BEEF SPECIES SYMPOSIUM: Difficulties associated with predicting forage intake by grazing beef cows S. W. Coleman, S. A. Gunter, J. E. Sprinkle and J. P. S. Neel J ANIM SCI 2014, 92:2775-2784. doi: 10.2527/jas.2013-7090 originally published online January 7, 2014
The online version of this article, along with updated information and services, is located on the World Wide Web at: http://www.journalofanimalscience.org/content/92/7/2775
www.asas.org
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Beef Species Symposium: Difficulties associated with predicting forage intake by grazing beef cows1,2 S. W. Coleman,* S. A. Gunter,†3 J. E. Sprinkle,‡ and J. P. S. Neel* *USDA-ARS, Grazinglands Research Laboratory, El Reno, OK 73801; †USDA-ARS, Southern Plains Range Research Station, Woodward, OK 73801; and ‡ Department of Animal Sciences, University of Arizona, Tucson 85721
ABSTRACT: The current NRC model to estimate DMI is based on a single equation related to metabolic size and net energy density of the diet; this equation was a significant improvement over previous models. However, observed DMI by grazing animals can be conceptualized by a function that includes animal demand, largely determined by metabolic or linear size, physiological state, genetics, or any combination of these. Even in the database used to generate the current NRC equation, DMI by cows is poorly predicted at the extremes. In fact, across a wide range of actual DMI, predicted DMI is rather flat, indicating an insensitivity of prediction, so the model requires further refinement. A broad-based database was developed that includes pasture and confinement studies with growing, nonlactating, and lactating cattle. New equations are presented
for consideration in the new model. It was found that the premise behind earlier NRC equations based on diet digestibility and BW are sound but that for cows, additional drivers based on milk production or calf performance were stronger than BW. Future models should be based on multiple variables, including functions for physiological state, animal suitability to the environment, and activity to modify the predicted DMI. Further, the model could possibly account for imbalances of protein to energy, particularly as they relate to ruminal function. Further, the issue of how reference data were collected (pen vs. pasture) and how the methods or constraints influence DMI must be evaluated. Overall, the new NRC model needs to be more robust in its ability to account for the wide variation in the environment, dietary characteristics, and metabolic demands.
Key words: beef, cattle, forage intake, pasture, rangelands © 2014 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2014.92:2775–2784 doi:10.2527/jas2013-7090 INTRODUCTION Predicting intake of grazing animals is a worthy goal, but difficulties abound with the impossibility of measuring and the difficulty of estimating the intake of grazing animals (Coleman et al., 1999). In their review, Coleman et al. (1999) noted that intake varies due to both feed quality and physical characteristics and also due to the animal’s physiological state. Therefore, any attempt to predict or model intake must take in both considerations 1Based on a presentation at the Beef Species Symposium titled “Nutrient Requirements of the Beef Female in Extensive Grazing Systems: Considerations for Revising the Beef NRC” at the Joint Annual Meeting, July 8–12, 2013, Indianapolis, IN. 2Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA. 3Corresponding author: [email protected] Received August 29, 2013. Accepted November 25, 2013.
and, at the very best, provide an estimate of the average intake under the specified conditions. The NRC committee included a model for intake in the NRC (1984, 1996) publication and, because of limitations, later modified the model (NRC, 2000). Major limitations of the model included the fact that all the experimental data that were included were from pen-fed animals, many on highconcentrate diets. Observational field data and peer-reviewed research have identified deficiencies in the 1996 and 2000 Nutrient Requirements of Beef Cattle (NRC, 1996, 2000) with respect to maintenance requirements and forage intake on rangelands. Lardy et al. (2004) reported that predicted energy deficits for summer-calving cows grazing either range or subirrigated meadow in November were “biologically unreasonable.” They further recommended that the “on pasture” function in the nutritional model “not be used because it unreasonably increases energy requirements.” In most instances, cow performance field data agree with the findings of Lardy
2775 Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2776
Coleman et al.
et al. (2004), although there have been some studies where predicted forage intake for dormant forage exceeded estimated forage intake indirectly measured with external markers (Cr marker calculated; Bodine and Purvis, 2003). THE NRC MODEL BASIS AND ASSUMPTIONS The forage intake models presented in NRC (1984, 1996, 2000) are based on the California NE System of Lofgreen and Garrett (1968) with various adjustments to the model for breed, physiological state, activity, heat loss or gain, milk production, and grazing forage allowance. The NEm calculation and the adjustment for activity (HiE) are problematic for managers of grazing cattle in a rangeland setting because the calculated values for NEm were originally generated with mostly idle animals in either a calorimeter or a feedlot. As the NRC (2000, p. 6) model states, “This expression was derived using data from, primarily, growing steers and heifers of British ancestry that were penned in generally non-stressful environments.” Further, the NRC (2000, p. 88) model only used intake data collected from growing and finishing animals in the feedlot. There were 3 validations of the NEm intake model with data from cattle consuming forage diets. Two validations were with grazing cattle, 1 in New Mexico (Funk et al., 1987; Krysl et al., 1987; Pordomingo et al., 1991) and 1 in Oklahoma (Gunter, 1993; Gunter et al., 1993). The only validation for mature cows was with nonlactating cows fed grass hay (Vona et al., 1984). The 2 main problems associated with using the NEm system in rangeland settings are, first, failure to adequately account for plasticity or adaptations that often occur with producing range cows as high-energy demand visceral tissues (Ferrell, 1988) shrink and expand in response to nutrient availability and demands (Forbes, 1986) in a much greater fashion than observed in feedlot systems and, second, failure to accurately assess activity levels and nutrient demands of cattle grazing rangelands, thus generating unreasonable maintenance requirements, especially with dormant forage. The objective of this paper is to evaluate the expansion of intake data that were used to develop relationships and to compare those relationships to the existing NRC equations. The expanded data included many grazing trials in which intake was estimated with various marker systems. DATABASE DEVELOPMENT A database was compiled from the literature with 50 studies containing a total of 482 group means, conducted both on pasture and in confinement, largely with all or very high forage diets. The only assumption made was that when forage intake was explicitly characterized as DM, then it was transformed to OM intake (OMI) by
Table 1. Summary of variables compiled in the database before any truncation Variable Forage CP, % Forage digestibility, % Initial BW, kg Average BW, kg ADG, kg Milk production, kg/d Forage OM intake, kg/d Total OM intake, kg/d Digestible DMI, kg/d DMI, % of BW Calf weaning wt, kg Calf ADG, kg Calf forage DMI, kg/d
n 326 448 319 484 180 160 482 482 448 482 142 128 55
Mean SD Minimum Maximum 11.5787 5.7269 1.3 28.5 59.8023 11.1053 30.7 84.8 419.9463 130.9662 146 669 414.8421 121.9692 146 675 0.2911 0.5753 -1.31 2.13 6.6100 3.2118 0.38 21.20 8.9710 3.8554 0.78 21.64 9.2786 3.7715 0.78 21.64 5.5183 2.5582 0.44 13.97 2.2239 0.6910 0.21 4.55 200.4652 38.8218 133.8 317.2 0.7256 0.1944 0.39 1.67 1.9189 1.2445 0.12 5.50
multiplying forage DMI by 0.92. These data are summarized in Table 1. Some of the animals from pasture research received supplemental feed. In those instances where cattle were supplemented, dietary digestibility (D) and CP concentration were assumed to be the same as that reported for forage samples; hence, no adjustment in diet nutritive value was made for supplement, but intake was recorded as the total forage and supplement. Although this procedure ignored potential associative effects (positive for increased DIP and negative for starch), the major objective of this project was to determine the feasibility of using data from grazing trials and compare it with the existing model, not to develop a final model to be used in the new NRC to be published. Table 2 presents OMI means delineated by method of intake measurement (n = 5; directly measured in confinement, Cr, Yb, total fecal collection, and alkanes) and physiological state (n = 3; growing, lactating, and nonlactating). Some steers were described as “mature” but were included with the “growing” animals. It was concluded that the maintenance machinery of producing cows was significantly different than that of growing animals and including mature steers with growing animals would stretch the inference space for that group. Growing animals were included in the analysis to test conclusions and comparisons with previous NRC (1996, 2000) models on animals more closely related to those used to develop the models. That premise should give us a relative basis to evaluate methods used to estimate OMI of cattle grazing pasture. The reasoning is that if data from growing animals grazing pasture fit the model but data from cows did not, then we could attribute the majority of the lack of fit to physiological state rather than methodology. In contrast, if pasture-based growing animals could not be predicted with the NRC (1996, 2000) models, then methodology would be suspect, although grazing animals of the same class may have different OMI demand than confined animals. Although all classes were
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2777
Forage intake by grazing beef cows
Table 2. Mean intake as affected by method of estimate and animal physiological state (n = 419)1 Growing Method Direct Fecal bags Alkanes Cr2O3 YbCl3
kg/d 7.2 ± 0.45 6.6 ± 1.60 — 7.5 ± 2.60 7.4 ± 2.56
Dry cows g·kg-1 BW·d-1 18.0 ± 2.63 19.0 ± 1.79 — 23.9 ± 6.17 19.5 ± 3.67
kg/d 8.7 ± 0.64 3.7 ± 0.28 12.8 ± 2.31 7.2 ± 1.48 11.7 ± 2.11
g·kg-1 BW·d-1 16.6 ± 1.26 12.3 ± 0.50 25.3 ± 3.65 14.5 ± 3.43 26.3 ± 6.46
Lactating cows kg/d g·kg-1 BW·d-1 11.9 ± 1.57 24.2 ± 3.94 — — 15.8 ± 0.83 32.6 ± 1.32 9.6 ± 2.03 21.4 ± 3.62 11.9 ± 2.10 29.0 ± 1.11
1Adjusted for random variance and covariance of intercept and slope of effects within each study (St-Pierre, 2001). Effects in model include BW, method, physiological state, method × physiological state, physiological state × BW, and herbage digestibility. Values are means ± standard deviation of data adjusted for covariance.
