| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
School of Veterinary Medicine, University of Wisconsin, Madison 53706-1102
2 Corresponding author: nbcook{at}wisc.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT ACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: feeding behavior • pen design • dairy cow
Dairy cattle demonstrate highly synchronized ‘allelomimetic’ behavior while feeding and resting at pasture, but they modify their behavior considerably when they are intensively housed. Feeding time is typically reduced from around 8 to 9 h/d when grazing to 4 to 5 h/d when housed in free stalls and fed a TMR (Miller and Wood-Gush, 1991). Several studies monitoring the proportion of a group feeding throughout the day in free-stall pens with at least 0.5 m of feed space per cow have demonstrated that it is rare for >80% of the group to be feeding at any one time (Wagner-Storch and Palmer, 2003; DeVries et al., 2004; Endres et al., 2005) and that these periods are confined to the time immediately after fresh feed delivery and return to the pen from the milking parlor (DeVries and von Keyserlingk, 2005). Currently, there are 2 main industry standards in free-stall pen design that provide for very different amounts of bunk space per cow (Smith et al., 2000). Pens with 3 rows of free stalls, when stocked at one cow per stall, provide approximately 0.46 m of bunk space per cow, less than the bunk allowances described in previous studies. In contrast, pens with 2 rows of free stalls, when similarly stocked, provide approximately 0.74 m of bunk space per cow. Therefore, the questions remain: what is a reasonable amount of bunk space to provide cows at different stages of lactation, and what factors limit group feeding behavior?
Factors that appear to limit access to feed include not only the amount of bunk space, but also social factors, such as the rate of aggressive displacements, which appear to increase when feed space is limited (Huzzey et al., 2006), and a behavioral requirement to maintain greater inter-cow distances while feeding (DeVries et al., 2004). When provided >360 m2 of pasture per cow, mean distance between neighbors was maintained within a range of 10 to 12 m per cow (Kondo et al., 1989). Clearly, in a free-stall environment, this ‘optimal spacing between neighbors’ will never be achieved, and a compromise must be made that balances economics with production and health. In a study using small groups of 6 cows in free-stall pens, all of the cows rarely fed at the same time after fresh feed delivery, even when 1.0 m of feed space per cow was provided (DeVries et al., 2004).
Awareness of significant differences in the use of the available feed space by high-yielding cows in pens with either 2 or 3 rows of free stalls would be helpful to farmers faced with the difficult decision of what type of free-stall building to construct. The aim of this observational study was to document the amount of bunk being used in free-stall pens with either 2 or 3 rows of stalls at peak utilization periods including 90-min periods after delivery of fresh TMR or return from milking.
Previously recorded videos of a 24-h period of activity of the high-yielding, mature cow group in 10 free-stall-housed dairy herds were used for the study. Details of herd selection and the methodology used for video capture have been described elsewhere (Cook et al., 2004a). In all 10 herds, cows were fed once daily after the morning milking and were milked either 2 times (3 herds) or 3 times (7 herds)/d. Five herds had a 2-row pen design, and 5 herds had a 3-row pen design. Free-stall base was either a rubber crumb-filled mattress bedded with sawdust (2 herds with 2-row pens and 3 herds with 3-row pens) or deep sand (3 herds with 2-row pens and 2 herds with 3-row pens).
A scan-sampling technique at 10-min intervals was used to record bunk utilization score, which was defined as the proportion of feed spaces in the pen that were filled at each time interval. A feed space was defined in 8 herds as a head-lock, which were all located 0.61 m on center. In 2 herds that used a post-and-rail feed barrier, the number of 0.61-m sections between supporting posts that were located at 3.05-m intervals were counted as available feed spaces. An occupied feed space was defined as one in which the head of a cow passed the plane of the feed bunk curb.
