- Email: [email protected]
Assessing cow comfort: effects of two floor types and two tie stall designs on the behaviour of lactating dairy cows
Applied Animal Behaviour Science 71 (2001) 105±117
Assessing cow comfort: effects of two ¯oor types and two tie stall designs on the behaviour of lactating dairy cows D.B. Haley*, A.M. de PassilleÂ, J. Rushen Dairy and Swine Research and Development Centre, Agriculture and Agri-Food Canada, P.O. Box 90, 2000 Route 108 East, Lennoxville, Que., Canada J1M 1Z3 Accepted 26 August 2000
Abstract Our objective was to test the effects of two types of stall ¯ooring and two tie stall designs on the behaviour of dairy cows. Using two cross-over designs, 16 lactating Holstein cows were housed for 3-week periods in each of four tie stalls to compare the effects of concrete versus mattress ¯ooring, and a narrow versus a wide stall front opening. When cows were housed on a mattress ¯ooring they increased their total lying time by 1.8 h per day (51.0 versus 43.4% time per day). The duration of individual lying bouts was longer on concrete ¯ooring (78 versus 62 min) and the same was true for the duration of individual bouts of standing (80 versus 48 min). On mattress ¯ooring cows stood up and lay down more often than on concrete (28 versus 20 per day, respectively). On concrete cows spent more time standing without eating (35.2 versus 28.1% of sampling observations). Cows consistently made use of the additional space provided with the wide stall front opening, placing their head in the manger area during 70.6% of lying observations compared to only 43.4% of observations in the narrow stall front. However, measures of general activity were not signi®cantly different for the two stall designs. Cows spent similar amounts of time eating and consumed similar quantities of feed in all four tie stalls. Our results suggest a lack of comfort may be apparent in reduced time spent lying and a subsequent increase in time spent standing without eating. A lower frequency of standing and lying and an increase in the bout duration of lying suggests that cows on concrete ¯ooring have dif®culty standing up and lying down. Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. Keywords: Cattle-housing; Cattle-resting behaviour; Comfort
* Corresponding author. Present address: Department of Large Animal Clinical Sciences, Western College of Veterinary Medicine, University of Saskatchewan, 52 Campus Drive, Saskatoon, Sask., Canada S7N 5B4. Tel.: 1-306-966-7056; fax: 1-306-966-7159. E-mail address: [email protected] (D.B. Haley).
0168-1591/01/$ ± see front matter Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 5 9 1 ( 0 0 ) 0 0 1 7 5 - 1
106
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
1. Introduction Lying is an important behaviour for cattle occupying approximately 50% of their daily time budget (Ruckebusch, 1974; Metz, 1985; Dechamps et al., 1989; Krohn and Munksgaard, 1993). For lactating animals, blood ¯ow to the mammary gland is increased during lying (Metcalf et al., 1992; Rulquin and Caudal, 1992) and when cows are deprived of lying, plasma concentration of growth hormone is reduced which is likely to affect milk production (Munksgaard and Lùvendahl, 1993). Disturbed rest can increase hypothalamopituitary activity (Ladewig and Smidt, 1989) and activity of the sympathetic nervous system in cattle (MuÈller et al., 1989). In order to ensure the welfare and productivity of dairy cattle, it is important to optimize their resting time. Housing facilities can substantially affect the activity and resting behaviour of dairy cows. For example, Haley et al. (2000) found that housing cows in large pens with a mattress ¯ooring increased lying time by 4 h per day compared to housing them in tie stalls with concrete ¯ooring. This occurred as a result of a reduction in the amount of time spent standing without eating. However, in that study the individual effects of spatial restriction, tethering and ¯oor softness were not differentiated experimentally. A variety of stall designs and ¯ooring are available, but little information is available on how they impact cow comfort. A number of studies have examined the effects of housing environments on standing and lying behaviour of cattle (for a review, see Lidfors (1989)). To reduce the incidence of mastitis, many dairy farmers are reducing their use of organic bedding material (Elbers et al., 1998), and a number of alternative ¯ooring materials are available (Nilsson, 1988). Some studies have examined the effects of ¯ooring properties on their thermal and frictional characteristics (Nilsson, 1988). However, few controlled studies are available comparing the effects of ¯ooring softness on cow behaviour. MuÈller et al. (1989) found cows in tie stalls without any bedding showed more investigation of the lying area, using sweeping head movements performed close to the ¯oor, than cows in groups with bedding. It has been suggested that this re¯ects hesitation by the animal to lie down. Young cattle perform more of this behaviour immediately after tethering compared to after a period of adaptation (Jensen, 1999). Many recommendations are available regarding the optimum size of tie stalls. However, Krohn and Munksgaard (1993) examined the effects of some tie stall design features and found an effect of the type of tether and stall front on the cows' use of space within the stall. The dynamic space required by the cow to lie down or stand up includes an additional 150 cm above and beyond the static size of the animal (Tschanz and KaÈmmer, 1977). This ``lunging space'' is likely to be reduced with certain types of tethers or stall front designs. The purpose of this study was to examine the effects of ¯oor softness and two different stall front designs on the activity and resting behaviour of lactating dairy cows. 2. Materials and methods 2.1. Animals We used 16 lactating Holstein cows with a mean (x S:T:D:) body weight of 625:1 61:8 kg and body length (de®ned as topline length from nose to tail) of
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
107
247:6 9:28 cm. Cows averaged 149:2 34:7 days of lactation at the beginning of the experiment and on average were in their second lactation (range 1±5). During the experimental period cows consumed an average of 35:35 0:70 kg of total mixed ration per day and produced 24:71 0:47 kg of milk per day. 2.2. Housing and management Cows were cared for according to the recommended code of practice for dairy cattle (Agriculture Canada, 1990). Prior to the experiment all cows were housed in tie stalls (``Comfort H-500'', J. Houle and Sons, Inc., Drummondville, Que.) (1:95 m 1:30 m, Fig. 1) on rubber-®lled geotextile mattress ¯ooring (Pasture Mat1, Promat, Ltd., Seaforth, Ont.) with a 2 cm covering of chopped straw. The barn was arti®cially ventilated and lights were on from 06.00 to 22.00 h (>80 lx). During the dark period red lights (<5 lx) were left on to enable ®lming with light-sensitive cameras. A total mixed ration was fed to ad libitum intake. Each day at 14.00 h a weighed quantity of fresh feed was delivered to each animal. Leftovers were removed daily and weighed on the same 5 days each week throughout the experiment. Individual feed intake was then calculated based on feed intake recorded 5 days each week (Monday±Friday). Water was drunk from ¯oat-activated water bowls shared by adjacent cows. Bowls were cleaned every 2 weeks. Cows were milked twice each day (06.00 and 18.00 h) in a milking parlour and individual milk yields were recorded electronically by Metatron milk meters (Westfalia, Elk Grove Village, IL).
Fig. 1. Diagram showing the tie stall design used to house cows prior to the experiment. The same stall front design was used for both treatments in the flooring comparison and also served as the narrow opening treatment in the stall front comparison.
108
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
2.3. Experimental All experimental methods for this study were approved by the Institutional Animal Care Committee at the Dairy and Swine Research and Development Centre in compliance with the guidelines of the Canadian Council on Animal Care. We compared the effects of concrete and mattress ¯ooring surfaces on the behaviour of cows kept in tie stalls. We also compared the behaviour of cows when they were housed in two tie stalls of different design: one with a narrow opening at the front of the stall (Fig. 1) and one with a wide front opening (Fig. 2). We conducted these studies as two separate cross-over designs. The same 16 cows were subjects for both comparisons and every cow spent three consecutive weeks in each of the four stalls. Half the cows started with the ¯ooring treatments and half with the stall front designs. After 6 weeks, once cows had spent 3 weeks in each of two stalls of one comparison, they were moved over to the other comparison. To facilitate the rotation of the animals during the experiment we grouped them into four blocks of four cows each, on the basis of body weight and body length. 2.3.1. Flooring comparison Stalls used to examine the effects of tie stall ¯ooring measured 1:80 m 1:30 m and had the same stall front design as Fig. 1. The only difference between the two stalls was the ¯ooring: half of the stalls had concrete ¯ooring and half had a geotextile mattress ¯ooring which consisted of multiple cells ®lled with uniformly sized rubber crumbs (Pasture Mat1, Promat, Ltd., Seaforth, Ont.). The stalls were in one row and we alternated pairs of stalls with regard to ¯ooring to balance for position within the row. Both concrete and mattress stalls were provided with a fresh 2 cm covering of chopped straw on a daily basis.