included to evaluate both methodology and physiological state, considerable confounding probably occurs in the data set because not all conditions existed for each method and physiological state. Moreover, methodology is often confounded with location and pasture conditions of the experiment. The confounding of location and pasture condition was not evaluated for this paper, but it would need to be considered before incorporating these data into a more enduring predictive model. From the totals (n = number of records for group means) in Table 1, we eliminated data for 1 experiment using Ru as a marker because only a single experiment from the United Kingdom (2 high-quality forages, perennial ryegrass or white clover pasture) used that marker (Beever et al., 1986). Also, to achieve convergence, any experiment with only 1 treatment group was eliminated (n = 3), and some of the records did not have all the necessary data, especially D, used for the model. The remaining data (n = 419) were used for an overall analysis in PROC MIXED (SAS Inst. Inc., Cary, NC; version 9.3) to evaluate the effect of BW and forage D on OMI after adjusting for random effects of intercept and slope (St-Pierre, 2001). Each experiment was used as a subject. Quadratic terms for all continuous variables were investigated as well as interactions with physiological state and with each other. Any term with P > 0.10 was discarded. The adjusted values were exported and are presented (Fig. 1 to 4). An additional analysis was conducted to evaluate the effect of herbage CP concentration on OMI (n = 290). To further evaluate the effect of milk production (n = 139) and calf performance (n = 116 for ADG or n = 134 for weaning weight), records with those values were analyzed using a similar model. RESULTS AND DISCUSSION The simple statistics for the database are presented in Table 1 and are inclusive of the entire compilation. All viable data were used initially to determine the relationships that were common with the earlier NRC equations (OMI, BW, and D). The effect of method on OMI was of interest, both from an evaluation standpoint and for adjusting values so that true relationships of intake with driver and constraint variables could be ascertained.
Method was a significant (P < 0.01) component and interacted with physiological state (Table 2). Method could be confounded with herbage nutritive value, location of experiment, season, or any combination and its effect on OMI. For instance, the Florida data for cows determined by alkane markers (S. W. Coleman, unpublished data) had greater forage D than most other experiments. One difference is that this D was determined in vivo with the internal alkane ratio, whereas most other methods relied on in vitro estimates of D, either on diet samples gathered from the pasture or on harvested forage. In the Florida experiment, in vitro values for clipped herbage were substantially less than D estimated in vivo using alkanes. In vivo values may be greater because of diet selectivity or potentially increased residence time of residue for Bahia grass (Nelson et al., 1976; Gregorini et al., 2007). Alkanes are noted for their incomplete recovery, but the data reported were adjusted for recovery rates of 75% (C31), 80% (C33), and 85% (C35; Dove and Mayes, 1991; Morais et al., 2011). Collection of total feces with fecal bags produced the lowest estimates, particularly on a BW basis. These low estimates would be anticipated because of the amount of interference with grazing time that is required to frequently change bags. Further, fecal loses in the pasture as a result of the bags being pushed aside during lying and rubbing activities are unaccountable. Although direct measurements must be considered the gold standard for accuracy, confinement for direct measurement cannot be considered the gold standard in terms of animal activity, harvesting energy, climatic conditions, etc. If requirements are based on intake determined solely for confined animals, then many energy expenditures will be omitted from the model. Further, because scatterplots indicated data from direct measurements were interspersed among the greater database, we found no reason not to combine indirect methods with direct methods for evaluating intake on a broader base of conditions and animals. The major limitation of most of the reports in the current database is that longevity of the experiment precluded good estimates of BW and BCS change, both of which are important for assessing demand, as well as determining efficiency. However, it is not appropriate to use the NRC (1996, 2000) equations in the assessment of individual
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2778
Coleman et al.
using metrics based on energy, particularly if one is trying to assess demand or the animal drivers for performance. However, voluntary OMI is a trade-off between physiological drivers that create demand and constraints contributed by the feed resource, its availability and distribution, ease of prehension, climatic factors, and interactions of all these factors. Grazing animals eat mostly forage and have tremendous ability to parse the available forage and select an optimized diet that most nearly meets the demand from what is offered (Coleman and Sollenberger, 2007; Launchbaugh and Dougherty, 2007). That selectivity has compromised efforts to quantify dietary composition and energy intake for many years (Gunter et al., 1993, 1995; Gregorini et al., 2007, 2011). Another issue that was raised with the earlier NRC (1996, 2000) equations involves the animal base on which to anchor measurements. Earlier versions used BW0.75 after Kleiber (1975) and Blaxter (1962). However, although, over the spectrum of vertebrates, BW0.75 may fit most animals, within a species the curve is more likely to follow linear or asymptotic relationships. In the current database, the quadratic functions for BW was indeed significant for predicting OMI, but opposite signs were observed for growing vs. mature animals, as shown in the equations below: Growing animals: OMI = 4.56 + 0.0053W – 0.00002W2 – 0.05531D + 0.00032W × D,
[1]
Dry cows: OMI = 12.5 – 0.0299W + 0.00002W2 – 0.05531D + 0.00032W × D, [2] Lactating cows: OMI = 27.9 – 0.0902W + 0.00009W2 – 0.05531D +0.00032W × D,
Figure 1. Relationship between OM intake and animal BW by physiological state. (a) Growing animals: OMI = 4.56 + 0.0053W – 0.00002W2 – 0.05531D + 0.00032W × D. (b) Dry cows: OMI = 12.5 – 0.0299W + 0.00002W2 – 0.05531D + 0.00032W × D. (c) Lactating cows: OMI = 27.9 – 0.0902W + 0.00009W2 – 0.05531D + 0.00032W × D. OMI = OM intake; W = BW, D = digestibility, and W × D = BW × D.
animal efficiency (e.g., residual feed intake [RFI] of G:F) because the model is more appropriately a predictor of the average animal under the prescribed conditions. It assumes average efficiency and may be used as the standard by which individual animal efficiency is compared, rather than the mean of the contemporary group as is done today (Koch et al., 1963; Archer et al., 1997). The data for predictors of OMI rather than digestible OMI or ME intake (MEI) were investigated. Arguments could be made for
[3]
where W = BW, D = digestibility, and W × D = BW × D. Body weight was more important for predicting the OMI for growing animals (Eq. [1]; Fig. 1a) and was more linear than in the other physiological states considered (Eq. [1] and [2]; Fig. 1a and 1c). Intake by nonlactating and lactating mature cows was less dependent on BW than for growing cattle. Because of the low statistical significance and large scatter of data about the regression lines noted in Fig. 1, we conclude that BW becomes less important in determining OMI in mature cattle than physiological state or other nonanalyzed factors. Be aware that the scatter data in all figures also include variation associated with the method used to estimate intake. For lactating cows, BW interacted (P < 0.01) with forage D when predicting OMI (Eq. [3]; Fig. 2). A significant issue with nonlactating and lactating cows is that OMI was quite variable across a number of studies in which
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
2779
Figure 2. Response surface resulting from the interaction between forage digestibility and animal BW compared with the dependent variable, OM intake. Some data points are obscured from view because they are below the response surface.
BW and forage D also varied greatly. On the basis of the analysis presented in Fig. 1 and the fact that the independent variables, BW and forage D, interact when predicting OMI, it is not surprising that OMI is poorly predicted with previous intake models (NRC, 1984, 1996, 2000). Because forage D is important when predicting OMI, MEI becomes more relevant since forage D is an important part of the mathematical equation for calculating MEI from DE (Agricultural Research Council, 1965; NRC, 1984). The magnitude of the intercept and linear β coefficient for BW from the equations for nonlactating (0.00002 in Eq. [2]) and lactating (0.00009 in Eq. [3]) cows demonstrate the increase in maintenance and production requirements of lactating cows. The negative coefficient for forage D must be considered part of the overall equation, as the positive interaction with BW compensates for the direct effect of the negative coefficient. The interaction of forage D with BW indicates that larger animals take greater advantage of forage with higher D and their OMI may be compromised when forage D is low and improved when forage D is high, provided forage availability is not limiting. To compound this issue, we found a significant positive correlation between cow BW and milk production (r2 = 0.39; Fig. 3). Therefore, the productivity demand is connected to the maintenance requirement for these cows and
hence OMI demand. Does higher lactation cause higher maintenance demand during nonlactating periods? This question is hard to assess due to compensatory mechanisms exhibited by lactating cows to regain BW lost during parturition and lactation, but earlier research indicates that biological types of cattle with a higher milk production have a greater maintenance requirement than types with a lower milk-producing ability even when nonlactating and nonpregnant (Ferrell and Jenkins, 1987). Lactation causes an increase in gastrointestinal tract size (Tulloh, 1966; Forbes, 1986) and increases OMI (Ferrell and Jenkins, 1985) compared to nonlactating cows regardless of pregnancy status. As the gastrointestinal tract increases in size in midlactation, forage intake peaks (Rosiere et al., 1980; Hunter and Siebert, 1986; Wagner et al., 1986), and fat stores are replenished (McNamara and Hillers, 1986). Adams et al. (1987) reported that when forage intakes are expressed on a metabolic BW basis, thinner cattle eat more than fatter cattle to regain body condition. Sprinkle et al. (2000) found that thinner lactating cows in early summer increased bite rate, and they calculated that lactating cows with a BCS of 3 would consume approximately 20% more forage than lactating cows having a BCS of 6, assuming similar bite sizes. Research with cattle grazing Bermuda grass shows that heifers with less ruminal fill
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2780
Coleman et al.