To examine differences in bunk utilization between herds, data were aligned by primary, secondary, and tertiary bunk utilization peaks. A primary peak had a start time coinciding with the maximum feed bunk utilization score occurring after fresh feed delivery, which for the herds included in the study occurred 1.5 to 5 h after the cows left the pen for the morning milking. A secondary peak started at peak bunk utilization score after return from the afternoon milking, and a tertiary peak started at peak bunk utilization score after return from the night milking in herds milking 3 times daily. Neither secondary nor tertiary peaks coincided with fresh feed delivery.
Comparison between pen designs was made using a one-way ANOVA in SAS (SAS Institute, 2001) with log-transformed data where appropriate for milking herd size; pen size; stocking density; rolling herd average milk production; feed space per cow; feed push-up frequency; peak bunk utilization for primary, secondary, and tertiary peaks; and mean bunk utilization score for the 24-h period.
Bunk utilization for a 90-min period from the start of the primary, secondary, and tertiary peaks was compared between 2-row and 3-row pens using PROC MIXED (SAS Institute, 2001) and 2 repeated measures covariance structures, namely compound symmetry and first-order autoregressive. The optimal model was chosen using assessment of the magnitude of Akaike’s information criterion. Bunk utilization was transformed (square or square root) where appropriate based on residual plots, to ensure an approximate normal distribution and to avoid heteroscedasticity. Significance for the effects of rows of stalls in the pen (1 df), time after bunk utilization peak (9 df), and their interactions (9 df) was set at P < 0.05. A similar repeated measures mixed model was used to compare bunk utilization between types of peak by pen row (2-row primary, 2-row secondary, 2-row tertiary, etc); significance was assessed at P < 0.05 for differences between least squares means using Fisher’s protected least significant difference. Endres et al. (2005) have reported differences in feeding behavior between post-and-rail feed barriers and head-locks. We repeated all of the analyses without the two 2-row herds utilizing the post-and-rail design. Although the level of significance was influenced by a reduction in sample size and the bunk utilization scores were slightly higher for herds with 2-row pens just using head-locks, numerical relationships between the 2-row and 3-row comparisons were unaltered. Therefore, the 2 herds that used post-and-rail systems were included in the final analyses. We did not report aggressive displacements, as this was a major factor influenced by bunk design in the study by Endres et al. (2005).
The 2-row pens were slightly more overstocked relative to stalls than 3-row pens (1.17 vs. 1.07 cows per stall, respectively), but feed space per cow was significantly greater at 0.58 m per cow compared with only 0.42 m per cow in 3-row pens (P < 0.001), indicative of the greater number of cows per available length of feed space inherent in 3-row pens than in 2-row pens. This reduction in feed space per cow in 3-row pens was reflected in the greater mean bunk utilization score for the 24-h period of 0.26, compared with only 0.22 in 2-row pens (P = 0.004). Although not statistically different, feed was pushed up 6.0 times/d in 3-row pens compared with an average of 4.6 feed push-ups/d in 2-row pens (P = 0.191; Table 1
). Although some researchers have noted a slight increase in feeding activity in cows experiencing more frequent feed push-ups (Menzi and Chase, 1994), a more recent study concluded that 2 additional feed push-ups/d did not significantly increase feeding activity when compared with a baseline schedule of 2 feedings and 2 feed push-ups/d (DeVries et al., 2003).
|
). Least squares means were 0.71 (0.10) for 3-row pens and 0.65 (0.10) for 2-row pens. Primary peak bunk utilization was not significantly different (P = 0.649, and there was no significant difference in the rate of decline in bunk utilization score over the ensuing 90-min period (P = 0.096) using the first-order autoregressive repeated measures mixed model, which was the preferred covariance structure for all of the time-dependent analyses (Figure 1a).