Fig. 2. Diagram showing the tie stall design with the wide front opening.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
109
2.3.2. Design of stall front comparison Stalls used to compare the effects of two stall front designs were 20 cm longer than those used in the ¯ooring comparison (2:00 m 1:30 m). Half of the stalls had a narrow opening between the stall front and the manger (30 cm, Fig. 1). The other half of the tie stalls were of a different design with a front partition (a wide opening, 120 cm, Fig. 2) (``Comfort Plus H-1100'', J. Houle and Sons, Inc., Drummondville, Que.) that provided more open access to the manger and presumably more lunging space to facilitate the cows' ability to change position from lying to standing and vice versa. Tether chains in the wide stall were positioned at a ®xed point vertically (45 cm from the ¯oor) while the tether in the narrow stalls were not ®xed and could slide vertically up and down the bars to which they were fastened. However, both chains were standardized to allow the same maximum perpendicular distance from the stall-manger border (80 cm). The stalls were in a single row and we alternated the stall fronts within the row. All stalls had the same mattress ¯ooring as described in the ¯ooring comparison and a fresh daily covering of chopped straw to a depth of 2 cm. 2.4. Behavioural observations During the 3 weeks spent in each housing condition all cows were videotaped for four 24 h periods: day 1 (the ®rst day in each system) and days 9, 13 and 17. Time-lapse video recorders (Panasonic AG-6730; Panasonic, Secaucus, NJ) were used in 24 h time mode, each recording the output from four video cameras (Panasonic WV-BL200; Panasonic, Secaucus, NJ) using a video multiplexer (Uniplex Sprite Video Multiplexer; Dedicated Microcomputers Ltd., Reston, VA). Two adjacent cows could be viewed on a single camera and for a complete view we positioned one camera in front and one camera behind each pair of animals. All video recordings were ®rst read by continuous observation. We recorded the general activity patterns of each animal (see Table 1) from all observation days and watched tapes at 12 normal recording speed using The Observer1 Video Tape Analysis System (Noldus Information Technology, Wageningen) and a video time code reader. To obtain more detailed information on resting postures and eating behaviour, all tapes were reviewed a second time using a scan sampling technique such that each cow was observed during a 3 min period every 12 min (120 observations per cow per day). This process was again done using The Observer1 Video Tape Analysis System. We found a very high positive correlation between actual lying time from continuous observations and estimated lying time using our scan sampling method (r p 0:99). With scan sampling, body position of the cow at the beginning of the 3 min window was recorded as either standing or lying. If a cow was standing, we recorded whether the cow ate or whether it stood for the whole period without eating, and whether or not the cow performed sweeping movements with the head during each 3 min observation period (see Table 1 for operational de®nitions). Head sweeping movements have been reported to occur prior to lying and reported to increase in tie stalls compared to loose housing (MuÈller et al., 1989). If the cow was lying at the start of the 3 min time window, we recorded the position of the head and leg positions (see Table 1) at the beginning of the time interval.
110
Table 1 Descriptions of behaviour recorded by continuous and scan sampling methods during 24 h observation periods Behaviour
Description
Continuous
Standing Standing in gutter Kneeling Lying right side Lying left side Lying
Weight supported by four legs Standing with one or both hind legs positioned in the dung gutter behind the stall (20 cm depth) The body is in a position with both knees bent and hind legs extended (minimum time 10 s) The body is resting on the ground with the right side in contact with the floor The body is resting on the ground with the left side in contact with the floor Lying right side lying left side
Scan sampling
Eating Not eating Head sweeping
The cow was observed rummaging in the feed manger followed by chewing movements The cow was standing but did not eat once during the 3 min observation window With the nose positioned 20 cm from the floor (judged relative to the 20 cm height of the stall-manger border) the head was moved from side-to-side in a continuous movement with at least two changes in direction and covering a lateral distance equal to or greater than the width of the animals shoulders The body is resting on the ground The body is resting flat on one side with the legs of the supported underside extended and the head resting on the ground The cow is lying on the sternum, the head is raised off the ground The cow is lying on the sternum, the head is turned back resting against the body The cow is lying on the sternum, the head is stretched resting on the floor Recorded if any part of the cow's nose was encroaching the manger area (beyond the stall-manger border) Recorded if the nose was behind the stall-manger boarder The exposed front leg is tucked under the body (full plantar-flexion at the humoral joint) The exposed front leg is extended in front of the body The exposed hind leg is positioned at an angle of 908 or greater in relation to the body axis The exposed hind leg is positioned at an angle between 45 and 908 in relation to the body axis The exposed hind leg is positioned at an angle of less than 458 in relation to the body axis
Lying on sternum Lying on side Head upright Head back Head on ground Head in manger Head in stall Front leg tucked Front leg exposed Hind leg extended Hind leg mid-position Hind leg tucked
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
Method
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
111
2.5. Statistical analysis From continuous observations we analyzed the frequency of bouts, mean bout durations and the total duration of time spent standing and lying by analysis of variance (SAS, 1990). The following factors were used in the analysis: block, cow identity (as a random factor nested within block), treatment and the interaction between treatment and block. The effect of block was tested against cow identity as an error term. Planned contrasts were used to examine differences between the two stall types and the two types of ¯ooring. From observations based on interval sampling, the frequency of observation for lying, standing, eating and standing not eating, as well as head and leg positions while lying were all analyzed using the same model. 3. Results 3.1. Flooring comparison Complete results for the ¯ooring comparison are presented in Tables 2±4. No signi®cant effect of the day of observation, and no interaction between the day of observation and the treatment was noted. When on mattress ¯ooring, cows increased their lying time by 18% compared to the time they spent lying on concrete ¯ooring, while on concrete cows spent more time standing. On concrete the cows spent 1.7 h more standing without eating than when they were kept on the mattress ¯ooring. There was no difference between concrete and mattress ¯ooring in the amount of time cows spent eating (4.4 h per day). We found no signi®cant difference between concrete and mattress ¯ooring with regard to the number of times cows were observed head sweeping. Cows changed position more frequently when on mattress ¯ooring than they did on concrete as measured by the total number of times cows got up and lay down. The average Table 2 Effect of concrete and mattress flooring in tie stalls on measures (mean S:E:) of lying and standing by lactating dairy cows (n 16) based on four 24 h observation periods P-valuesa
Behaviour
Concrete
Mattress
Lying Total time lying (h per day) Frequency (per day) Bout duration (min)
10.42 0.42 9.05 0.94 77.71 6.70
12.25 0.33 13.13 1.12 61.92 5.08
Standing Total standing time (h per day) Frequency (per day) Bout duration (min) Standing not eating (% of total observations) Standing eating (% of total observations) Head sweeping (% of total observations)
12.87 10.77 80.14 35.24 18.74 3.01
11.04 14.84 48.35 28.11 18.32 3.10
a
d.f. 1, 36.
