Housing dairy cows offers the possibility to control many aspects of their lives, including accurate rationing, which is especially important for high yielding cows, and rapid health care. In addition, some parasitic diseases are largely controlled by removing cows from pasture. However, housing cows is associated with an increased prevalence of several serious diseases, e. g. mastitis and lameness. In housing systems cows can less readily synchronise their behaviour with other cows, maintain adequate personal space and express oestrus behaviour, compared to cows at pasture. Soft ground and space at pasture facilitate natural locomotion, lying down/standing up motions and resting, without the behavioural abnormalities that may occur inside cubicle houses. Although an inability to perform natural behaviour often impairs health and welfare in housed cows, this is not always the case, and so the precise welfare implications of housing with regards to some of the different types of (natural) behaviour remain tentative. The present findings suggest that dairy cow production based on intensively housed cows is less desirable from the perspective of animal behaviour and health, and hence welfare. However, increasingly larger and more productive dairy herds are utilising intensive housing systems because they facilitate mechanised management systems and a reduction in labour requirements, and the negative implications for welfare are worthy of detailed consideration.

Cows in automated milking systems with free access to feeders sometimes show a reduced use of the robotic milkers, while forced traffic where cows have to pass through the robot to reach the feeders may reduce feeding time and frequency. We examined two groups of 35 lactating cows. For 21 d, one group had free access to feeders, while the other group had to pass through the robot to reach feed. For a subsequent 21 d period, the treatments were reversed. The behaviour of the cows was observed for 3 periods of 24 h using 10 min scan sampling. Milking data were obtained from the robot milkers. The average number of visits to the robot, either with or without milking, did not differ between the two types of traffic. On average cows lay down for 12.3 h/d, with no significant difference between free and forced traffic. The type of traffic did not affect the duration of time spent in the feeding area. With both forced and free traffic, there was marked diurnal variation in feeding, with relatively little feeding occurring between 2200 h and 0800 h. Use of forced traffic did not reduce feeding time. It did not improve access to the robot milker or eliminate diurnal variation. Differences between cows in visits to the robot may reflect differences in their motivation to rest and eat concentrates.

Recent research on how to combine automatic milking (AM) systems and grazing is presented with the objective to summarise the results obtained in an EU-project and discuss these results in relation to present knowledge in the field. Experiments with 24-hour grazing conducted in Sweden showed no difference in water intake between cows with drinking water indoors, and cows with water both indoors and on pasture. Ad libitum supply of silage did not increase milk yield or lower the milking interval compared with 3 kg dry matter*cow*day-1. Free cow traffic (access to feed supplements) gave a significantly lower number of voluntary milkings compared with controlled access. Grazing near the barn (50m) gave a higher milk yield compared with grazing further away (260m). Dutch surveys and studies on AM-farms with grazing showed that selection gates contributed to a lower number of manually fetched animals. In an experiment, the effect of pasture distance (<150m versus >500m) and supplementation level (6 versus 10 kg DM) on milking interval, pasturing time and milk yield was studied. No significant effects of distance or level of roughage supplements on milk yield and milking interval were observed. Case studies on farms in Denmark showed that a relatively high milking frequency was obtained (variation 2.1-2.8) but most cows had to be fetched from the pasture. In periods when many cows were at pasture a decrease in number of milkings per hour often occurred. Cow behaviour was more synchronised at pasture than when in the barn, and only few cows entered the barn alone. In conclusion, automatic milking can successfully be combined with pasture and grazing. High levels of supplementary feeding did not lead to higher production when pasture was available in sufficient amounts but supplementation can be used strategically as a management tool to obtain smooth cow traffic.

The effect of handling and hand-feeding of calves with or without their dams present on the calves' subsequent response to humans was investigated using 5 groups of 10 Danish Friesian calves, each of which received a different treatment. The treatments were: calf separated from its dam immediately after birth, housed in a single pen and handled (group SH) or not handled (group S) calf adjacent to cow and handled (group CH) or not handled (C) for the first 4 days after calving. The fifth group was housed together with their dams for the first 4 days of life and then moved to single pens and handled there for the next 4 days (group C-SH). The handling treatment was carried out three times daily for 4 days and consisted of 6 min of hand-feeding with milk from a teat-bucket and patting, stroking and talking to the calf. After separation, all calves were housed in single pens. Human contact was minimised except during the treatment periods.The approach behaviour of each calf to an unknown person was tested at days 20, 40 and 55 in their home pen. Flight distance was determined at day 55. The test at day 55 was carried out in a large single pen where the calves had been housed for 24 h prior to the test. The latency to interact (in seconds) with the person was significantly shorter (P < 0,001) in all three tests for SH (13 +/- 5, 29 +/- 7 and 44 +/- 8, at days 20. 40 and 55, respectively) and C-SH (28 6, 47 17 and 70 +/- 17) compared to S (240 +/- 25, 238 +/- 33 and 173 +/- 5). C (240 +/- 28, 202 +/- 36 and 167 +/- 13) and CH (233 +/- 36, 271 +/- 29 and 128 +/- 19). The results of the latter three groups did not differ significantly. Treatment also affected the position and the orientation of the calf in the pen during the tests. Calves in SH and C-SH stood at the front of the pen and faced the person more often than calves in S, C and CH. Again, there were no significant differences between the results of the latter three groups.The results indicate that the presence of the dam limits the effect of the handling treatment on young calves' motivation to interact with humans. It is likely that primary socialization occurs with the dam, preventing a secondary socialization developing with humans, until the calf is isolated from the dam. (C) 2002 Elsevier Science B.V. All rights reserved.