|
|
); there was a significant pen type by time interaction (P < 0.001). Bunk utilization score started lower and fell slower in 2-row pens compared with 3-row pens. For tertiary peaks in bunk utilization, the effect of pen type was significant (P < 0.001); 2-row pens again were lower than 3-row pens, but the rate of decline was similar (Figure 1c). There was no significant interaction between pen type and time (P = 0.94). Within pen row type, changes in bunk utilization after different types of peak use were significantly different between primary and secondary peaks for 3-row pens (P = 0.007), but not for 2-row pens (P = 0.099). There were no significant differences between primary and tertiary peaks (P = 0.284) and secondary and tertiary peaks (P = 0.151) in 3-row pens, whereas those comparisons were significantly different in 2-row pens (P < 0.001). Bunk utilization score is a crude monitor and only sensitive to large differences in feeding behavior between treatments when data are gathered for a single 24-h period. However, bunk utilization score does provide more detailed information on the amount of space occupied and available at the bunk at peak feeding times in comparison with analyses that report only the proportion of the group feeding. The fact that there were no significant differences in peak and rate of decline in bunk use after primary peaks between pen types confirms that fresh feed delivery is an important feeding stimulus for high-yielding dairy cows. At times of fresh feed delivery, access to the bunk should be maximized as much as possible. However, complete occupancy of the bunk was rare even when fresh feed was delivered in 3-row pens, for which mean feed space per cow was limited to only 0.42 m. Bunk utilization, as a reflection of the ability of a group of dairy cows to feed simultaneously, given the constraints of bunk length and design, stocking density, and pen design, appears to be moderated by several competing influences.
Feed spaces were defined in the current study as being 0.61 m wide, the width of head locks that are commonly installed in North American free-stall barns. Although poorly documented, it is apparent that this may be narrower than the width of many mature Holstein dairy cows. This feed space width may be too narrow to maximize bunk utilization score, and more research is required to determine the most appropriate dimensions and design of each feed space for different sized dairy cows at different stages of their life cycle.
Bunk utilization scores following tertiary peaks were significantly different from other peaks in 2-row pens, but not in 3-row pens. Thus, there appear to be some differences in the drives that determine peak bunk use between 2-row and 3-row pens at times other than fresh feed delivery. Feed space per cow provided in the 2-row pens (0.52 to 0.63 m) when stocked at the densities observed in the herds under observation in this study appeared to be sufficient to prevent crowding at the bunk during peak feeding periods immediately after milking. In contrast, the limited feed space available in the 3-row pens (0.38 to 0.45 m) resulted in increased bunk use after milking compared with 2-row pens. This may reflect feeding activity by subordinate cows that failed to access the bunk when more dominant cows were feeding, as has been shown by Huzzey et al. (2006). Feed sorting behavior in dairy cattle, documented by Leonardi et al. (2005), which leads to differences in the constituents of the TMR over time (DeVries et al., 2005), is obviously of concern in situations where some cows are prevented from accessing feed when they wish. Cook et al. (2004b) argued that altered feeding patterns in individual cows that limit intake when they are hungry may trigger bouts of subacute ruminal acidosis. Therefore, in 3-row pens, it would seem appropriate to recommend pushing up feed more frequently, particularly when cows return to the pen from milking and within 1 to 2 h of fresh feed delivery. Altering feeding behavior by delivering fresh feed more frequently, at times other than milking time, as suggested by DeVries and von Keyserlingk (2005), could also be an important management adaptation to restricted feed space. In contrast, the feed space per cow provided by a 2-row pen appears to improve access to feed at times of peak bunk utilization, perhaps allowing cows to demonstrate other important feeding behaviors, such as avoidance of direct confrontation between cows of different social rank and maintenance of greater inter-cow distances while feeding. Whether these behaviors have any beneficial effects on health and productivity in 2-row pens has yet to be determined.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTACKNOWLEDGEMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Received for publication June 24, 2005. Accepted for publication September 2, 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. E. Fernandez, R. A. Marino, and X. C. Carreira Algorithms for dairy barn design: Maternity and milking areas J Dairy Sci, May 1, 2009; 92(5): 2276 - 2296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. B. Walsh, D. F. Kelton, T. F. Duffield, K. E. Leslie, J. S. Walton, and S. J. LeBlanc Prevalence and Risk Factors for Postpartum Anovulatory Condition in Dairy Cows J Dairy Sci, January 1, 2007; 90(1): 315 - 324. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