0.42 0.93 7.53 2.01 0.72 0.66
3.68 1.15 3.68 1.38 0.54 0.61
0.001 0.001 0.01 0.001 0.001 0.001 0.01 >0.10 >0.10
112
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
Table 3 Effect of concrete and mattress flooring in tie stalls on the head position (mean S:E:) taken by lactating dairy cows (n 16) while lying based on four 24 h observation periods Head position (% of total observations)
Concrete
Mattress
P-valuesa
Back against body On ground Upright
4.28 0.58 1.51 0.45 44.36 1.50
4.80 0.43 1.46 0.46 37.36 1.84
>0.10 >0.10 0.001
a
d.f. 1, 36.
duration of individual standing bouts when cows were on concrete was 60% longer than the duration of standing bouts when the cows were on mattress ¯ooring. The difference in average lying bout duration was also signi®cantly longer when the cows were on concrete (Table 2). Cows rested with their head back against the body or resting on the ground for similar amounts of time on concrete or mattress ¯ooring. Cows on the mattress ¯ooring spent more time lying with their head upright, which was the most common head position (Table 3). The only signi®cant differences in the leg positions between treatments were with the most commonly observed positions; cows on mattresses spent more time lying down with their front legs tucked underneath the body and hind legs in the middle range position (Table 4). 3.2. Stall front design comparison No signi®cant effect of the day of observation, and no interaction between the day of observation and the treatment was noted. We found no signi®cant differences between stall front designs in the measures related to general activity of the cows. On an average cows spent 53.85% of their time lying down (narrow 12:57 0:42 versus wide 13:28 0:44 h per day; NS). On an average cows had 12.5 lying bouts each day which lasted a mean Table 4 Effect of concrete and mattress flooring in tie stalls on the mean (S.E.) percentage of observations for the various leg positions of lactating dairy cows (n 16) while lying based on four 24 h periods Leg positiona (% of total observations)
Concrete
Mattress
P-valuesb
Front leg Extended Tucked
2.77 0.76 40.40 1.77
2.34 0.86 48.28 1.22
>0.10 0.001
Hind leg Extended Middle Tucked
3.37 179 33.51 2.13 6.29 1.04
4.17 1.35 39.39 1.68 7.07 1.23
>0.10 0.05 >0.10
6.14 1.04 0.54 0.29
6.92 1.22 0.68 0.43
>0.10 >0.10
Combinations Both exposed legs tucked Both exposed legs extended a b
Only positions of the exposed front and hind legs were recorded. d.f. 1, 36.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
113
Table 5 Effect of narrow and wide tie stall fronts on the mean (S.E.) percentage of observations for the various leg positions of lactating dairy cows (n 16) while lying based on four 24 h periods Leg positiona (% of total observations)
Narrow
Wide
P-valuesb
Front leg Extended Tucked
1.33 0.54 51.02 1.60
0.66 0.21 54.20 1.79
>0.10 >0.10
Hind leg Extended Middle Tucked
4.94 1.38 39.39 1.68 8.70 1.85
5.95 1.56 38.72 2.29 10.37 2.38
>0.10 >0.10 >0.10
8.61 1.83 0.31 0.13
10.34 2.37 0.13 0.04
>0.10 >0.10
Combinations Both exposed legs tucked Both exposed legs extended a b
Only positions of the exposed front and hind legs were noted. d.f. 1, 36.
time of 66 min with no signi®cant differences between the two stall front designs. There were no signi®cant differences in total time spent standing (narrow 10:67 0:42 versus wide 9:98 0:45 h per day; NS), the frequency of standing (narrow 14:14 0:91 versus wide 15:88 1:40 h per day; NS) or the mean duration of standing bouts (narrow 48:71 4:11 versus wide 43:11 4:79 h per day; NS). No differences between stall designs were found for the percentage of observations where cows were standing and not eating (narrow 25:89 1:74 versus wide 24:78 2:26 h per day; NS), standing eating (narrow 18:70 0:78 versus wide 17:28 0:73 h per day; NS) or the percentage of observations where head sweeping was observed (narrow 2:55 0:43 versus wide 1:66 0:33 h per day; NS). Stall front opening did not in¯uence the leg positions assumed by cows while lying (Table 5). Stall front design, however, had signi®cant effects on where the cows placed their head while lying. Cows used the additional space made available to them by the wide stall front opening, positioning their head in the manger area while lying signi®cantly more than cows in stalls with a narrow stall front (Table 6). Accordingly, cows in stalls with narrow fronts Table 6 Effect of narrow and wide tie stall fronts on the head position of lactating dairy cows (n 16) while lying (mean percentage S.E.) based on four 24 h observations Head position (% of total observations)
Narrow
Wide
Area of head placement In manger In stall
20.65 4.18 26.88 3.89
34.91 3.71 26.88 3.89
Head position Back against body On ground Upright
4.82 0.44 0.67 0.24 46.81 1.77
5.42 0.80 0.30 0.10 49.10 1.80
a
d.f. 1, 36.
P-valuesa 0.001 0.001 >0.10 >0.10 >0.10
114
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
spent signi®cantly more time with their head completely behind the stall-manger partition (Table 6). The relative position of the head to the body during lying did not differ signi®cantly between the two stall front designs (Table 6). 4. Discussion Stall ¯ooring had marked effects on the cow's behaviour. In the present experiment, the daily percentage of time spent lying (concrete 43:4%, mattress 51:0%) were within the range of values reported in previous studies (Dechamps et al., 1989; Krohn and Munksgaard, 1993). With the softer mattress ¯ooring cows increased their lying time by an average of 1.8 h per day compared to when they were housed on concrete ¯ooring. Cows on concrete increased the time they spent standing without eating, while time spent eating was unaffected. Despite the fact that the cows on softer ¯ooring lay down for longer over the whole day, the average duration of individual bouts of lying was longer when cows were housed on concrete. On the mattress ¯ooring, cows increased their total lying time by lying down more frequently, but for shorter periods of time. MuÈller et al. (1989) and Ladewig and Smidt (1989) also found reduced frequency of position changes and an increased duration of lying bout length when bulls were tethered on concrete ¯ooring compared to when they were loose housed on straw. We also found the same for bouts of standing: cows stood up more frequently and remained standing for shorter periods when they were housed on mattress ¯ooring. These results suggest that the cows were more reluctant to stand up and lay down on the harder concrete ¯ooring surface because of discomfort experienced during the process of standing up and lying down. Metzner (1978) calculated that for a 600 kg animal, the maximal load force applied to the front knees is approximately 4 kN. Softer ¯ooring deforms under pressure which increases the contact area and dampens the force impact on the knees (Dumelow, 1995). Previously, we found that cows kept on concrete ¯ooring for 16 weeks had a higher incidence of swelling in their front knees compared to cows housed on rubber mats (Rushen et al., 1998; Haley et al., 1999). Physical changes such as these can result from force trauma and provide further evidence that changing position on concrete surfaces may be painful. Aside from hardness, other characteristics of the ¯ooring surface may also be partly responsible for the changes in behaviour. For example, differences in the coef®cient of friction between the two ¯ooring surfaces may have affected cow behaviour by altering the ease with which cows could change position. Concrete ¯ooring can be quite slippery when wet. In a previous study, Haley et al. (2000) compared cow behaviour in two types of housing, and found untethered cows loose-housed individually in box stalls with a mattress ¯ooring spent 61.4% of their time lying, 4.2 h more time per day than cows tethered in tie stalls with concrete ¯ooring. The present experiment suggests that the effects of softer ¯ooring alone cannot explain the results of Haley et al. (2000) since the difference in lying time between mattress and concrete ¯ooring was only 1.8 h. This suggests that tethering and/or some aspect of spatial restriction, associated with the tie stall was also reducing lying time. With regard to the body positions of cows during lying, the only effect of ¯ooring type was on the most commonly observed postures; lying in a sternal position with the head
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
115