Figure 3. (a) Relationship between OM intake and forage digestibility by physiological state by beef cattle grazing pasture. Regression lines are from overall model and includes intake method. (b) The same as (a) with NRC predicted values included (small solid symbols). Energy value of milk added to cows with known milk production assuming 5% BF and 8.3% SNF.
increased bite size and rate and bites per feeding station, indicating a greater drive to consume forage (Gregorini et al., 2007). These previous research reports indicate that cows with lower energy stores, such as a cow with a BCS of 3 compared with a BCS of 6, may increase forage intake through changes in grazing behavior. The simple relationship of OMI to D is shown in Fig. 3a alongside the values predicted by the NRC (1996) model in Fig. 3b. The predicted values were calculated by the standard equation for growing animals and by the cow equation for both nonlactating and lactating cows. For lactating cows with milk production data, the feed required to supply the energy value of milk was added to the predicted value. The data from the current database are reasonably well predicted by the equations for forages with D values within the range addressed in the NRC (1996) publication (Fig. 3b). Some lactating cows with greater milk production and cattle grazing herbage with low D were substantially over- and underpredicted, respectively. The fact that the NRC (1996) intake models poorly predict actual intakes by cattle consuming diets with extremely low D or very high levels of production was also noted by other researchers (Lardy et al., 2004), indicating that the forage intake model may need switches to adjust for extreme levels of production or a wider range of environments. An alternative to switches would be to develop the equations on the basis of a more robust data set such as the one reported here. Cow performance is best measured by calf production, particularly weaning weight, but the direct output from the cow to the calf is milk. Milk production was a positive driver for cow OMI (Fig. 4) and accounted for 56% of the variation in adjusted intake, coinciding with other reported values (Wagner, 1985). The overall equation was
OMI = 71.6 + 0.015BW – 2.4D +0.021D2 – 11.7MP + 0.42MP × D – 0.0036 MP × D2,
[4]
where MP = milk production (as reported) and D = digestibility. Milk production is difficult to measure even though it is listed in most registration EPD. However, the calculation is not based on milk production but the residual after the growth potentials of both sire and dam were accounted for with the best linear unbiased prediction models. Therefore, it makes little sense to try to include milk in a general prediction equation for an important economic expense like feed intake. A surrogate could be calf weaning weight, but in these data, calf ADG is more closely related to milk production (Fig. 5). With the current database, calf preweaning ADG was a good predictor of OMI and explained 64% of the variation, which is even a better predictor than milk production and is much easier to measure in the field. Equation [5] quantifies the overall relationship: OMI = 251 – 0.06BW + 0.00008BW2 – 7.6D + 0.062D2 – 265G + 8.7G × D – 0.07G × D2, [5] where W = cow BW, D = digestibility, and G = calf preweaning ADG. Calf performance is a good integrator of cow performance output by combining cow size and growth potential with milk production. Hence, it is logical that calf performance, either weaning weight or ADG, might be more closely related to demand than milk production per se. Also, measurement of milk production is perhaps less precise than calf performance, especially since in these data both total-milk-out and weighsuckle-weigh techniques were used. Further, very few of the references included data as to where in the lactation cycle the data were collected. On the basis of simple statistics (R2 and residual SE), calf weaning weight was
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
Figure 4. Relationship between OM intake by lactating cows and milk production.
not as good as ADG for predicting OMI by cows, but the equation is included here because that is probably the most available estimate of productivity. OMI = 266– 0.08W + 0.00009W2 – 8.1D + 0.067D2 – 1.06WW + 0.036WW × D – 0.00029WW × D2, [6] where W = cow BW, D = digestibility, and WW = calf weaning weight. In the very few studies where calf forage intake was recorded, there was little effect on ADG or weaning weight (Ansotegui et al., 1991). Milk production was far more important statistically and may indicate that the value of milk production for calf productivity does not differ significantly across regions or forage bases. In all these relationships, there are quadratic interactions with herbage D. In all cases where we have included the significant quadratic coefficient, a nonlinear, asymptotic equation likely would provide a better and safer overall fit. We challenge the committee in developing the new model to include nonlinear forms such as those shown in Fig. 6 and 7. The relationship in Fig. 6 for the effect of forage CP concentration was taken from the data of Moore et al. (1999). Only a few subsets of the data in the current database indicated a strong relationship between CP and OMI, probably because there were few studies when RDP was limiting. When RDP is limited, forage intake and D may be increased when protein supplements are fed. Moore et al. (1999) developed forage intake prediction equations for nonlactating cattle receiving energy or protein supplements. This research reported that forage intake is normally increased with protein supplementation when the forage CP concentration is less than 10%. However, over the long term, cows may adapt to a low CP diet and compensate with more efficient N cycling (Coleman and Barth, 1977). Also, the relationship for relative DMI (proportion of maximum intake) compared with forage standing crop (forage weight per unit of land) that Rayburn (1986) published was
2781
Figure 5. Relationship between OM intake by the cows and calf preweaning ADG.
used by the authors of the NRC (1996) and is generally true for pastures with dense sods. When swards become sparser, as we see on the native rangelands of the western United States, the relationship mentioned previously does not hold true. For example, Illius and Gordon (1987) show that in dense sods the plant structure (height) plays a more minor role in determining relative forage intake, and intake is more closely related to herbage density. However, when pastures are sparse, such as we see on arid rangelands, intake is greatly affected by plant height and cattle BW. A particularly interesting relationship in this research was that heavy cattle required taller plants to achieve the same relative level of intake as lighter cattle (Illius and Gordon, 1987). The researchers showed that lighter cattle (hence, smaller mouths) achieved a relatively greater bite size on the shorter pastures than the heavier cattle and a relatively greater level of forage intake. The relationship of milk production, BW loss or gain, and forage intake for grazing cattle in more limiting
Figure 6. Relationship of OM intake to forage crude protein. Adapted from Moore et al. (1999).
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2782
Coleman et al.
Figure 7. Relationship of relative (proportion of maximum) intake to herbage mass. This applies to densely populated pastures, normally introduced species (adapted from Rayburn, 1986).
environments is depicted in Fig. 8. All of these point to the dynamic nature of energy intake for cattle grazing in forage-based systems, particularly those that vary considerably in forage quality and quantity during the production cycle. Examination of the conceptual model in Fig. 8 identified a curvilinear function for forage intake not properly accounted for in the NRC (1996) model as the lactating cow advances in the production cycle. As mentioned previously, the NRC (1996) model only considered feedlot growing steers and heifers in developing the database for feed intake. Therefore, not only did that database fail to account for the maintenance requirements of grazing beef cattle, it also failed to account for the expansion and shrinking of the gastrointestinal tract for lactating beef cattle. This phenomenon is similar to that observed for animals experiencing compensatory gain. It has been found that much of the compensatory gain in previously stressed animals is due to decreased maintenance requirements brought about by shrunken organ mass (Hogg, 1989). Fifteen percent to 33% of the realized BW gains in compensatory gain cattle are in increased gut fill, a direct indicator of increasing organ size (Hogg, 1989). With the recent advent of residual feed intake determinations, repeated forage intakes are being individually obtained in drylot feeding situations with producing beef cows. There is an opportunity to evaluate lactating range cows in a feedlot with this specialized equipment and then to follow the same cows in a rangeland or pasture with an external marker to determine if forage intake while grazing is related to forage intake previously measured in the feedlot. The forage intake, as calculated by the external marker, could be validated in the feedlot before using it in the field. Clipped forage from the rangeland or pastures could be fed in the pen, and forage species composition of the rangeland and pasture could easily be estimated using such monitoring techniques as dry-weight rank (Gillen
Figure 8. Generalized energy, forage intake, and milk production curve for a lactating grazing beef cow (adapted from Coppock, 1985).
and Smith, 1986). The additional forage intake for each cow over that consumed in the feedlot could account for the energetic cost of harvesting forage and could provide more reliable estimates of forage intake with free-ranging cattle. Granted, this research would need to be replicated at several locations around the world with varying topography, but it would generate valuable data to estimate the energetic cost of harvesting forage by grazing cattle. Utilizing the pen-fed and free-ranging forage intake comparisons for the same cows, as described above, would also provide information concerning any penalty (Fig. 6) that may exist with unimproved pastures if data were collected in locations with a variety of forage abundances. RECOMMENDATIONS FOR FUTURE INTAKE MODELS In future models predicting forage intake by grazing beef cattle, the Committee on Nutrient Requirements of Beef Cattle appointed by the National Academy of Sciences is requested to consider the following recommendations during their revision of the NRC (2000): 1) because many variables influence forage intake, continue to have entry screens options that provide adjustments for changes in environmental conditions, 2) integrate forage intake estimates for grazing cattle from studies using markers (both internal and external) into the databases used to develop the next NRC forage intake prediction model, 3) develop different equations for growing, nonlactating, and lactating beef cattle, 4) because of the difficulty in obtaining milk production, provide an alternative entry that can be used in place of peak milk production (a good surrogate to replace the measurement of milk production would be calf preweaning ADG), 5) consider integrating some of the equations that have been developed for predicting forage intake for supplemented grazing cattle (Moore et al., 1999), 6) for the future, encourage the validation of the activity cost of grazing by using paired comparisons for individual cows consuming range or
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
pasture forage in the feedlot using individual intake stall equipment (e.g., GrowSafe Systems Ltd., Airdrie, AB, Canada, or Insentec B.V., Marknesse, The Netherlands) and then in the field (external markers), and 7) similarly, develop a new regression equation for forage allowance and intake with the above pairwise comparisons for unimproved pastures. SUMMARY AND CONCLUSIONS As stated in the latest issue of Nutrient Requirements of Beef Cattle (NRC, 2000, p. 95), “Further research is needed to develop more accurate means of predicting intake by beef cattle fed all forage diets.” Some of that research has been completed since the last update; however, much remains to be done. Forage intake by grazing beef cows is not static and not determined simply by BW. Forage OMI varies with physiological state, milk production, a shrinking and expanding gastrointestinal tract, temperature, and forage quality, to name a few. Valiant efforts were made by the NRC committee to develop the intake prediction equations in the latest NRC (1996, 2000) model. Science builds on the discoveries and efforts of past scientists. The next model developed for grazing beef cattle will be an important improvement over the current model, but we acknowledge that it will also have shortcomings given the plethora of biological types of cattle grazing in a variety of environments. LITERATURE CITED Adams, D. C., R. E. Short, and B. W. Knapp. 1987. Body size and body condition effects on performance and behavior of grazing beef cows. Nutr. Rep. Int. 35:269–277. Agricultural Research Council. 1965. The nutrient requirements of farm livestock. No. 2. Ruminants. Agric. Res. Counc., London. Ansotegui, R. P., K. M. Havstad, J. D. Wallace, and D. M. Halford. 1991. Effects of milk intake on forage intake and performance of suckling range calves. J. Anim. Sci. 69:899–904. Archer, J. A., P. F. Arthur, R. M. Herd, P. F. Parnell, and W. S. Pitchford. 1997. Optimum postweaning test for measurement of growth rate, feed intake, and feed efficiency in British breed cattle. J. Anim. Sci. 75:2024–2032. Beever, D. E., H. R. Losada, S. B. Cammell, R. T. Evans, and M. J. Haines. 1986. Effect of forage species and season on nutrient digestion and supply in grazing cattle. Br. J. Nutr. 56:209–225. Blaxter, K. L. 1962. The energy metabolism of ruminants. C. C. Thomas, Springfield, IL. Bodine, T. N., and H. T. Purvis II. 2003. Effects of supplemental energy and/or degradable intake protein on performance, grazing behavior, intake, digestibility, and fecal and blood indices by beef steers grazed on dormant native tallgrass prairie. J. Anim. Sci. 81:304–317. Coleman, S. W., and K. M. Barth. 1977. Utilization of supplemental NPN and energy sources by beef steers consuming low-protein hays. J. Anim. Sci. 45:1180–1186. Coleman, S. W., H. Lippke, and M. Gill. 1999. Estimating the nutritive potential of forages. In: H. G. Jung and G. C. Fahey Jr., editors, Nutritional ecology of herbivores: Proc. Vth Int. Symp. Nutr. Herbivores. Am. Soc. Anim. Sci., Savoy, IL. p. 647–695.
