N. Hoffman et al. (2001) found that growth of Holstein heifers was optimized when the dietary CP concentration was 0.13 g/g of DM. The average dietary CP concentration was 0.112 g/g of DM in the HEIFER data set (Table 2). Feeding heifers diets with a CP concentration of 0.13 g/g of DM would be expected to increase NE compared with the average in the HEIFER data set.
Development of prediction equations for PE in the HEIFER data set was not possible due to the limited data available and the variation within the data set.
Previously developed prediction equations for ME and NE in heifers were evaluated. When the equation for ME [ME = (0.0499 x BW) + (44.2 x Dietary CP, g/g of DM) + (5.86 x Dietary NDF, g/g of DM) – 5.918] (Wilkerson et al., 1997) was evaluated, there were no mean or linear biases, indicating that this previously published equation adequately described the heifers in this data set. Evaluation of a previously published (Wilkerson et al., 1997) NE equation [NE = (0.000471 x BW) + (0.867 x Dietary CP, g/g of DM) – (0.0109 x Dietary NDF, g/g of DM) – 0.159] from heifers resulted in no linear bias, but there was a trend (P < 0.07) toward a mean bias. The mean bias for the equation was 0.1212 kg/d, indicating an underprediction of NE from animals in the HEIFER data set.
Calves
The data set for calves included 46 observations ranging in BW from 86 to 205 kg (Table 2). Development of equations for excretion estimates from calves used a small data set because of a shortage of calf data from total collection metabolism studies. As with heifers, estimates of excretion for calves were not available in the 2001 ASAE manure excretion estimates for dairy cattle. Use of the 2001 ASAE manure excretion estimates for calves would not be expected to be accurate for estimating ME or nutrient excretion.
Average ME from the CALF data set was approximately half the ME of the HEIFER data set (12.1 kg/d less), even though the average calf BW was about one-third as much as the heifers (Tables 2 and 5). The most accurate equation for predicting ME of calves included DMI [17].
 | ([17]) |
Although DMI was the best predictor of ME for the CALF data set, BW was also a predictor of ME (P < 0.01) [18].
 | ([18]) |
Prediction of DME was similar to ME for calves. Equation [19] shows the relationship between DME and DMI for the CALF data set, with greater DME occurring as DMI increases.
 | ([19]) |
Nitrogen excretion was directly related to CP intake in the CALF data set [20] as it was for the other classes of dairy animals. In the CALF data set, the coefficient for CP intake was greater than was seen in the HEIFER data set (112.6 and 78.4, respectively) for the simple linear equation to predict NE (Figures 3 and 4).
|
 | ([20]) |
A relationship between P intake and PE was seen in the CALF data set (Figure 5) even though there was not a relationship (P > 0.25) in the HEIFER data set. Phosphorus excretion increased with greater P intakes in the CALF data set [21].
|
 | ([21]) |
Although there was not a direct relationship between PE and P intake in the HEIFER data set, when equation [21] was evaluated in the HEIFER data set, there were no mean or linear biases. Because [21] accurately predicted the PE from HEIFERS, this equation could be used for nonlactating heifers regardless of BW.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication December 22, 2004. Accepted for publication June 14, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. Huhtanen, J. I. Nousiainen, M. Rinne, K. Kytola, and H. Khalili Utilization and Partition of Dietary Nitrogen in Dairy Cows Fed Grass Silage-Based Diets J Dairy Sci, September 1, 2008; 91(9): 3589 - 3599. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Capper, E. Castaneda-Gutierrez, R. A. Cady, and D. E. Bauman The environmental impact of recombinant bovine somatotropin (rbST) use in dairy production PNAS, July 15, 2008; 105(28): 9668 - 9673. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Hollmann, K. F. Knowlton, and M. D. Hanigan Evaluation of Solids, Nitrogen, and Phosphorus Excretion Models for Lactating Dairy Cows J Dairy Sci, March 1, 2008; 91(3): 1245 - 1257. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Powell, G. A. Broderick, and T. H. Misselbrook Seasonal Diet Affects Ammonia Emissions from Tie-Stall Dairy Barns J Dairy Sci, February 1, 2008; 91(2): 857 - 869. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Powell, T. H. Misselbrook, and M. D. Casler Season and Bedding Impacts on Ammonia Emissions from Tie-stall Dairy Barns J. Environ. Qual., January 4, 2008; 37(1): 7 - 15. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. L. Moody, G. I. Zanton, J. M. Daubert, and A. J. Heinrichs Nutrient Utilization of Differing Forage-to-Concentrate Ratios by Growing Holstein Heifers J Dairy Sci, December 1, 2007; 90(12): 5580 - 5586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. R. Hill, K. F. Knowlton, R. E. James, R. E. Pearson, G. L. Bethard, and K. J. Pence Nitrogen and Phosphorus Retention and Excretion in Late-Gestation Dairy Heifers J Dairy Sci, December 1, 2007; 90(12): 5634 - 5642. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Yan, J. P. Frost, T. W. J. Keady, R. E. Agnew, and C. S. Mayne Prediction of nitrogen excretion in feces and urine of beef cattle offered diets containing grass silage J Anim Sci, August 1, 2007; 85(8): 1982 - 1989. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. M. Powell, D. B. Jackson-Smith, D. F. McCrory, H. Saam, and M. Mariola Nutrient Management Behavior on Wisconsin Dairy Farms Agron. J., January 1, 2007; 99(1): 211 - 219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Yan, J. P. Frost, R. E. Agnew, R. C. Binnie, and C. S. Mayne Relationships among manure nitrogen output and dietary and animal factors in lactating dairy cows. J Dairy Sci, October 1, 2006; 89(10): 3981 - 3991. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M. Powell, D. B. Jackson-Smith, D. F. McCrory, H. Saam, and M. Mariola Validation of feed and manure data collected on Wisconsin dairy farms. J Dairy Sci, June 1, 2006; 89(6): 2268 - 2278. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. E. Cabrera, A. de Vries, and P. E. Hildebrand Prediction of Nitrogen Excretion in Dairy Farms Located in North Florida: A Comparison of Three Models J Dairy Sci, May 1, 2006; 89(5): 1830 - 1841. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||