upright, front legs tucked under the body and the hind legs in the middle position. There were no differences in the amount of time cows spent lying with their head resting on the ¯oor or time spent lying in lateral recumbency (¯at on their side). Both of these positions were detected very infrequently. Thus, the hardness of the ¯ooring did not appear to prevent cows from assuming particular positions while lying. However, it may be worth noting that this experiment was conducted during the summer months, and given differences in the thermal properties of the ¯ooring materials (Nilsson, 1988) the present results may be different during cold winter months. Previous research has shown cattle that were tethered or kept in stanchions on concrete perform more head sweeping movements than loose-housed cattle and it has been said this is investigative behaviour and is indicative of hesitation to lie down (Krohn and Munksgaard, 1993, ``examining behaviour''; MuÈller et al., 1989). As a support of this hypothesis, Krohn and Munksgaard (1993) also found tethered cattle showed more interrupted attempts at lying. We had hypothesized that when cows were kept on concrete ¯ooring they would perform more head sweeping, but we did not ®nd this: based on our operational de®nition of the behaviour, ¯oor type had no signi®cant effect on head sweeping movements. The ¯ooring type did not affect the time the cows spent eating. The priority of feeding behaviour, especially for high producing lactating cows, may have limited the possibility of detecting short-term differences in feeding due to housing effects. The design of the stall front affected the way cows used the stall space during lying. When additional space was made available to cows by providing them with a wide stall front opening, cows used the space, placing their head in the manger area during 35% of observations compared to only 21% of observations in stalls with a narrow opening. Detailed analysis of head positions in tie stalls has shown that in stalls with more open fronts, the mean position of the cow's head is 15±20 cm closer to the front of the stall than in stalls with less open fronts and the time that their head is in the neighbouring cow's stall is greatly reduced (Jahodova, 1997). Despite this, the design of the stall front did not signi®cantly affect overall activity nor did it affect the type of postures that the cows adopted when lying. From casual observation, it appeared that cows in stalls with a narrow front had more dif®culty standing up and often had to contort their neck into strange positions compared to normal lunging behaviour. However, we did not ®nd any effect of the stall front on the frequency with which the cows stood up or lay down. One possible reason we did not ®nd any signi®cant difference in lying time between the two stall fronts is that during previous lactations, the cows in this study were housed in stalls with a narrow front design. Prior experience may have reduced the probability of ®nding signi®cant effects of stall front since cows may have previously adapted to the differences in spatial restriction imposed by the narrow stall front. Using subjects that were naive to either stall front design, or to tethering, or both, may have yielded different results. Recent work by Jensen (1999) has brought attention to the dif®culty experienced by heifers in adapting to tethering. Alternative stall front designs that offer more lunging space may be of particular bene®t to cattle during initial tethering. Although lying times were the same for narrow and wide stall front openings, analysis of the biomechanical movements when cows change position from lying to standing (and vice
116
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
versa) may show differences in other measurable aspects of lying behaviour (e.g. time required to stand up). These measures may re¯ect important differences which are affected by stall front design. Based on the camera angles used for ®lming in the present study we were unable to do any temporal analysis of standing and lying movements. 5. Conclusions Increasing the softness of ¯ooring in tie stalls with a geotextile mattress substantially increases the time that the cows spend lying down, which is an advantage for cow comfort, and probably overall production. The effect occurs mainly because cows appear to be reluctant to stand up or lie down on hard concrete ¯ooring, rather than because of obvious discomfort when resting. A more open front for the tie stall appears to allow the cow to rest with its head in the manger, and so will affect the cow's use of the space within the stall. The physical dimensions of a tie stall (length and breadth) alone cannot ensure that the animal has adequate space if the front of the stall is poorly designed. However, we found no evidence to indicate that this affects overall activity, so we cannot conclude any effects on cow comfort. Acknowledgements This research received ®nancial support from the FeÂdeÂration des producteurs de lait du QueÂbec and the Agriculture and Agri-Food Canada MII fund. We thank Isabelle Blanchet, Aleksija Neimanis, Marjolaine St-Louis and Tristan Welp for their technical assistance and the staff at the Dairy Centre for looking after the animals. References Agriculture Canada, 1990. Recommended code of practice for the care and handling of dairy cattle. Dechamps, P., Nicks, B., Canart, B., Gielen, M., Istasse, L., 1989. A note on resting behaviour of cows before and after calving in two different housing systems. Appl. Anim. Behav. Sci. 23, 99±105. Dumelow, J., 1995. Testing cubicle mats for dairy cows. Agric. Eng. 50, 17±21. Elbers, A.R.W., Miltenburg, J.D., De Lang, D., Crauwels, A.P.P., Barkema, H.W., Schukken, Y.H., 1998. Risk factors for clinical mastitis in a random sample of dairy herds from the southern part of The Netherlands. J. Dairy Sci. 81, 420±426. Haley, D.B., Rushen, J., de PassilleÂ, A.M., 1999. Effects of softer flooring on the behaviour, health and productivity of dairy cows in tie stall housing. In: Proceedings of the 33rd International Congress of the International Society for Applied Ethology, Lillehammer, Norway, 17±21 August, p. 127. Haley, D.B., Rushen, J., de PassilleÂ, A.M., 2000. Behavioural indicators of cow comfort: activity and resting behaviour of dairy cows in two types of housing. Can. J. Anim. Sci 80, 257±263. Jahodova, J., 1997. Effect of housing conditions on welfare of dairy cattle. M.Sc. Thesis, University of Edinburgh, Edinburgh. Jensen, M.B., 1999. Adaptation to tethering in yearling dairy heifers assessed by the use of lying down behaviour. Appl. Anim. Behav. Sci. 62, 115±123. Krohn, C.C., Munksgaard, L., 1993. Behaviour of dairy cows kept in extensive (loose housing/pasture) or intensive (tie-stall) environments. II. Lying and lying-down behaviour. Appl. Anim. Behav. Sci. 37, 1±16.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
117
Ladewig, J., Smidt, D., 1989. Behavior, episodic secretion of cortisol, and adrenocortical reactivity in bulls subjected to tethering. Horm. Behav. 23, 344±360. Lidfors, L., 1989. The use of getting up and lying down movements in the evaluation of cattle environments. Vet. Res. Commun. 13, 307±324. Metcalf, J.A., Roberts, S.J., Sutton, J.D., 1992. Variations in blood flow to and from the bovine mammary gland measured using transit time ultrasound and dye dilution. Res. Vet. Sci. 53, 59±63. Metz, J.H.M., 1985. The reaction of cows to a short-term deprivation of lying. Appl. Anim. Behav. Sci. 13, 301± 307. Metzner, R., 1978. Analyse tierischer BewegungsablaÈufe zur Gestatung artemaÈber Rinderkrippen. Landtechnik 9, 397±404. MuÈller, C., Ladewig, J., Thielscher, H.H., Schmidt, D., 1989. Behavior and heart rate of heifers housed in tether stanchions without straw. Physiol. Behav. 46, 751±754. Munksgaard, L., Lùvendahl, P., 1993. Effects of social and physical stressors on growth hormone levels in dairy cows. Can. J. Anim. Sci. 73, 847±853. Nilsson, C., 1988. Technical design with respect to the biological needs of animals in reference to the thermal, friction and abrasive characteristics and the softness of the flooring material. Swedish University of Agricultural Sciences, Lund. Ruckebusch, Y., 1974. Sleep deprivation in cattle. Brain Res. 78, 495±499. Rulquin, H., Caudal, J.P., 1992. Effects of lying or standing on mammary blood flow and heart rate in dairy cows. Ann. Zootech. 41, 101. Rushen, J., de PassilleÂ, A.M., Haley, D.B., 1998. The effects on productivity, health and behaviour of housing lactating cows on NRI ``Cloud 9'' rubber mats. Report to NRI, Inc. SAS, 1990. SAS1 User's Guide: Statistics, Ver. 6, 4th Edition. SAS Institute, Inc., Cary, NC, 1990. Tschanz, B., KaÈmmer, P., 1977. Verhaltensbiologische AnsaÈtze zur beurteilung von Liegeboxen. Der TierzuÈchter 29, 151±153.
