| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Animal Environment and Health, Swedish University of Agricultural Science, Skara, Sweden
Swedish Dairy Association, Box 234, 532 23 Skara, Sweden
1 Corresponding author: Evgenij.Telezhenko{at}hmh.slu.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: dairy cattle • preference test • rubber mat • lameness
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The softness of the lying areas has always been the main issue regarding cow comfort (Tucker and Weary, 2001), but more attention is now being paid to cow comfort in the walking areas (Rushen and de Passillé, 2006). Besides providing better comfort, it is assumed that elastic rubber mats in the walking areas may have favorable effects on claw health due to decreased pressure on the claw soles (Cook et al., 2004). Studies of a computer-based finite element model of a bovine claw showed that the maximum pressure peaks were 3 times higher on a hard surface than on a soft one (Hinterhofer et al., 2005). In tie stalls, rubber mats significantly reduced sole lesions compared with concrete (Bergsten and Frank, 1996). The effect of rubber flooring on claw health in walking areas for loose-housed cattle was widely discussed, but there was no strong evidence that rubber mats have a significant effect on reducing claw lesions (Vokey et al., 2001; Vanegas et al., 2006). Nevertheless, many dairy farms with free-stall housing currently use rubber flooring in selected areas where cows spend a considerable amount of time standing or walking.
Given the possible beneficial effect of rubber flooring on claw health (Benz, 2002), should we expect that cows given only partial access to rubber covering would instantly prefer to walk and stand on the rubber mats? Answering this question requires the use of a preference or choice test. Such a test can increase our understanding of the requirements of animals, and thus how their welfare might be improved (Fraser and Matthews, 1997).
The objective of this study was to investigate whether dairy cows prefer to stand and walk on rubber mats rather than concrete floor under the conditions of a normal commercial dairy farm.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
).
|
Rubber mats of different qualities (Gummiwerk Kraiburg Elastik, Tittmoning, Germany) were used to modify the existing solid and slatted concrete floors of the different test sections. The soft rubber mats (KEN; 20 mm thick) and extra soft rubber mats (KSL; 30 mm thick) were used to study standing preferences (experiment 1), and slatted rubber mats (KURA-S) and solid rubber mats (KURA-P; both 22 mm thick) were used to study walking preferences (experiment 2). The KURA-S rubber mats had perforated slots (123 x 38 mm) with a void ratio of approximately 15%, but were otherwise identical to the KURA-P mats. Both KURA-S and KURA-P mats had studded profiles (5-mm studs) on the bottom side of the mat, to provide extra softness. The KEN mats had a design and softness similar to the KURA mats, whereas the KSL mats had larger studs on the bottom (15 mm) and consequently greater softness.
Study Design
Experiment 1.
Floor preference for standing was tested in the holding pen before the cows entered the milking parlor. The holding pen was divided lengthwise into 2 sections (Figure 1
), identical regarding size and design and placement of waterers, but the right side of the parlor was connected with the entrance to the calving pens and was oriented to the north. Before comparing the different flooring materials, the control data, using concrete in both sections, were counted as the proportion of cows standing first on the right side of the holding pen and then on the left side, during 2 d of observation per side (Table 1). The following floor materials were then tested: KSL vs. KEN, KEN vs. solid concrete, and KSL vs. solid concrete. Every alternative was tested for 4 d on each side of the holding pen (Table 1). All observations of cow distribution in the sections were made from video recordings, captured by a video camera positioned 3.5 m above the holding pen, in association with the afternoon milking. Because of twice-daily milking, the cows had actual contacts twice a day with each floor combination. Because observations were made only once daily, the recording of cow distribution was carried out at first, third, fifth, and seventh contact with the particular floor on the particular side of the holding pen. At every shift of the milking side in the milking parlor, the number of animals in the holding pen was reduced by 12; the number of cows standing in each section was counted 5 min after this shift had been made. That resulted in an observation approximately every 7 min. A new management group was introduced after the holding pen was emptied sequentially until the entire herd was milked. The distribution of cows between sections was observed for the entire herd; however, only observations with more than 3.2 m2 of free space per cow within the holding pen were used for the analysis, and so recordings of animal distribution were started with approximately 40 animals in the holding pen for each management group. This cut-off point for stocking density was chosen to give a theoretical possibility of all the cows standing in the same flooring section in the holding pen. Because the number of cows within a group was not always divisible by 12, the number of observations varied from 3 to 4 observations per management group and day.
|
Experiment 2.