The effects of handling at different times after birth on the subsequent response to humans was investigated using 40 Danish Friesian calves removed from their dams immediately after birth. Treatments were feeding and handling during days 1-4 after birth (H1-4), during days 6-9 (H6-9), or during days 11-14 (H11-14), and no handling (C: control). The handling treatment was carried out three times daily for 4 days and consisted of 6 min of hand-feeding with milk from a teat-bucket and patting, stroking and talking to the calf. All calves were housed in single pens. Human contact was minimised except during the treatment periods.The approach behaviour of each calf to an unknown person was tested at days 20, 40 and 55 in their home pen. The test at day 55 was carried out in a large single pen where the calves had been housed for 24 h prior to the test. Latency to interact with the person was shorter in all three tests for calves in H1-4 compared to C. There was no significant difference between H11-14 and C. Group H6-9 differed only from the control group at day 20. Treatment also affected the position and the orientation of the calf in the pen during the tests. Calves in H1-4 stood at the front of the pen and faced the person more often than calves in C. There was no difference between H11-14 and the C. All the three handled treatments had a shorter flight distance at day 55 compared to the control group. The results indicate that handling and hand-feeding, especially during the first 4 days after birth, increases the motivation of young calves to approach a human compared with calves that receive minimal human contact. (C) 2001 Elsevier Science B.V. All rights reserved.

This article discusses the consequences of different suckling systems in the industrial countries for the milk production, udder health, reproduction and behaviour of high producing dairy cows and the effects on the gain, health and behaviour of the calves. The suckling systems are divided into three different categories depending on the purpose and duration of the suckling period. Long-term suckling with or without additional milking covering the period where the calf has a nutritional need for milk, and short-term suckling, where cow and calf are kept together in the colostrum period only.Long-term suckling without additional milking in early lactation can in some situations stimulate the subsequent milk production to a greater extent than milking alone. No clear or significant differences can be found between restricted and free suckling systems. Most experiments show that suckling decreases the risk of mastitis in the suckling period and in some cases even for some time after the suckling has been terminated. Suckling and milking during the same period is not advantageous in production turns because of a very poor ejection of milk. Long-term suckling can increase the post-partum interval until first heat, in some cases until the end of the suckling period. However, as the cows appear to be more fertile, the net effect on reproduction is small. The suckled calves are usually healthy with a high daily gain. Short-term suckling have more advantages than disadvantages on production, health and behaviour of both the cow and the calf compared to an immediate separation after birth.

The effect of housing during the first three months of life till the end of the first lactation on the human-animal relationship and milk production was examined in an experiment with four treatments of 20 female dairy calves each. For the first 12 weeks of their lives, the calves were housed in either open single pens (SOpen) or closed single pens (SClosed), in groups of five calves (GCalf) or in groups of five cows and calves (GCow). Calves housed in a closed single pen approached humans more than calves housed in an open single pen. There was no difference in the ease of handling. The human-animal relationship was not affected by type of housing. Calves housed in groups of calves and cows sought less human contact, but were not difficult to handle. Milk production was not affected by housing method during early rearing. Therefore, group housing, which allows full socialization with other animals and involves better welfare than individual housing, can be recommended, as no negative effect was found on later handling and milking.

The experiment was carried out with 57 female dairy calves (Danish Holstein Friesian) divided into three groups and treated differently during the first four days of life (colostrum period). Treatments were: single box and no contact with the dam (group S0), cow and calf together in a maternity pen and no suckling (C0), and cow and calf together and suckling (C4). In the post-treatment period from day five until the end of their 24th week of lactation, all three groups received the same treatment (tie stalls in winter and pasture in summer). The results from the treatment period showed that calves, which were together with their dams (groups C0 and C4), grew about 100% more per day than calves in group S0. The calves in group S0 were licked less and had the highest duration of non-nutritive sucking on equipment.In the post-treatment period, non-nutritive sucking on equipment was lower in both duration and frequency for group C4 than for the two other groups during the observations on days 21, 42 and 70. The duration of physical contact with the human in a voluntary human approach test in an open arena in weeks 2, 10 and 25 was lower for group C4 than for the two other groups. A social behaviour test in weeks 3 and 11 showed that calves from group C4 preferred to spend more time near an unfamiliar heifer than calves from group S0. The results of a forced human approach test on pasture when the heifers were 15-18 months-old showed that heifers from group C4 were more difficult to approach than those from groups S0 and C0. There was no significant difference between the groups during the rearing period in daily gain, diarrhoea or pneumonia, or in milk production in first lactation.