Assessing cow comfort: effects of two ¯oor types and two tie stall designs on the behaviour of lactating dairy cows D.B. Haley*, A.M. de PassilleÂ, J. Rushen Dairy and Swine Research and Development Centre, Agriculture and Agri-Food Canada, P.O. Box 90, 2000 Route 108 East, Lennoxville, Que., Canada J1M 1Z3 Accepted 26 August 2000
Abstract Our objective was to test the effects of two types of stall ¯ooring and two tie stall designs on the behaviour of dairy cows. Using two cross-over designs, 16 lactating Holstein cows were housed for 3-week periods in each of four tie stalls to compare the effects of concrete versus mattress ¯ooring, and a narrow versus a wide stall front opening. When cows were housed on a mattress ¯ooring they increased their total lying time by 1.8 h per day (51.0 versus 43.4% time per day). The duration of individual lying bouts was longer on concrete ¯ooring (78 versus 62 min) and the same was true for the duration of individual bouts of standing (80 versus 48 min). On mattress ¯ooring cows stood up and lay down more often than on concrete (28 versus 20 per day, respectively). On concrete cows spent more time standing without eating (35.2 versus 28.1% of sampling observations). Cows consistently made use of the additional space provided with the wide stall front opening, placing their head in the manger area during 70.6% of lying observations compared to only 43.4% of observations in the narrow stall front. However, measures of general activity were not signi®cantly different for the two stall designs. Cows spent similar amounts of time eating and consumed similar quantities of feed in all four tie stalls. Our results suggest a lack of comfort may be apparent in reduced time spent lying and a subsequent increase in time spent standing without eating. A lower frequency of standing and lying and an increase in the bout duration of lying suggests that cows on concrete ¯ooring have dif®culty standing up and lying down. Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. Keywords: Cattle-housing; Cattle-resting behaviour; Comfort
* Corresponding author. Present address: Department of Large Animal Clinical Sciences, Western College of Veterinary Medicine, University of Saskatchewan, 52 Campus Drive, Saskatoon, Sask., Canada S7N 5B4. Tel.: 1-306-966-7056; fax: 1-306-966-7159. E-mail address: [email protected] (D.B. Haley).
0168-1591/01/$ ± see front matter Crown Copyright # 2001 Published by Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 5 9 1 ( 0 0 ) 0 0 1 7 5 - 1
106
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
1. Introduction Lying is an important behaviour for cattle occupying approximately 50% of their daily time budget (Ruckebusch, 1974; Metz, 1985; Dechamps et al., 1989; Krohn and Munksgaard, 1993). For lactating animals, blood ¯ow to the mammary gland is increased during lying (Metcalf et al., 1992; Rulquin and Caudal, 1992) and when cows are deprived of lying, plasma concentration of growth hormone is reduced which is likely to affect milk production (Munksgaard and Lùvendahl, 1993). Disturbed rest can increase hypothalamopituitary activity (Ladewig and Smidt, 1989) and activity of the sympathetic nervous system in cattle (MuÈller et al., 1989). In order to ensure the welfare and productivity of dairy cattle, it is important to optimize their resting time. Housing facilities can substantially affect the activity and resting behaviour of dairy cows. For example, Haley et al. (2000) found that housing cows in large pens with a mattress ¯ooring increased lying time by 4 h per day compared to housing them in tie stalls with concrete ¯ooring. This occurred as a result of a reduction in the amount of time spent standing without eating. However, in that study the individual effects of spatial restriction, tethering and ¯oor softness were not differentiated experimentally. A variety of stall designs and ¯ooring are available, but little information is available on how they impact cow comfort. A number of studies have examined the effects of housing environments on standing and lying behaviour of cattle (for a review, see Lidfors (1989)). To reduce the incidence of mastitis, many dairy farmers are reducing their use of organic bedding material (Elbers et al., 1998), and a number of alternative ¯ooring materials are available (Nilsson, 1988). Some studies have examined the effects of ¯ooring properties on their thermal and frictional characteristics (Nilsson, 1988). However, few controlled studies are available comparing the effects of ¯ooring softness on cow behaviour. MuÈller et al. (1989) found cows in tie stalls without any bedding showed more investigation of the lying area, using sweeping head movements performed close to the ¯oor, than cows in groups with bedding. It has been suggested that this re¯ects hesitation by the animal to lie down. Young cattle perform more of this behaviour immediately after tethering compared to after a period of adaptation (Jensen, 1999). Many recommendations are available regarding the optimum size of tie stalls. However, Krohn and Munksgaard (1993) examined the effects of some tie stall design features and found an effect of the type of tether and stall front on the cows' use of space within the stall. The dynamic space required by the cow to lie down or stand up includes an additional 150 cm above and beyond the static size of the animal (Tschanz and KaÈmmer, 1977). This ``lunging space'' is likely to be reduced with certain types of tethers or stall front designs. The purpose of this study was to examine the effects of ¯oor softness and two different stall front designs on the activity and resting behaviour of lactating dairy cows. 2. Materials and methods 2.1. Animals We used 16 lactating Holstein cows with a mean (x S:T:D:) body weight of 625:1 61:8 kg and body length (de®ned as topline length from nose to tail) of
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
107
247:6 9:28 cm. Cows averaged 149:2 34:7 days of lactation at the beginning of the experiment and on average were in their second lactation (range 1±5). During the experimental period cows consumed an average of 35:35 0:70 kg of total mixed ration per day and produced 24:71 0:47 kg of milk per day. 2.2. Housing and management Cows were cared for according to the recommended code of practice for dairy cattle (Agriculture Canada, 1990). Prior to the experiment all cows were housed in tie stalls (``Comfort H-500'', J. Houle and Sons, Inc., Drummondville, Que.) (1:95 m 1:30 m, Fig. 1) on rubber-®lled geotextile mattress ¯ooring (Pasture Mat1, Promat, Ltd., Seaforth, Ont.) with a 2 cm covering of chopped straw. The barn was arti®cially ventilated and lights were on from 06.00 to 22.00 h (>80 lx). During the dark period red lights (<5 lx) were left on to enable ®lming with light-sensitive cameras. A total mixed ration was fed to ad libitum intake. Each day at 14.00 h a weighed quantity of fresh feed was delivered to each animal. Leftovers were removed daily and weighed on the same 5 days each week throughout the experiment. Individual feed intake was then calculated based on feed intake recorded 5 days each week (Monday±Friday). Water was drunk from ¯oat-activated water bowls shared by adjacent cows. Bowls were cleaned every 2 weeks. Cows were milked twice each day (06.00 and 18.00 h) in a milking parlour and individual milk yields were recorded electronically by Metatron milk meters (Westfalia, Elk Grove Village, IL).
Fig. 1. Diagram showing the tie stall design used to house cows prior to the experiment. The same stall front design was used for both treatments in the flooring comparison and also served as the narrow opening treatment in the stall front comparison.
108
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
2.3. Experimental All experimental methods for this study were approved by the Institutional Animal Care Committee at the Dairy and Swine Research and Development Centre in compliance with the guidelines of the Canadian Council on Animal Care. We compared the effects of concrete and mattress ¯ooring surfaces on the behaviour of cows kept in tie stalls. We also compared the behaviour of cows when they were housed in two tie stalls of different design: one with a narrow opening at the front of the stall (Fig. 1) and one with a wide front opening (Fig. 2). We conducted these studies as two separate cross-over designs. The same 16 cows were subjects for both comparisons and every cow spent three consecutive weeks in each of the four stalls. Half the cows started with the ¯ooring treatments and half with the stall front designs. After 6 weeks, once cows had spent 3 weeks in each of two stalls of one comparison, they were moved over to the other comparison. To facilitate the rotation of the animals during the experiment we grouped them into four blocks of four cows each, on the basis of body weight and body length. 2.3.1. Flooring comparison Stalls used to examine the effects of tie stall ¯ooring measured 1:80 m 1:30 m and had the same stall front design as Fig. 1. The only difference between the two stalls was the ¯ooring: half of the stalls had concrete ¯ooring and half had a geotextile mattress ¯ooring which consisted of multiple cells ®lled with uniformly sized rubber crumbs (Pasture Mat1, Promat, Ltd., Seaforth, Ont.). The stalls were in one row and we alternated pairs of stalls with regard to ¯ooring to balance for position within the row. Both concrete and mattress stalls were provided with a fresh 2 cm covering of chopped straw on a daily basis.