Experiment 2 was conducted 1 yr after experiment 1. Walking preference with the different floorings was tested on the 12- x 3-m walkway as the cows left the milking parlor (Figure 1). Only cows from the 2 high-producing management groups (n = 160) were included in the experiment, because they used the whole length of the walkway on their way to the free-stalls. The walkway was divided lengthwise into 2 equal parts, with no differences between sides or surroundings other than the geographical orientation. The following floor materials were tested: KURA-S vs. slatted concrete, KURA-P vs. slatted concrete, and KURA-P vs. KURA-S. Each flooring material was tested for 4 d on each side of the walkway (Table 1). A video camera was placed 4 m above the walkway, and the positions of the cows were recorded continuously during their free passage through the walkway after the afternoon milking. The number of observed contacts with a particular floor on the particular side was the same as described for experiment 1. The traffic flow was approximately 70 cows/h.
The floor choices during walking were characterized as follows:
Data on cows changing side twice was recorded, but was omitted from the subsequent analysis due to its low occurrence. Observations of cow allocation in which any interaction with personnel occurred were omitted.
The "test side" was the side covered with rubber mats when compared with concrete floor, and the side covered with solid rubber mats when compared with slatted rubber mats. For the control data, the right side of the walkway was considered as the test side for the first 2 d, and the left side for the following 2 d.
Occasionally, some cows passed along the walkway alone, that is, with no other cow on the walkway in front or behind them. The floor choices of these "single" cows were recorded separately in the same manner. Lame cows were identified during every observation, and their floor choice was recorded separately. Only cows with moderate and severe lameness (i.e., visible gait asymmetry) were considered lame. Across all the days of observation, about 17% of the cows were categorized as lame.
Statistical Analysis
The analyses of cow preference were carried out on a group level. The first day of observation of each floor section on each side was omitted from the analysis in both experiments due to possible carryover effects. The proportions of cows standing (experiment 1) or walking (experiment 2) on the relevant side were transformed by arcsine of square root of the proportion for the variance stabilizing (Dohoo et al., 2003). A repeated-measures ANOVA was used to analyze the transformed data in both experiments (PROC MIXED procedure; SAS 8.02, SAS Institute, 2003).
The model for experiment 1 included the following sources of variation: the effect of floor combination (extra soft vs. soft, soft vs. concrete, extra soft vs. concrete, and concrete vs. concrete; df = 3); the effect of the side of the holding pen (df = 1); the effect of the number of contacts with a floor on the same side of the holding pen (df = 2); and the effect of the management group to which the cows belonged (defined as a subject in the model, df = 2). Because the proportion of cows standing on a particular side might be based on different numbers of observations within each group and day, the data were weighted by the number of observations for each day and group.
Due to a significant interaction between floor and number of contacts with the floor in experiment 1, the effect of floor was built as 10 categories: control observations and interaction term between floor combination and the number of the contact (df = 9). Other effects in the model were the effect of the side of the walkway (df = 1); the effect of cows walking alone or together with other animals (df = 1); and the effect of the management group to which the cows belonged (defined as a subject in the model, df = 1). Again, as the proportion of cows walking on a particular side might be based on observations of different numbers of animals within each group and day, the data were weighted by the number of observed animals for each group and day. There was a negative correlation between the distributions of lame and nonlame cows (r = – 0.19; P < 0.05); therefore, to reduce the risk of type II error (Sokal and Rohlf, 1995), these 2 subgroups were analyzed separately using the same model. The other possible interactions were tested in both models, but were omitted because of their non-significance. The first-order autoregressive correlation structure for the management group within a tested floor combination and side was used for models in both experiments. The analyses of residuals for models of both experiments revealed no significant violation of the assumptions of normality and homoscedasticity. Multiple comparisons of the least squares means for "floor combination" (experiment 1) and "floor combination x number of contact" (experiment 2) with the control observations were performed using Dunnett adjustment (Dunnett’s 2-tailed t-test). The differences were considered significant at P < 0.05; P values at P 
0.1 were considered as a tendency.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
). There was a tendency for cows to prefer extra soft rubber mats over soft rubber mats. There was no significant effect of the number of contacts with a particular flooring (P = 0.51, F = 0.69, df = 2) and no significant interaction between the number of contacts and a particular flooring (P = 0.83, F = 0.42, df = 5).
|
Experiment 2: Preference for Walking on Different Floorings
Nonlame Cows.