2783
Coleman, S. W., and L. E. Sollenberger. 2007. Plant-herbivore interactions. In: R. F. Barnes, C. J. Nelson, and C. W. Flick, editors, Forages. Vol. 2. 6th ed. Blackwell, Ames, IA. p. 123–136. Coppock, C. E. 1985. Energy nutrition and metabolism of the lactating dairy cow. J. Dairy Sci. 68:3403–3410. Dove, H., and R. W. Mayes. 1991. The use of plant wax alkanes as marker substances in studies of the nutrition of herbivores: A review. Aust. J. Agric. Sci. 42:913–952. Ferrell, C. L. 1988. Energy metabolism. In: D. C. Church, editor, The ruminant animal: Digestive physiology and nutrition. Prentice Hall, Englewood Cliffs, NJ. p. 250–268. Ferrell, C. L., and T. G. Jenkins. 1985. Cow type and the nutritional environment: Nutritional aspects. J. Anim. Sci. 61:725–741. Ferrell, C. L., and T. G. Jenkins. 1987. Influence of biological types on energy requirements. In: M. K. Judkins, D. C. Clanton, M. K. Petersen, and J. D. Wallace, editors, Proc. Grazing Livest. Nutr. Conf. Univ. of Wyoming, Laramie. p. 1–7. Forbes, J. M. 1986. The effects of sex hormones, pregnancy and lactation on digestion, metabolism, and voluntary food intake. In: L. P. Milligan, W. L. Grovum, and A. Dobson, editors, Control of digestion and metabolism in ruminants. Prentice Hall, Englewood Cliffs, NJ. p. 420–435. Funk, M. A., M. L. Galyean, M. E. Branine, and L. J. Krysl. 1987. Steers grazing blue grama rangeland throughout the growing season. I. Dietary composition, intake, digesta kinetics and ruminal fermentation. J. Anim. Sci. 65:1342–1353. Gregorini, P., S. A. Gunter, M. T. Bowman, J. D. Caldwell, C. A. Masino, W. K. Coblentz, and P. A. Beck. 2011. Effect of herbage depletion on short-term foraging dynamics and diet quality of steers grazing wheat pastures. J. Anim. Sci. 89:3824–3830. Gregorini, P., S. A. Gunter, C. A. Masino, and P. A. Beck. 2007. Effects of ruminal fill on short-term herbage intake and grazing dynamics by beef heifers. Grass Forage Sci. 62:346–354. Gillen, R. L., and E. L. Smith. 1986. Evaluation of the dry-weightrank method for determining species composition in tallgrass prairie. J. Range Manage. 39:283–285. Gunter, S. A. 1993. Nutrient intake and digestion by cattle grazing midgrass prairie rangeland and plains bluestem pasture. PhD Diss. Oklahoma State Univ., Stillwater. Gunter, S. A., F. T. McCollum III, R. L. Gillen, and L. J. Krysl. 1993. Forage intake and digestion by cattle grazing midgrass prairie rangeland or sideoats grama/sweetclover pasture. J. Anim. Sci. 71:3432–3441. Gunter, S. A., F. T. McCollum III, R. L. Gillen, and L. J. Krysl. 1995. Diet quality and ruminal digestion in beef cattle grazing midgrass prairie rangeland or plains bluestem pasture throughout the summer. J. Anim. Sci. 73:1174–1186. Hogg, B. W. 1989. Compensatory growth in ruminants. In: A. M. Pearson, editor, Muscle and meat biochemistry. Academic Press, San Diego. Hunter, R. A., and B. D. Siebert. 1986. The effects of genotype, age, pregnancy, lactation, and rumen characteristics on voluntary intake of roughage diets by cattle. Aust. J. Agric. Res. 37:549–560. Illius, A., and I. Gordon. 1987. The allometry of food intake in grazing ruminants. J. Anim. Ecol. 56:989–999. Kleiber, M. 1975. The fire of life. Krieger, Huntington, NY. Koch, R. M., L. A. Swinger, D. Chambers, and K. E. Gregory. 1963. Efficiency of feed use in beef cattle. J. Anim. Sci. 22:486–494. Krysl, L. J., M. L. Galyean, M. B. Judkins, M. E. Branine, and R. E. Estell. 1987. Digestive physiology of steers grazing fertilized and non-fertilized blue grama rangeland. J. Range Manage. 40:493–501. Lardy, G. P., D. C. Adams, T. J. Klopfenstein, and H. H. Patterson. 2004. Building beef cow nutritional programs with the 1996 NRC beef cattle requirements model. J. Anim. Sci. 82(E. Suppl.):E83–E92. Launchbaugh, K. L., and C. T. Dougherty. 2007. Grazing animal Behavior. In: R. F. Barnes, C. J. Nelson, and C. W. Flick, editors, Forages. Vol. 2. 6th ed. Blackwell, Ames, IA. p. 675–686. Lofgreen, G. P., and W. N. Garrett. 1968. A system for expressing net energy requirements and feed values for growing and finishing cattle. J. Anim. Sci. 27:793–806.
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2784
Coleman et al.
McNamara, J. P., and J. K. Hillers. 1986. Regulation of bovine adipose tissue metabolism during lactation. 1. Lipid synthesis in response to increased milk production and decreased energy intake. J. Dairy Sci. 69:3032–3041. Moore, J. E., M. H. Brant, W. E. Kunkle, and D. I. Hopkins. 1999. Effects of supplementation on voluntary forage intake, diet digestibility, and animal performance. J. Anim. Sci. 77:122–135. Morais, J. A. S., T. T. Berchielli, A. de Vega, M. F. S. Queiroz, A. Keli, R. A. Reis, L. M. A. Bertipaglia, and S. F. Souza. 2011. The validity of n-alkanes to estimate intake and digestibility in Nellore beef cattle fed a tropical grass (Brachiaria brizantha cv. Marandu). Livest. Sci. 135:184–192. Nelson, B. D., C. R. Montgomery, P. E. Schilling, and L. Mason. 1976. Effects of fermentation time on in vivo/in vitro relationships. J. Dairy Sci. 59:270–277. NRC. 1984. Nutrient requirements of beef cattle. 6th ed. Natl. Acad. Press, Washington, DC. NRC. 1996. Nutrient requirements of beef cattle. 7th ed. Natl. Acad. Press, Washington, DC. NRC. 2000. Nutrient requirements of beef cattle. 7th revised ed. Natl. Acad. Press, Washington, DC. Pordomingo, A. J., J. D. Wallace, A. S. Freeman, and M. L. Galyean. 1991. Supplemental corn grain for steers grazing native rangeland during summer. J. Anim. Sci. 69:1678–1687.
Rayburn, E. B. 1986. Quantitative aspects of pasture management. Seneca Trail RC&D Tech. Manual. Seneca Trail RC&D, Franklinville, NY. Rosiere, R. E., J. D. Wallace, and R. D. Pieper. 1980. Forage intake in two-year old cows and heifers grazing blue grama summer range. J. Range Manage. 33:71–73. Sprinkle, J. E., J. W. Holloway, B. G. Warrington, W. C. Ellis, J. W. Stuth, T. D. A. Forbes, and L. W. Greene. 2000. Digesta kinetics, energy intake, grazing behavior, and body temperature of grazing beef cattle differing in adaptation to heat. J. Anim. Sci. 78:1608–1624. St-Pierre, N. R. 2001. Invited review: Integrating quantitative findings from multiple studies using mixed model methodology. J. Dairy Sci. 84:741–755. Tulloh, N. M. 1966. Physical studies of the alimentary tract of grazing cattle. IV. Dimensions of the tract in lactating and non-lactating cows. N. Z. J. Agric. Res. 9:999–1008. Vona, L. C., G. A. Jung, R. L. Reid, and W. C. Sharp. 1984. Nutritive value of warm-season grass hays for beef cattle and sheep: Digestibility, intake and mineral utilization. J. Anim. Sci. 59:1582–1593. Wagner, M. W. 1985. Forage intake and milk production of rangeland beef cows with varying degrees of crossbred influence. MS Thesis. Montana State Univ., Bozeman. Wagner, M. W., K. M. Havstad, D. E. Doornbos, and E. L. Ayers. 1986. Forage intake of rangeland beef cows with varying degrees of crossbred influence. J. Anim. Sci. 63:1484–1490.