Fig. 2. Diagram showing the tie stall design with the wide front opening.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
109
2.3.2. Design of stall front comparison Stalls used to compare the effects of two stall front designs were 20 cm longer than those used in the ¯ooring comparison (2:00 m 1:30 m). Half of the stalls had a narrow opening between the stall front and the manger (30 cm, Fig. 1). The other half of the tie stalls were of a different design with a front partition (a wide opening, 120 cm, Fig. 2) (``Comfort Plus H-1100'', J. Houle and Sons, Inc., Drummondville, Que.) that provided more open access to the manger and presumably more lunging space to facilitate the cows' ability to change position from lying to standing and vice versa. Tether chains in the wide stall were positioned at a ®xed point vertically (45 cm from the ¯oor) while the tether in the narrow stalls were not ®xed and could slide vertically up and down the bars to which they were fastened. However, both chains were standardized to allow the same maximum perpendicular distance from the stall-manger border (80 cm). The stalls were in a single row and we alternated the stall fronts within the row. All stalls had the same mattress ¯ooring as described in the ¯ooring comparison and a fresh daily covering of chopped straw to a depth of 2 cm. 2.4. Behavioural observations During the 3 weeks spent in each housing condition all cows were videotaped for four 24 h periods: day 1 (the ®rst day in each system) and days 9, 13 and 17. Time-lapse video recorders (Panasonic AG-6730; Panasonic, Secaucus, NJ) were used in 24 h time mode, each recording the output from four video cameras (Panasonic WV-BL200; Panasonic, Secaucus, NJ) using a video multiplexer (Uniplex Sprite Video Multiplexer; Dedicated Microcomputers Ltd., Reston, VA). Two adjacent cows could be viewed on a single camera and for a complete view we positioned one camera in front and one camera behind each pair of animals. All video recordings were ®rst read by continuous observation. We recorded the general activity patterns of each animal (see Table 1) from all observation days and watched tapes at 12 normal recording speed using The Observer1 Video Tape Analysis System (Noldus Information Technology, Wageningen) and a video time code reader. To obtain more detailed information on resting postures and eating behaviour, all tapes were reviewed a second time using a scan sampling technique such that each cow was observed during a 3 min period every 12 min (120 observations per cow per day). This process was again done using The Observer1 Video Tape Analysis System. We found a very high positive correlation between actual lying time from continuous observations and estimated lying time using our scan sampling method (r p 0:99). With scan sampling, body position of the cow at the beginning of the 3 min window was recorded as either standing or lying. If a cow was standing, we recorded whether the cow ate or whether it stood for the whole period without eating, and whether or not the cow performed sweeping movements with the head during each 3 min observation period (see Table 1 for operational de®nitions). Head sweeping movements have been reported to occur prior to lying and reported to increase in tie stalls compared to loose housing (MuÈller et al., 1989). If the cow was lying at the start of the 3 min time window, we recorded the position of the head and leg positions (see Table 1) at the beginning of the time interval.
110
Table 1 Descriptions of behaviour recorded by continuous and scan sampling methods during 24 h observation periods Behaviour
Description
Continuous
Standing Standing in gutter Kneeling Lying right side Lying left side Lying
Weight supported by four legs Standing with one or both hind legs positioned in the dung gutter behind the stall (20 cm depth) The body is in a position with both knees bent and hind legs extended (minimum time 10 s) The body is resting on the ground with the right side in contact with the floor The body is resting on the ground with the left side in contact with the floor Lying right side lying left side
Scan sampling
Eating Not eating Head sweeping
The cow was observed rummaging in the feed manger followed by chewing movements The cow was standing but did not eat once during the 3 min observation window With the nose positioned 20 cm from the floor (judged relative to the 20 cm height of the stall-manger border) the head was moved from side-to-side in a continuous movement with at least two changes in direction and covering a lateral distance equal to or greater than the width of the animals shoulders The body is resting on the ground The body is resting flat on one side with the legs of the supported underside extended and the head resting on the ground The cow is lying on the sternum, the head is raised off the ground The cow is lying on the sternum, the head is turned back resting against the body The cow is lying on the sternum, the head is stretched resting on the floor Recorded if any part of the cow's nose was encroaching the manger area (beyond the stall-manger border) Recorded if the nose was behind the stall-manger boarder The exposed front leg is tucked under the body (full plantar-flexion at the humoral joint) The exposed front leg is extended in front of the body The exposed hind leg is positioned at an angle of 908 or greater in relation to the body axis The exposed hind leg is positioned at an angle between 45 and 908 in relation to the body axis The exposed hind leg is positioned at an angle of less than 458 in relation to the body axis
Lying on sternum Lying on side Head upright Head back Head on ground Head in manger Head in stall Front leg tucked Front leg exposed Hind leg extended Hind leg mid-position Hind leg tucked
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
Method
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
111
2.5. Statistical analysis From continuous observations we analyzed the frequency of bouts, mean bout durations and the total duration of time spent standing and lying by analysis of variance (SAS, 1990). The following factors were used in the analysis: block, cow identity (as a random factor nested within block), treatment and the interaction between treatment and block. The effect of block was tested against cow identity as an error term. Planned contrasts were used to examine differences between the two stall types and the two types of ¯ooring. From observations based on interval sampling, the frequency of observation for lying, standing, eating and standing not eating, as well as head and leg positions while lying were all analyzed using the same model. 3. Results 3.1. Flooring comparison Complete results for the ¯ooring comparison are presented in Tables 2±4. No signi®cant effect of the day of observation, and no interaction between the day of observation and the treatment was noted. When on mattress ¯ooring, cows increased their lying time by 18% compared to the time they spent lying on concrete ¯ooring, while on concrete cows spent more time standing. On concrete the cows spent 1.7 h more standing without eating than when they were kept on the mattress ¯ooring. There was no difference between concrete and mattress ¯ooring in the amount of time cows spent eating (4.4 h per day). We found no signi®cant difference between concrete and mattress ¯ooring with regard to the number of times cows were observed head sweeping. Cows changed position more frequently when on mattress ¯ooring than they did on concrete as measured by the total number of times cows got up and lay down. The average Table 2 Effect of concrete and mattress flooring in tie stalls on measures (mean S:E:) of lying and standing by lactating dairy cows (n 16) based on four 24 h observation periods P-valuesa
Behaviour
Concrete
Mattress
Lying Total time lying (h per day) Frequency (per day) Bout duration (min)
10.42 0.42 9.05 0.94 77.71 6.70
12.25 0.33 13.13 1.12 61.92 5.08
Standing Total standing time (h per day) Frequency (per day) Bout duration (min) Standing not eating (% of total observations) Standing eating (% of total observations) Head sweeping (% of total observations)
12.87 10.77 80.14 35.24 18.74 3.01
11.04 14.84 48.35 28.11 18.32 3.10
a
d.f. 1, 36.