When the rubber mats were tested against concrete floor, the proportion of WTS was significantly higher than the control level, and increased gradually with time of observation on the same side (Figure 3). There was a drop in the proportion of cows walking on the rubber flooring when it was switched to the opposite side (and so the first contact was not used in the analysis). When the solid and slatted rubber mats were tested, the proportion of WTS increased significantly up to the fifth contact, but somewhat decreased at the seventh contact; this did not differ significantly from the control level.
|
The effect of side was significant for all of the 4 categories of animal distribution (WTS: P = 0.003, F = 9.8, df = 1; WOS: P < 0.001, F = 26.9, df = 1; CTS: P = 0.002, F = 11.1, df = 1; COS: P = 0.02, F = 5.94, df = 1). The animals preferred walking on the left side of the walkway. No significant interaction was found between side and floor combination.
The proportion of cows walking on the test side was greater in cows walking alone than in cows walking within a group of herd mates (P = 0.008, F = 7.57, df = 1). The proportion of COS decreased if the animals walked alone (P = 0.01, F = 8.2, df = 1). The presence of more than 1 cow in the walkway did not have a significant effect on other choices along with interactions with floor combination.
Lame Cows.
Among the lame cows, the floor combination of solid and slatted rubber mats did not show any significant effect on floor choice. When the slatted rubber was tested against slatted concrete, the proportion of WTS decreased slightly at the third contact compared with the control level (Figure 4). The proportion increased gradually thereafter, and became significantly higher only at the seventh contact with the floorings. A similar pattern was observed when solid rubber was tested against slatted concrete floor (Figure 4).
|
The proportion of lame cows was more or less constant during the observation period. No significant difference in proportion of lame cows was found between the days when the entire floor was covered with rubber and the days when only concrete was tested. In lame animals, the effect of side was significant for WTS (P < 0.001, F = 43.5, df = 1), WOS (P < 0.001, F = 14.9, df = 1), CTS (P < 0.001, F = 32.2, df = 1) and COS (P = 0.005, F = 8.2, df = 1). Lame animals preferred to walk on the right side of the walkway, but there was no significant interaction between side and floor combination.
The proportion of lame cows walking alone on the test side tended to be larger than the proportion of cows walking together with other animals (P = 0.075, F = 3.26, df = 1). The proportion of COS for lame cows was lower if the animals walked alone (P = 0.04, F = 4.2, df = 1). Other floor section choices were not significantly affected by the presence of other animals on the walkway.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The preference tests in the present study were carried out under conditions of "normal cow life" in a commercial dairy herd with free-stall housing. The complexity of relationships between cows within a herd makes it difficult to evaluate correlations between individuals within a group. The cow is a social animal that may follow herd mates when choosing which flooring to walk on, resulting in a positive correlation between individuals. At the same time, competition for limited space could result in a negative correlation between the observed individuals. To secure reliable conclusions we analyzed the data on a group level, which resulted in a conservative test (Iason and Elston, 2002).
The preference tests are sensitive to the conditions under which the animals were reared, but it was demonstrated that previous experience would not necessarily play a crucial role in the choices an animal makes (Dawkins, 1983). Nevertheless, the animals in this study were given a short time (8 d or 16 contacts) to become accustomed to the rubber flooring at the beginning of the experiments, together with testing preference between different types of the rubber mats.
In experiment 1, the cows in the holding pen seemed to need a greater stimulus to move, because almost every position change between the floorings seemed associated with personnel moving animals to the parlor for milking. Furthermore, the preference for the softer flooring in experiment 1 was not dependent on the number of contacts with a particular floor position. The explanation could be that cows already standing on the softer flooring had a higher willingness to continue to stay there than those on the opposite (harder) side. We could not identify lame cows during experiment 1, and the possible latent effect of lameness is present in the results. The preferences were based on observations of the holding pen when at lower stocking density, and it was quite likely that most of the lame cows were still in the holding pen, because lame cows often enter the parlor late (Hassall et al., 1993). It is difficult to judge how a higher proportion of lame cows would affect the group preference for soft flooring. It is possible that lame cows had stronger motivation to stand on the soft flooring, but the negative correlation between the distributions of lame and nonlame cows in experiment 2 (when healthy cows walked on the test floor, it was less likely that lame cows walked there) suggests that lame cows may avoid the rubber flooring because of competition with other cows.
Experiment 2 was motivated by the difficulty of appreciating the effects of lameness and human interference on the flooring preferences in experiment 1. In experiment 2, cows walked freely, with motivation to move toward eating and resting areas; in this scenario, detection of lameness was possible. Because subtle lameness might be masked by the soft flooring (Telezhenko and Bergsten, 2005), we only categorized animals as lame if they had a very clear locomotion distortion that was visible even on the rubber mats.