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
References
This article cites 34 articles, 17 of which you can access for free at: http://www.journalofanimalscience.org/content/92/7/2775#BIBL
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
The online version of this article, along with updated information and services, is located on the World Wide Web at: http://www.journalofanimalscience.org/content/92/7/2775
www.asas.org
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Beef Species Symposium: Difficulties associated with predicting forage intake by grazing beef cows1,2 S. W. Coleman,* S. A. Gunter,†3 J. E. Sprinkle,‡ and J. P. S. Neel* *USDA-ARS, Grazinglands Research Laboratory, El Reno, OK 73801; †USDA-ARS, Southern Plains Range Research Station, Woodward, OK 73801; and ‡ Department of Animal Sciences, University of Arizona, Tucson 85721
ABSTRACT: The current NRC model to estimate DMI is based on a single equation related to metabolic size and net energy density of the diet; this equation was a significant improvement over previous models. However, observed DMI by grazing animals can be conceptualized by a function that includes animal demand, largely determined by metabolic or linear size, physiological state, genetics, or any combination of these. Even in the database used to generate the current NRC equation, DMI by cows is poorly predicted at the extremes. In fact, across a wide range of actual DMI, predicted DMI is rather flat, indicating an insensitivity of prediction, so the model requires further refinement. A broad-based database was developed that includes pasture and confinement studies with growing, nonlactating, and lactating cattle. New equations are presented
for consideration in the new model. It was found that the premise behind earlier NRC equations based on diet digestibility and BW are sound but that for cows, additional drivers based on milk production or calf performance were stronger than BW. Future models should be based on multiple variables, including functions for physiological state, animal suitability to the environment, and activity to modify the predicted DMI. Further, the model could possibly account for imbalances of protein to energy, particularly as they relate to ruminal function. Further, the issue of how reference data were collected (pen vs. pasture) and how the methods or constraints influence DMI must be evaluated. Overall, the new NRC model needs to be more robust in its ability to account for the wide variation in the environment, dietary characteristics, and metabolic demands.
Key words: beef, cattle, forage intake, pasture, rangelands © 2014 American Society of Animal Science. All rights reserved. J. Anim. Sci. 2014.92:2775–2784 doi:10.2527/jas2013-7090 INTRODUCTION Predicting intake of grazing animals is a worthy goal, but difficulties abound with the impossibility of measuring and the difficulty of estimating the intake of grazing animals (Coleman et al., 1999). In their review, Coleman et al. (1999) noted that intake varies due to both feed quality and physical characteristics and also due to the animal’s physiological state. Therefore, any attempt to predict or model intake must take in both considerations 1Based on a presentation at the Beef Species Symposium titled “Nutrient Requirements of the Beef Female in Extensive Grazing Systems: Considerations for Revising the Beef NRC” at the Joint Annual Meeting, July 8–12, 2013, Indianapolis, IN. 2Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA. 3Corresponding author: [email protected] Received August 29, 2013. Accepted November 25, 2013.
and, at the very best, provide an estimate of the average intake under the specified conditions. The NRC committee included a model for intake in the NRC (1984, 1996) publication and, because of limitations, later modified the model (NRC, 2000). Major limitations of the model included the fact that all the experimental data that were included were from pen-fed animals, many on highconcentrate diets. Observational field data and peer-reviewed research have identified deficiencies in the 1996 and 2000 Nutrient Requirements of Beef Cattle (NRC, 1996, 2000) with respect to maintenance requirements and forage intake on rangelands. Lardy et al. (2004) reported that predicted energy deficits for summer-calving cows grazing either range or subirrigated meadow in November were “biologically unreasonable.” They further recommended that the “on pasture” function in the nutritional model “not be used because it unreasonably increases energy requirements.” In most instances, cow performance field data agree with the findings of Lardy
2775 Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2776
Coleman et al.
et al. (2004), although there have been some studies where predicted forage intake for dormant forage exceeded estimated forage intake indirectly measured with external markers (Cr marker calculated; Bodine and Purvis, 2003). THE NRC MODEL BASIS AND ASSUMPTIONS The forage intake models presented in NRC (1984, 1996, 2000) are based on the California NE System of Lofgreen and Garrett (1968) with various adjustments to the model for breed, physiological state, activity, heat loss or gain, milk production, and grazing forage allowance. The NEm calculation and the adjustment for activity (HiE) are problematic for managers of grazing cattle in a rangeland setting because the calculated values for NEm were originally generated with mostly idle animals in either a calorimeter or a feedlot. As the NRC (2000, p. 6) model states, “This expression was derived using data from, primarily, growing steers and heifers of British ancestry that were penned in generally non-stressful environments.” Further, the NRC (2000, p. 88) model only used intake data collected from growing and finishing animals in the feedlot. There were 3 validations of the NEm intake model with data from cattle consuming forage diets. Two validations were with grazing cattle, 1 in New Mexico (Funk et al., 1987; Krysl et al., 1987; Pordomingo et al., 1991) and 1 in Oklahoma (Gunter, 1993; Gunter et al., 1993). The only validation for mature cows was with nonlactating cows fed grass hay (Vona et al., 1984). The 2 main problems associated with using the NEm system in rangeland settings are, first, failure to adequately account for plasticity or adaptations that often occur with producing range cows as high-energy demand visceral tissues (Ferrell, 1988) shrink and expand in response to nutrient availability and demands (Forbes, 1986) in a much greater fashion than observed in feedlot systems and, second, failure to accurately assess activity levels and nutrient demands of cattle grazing rangelands, thus generating unreasonable maintenance requirements, especially with dormant forage. The objective of this paper is to evaluate the expansion of intake data that were used to develop relationships and to compare those relationships to the existing NRC equations. The expanded data included many grazing trials in which intake was estimated with various marker systems. DATABASE DEVELOPMENT A database was compiled from the literature with 50 studies containing a total of 482 group means, conducted both on pasture and in confinement, largely with all or very high forage diets. The only assumption made was that when forage intake was explicitly characterized as DM, then it was transformed to OM intake (OMI) by
Table 1. Summary of variables compiled in the database before any truncation Variable Forage CP, % Forage digestibility, % Initial BW, kg Average BW, kg ADG, kg Milk production, kg/d Forage OM intake, kg/d Total OM intake, kg/d Digestible DMI, kg/d DMI, % of BW Calf weaning wt, kg Calf ADG, kg Calf forage DMI, kg/d
n 326 448 319 484 180 160 482 482 448 482 142 128 55
Mean SD Minimum Maximum 11.5787 5.7269 1.3 28.5 59.8023 11.1053 30.7 84.8 419.9463 130.9662 146 669 414.8421 121.9692 146 675 0.2911 0.5753 -1.31 2.13 6.6100 3.2118 0.38 21.20 8.9710 3.8554 0.78 21.64 9.2786 3.7715 0.78 21.64 5.5183 2.5582 0.44 13.97 2.2239 0.6910 0.21 4.55 200.4652 38.8218 133.8 317.2 0.7256 0.1944 0.39 1.67 1.9189 1.2445 0.12 5.50
multiplying forage DMI by 0.92. These data are summarized in Table 1. Some of the animals from pasture research received supplemental feed. In those instances where cattle were supplemented, dietary digestibility (D) and CP concentration were assumed to be the same as that reported for forage samples; hence, no adjustment in diet nutritive value was made for supplement, but intake was recorded as the total forage and supplement. Although this procedure ignored potential associative effects (positive for increased DIP and negative for starch), the major objective of this project was to determine the feasibility of using data from grazing trials and compare it with the existing model, not to develop a final model to be used in the new NRC to be published. Table 2 presents OMI means delineated by method of intake measurement (n = 5; directly measured in confinement, Cr, Yb, total fecal collection, and alkanes) and physiological state (n = 3; growing, lactating, and nonlactating). Some steers were described as “mature” but were included with the “growing” animals. It was concluded that the maintenance machinery of producing cows was significantly different than that of growing animals and including mature steers with growing animals would stretch the inference space for that group. Growing animals were included in the analysis to test conclusions and comparisons with previous NRC (1996, 2000) models on animals more closely related to those used to develop the models. That premise should give us a relative basis to evaluate methods used to estimate OMI of cattle grazing pasture. The reasoning is that if data from growing animals grazing pasture fit the model but data from cows did not, then we could attribute the majority of the lack of fit to physiological state rather than methodology. In contrast, if pasture-based growing animals could not be predicted with the NRC (1996, 2000) models, then methodology would be suspect, although grazing animals of the same class may have different OMI demand than confined animals. Although all classes were
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2777
Forage intake by grazing beef cows
Table 2. Mean intake as affected by method of estimate and animal physiological state (n = 419)1 Growing Method Direct Fecal bags Alkanes Cr2O3 YbCl3
kg/d 7.2 ± 0.45 6.6 ± 1.60 — 7.5 ± 2.60 7.4 ± 2.56
Dry cows g·kg-1 BW·d-1 18.0 ± 2.63 19.0 ± 1.79 — 23.9 ± 6.17 19.5 ± 3.67
kg/d 8.7 ± 0.64 3.7 ± 0.28 12.8 ± 2.31 7.2 ± 1.48 11.7 ± 2.11
g·kg-1 BW·d-1 16.6 ± 1.26 12.3 ± 0.50 25.3 ± 3.65 14.5 ± 3.43 26.3 ± 6.46
Lactating cows kg/d g·kg-1 BW·d-1 11.9 ± 1.57 24.2 ± 3.94 — — 15.8 ± 0.83 32.6 ± 1.32 9.6 ± 2.03 21.4 ± 3.62 11.9 ± 2.10 29.0 ± 1.11
1Adjusted for random variance and covariance of intercept and slope of effects within each study (St-Pierre, 2001). Effects in model include BW, method, physiological state, method × physiological state, physiological state × BW, and herbage digestibility. Values are means ± standard deviation of data adjusted for covariance.