0.42 0.93 7.53 2.01 0.72 0.66
3.68 1.15 3.68 1.38 0.54 0.61
0.001 0.001 0.01 0.001 0.001 0.001 0.01 >0.10 >0.10
112
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
Table 3 Effect of concrete and mattress flooring in tie stalls on the head position (mean S:E:) taken by lactating dairy cows (n 16) while lying based on four 24 h observation periods Head position (% of total observations)
Concrete
Mattress
P-valuesa
Back against body On ground Upright
4.28 0.58 1.51 0.45 44.36 1.50
4.80 0.43 1.46 0.46 37.36 1.84
>0.10 >0.10 0.001
a
d.f. 1, 36.
duration of individual standing bouts when cows were on concrete was 60% longer than the duration of standing bouts when the cows were on mattress ¯ooring. The difference in average lying bout duration was also signi®cantly longer when the cows were on concrete (Table 2). Cows rested with their head back against the body or resting on the ground for similar amounts of time on concrete or mattress ¯ooring. Cows on the mattress ¯ooring spent more time lying with their head upright, which was the most common head position (Table 3). The only signi®cant differences in the leg positions between treatments were with the most commonly observed positions; cows on mattresses spent more time lying down with their front legs tucked underneath the body and hind legs in the middle range position (Table 4). 3.2. Stall front design comparison No signi®cant effect of the day of observation, and no interaction between the day of observation and the treatment was noted. We found no signi®cant differences between stall front designs in the measures related to general activity of the cows. On an average cows spent 53.85% of their time lying down (narrow 12:57 0:42 versus wide 13:28 0:44 h per day; NS). On an average cows had 12.5 lying bouts each day which lasted a mean Table 4 Effect of concrete and mattress flooring in tie stalls on the mean (S.E.) percentage of observations for the various leg positions of lactating dairy cows (n 16) while lying based on four 24 h periods Leg positiona (% of total observations)
Concrete
Mattress
P-valuesb
Front leg Extended Tucked
2.77 0.76 40.40 1.77
2.34 0.86 48.28 1.22
>0.10 0.001
Hind leg Extended Middle Tucked
3.37 179 33.51 2.13 6.29 1.04
4.17 1.35 39.39 1.68 7.07 1.23
>0.10 0.05 >0.10
6.14 1.04 0.54 0.29
6.92 1.22 0.68 0.43
>0.10 >0.10
Combinations Both exposed legs tucked Both exposed legs extended a b
Only positions of the exposed front and hind legs were recorded. d.f. 1, 36.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
113
Table 5 Effect of narrow and wide tie stall fronts on the mean (S.E.) percentage of observations for the various leg positions of lactating dairy cows (n 16) while lying based on four 24 h periods Leg positiona (% of total observations)
Narrow
Wide
P-valuesb
Front leg Extended Tucked
1.33 0.54 51.02 1.60
0.66 0.21 54.20 1.79
>0.10 >0.10
Hind leg Extended Middle Tucked
4.94 1.38 39.39 1.68 8.70 1.85
5.95 1.56 38.72 2.29 10.37 2.38
>0.10 >0.10 >0.10
8.61 1.83 0.31 0.13
10.34 2.37 0.13 0.04
>0.10 >0.10
Combinations Both exposed legs tucked Both exposed legs extended a b
Only positions of the exposed front and hind legs were noted. d.f. 1, 36.
time of 66 min with no signi®cant differences between the two stall front designs. There were no signi®cant differences in total time spent standing (narrow 10:67 0:42 versus wide 9:98 0:45 h per day; NS), the frequency of standing (narrow 14:14 0:91 versus wide 15:88 1:40 h per day; NS) or the mean duration of standing bouts (narrow 48:71 4:11 versus wide 43:11 4:79 h per day; NS). No differences between stall designs were found for the percentage of observations where cows were standing and not eating (narrow 25:89 1:74 versus wide 24:78 2:26 h per day; NS), standing eating (narrow 18:70 0:78 versus wide 17:28 0:73 h per day; NS) or the percentage of observations where head sweeping was observed (narrow 2:55 0:43 versus wide 1:66 0:33 h per day; NS). Stall front opening did not in¯uence the leg positions assumed by cows while lying (Table 5). Stall front design, however, had signi®cant effects on where the cows placed their head while lying. Cows used the additional space made available to them by the wide stall front opening, positioning their head in the manger area while lying signi®cantly more than cows in stalls with a narrow stall front (Table 6). Accordingly, cows in stalls with narrow fronts Table 6 Effect of narrow and wide tie stall fronts on the head position of lactating dairy cows (n 16) while lying (mean percentage S.E.) based on four 24 h observations Head position (% of total observations)
Narrow
Wide
Area of head placement In manger In stall
20.65 4.18 26.88 3.89
34.91 3.71 26.88 3.89
Head position Back against body On ground Upright
4.82 0.44 0.67 0.24 46.81 1.77
5.42 0.80 0.30 0.10 49.10 1.80
a
d.f. 1, 36.
P-valuesa 0.001 0.001 >0.10 >0.10 >0.10
114
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
spent signi®cantly more time with their head completely behind the stall-manger partition (Table 6). The relative position of the head to the body during lying did not differ signi®cantly between the two stall front designs (Table 6). 4. Discussion Stall ¯ooring had marked effects on the cow's behaviour. In the present experiment, the daily percentage of time spent lying (concrete 43:4%, mattress 51:0%) were within the range of values reported in previous studies (Dechamps et al., 1989; Krohn and Munksgaard, 1993). With the softer mattress ¯ooring cows increased their lying time by an average of 1.8 h per day compared to when they were housed on concrete ¯ooring. Cows on concrete increased the time they spent standing without eating, while time spent eating was unaffected. Despite the fact that the cows on softer ¯ooring lay down for longer over the whole day, the average duration of individual bouts of lying was longer when cows were housed on concrete. On the mattress ¯ooring, cows increased their total lying time by lying down more frequently, but for shorter periods of time. MuÈller et al. (1989) and Ladewig and Smidt (1989) also found reduced frequency of position changes and an increased duration of lying bout length when bulls were tethered on concrete ¯ooring compared to when they were loose housed on straw. We also found the same for bouts of standing: cows stood up more frequently and remained standing for shorter periods when they were housed on mattress ¯ooring. These results suggest that the cows were more reluctant to stand up and lay down on the harder concrete ¯ooring surface because of discomfort experienced during the process of standing up and lying down. Metzner (1978) calculated that for a 600 kg animal, the maximal load force applied to the front knees is approximately 4 kN. Softer ¯ooring deforms under pressure which increases the contact area and dampens the force impact on the knees (Dumelow, 1995). Previously, we found that cows kept on concrete ¯ooring for 16 weeks had a higher incidence of swelling in their front knees compared to cows housed on rubber mats (Rushen et al., 1998; Haley et al., 1999). Physical changes such as these can result from force trauma and provide further evidence that changing position on concrete surfaces may be painful. Aside from hardness, other characteristics of the ¯ooring surface may also be partly responsible for the changes in behaviour. For example, differences in the coef®cient of friction between the two ¯ooring surfaces may have affected cow behaviour by altering the ease with which cows could change position. Concrete ¯ooring can be quite slippery when wet. In a previous study, Haley et al. (2000) compared cow behaviour in two types of housing, and found untethered cows loose-housed individually in box stalls with a mattress ¯ooring spent 61.4% of their time lying, 4.2 h more time per day than cows tethered in tie stalls with concrete ¯ooring. The present experiment suggests that the effects of softer ¯ooring alone cannot explain the results of Haley et al. (2000) since the difference in lying time between mattress and concrete ¯ooring was only 1.8 h. This suggests that tethering and/or some aspect of spatial restriction, associated with the tie stall was also reducing lying time. With regard to the body positions of cows during lying, the only effect of ¯ooring type was on the most commonly observed postures; lying in a sternal position with the head
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
115