The proportion of cows walking only on the rubber mats increased gradually during the 4 d of testing on the same side. In the case of nonlame cows, the application of slatted rubber mats resulted in a 25% increase in the proportion of cows walking on the same side at the seventh contact. Use of solid rubber mats at the seventh contact caused a 30% increase. Dairy cows often persist in their choice of a particular side of milking parlor (Paranhos da Costa and Broom, 2001). We could see that cows preferred a particular side of the walkway or the holding pen regardless of flooring. Once the cows got used to soft flooring on one side, they were reluctant to change their habits even when the flooring was shifted to the opposite side. During experiment 2, when rubber mats were switched to the other side after 4 d, at first the majority of cows continued to walk on the side that had been floored with rubber, a result in concordance with a study by Grandin et al. (1994). Observations during the first contact with the test flooring were therefore omitted from the analysis due to the above-mentioned carryover effect. Nevertheless, the carryover effect did not disappear entirely even at the third and fifth contact, because the proportion of cows walking on the rubber mats on the "new" side increased progressively, in the same way as during the testing on the first side. The choice for soft flooring in lame cows became more distinct during the rubber mat exposure on the same side as well.
The social interactions among the animals affected their preferences for a particular floor. When cows walked alone through the walkway (experiment 2), the proportion of those walking on the rubber mats was higher than when the cows walked in a group with other cows. The explanation for this could be that in those cases when cows crowded onto the section with rubber flooring, other cows had to use the concrete side of the walkway, which would probably not be their primary choice under other conditions. It has been established that high-density husbandry causes increased social aggressiveness in cows (Bouissou et al., 2001). Therefore, the real preference of the cows for the softer floorings could have been underestimated when they were tested in groups. Yet, it was difficult to appreciate the real effect of the group, due to a lack of information regarding the previous experiences (and choices) of single walking animals.
The possible underestimation of the preference for soft flooring is especially plausible for the lame cows; they did not choose to walk on the soft side at the beginning of testing, most probably because of an increased proportion of other cows walking on that side. It was shown that cows with low social rank are more likely to become lame (Galindo and Broom, 2000). Presumably, the low-ranked lame cows in our study could not compete for softer flooring on the same terms as nonlame cows. Competition between cows for softer lying areas was reported by Herlin (1997), who described how cows displaced other animals significantly more from the most comfortable lying surfaces. Similar to nonlame cows, lame animals chose the side with rubber mats more often, and did not change to the side with concrete floor when there were no other cows on the walkway. However, even when walking alone, lame cows did not always choose the rubber mats. This might be explained by previous experience of some lame animals from earlier days of testing, when they were not allowed to walk on that side. The other reason for not choosing the rubber mats could be that the soft surface does not contribute to pain relief for certain types of lameness. Telezhenko and Bergsten (2005) showed decreased step asymmetry when lame cows walked on soft rubber mats; however, the origin of lameness was not established in that study.
When solid rubber mats were compared with slatted rubber mats (experiment 2), the proportion of cows choosing the solid part of the floor was somewhat larger. Cows may have a preference for solid flooring for the same reason they avoid crossing a "cattle guard." Albright and Arave (1997) showed that cattle avoided a cattle guard even if it was just painted across a paved road.
The probable underestimation of the preference for soft flooring in experiment 2 may have arisen from the short period of adaptation to the new floor placement, and the limited space of the walkway. An earlier demonstration of cow preference for a soft track surface by Gregory and Taylor (2002) showed that, when walking in groups, 96% of the cows preferred to walk on wood-chips instead of on hard-core track.
Generally, it seems as if soft flooring is appreciated by dairy cows. Gregory and Taylor (2002) and Tucker et al. (2006) have shown that cows prefer soft flooring on areas other than the laying surface. Nonetheless, other aspects of welfare should be taken into account, such as effects of soft floorings on claw status. Kremer (2006) criticized the use of rubber flooring, because it resulted in claw horn overgrowth due to the low abrasiveness of rubber mats. However, other studies showed that cows kept on rubber flooring showed not only decreased claw horn wear, but also decreased claw horn growth (Vanegas et al., 2006) and that net growth rate on rubber mats did not exceed that on concrete (Vokey et al., 2001). Another critical point concerning rubber matting may be when the addition of rubber mats in alleys in combination with compromised stall comfort increases standing time and may have a negative effect on claw health (Cook et al., 2004).
The results of the present study may be important for the deployment of new soft flooring materials. When rubber mats are used to partially cover the floor in free-stall barns, lame cows (those that probably need the improved flooring most) may not be able to make use of this resource because of competition with nonlame, higher-ranked cows.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication December 21, 2006. Accepted for publication April 19, 2007.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