included to evaluate both methodology and physiological state, considerable confounding probably occurs in the data set because not all conditions existed for each method and physiological state. Moreover, methodology is often confounded with location and pasture conditions of the experiment. The confounding of location and pasture condition was not evaluated for this paper, but it would need to be considered before incorporating these data into a more enduring predictive model. From the totals (n = number of records for group means) in Table 1, we eliminated data for 1 experiment using Ru as a marker because only a single experiment from the United Kingdom (2 high-quality forages, perennial ryegrass or white clover pasture) used that marker (Beever et al., 1986). Also, to achieve convergence, any experiment with only 1 treatment group was eliminated (n = 3), and some of the records did not have all the necessary data, especially D, used for the model. The remaining data (n = 419) were used for an overall analysis in PROC MIXED (SAS Inst. Inc., Cary, NC; version 9.3) to evaluate the effect of BW and forage D on OMI after adjusting for random effects of intercept and slope (St-Pierre, 2001). Each experiment was used as a subject. Quadratic terms for all continuous variables were investigated as well as interactions with physiological state and with each other. Any term with P > 0.10 was discarded. The adjusted values were exported and are presented (Fig. 1 to 4). An additional analysis was conducted to evaluate the effect of herbage CP concentration on OMI (n = 290). To further evaluate the effect of milk production (n = 139) and calf performance (n = 116 for ADG or n = 134 for weaning weight), records with those values were analyzed using a similar model. RESULTS AND DISCUSSION The simple statistics for the database are presented in Table 1 and are inclusive of the entire compilation. All viable data were used initially to determine the relationships that were common with the earlier NRC equations (OMI, BW, and D). The effect of method on OMI was of interest, both from an evaluation standpoint and for adjusting values so that true relationships of intake with driver and constraint variables could be ascertained.
Method was a significant (P < 0.01) component and interacted with physiological state (Table 2). Method could be confounded with herbage nutritive value, location of experiment, season, or any combination and its effect on OMI. For instance, the Florida data for cows determined by alkane markers (S. W. Coleman, unpublished data) had greater forage D than most other experiments. One difference is that this D was determined in vivo with the internal alkane ratio, whereas most other methods relied on in vitro estimates of D, either on diet samples gathered from the pasture or on harvested forage. In the Florida experiment, in vitro values for clipped herbage were substantially less than D estimated in vivo using alkanes. In vivo values may be greater because of diet selectivity or potentially increased residence time of residue for Bahia grass (Nelson et al., 1976; Gregorini et al., 2007). Alkanes are noted for their incomplete recovery, but the data reported were adjusted for recovery rates of 75% (C31), 80% (C33), and 85% (C35; Dove and Mayes, 1991; Morais et al., 2011). Collection of total feces with fecal bags produced the lowest estimates, particularly on a BW basis. These low estimates would be anticipated because of the amount of interference with grazing time that is required to frequently change bags. Further, fecal loses in the pasture as a result of the bags being pushed aside during lying and rubbing activities are unaccountable. Although direct measurements must be considered the gold standard for accuracy, confinement for direct measurement cannot be considered the gold standard in terms of animal activity, harvesting energy, climatic conditions, etc. If requirements are based on intake determined solely for confined animals, then many energy expenditures will be omitted from the model. Further, because scatterplots indicated data from direct measurements were interspersed among the greater database, we found no reason not to combine indirect methods with direct methods for evaluating intake on a broader base of conditions and animals. The major limitation of most of the reports in the current database is that longevity of the experiment precluded good estimates of BW and BCS change, both of which are important for assessing demand, as well as determining efficiency. However, it is not appropriate to use the NRC (1996, 2000) equations in the assessment of individual
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2778
Coleman et al.
using metrics based on energy, particularly if one is trying to assess demand or the animal drivers for performance. However, voluntary OMI is a trade-off between physiological drivers that create demand and constraints contributed by the feed resource, its availability and distribution, ease of prehension, climatic factors, and interactions of all these factors. Grazing animals eat mostly forage and have tremendous ability to parse the available forage and select an optimized diet that most nearly meets the demand from what is offered (Coleman and Sollenberger, 2007; Launchbaugh and Dougherty, 2007). That selectivity has compromised efforts to quantify dietary composition and energy intake for many years (Gunter et al., 1993, 1995; Gregorini et al., 2007, 2011). Another issue that was raised with the earlier NRC (1996, 2000) equations involves the animal base on which to anchor measurements. Earlier versions used BW0.75 after Kleiber (1975) and Blaxter (1962). However, although, over the spectrum of vertebrates, BW0.75 may fit most animals, within a species the curve is more likely to follow linear or asymptotic relationships. In the current database, the quadratic functions for BW was indeed significant for predicting OMI, but opposite signs were observed for growing vs. mature animals, as shown in the equations below: Growing animals: OMI = 4.56 + 0.0053W – 0.00002W2 – 0.05531D + 0.00032W × D,
[1]
Dry cows: OMI = 12.5 – 0.0299W + 0.00002W2 – 0.05531D + 0.00032W × D, [2] Lactating cows: OMI = 27.9 – 0.0902W + 0.00009W2 – 0.05531D +0.00032W × D,
Figure 1. Relationship between OM intake and animal BW by physiological state. (a) Growing animals: OMI = 4.56 + 0.0053W – 0.00002W2 – 0.05531D + 0.00032W × D. (b) Dry cows: OMI = 12.5 – 0.0299W + 0.00002W2 – 0.05531D + 0.00032W × D. (c) Lactating cows: OMI = 27.9 – 0.0902W + 0.00009W2 – 0.05531D + 0.00032W × D. OMI = OM intake; W = BW, D = digestibility, and W × D = BW × D.
animal efficiency (e.g., residual feed intake [RFI] of G:F) because the model is more appropriately a predictor of the average animal under the prescribed conditions. It assumes average efficiency and may be used as the standard by which individual animal efficiency is compared, rather than the mean of the contemporary group as is done today (Koch et al., 1963; Archer et al., 1997). The data for predictors of OMI rather than digestible OMI or ME intake (MEI) were investigated. Arguments could be made for
[3]
where W = BW, D = digestibility, and W × D = BW × D. Body weight was more important for predicting the OMI for growing animals (Eq. [1]; Fig. 1a) and was more linear than in the other physiological states considered (Eq. [1] and [2]; Fig. 1a and 1c). Intake by nonlactating and lactating mature cows was less dependent on BW than for growing cattle. Because of the low statistical significance and large scatter of data about the regression lines noted in Fig. 1, we conclude that BW becomes less important in determining OMI in mature cattle than physiological state or other nonanalyzed factors. Be aware that the scatter data in all figures also include variation associated with the method used to estimate intake. For lactating cows, BW interacted (P < 0.01) with forage D when predicting OMI (Eq. [3]; Fig. 2). A significant issue with nonlactating and lactating cows is that OMI was quite variable across a number of studies in which
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
2779
Figure 2. Response surface resulting from the interaction between forage digestibility and animal BW compared with the dependent variable, OM intake. Some data points are obscured from view because they are below the response surface.
BW and forage D also varied greatly. On the basis of the analysis presented in Fig. 1 and the fact that the independent variables, BW and forage D, interact when predicting OMI, it is not surprising that OMI is poorly predicted with previous intake models (NRC, 1984, 1996, 2000). Because forage D is important when predicting OMI, MEI becomes more relevant since forage D is an important part of the mathematical equation for calculating MEI from DE (Agricultural Research Council, 1965; NRC, 1984). The magnitude of the intercept and linear β coefficient for BW from the equations for nonlactating (0.00002 in Eq. [2]) and lactating (0.00009 in Eq. [3]) cows demonstrate the increase in maintenance and production requirements of lactating cows. The negative coefficient for forage D must be considered part of the overall equation, as the positive interaction with BW compensates for the direct effect of the negative coefficient. The interaction of forage D with BW indicates that larger animals take greater advantage of forage with higher D and their OMI may be compromised when forage D is low and improved when forage D is high, provided forage availability is not limiting. To compound this issue, we found a significant positive correlation between cow BW and milk production (r2 = 0.39; Fig. 3). Therefore, the productivity demand is connected to the maintenance requirement for these cows and
hence OMI demand. Does higher lactation cause higher maintenance demand during nonlactating periods? This question is hard to assess due to compensatory mechanisms exhibited by lactating cows to regain BW lost during parturition and lactation, but earlier research indicates that biological types of cattle with a higher milk production have a greater maintenance requirement than types with a lower milk-producing ability even when nonlactating and nonpregnant (Ferrell and Jenkins, 1987). Lactation causes an increase in gastrointestinal tract size (Tulloh, 1966; Forbes, 1986) and increases OMI (Ferrell and Jenkins, 1985) compared to nonlactating cows regardless of pregnancy status. As the gastrointestinal tract increases in size in midlactation, forage intake peaks (Rosiere et al., 1980; Hunter and Siebert, 1986; Wagner et al., 1986), and fat stores are replenished (McNamara and Hillers, 1986). Adams et al. (1987) reported that when forage intakes are expressed on a metabolic BW basis, thinner cattle eat more than fatter cattle to regain body condition. Sprinkle et al. (2000) found that thinner lactating cows in early summer increased bite rate, and they calculated that lactating cows with a BCS of 3 would consume approximately 20% more forage than lactating cows having a BCS of 6, assuming similar bite sizes. Research with cattle grazing Bermuda grass shows that heifers with less ruminal fill
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2780
Coleman et al.