upright, front legs tucked under the body and the hind legs in the middle position. There were no differences in the amount of time cows spent lying with their head resting on the ¯oor or time spent lying in lateral recumbency (¯at on their side). Both of these positions were detected very infrequently. Thus, the hardness of the ¯ooring did not appear to prevent cows from assuming particular positions while lying. However, it may be worth noting that this experiment was conducted during the summer months, and given differences in the thermal properties of the ¯ooring materials (Nilsson, 1988) the present results may be different during cold winter months. Previous research has shown cattle that were tethered or kept in stanchions on concrete perform more head sweeping movements than loose-housed cattle and it has been said this is investigative behaviour and is indicative of hesitation to lie down (Krohn and Munksgaard, 1993, ``examining behaviour''; MuÈller et al., 1989). As a support of this hypothesis, Krohn and Munksgaard (1993) also found tethered cattle showed more interrupted attempts at lying. We had hypothesized that when cows were kept on concrete ¯ooring they would perform more head sweeping, but we did not ®nd this: based on our operational de®nition of the behaviour, ¯oor type had no signi®cant effect on head sweeping movements. The ¯ooring type did not affect the time the cows spent eating. The priority of feeding behaviour, especially for high producing lactating cows, may have limited the possibility of detecting short-term differences in feeding due to housing effects. The design of the stall front affected the way cows used the stall space during lying. When additional space was made available to cows by providing them with a wide stall front opening, cows used the space, placing their head in the manger area during 35% of observations compared to only 21% of observations in stalls with a narrow opening. Detailed analysis of head positions in tie stalls has shown that in stalls with more open fronts, the mean position of the cow's head is 15±20 cm closer to the front of the stall than in stalls with less open fronts and the time that their head is in the neighbouring cow's stall is greatly reduced (Jahodova, 1997). Despite this, the design of the stall front did not signi®cantly affect overall activity nor did it affect the type of postures that the cows adopted when lying. From casual observation, it appeared that cows in stalls with a narrow front had more dif®culty standing up and often had to contort their neck into strange positions compared to normal lunging behaviour. However, we did not ®nd any effect of the stall front on the frequency with which the cows stood up or lay down. One possible reason we did not ®nd any signi®cant difference in lying time between the two stall fronts is that during previous lactations, the cows in this study were housed in stalls with a narrow front design. Prior experience may have reduced the probability of ®nding signi®cant effects of stall front since cows may have previously adapted to the differences in spatial restriction imposed by the narrow stall front. Using subjects that were naive to either stall front design, or to tethering, or both, may have yielded different results. Recent work by Jensen (1999) has brought attention to the dif®culty experienced by heifers in adapting to tethering. Alternative stall front designs that offer more lunging space may be of particular bene®t to cattle during initial tethering. Although lying times were the same for narrow and wide stall front openings, analysis of the biomechanical movements when cows change position from lying to standing (and vice
116
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
versa) may show differences in other measurable aspects of lying behaviour (e.g. time required to stand up). These measures may re¯ect important differences which are affected by stall front design. Based on the camera angles used for ®lming in the present study we were unable to do any temporal analysis of standing and lying movements. 5. Conclusions Increasing the softness of ¯ooring in tie stalls with a geotextile mattress substantially increases the time that the cows spend lying down, which is an advantage for cow comfort, and probably overall production. The effect occurs mainly because cows appear to be reluctant to stand up or lie down on hard concrete ¯ooring, rather than because of obvious discomfort when resting. A more open front for the tie stall appears to allow the cow to rest with its head in the manger, and so will affect the cow's use of the space within the stall. The physical dimensions of a tie stall (length and breadth) alone cannot ensure that the animal has adequate space if the front of the stall is poorly designed. However, we found no evidence to indicate that this affects overall activity, so we cannot conclude any effects on cow comfort. Acknowledgements This research received ®nancial support from the FeÂdeÂration des producteurs de lait du QueÂbec and the Agriculture and Agri-Food Canada MII fund. We thank Isabelle Blanchet, Aleksija Neimanis, Marjolaine St-Louis and Tristan Welp for their technical assistance and the staff at the Dairy Centre for looking after the animals. References Agriculture Canada, 1990. Recommended code of practice for the care and handling of dairy cattle. Dechamps, P., Nicks, B., Canart, B., Gielen, M., Istasse, L., 1989. A note on resting behaviour of cows before and after calving in two different housing systems. Appl. Anim. Behav. Sci. 23, 99±105. Dumelow, J., 1995. Testing cubicle mats for dairy cows. Agric. Eng. 50, 17±21. Elbers, A.R.W., Miltenburg, J.D., De Lang, D., Crauwels, A.P.P., Barkema, H.W., Schukken, Y.H., 1998. Risk factors for clinical mastitis in a random sample of dairy herds from the southern part of The Netherlands. J. Dairy Sci. 81, 420±426. Haley, D.B., Rushen, J., de PassilleÂ, A.M., 1999. Effects of softer flooring on the behaviour, health and productivity of dairy cows in tie stall housing. In: Proceedings of the 33rd International Congress of the International Society for Applied Ethology, Lillehammer, Norway, 17±21 August, p. 127. Haley, D.B., Rushen, J., de PassilleÂ, A.M., 2000. Behavioural indicators of cow comfort: activity and resting behaviour of dairy cows in two types of housing. Can. J. Anim. Sci 80, 257±263. Jahodova, J., 1997. Effect of housing conditions on welfare of dairy cattle. M.Sc. Thesis, University of Edinburgh, Edinburgh. Jensen, M.B., 1999. Adaptation to tethering in yearling dairy heifers assessed by the use of lying down behaviour. Appl. Anim. Behav. Sci. 62, 115±123. Krohn, C.C., Munksgaard, L., 1993. Behaviour of dairy cows kept in extensive (loose housing/pasture) or intensive (tie-stall) environments. II. Lying and lying-down behaviour. Appl. Anim. Behav. Sci. 37, 1±16.
D.B. Haley et al. / Applied Animal Behaviour Science 71 (2001) 105±117
117
Ladewig, J., Smidt, D., 1989. Behavior, episodic secretion of cortisol, and adrenocortical reactivity in bulls subjected to tethering. Horm. Behav. 23, 344±360. Lidfors, L., 1989. The use of getting up and lying down movements in the evaluation of cattle environments. Vet. Res. Commun. 13, 307±324. Metcalf, J.A., Roberts, S.J., Sutton, J.D., 1992. Variations in blood flow to and from the bovine mammary gland measured using transit time ultrasound and dye dilution. Res. Vet. Sci. 53, 59±63. Metz, J.H.M., 1985. The reaction of cows to a short-term deprivation of lying. Appl. Anim. Behav. Sci. 13, 301± 307. Metzner, R., 1978. Analyse tierischer BewegungsablaÈufe zur Gestatung artemaÈber Rinderkrippen. Landtechnik 9, 397±404. MuÈller, C., Ladewig, J., Thielscher, H.H., Schmidt, D., 1989. Behavior and heart rate of heifers housed in tether stanchions without straw. Physiol. Behav. 46, 751±754. Munksgaard, L., Lùvendahl, P., 1993. Effects of social and physical stressors on growth hormone levels in dairy cows. Can. J. Anim. Sci. 73, 847±853. Nilsson, C., 1988. Technical design with respect to the biological needs of animals in reference to the thermal, friction and abrasive characteristics and the softness of the flooring material. Swedish University of Agricultural Sciences, Lund. Ruckebusch, Y., 1974. Sleep deprivation in cattle. Brain Res. 78, 495±499. Rulquin, H., Caudal, J.P., 1992. Effects of lying or standing on mammary blood flow and heart rate in dairy cows. Ann. Zootech. 41, 101. Rushen, J., de PassilleÂ, A.M., Haley, D.B., 1998. The effects on productivity, health and behaviour of housing lactating cows on NRI ``Cloud 9'' rubber mats. Report to NRI, Inc. SAS, 1990. SAS1 User's Guide: Statistics, Ver. 6, 4th Edition. SAS Institute, Inc., Cary, NC, 1990. Tschanz, B., KaÈmmer, P., 1977. Verhaltensbiologische AnsaÈtze zur beurteilung von Liegeboxen. Der TierzuÈchter 29, 151±153.