Figure 3. (a) Relationship between OM intake and forage digestibility by physiological state by beef cattle grazing pasture. Regression lines are from overall model and includes intake method. (b) The same as (a) with NRC predicted values included (small solid symbols). Energy value of milk added to cows with known milk production assuming 5% BF and 8.3% SNF.
increased bite size and rate and bites per feeding station, indicating a greater drive to consume forage (Gregorini et al., 2007). These previous research reports indicate that cows with lower energy stores, such as a cow with a BCS of 3 compared with a BCS of 6, may increase forage intake through changes in grazing behavior. The simple relationship of OMI to D is shown in Fig. 3a alongside the values predicted by the NRC (1996) model in Fig. 3b. The predicted values were calculated by the standard equation for growing animals and by the cow equation for both nonlactating and lactating cows. For lactating cows with milk production data, the feed required to supply the energy value of milk was added to the predicted value. The data from the current database are reasonably well predicted by the equations for forages with D values within the range addressed in the NRC (1996) publication (Fig. 3b). Some lactating cows with greater milk production and cattle grazing herbage with low D were substantially over- and underpredicted, respectively. The fact that the NRC (1996) intake models poorly predict actual intakes by cattle consuming diets with extremely low D or very high levels of production was also noted by other researchers (Lardy et al., 2004), indicating that the forage intake model may need switches to adjust for extreme levels of production or a wider range of environments. An alternative to switches would be to develop the equations on the basis of a more robust data set such as the one reported here. Cow performance is best measured by calf production, particularly weaning weight, but the direct output from the cow to the calf is milk. Milk production was a positive driver for cow OMI (Fig. 4) and accounted for 56% of the variation in adjusted intake, coinciding with other reported values (Wagner, 1985). The overall equation was
OMI = 71.6 + 0.015BW – 2.4D +0.021D2 – 11.7MP + 0.42MP × D – 0.0036 MP × D2,
[4]
where MP = milk production (as reported) and D = digestibility. Milk production is difficult to measure even though it is listed in most registration EPD. However, the calculation is not based on milk production but the residual after the growth potentials of both sire and dam were accounted for with the best linear unbiased prediction models. Therefore, it makes little sense to try to include milk in a general prediction equation for an important economic expense like feed intake. A surrogate could be calf weaning weight, but in these data, calf ADG is more closely related to milk production (Fig. 5). With the current database, calf preweaning ADG was a good predictor of OMI and explained 64% of the variation, which is even a better predictor than milk production and is much easier to measure in the field. Equation [5] quantifies the overall relationship: OMI = 251 – 0.06BW + 0.00008BW2 – 7.6D + 0.062D2 – 265G + 8.7G × D – 0.07G × D2, [5] where W = cow BW, D = digestibility, and G = calf preweaning ADG. Calf performance is a good integrator of cow performance output by combining cow size and growth potential with milk production. Hence, it is logical that calf performance, either weaning weight or ADG, might be more closely related to demand than milk production per se. Also, measurement of milk production is perhaps less precise than calf performance, especially since in these data both total-milk-out and weighsuckle-weigh techniques were used. Further, very few of the references included data as to where in the lactation cycle the data were collected. On the basis of simple statistics (R2 and residual SE), calf weaning weight was
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
Figure 4. Relationship between OM intake by lactating cows and milk production.
not as good as ADG for predicting OMI by cows, but the equation is included here because that is probably the most available estimate of productivity. OMI = 266– 0.08W + 0.00009W2 – 8.1D + 0.067D2 – 1.06WW + 0.036WW × D – 0.00029WW × D2, [6] where W = cow BW, D = digestibility, and WW = calf weaning weight. In the very few studies where calf forage intake was recorded, there was little effect on ADG or weaning weight (Ansotegui et al., 1991). Milk production was far more important statistically and may indicate that the value of milk production for calf productivity does not differ significantly across regions or forage bases. In all these relationships, there are quadratic interactions with herbage D. In all cases where we have included the significant quadratic coefficient, a nonlinear, asymptotic equation likely would provide a better and safer overall fit. We challenge the committee in developing the new model to include nonlinear forms such as those shown in Fig. 6 and 7. The relationship in Fig. 6 for the effect of forage CP concentration was taken from the data of Moore et al. (1999). Only a few subsets of the data in the current database indicated a strong relationship between CP and OMI, probably because there were few studies when RDP was limiting. When RDP is limited, forage intake and D may be increased when protein supplements are fed. Moore et al. (1999) developed forage intake prediction equations for nonlactating cattle receiving energy or protein supplements. This research reported that forage intake is normally increased with protein supplementation when the forage CP concentration is less than 10%. However, over the long term, cows may adapt to a low CP diet and compensate with more efficient N cycling (Coleman and Barth, 1977). Also, the relationship for relative DMI (proportion of maximum intake) compared with forage standing crop (forage weight per unit of land) that Rayburn (1986) published was
2781
Figure 5. Relationship between OM intake by the cows and calf preweaning ADG.
used by the authors of the NRC (1996) and is generally true for pastures with dense sods. When swards become sparser, as we see on the native rangelands of the western United States, the relationship mentioned previously does not hold true. For example, Illius and Gordon (1987) show that in dense sods the plant structure (height) plays a more minor role in determining relative forage intake, and intake is more closely related to herbage density. However, when pastures are sparse, such as we see on arid rangelands, intake is greatly affected by plant height and cattle BW. A particularly interesting relationship in this research was that heavy cattle required taller plants to achieve the same relative level of intake as lighter cattle (Illius and Gordon, 1987). The researchers showed that lighter cattle (hence, smaller mouths) achieved a relatively greater bite size on the shorter pastures than the heavier cattle and a relatively greater level of forage intake. The relationship of milk production, BW loss or gain, and forage intake for grazing cattle in more limiting
Figure 6. Relationship of OM intake to forage crude protein. Adapted from Moore et al. (1999).
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2782
Coleman et al.
Figure 7. Relationship of relative (proportion of maximum) intake to herbage mass. This applies to densely populated pastures, normally introduced species (adapted from Rayburn, 1986).
environments is depicted in Fig. 8. All of these point to the dynamic nature of energy intake for cattle grazing in forage-based systems, particularly those that vary considerably in forage quality and quantity during the production cycle. Examination of the conceptual model in Fig. 8 identified a curvilinear function for forage intake not properly accounted for in the NRC (1996) model as the lactating cow advances in the production cycle. As mentioned previously, the NRC (1996) model only considered feedlot growing steers and heifers in developing the database for feed intake. Therefore, not only did that database fail to account for the maintenance requirements of grazing beef cattle, it also failed to account for the expansion and shrinking of the gastrointestinal tract for lactating beef cattle. This phenomenon is similar to that observed for animals experiencing compensatory gain. It has been found that much of the compensatory gain in previously stressed animals is due to decreased maintenance requirements brought about by shrunken organ mass (Hogg, 1989). Fifteen percent to 33% of the realized BW gains in compensatory gain cattle are in increased gut fill, a direct indicator of increasing organ size (Hogg, 1989). With the recent advent of residual feed intake determinations, repeated forage intakes are being individually obtained in drylot feeding situations with producing beef cows. There is an opportunity to evaluate lactating range cows in a feedlot with this specialized equipment and then to follow the same cows in a rangeland or pasture with an external marker to determine if forage intake while grazing is related to forage intake previously measured in the feedlot. The forage intake, as calculated by the external marker, could be validated in the feedlot before using it in the field. Clipped forage from the rangeland or pastures could be fed in the pen, and forage species composition of the rangeland and pasture could easily be estimated using such monitoring techniques as dry-weight rank (Gillen
Figure 8. Generalized energy, forage intake, and milk production curve for a lactating grazing beef cow (adapted from Coppock, 1985).
and Smith, 1986). The additional forage intake for each cow over that consumed in the feedlot could account for the energetic cost of harvesting forage and could provide more reliable estimates of forage intake with free-ranging cattle. Granted, this research would need to be replicated at several locations around the world with varying topography, but it would generate valuable data to estimate the energetic cost of harvesting forage by grazing cattle. Utilizing the pen-fed and free-ranging forage intake comparisons for the same cows, as described above, would also provide information concerning any penalty (Fig. 6) that may exist with unimproved pastures if data were collected in locations with a variety of forage abundances. RECOMMENDATIONS FOR FUTURE INTAKE MODELS In future models predicting forage intake by grazing beef cattle, the Committee on Nutrient Requirements of Beef Cattle appointed by the National Academy of Sciences is requested to consider the following recommendations during their revision of the NRC (2000): 1) because many variables influence forage intake, continue to have entry screens options that provide adjustments for changes in environmental conditions, 2) integrate forage intake estimates for grazing cattle from studies using markers (both internal and external) into the databases used to develop the next NRC forage intake prediction model, 3) develop different equations for growing, nonlactating, and lactating beef cattle, 4) because of the difficulty in obtaining milk production, provide an alternative entry that can be used in place of peak milk production (a good surrogate to replace the measurement of milk production would be calf preweaning ADG), 5) consider integrating some of the equations that have been developed for predicting forage intake for supplemented grazing cattle (Moore et al., 1999), 6) for the future, encourage the validation of the activity cost of grazing by using paired comparisons for individual cows consuming range or
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
Forage intake by grazing beef cows
pasture forage in the feedlot using individual intake stall equipment (e.g., GrowSafe Systems Ltd., Airdrie, AB, Canada, or Insentec B.V., Marknesse, The Netherlands) and then in the field (external markers), and 7) similarly, develop a new regression equation for forage allowance and intake with the above pairwise comparisons for unimproved pastures. SUMMARY AND CONCLUSIONS As stated in the latest issue of Nutrient Requirements of Beef Cattle (NRC, 2000, p. 95), “Further research is needed to develop more accurate means of predicting intake by beef cattle fed all forage diets.” Some of that research has been completed since the last update; however, much remains to be done. Forage intake by grazing beef cows is not static and not determined simply by BW. Forage OMI varies with physiological state, milk production, a shrinking and expanding gastrointestinal tract, temperature, and forage quality, to name a few. Valiant efforts were made by the NRC committee to develop the intake prediction equations in the latest NRC (1996, 2000) model. Science builds on the discoveries and efforts of past scientists. The next model developed for grazing beef cattle will be an important improvement over the current model, but we acknowledge that it will also have shortcomings given the plethora of biological types of cattle grazing in a variety of environments. LITERATURE CITED Adams, D. C., R. E. Short, and B. W. Knapp. 1987. Body size and body condition effects on performance and behavior of grazing beef cows. Nutr. Rep. Int. 35:269–277. Agricultural Research Council. 1965. The nutrient requirements of farm livestock. No. 2. Ruminants. Agric. Res. Counc., London. Ansotegui, R. P., K. M. Havstad, J. D. Wallace, and D. M. Halford. 1991. Effects of milk intake on forage intake and performance of suckling range calves. J. Anim. Sci. 69:899–904. Archer, J. A., P. F. Arthur, R. M. Herd, P. F. Parnell, and W. S. Pitchford. 1997. Optimum postweaning test for measurement of growth rate, feed intake, and feed efficiency in British breed cattle. J. Anim. Sci. 75:2024–2032. Beever, D. E., H. R. Losada, S. B. Cammell, R. T. Evans, and M. J. Haines. 1986. Effect of forage species and season on nutrient digestion and supply in grazing cattle. Br. J. Nutr. 56:209–225. Blaxter, K. L. 1962. The energy metabolism of ruminants. C. C. Thomas, Springfield, IL. Bodine, T. N., and H. T. Purvis II. 2003. Effects of supplemental energy and/or degradable intake protein on performance, grazing behavior, intake, digestibility, and fecal and blood indices by beef steers grazed on dormant native tallgrass prairie. J. Anim. Sci. 81:304–317. Coleman, S. W., and K. M. Barth. 1977. Utilization of supplemental NPN and energy sources by beef steers consuming low-protein hays. J. Anim. Sci. 45:1180–1186. Coleman, S. W., H. Lippke, and M. Gill. 1999. Estimating the nutritive potential of forages. In: H. G. Jung and G. C. Fahey Jr., editors, Nutritional ecology of herbivores: Proc. Vth Int. Symp. Nutr. Herbivores. Am. Soc. Anim. Sci., Savoy, IL. p. 647–695.
2783
Coleman, S. W., and L. E. Sollenberger. 2007. Plant-herbivore interactions. In: R. F. Barnes, C. J. Nelson, and C. W. Flick, editors, Forages. Vol. 2. 6th ed. Blackwell, Ames, IA. p. 123–136. Coppock, C. E. 1985. Energy nutrition and metabolism of the lactating dairy cow. J. Dairy Sci. 68:3403–3410. Dove, H., and R. W. Mayes. 1991. The use of plant wax alkanes as marker substances in studies of the nutrition of herbivores: A review. Aust. J. Agric. Sci. 42:913–952. Ferrell, C. L. 1988. Energy metabolism. In: D. C. Church, editor, The ruminant animal: Digestive physiology and nutrition. Prentice Hall, Englewood Cliffs, NJ. p. 250–268. Ferrell, C. L., and T. G. Jenkins. 1985. Cow type and the nutritional environment: Nutritional aspects. J. Anim. Sci. 61:725–741. Ferrell, C. L., and T. G. Jenkins. 1987. Influence of biological types on energy requirements. In: M. K. Judkins, D. C. Clanton, M. K. Petersen, and J. D. Wallace, editors, Proc. Grazing Livest. Nutr. Conf. Univ. of Wyoming, Laramie. p. 1–7. Forbes, J. M. 1986. The effects of sex hormones, pregnancy and lactation on digestion, metabolism, and voluntary food intake. In: L. P. Milligan, W. L. Grovum, and A. Dobson, editors, Control of digestion and metabolism in ruminants. Prentice Hall, Englewood Cliffs, NJ. p. 420–435. Funk, M. A., M. L. Galyean, M. E. Branine, and L. J. Krysl. 1987. Steers grazing blue grama rangeland throughout the growing season. I. Dietary composition, intake, digesta kinetics and ruminal fermentation. J. Anim. Sci. 65:1342–1353. Gregorini, P., S. A. Gunter, M. T. Bowman, J. D. Caldwell, C. A. Masino, W. K. Coblentz, and P. A. Beck. 2011. Effect of herbage depletion on short-term foraging dynamics and diet quality of steers grazing wheat pastures. J. Anim. Sci. 89:3824–3830. Gregorini, P., S. A. Gunter, C. A. Masino, and P. A. Beck. 2007. Effects of ruminal fill on short-term herbage intake and grazing dynamics by beef heifers. Grass Forage Sci. 62:346–354. Gillen, R. L., and E. L. Smith. 1986. Evaluation of the dry-weightrank method for determining species composition in tallgrass prairie. J. Range Manage. 39:283–285. Gunter, S. A. 1993. Nutrient intake and digestion by cattle grazing midgrass prairie rangeland and plains bluestem pasture. PhD Diss. Oklahoma State Univ., Stillwater. Gunter, S. A., F. T. McCollum III, R. L. Gillen, and L. J. Krysl. 1993. Forage intake and digestion by cattle grazing midgrass prairie rangeland or sideoats grama/sweetclover pasture. J. Anim. Sci. 71:3432–3441. Gunter, S. A., F. T. McCollum III, R. L. Gillen, and L. J. Krysl. 1995. Diet quality and ruminal digestion in beef cattle grazing midgrass prairie rangeland or plains bluestem pasture throughout the summer. J. Anim. Sci. 73:1174–1186. Hogg, B. W. 1989. Compensatory growth in ruminants. In: A. M. Pearson, editor, Muscle and meat biochemistry. Academic Press, San Diego. Hunter, R. A., and B. D. Siebert. 1986. The effects of genotype, age, pregnancy, lactation, and rumen characteristics on voluntary intake of roughage diets by cattle. Aust. J. Agric. Res. 37:549–560. Illius, A., and I. Gordon. 1987. The allometry of food intake in grazing ruminants. J. Anim. Ecol. 56:989–999. Kleiber, M. 1975. The fire of life. Krieger, Huntington, NY. Koch, R. M., L. A. Swinger, D. Chambers, and K. E. Gregory. 1963. Efficiency of feed use in beef cattle. J. Anim. Sci. 22:486–494. Krysl, L. J., M. L. Galyean, M. B. Judkins, M. E. Branine, and R. E. Estell. 1987. Digestive physiology of steers grazing fertilized and non-fertilized blue grama rangeland. J. Range Manage. 40:493–501. Lardy, G. P., D. C. Adams, T. J. Klopfenstein, and H. H. Patterson. 2004. Building beef cow nutritional programs with the 1996 NRC beef cattle requirements model. J. Anim. Sci. 82(E. Suppl.):E83–E92. Launchbaugh, K. L., and C. T. Dougherty. 2007. Grazing animal Behavior. In: R. F. Barnes, C. J. Nelson, and C. W. Flick, editors, Forages. Vol. 2. 6th ed. Blackwell, Ames, IA. p. 675–686. Lofgreen, G. P., and W. N. Garrett. 1968. A system for expressing net energy requirements and feed values for growing and finishing cattle. J. Anim. Sci. 27:793–806.
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
2784
Coleman et al.
McNamara, J. P., and J. K. Hillers. 1986. Regulation of bovine adipose tissue metabolism during lactation. 1. Lipid synthesis in response to increased milk production and decreased energy intake. J. Dairy Sci. 69:3032–3041. Moore, J. E., M. H. Brant, W. E. Kunkle, and D. I. Hopkins. 1999. Effects of supplementation on voluntary forage intake, diet digestibility, and animal performance. J. Anim. Sci. 77:122–135. Morais, J. A. S., T. T. Berchielli, A. de Vega, M. F. S. Queiroz, A. Keli, R. A. Reis, L. M. A. Bertipaglia, and S. F. Souza. 2011. The validity of n-alkanes to estimate intake and digestibility in Nellore beef cattle fed a tropical grass (Brachiaria brizantha cv. Marandu). Livest. Sci. 135:184–192. Nelson, B. D., C. R. Montgomery, P. E. Schilling, and L. Mason. 1976. Effects of fermentation time on in vivo/in vitro relationships. J. Dairy Sci. 59:270–277. NRC. 1984. Nutrient requirements of beef cattle. 6th ed. Natl. Acad. Press, Washington, DC. NRC. 1996. Nutrient requirements of beef cattle. 7th ed. Natl. Acad. Press, Washington, DC. NRC. 2000. Nutrient requirements of beef cattle. 7th revised ed. Natl. Acad. Press, Washington, DC. Pordomingo, A. J., J. D. Wallace, A. S. Freeman, and M. L. Galyean. 1991. Supplemental corn grain for steers grazing native rangeland during summer. J. Anim. Sci. 69:1678–1687.
Rayburn, E. B. 1986. Quantitative aspects of pasture management. Seneca Trail RC&D Tech. Manual. Seneca Trail RC&D, Franklinville, NY. Rosiere, R. E., J. D. Wallace, and R. D. Pieper. 1980. Forage intake in two-year old cows and heifers grazing blue grama summer range. J. Range Manage. 33:71–73. Sprinkle, J. E., J. W. Holloway, B. G. Warrington, W. C. Ellis, J. W. Stuth, T. D. A. Forbes, and L. W. Greene. 2000. Digesta kinetics, energy intake, grazing behavior, and body temperature of grazing beef cattle differing in adaptation to heat. J. Anim. Sci. 78:1608–1624. St-Pierre, N. R. 2001. Invited review: Integrating quantitative findings from multiple studies using mixed model methodology. J. Dairy Sci. 84:741–755. Tulloh, N. M. 1966. Physical studies of the alimentary tract of grazing cattle. IV. Dimensions of the tract in lactating and non-lactating cows. N. Z. J. Agric. Res. 9:999–1008. Vona, L. C., G. A. Jung, R. L. Reid, and W. C. Sharp. 1984. Nutritive value of warm-season grass hays for beef cattle and sheep: Digestibility, intake and mineral utilization. J. Anim. Sci. 59:1582–1593. Wagner, M. W. 1985. Forage intake and milk production of rangeland beef cows with varying degrees of crossbred influence. MS Thesis. Montana State Univ., Bozeman. Wagner, M. W., K. M. Havstad, D. E. Doornbos, and E. L. Ayers. 1986. Forage intake of rangeland beef cows with varying degrees of crossbred influence. J. Anim. Sci. 63:1484–1490.
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
References
This article cites 34 articles, 17 of which you can access for free at: http://www.journalofanimalscience.org/content/92/7/2775#BIBL
Downloaded from www.journalofanimalscience.org at Universite Laval on June 25, 2014