To investigate the effect of social contact during early development on open-field responses and social responses, 80 female dairy calves were housed either in open single pens (SOpen), closed single pens (SClosed), calf-groups (GCalf), or groups of calves and cows (GCow) until 3 months of age. During the first 3 months, the calves were open-field tested at 2 and 10 weeks of age. Calves isolated in closed single pens (SClosed) performed more exploration during the open-field tests at 2 and 10 weeks of age than did calves housed in open single pens (SOpen). During the open-field test after the experimental period at 6 months of age, previously isolated calves (SClosed) had a longer latency to enter an open-field arena, and during a social test at this stage group-reared calves (GCalf and GCow) sniffed and mounted other calves more than calves housed in single pens (SOpen and SClosed). The results suggest that isolation increases the motivation to explore a novel environment, and that housing in groups facilitates the development of normal social responses.

The effect of feeding and handling on the response of young cattle to humans was investigated using 40 Danish Friesian calves removed from their dam immediately after birth. From day 3 to 17 of age calves were either: fed by humans and handled (stroking); fed by humans but not handled; fed without visual contact with humans and handled; or fed without visual contact with humans and not handled (control). Observations during the handling and/or feeding treatments revealed that calves fed with a human present performed more bunting behaviour but performed less play behaviour than those handled but not fed. The approach behaviour of each calf to an unknown person was assessed at days 3, 17, 32, and 62 in both their home pen and in an arena. In the home pen, handling had no effect on latency to interact with the person, but at days 17, 32, and 62 calves fed by humans were quicker to interact with a person than those fed without a human present. When tested in the arena, no consistent significant treatment effects were found at any age in latencies to approach or interact with the person. In a third test, the approach behaviour towards a person when social companions were present was assessed. With the human present only, time spent within 1 m of the person did not differ with age or treatment (17 to 62 days). But when two other calves were present, latency to approach the person increased (p<0.05) and time spent near the person decreased (p<0.05) with age. It is concluded that feeding has a greater influence on the responses of young calves towards humans than handling. However, this appears to be limited to the location in which the feeding took place. Despite receiving no additional handling, calves that were fed without a human present readily approached and interacted with an unknown person and spent a large proportion of time near the person in the arena tests, suggesting that handling in the first 2 days after birth may be very important in the development of the subsequent interactions between humans and cattle.

The effects of social contact and space allowance on the expression of play behaviour in domestic calves were studied. Forty-eight female dairy calves in three groups were housed in one of four pen types: (1) small single pen (0.9 m×1.5 m); (2) large single pen (1.8 m×3.0 m); (3) small group pen (1.8 m×3.0 m for 4 calves); and (4) large group pen (3.0 m×5.4 m for 4 calves). The behaviour of all calves was video-recorded for 8 h in week 2 and for 24 h in weeks 4 and 6 of the experiment. Data for play behaviour were obtained from each individual for all hours of observation. In weeks 4 and 6, space allowance affected the quantity of locomotor play. A low space allowance reduced locomotor play in both individually and group-reared calves. The quality of locomotor play was also affected. Elements of locomotor play that involve much movement were either absent or rarely seen in the small single pens. Furthermore, calves in single pens were less active than calves in group pens. The results of this study indicate that sufficient space is essential for the expression of play behaviour in domestic calves. It is suggested that play behaviour may be used to indicate the presence of good welfare in calves and in juveniles of other farm animal species. The use of a measure to indicate the presence of good welfare in addition to measures to indicate the absence of poor welfare may be a step towards a better assessment of welfare in farm animals.

32 dairy heifer calves were housed in one of four pen types: (1) a small single pen (0.9 m × 1.5 m); (2) a large single pen (1.8 m × 3.0 m); (3) a small group-pen (1.8 m × 3.0 m for four calves); and (4) a large group-pen (3.0 m × 5.4 m for four calves) until three months of age. Thereafter, all calves were housed individually in tether stalls which were 0.9m wide. Each calf was subjected to two test-sessions: one before tethering at three months of age and one at six months of age. Each test session included two tests. The first test was an open-field test with an unfamiliar calf present and the second was an open-field test in an empty arena. Behaviour and heart rate were recorded. Based on the data for the open-field test with an unfamiliar calf present, two principal components reflecting exploratory motivation and fear of unfamiliar calves, respectively, were defined. At three months of age, individually reared calves had higher scores for the principal component reflecting fear of unfamiliar calves. Principal components reflecting exploratory motivation and fear, respectively, were also defined based on data for the open-field test in an empty arena. At three months of age, individually reared calves had higher scores for the principal component reflecting fear in this test. The results suggest that individually reared calves are more fearful when introduced to a novel social situation and when isolated in a novel arena. Differences in space allowance did not affect the measured responses during the tests at three months of age. At six months of age, no effects of either social rearing or space were found.

An experiment was conducted on three Danish commercial dairy farms using 70 Danish Friesian heifers over a period of 141 to 162 days beginning in November. Average body weight of the heifers was 315 kg at the beginning of the experiment. Fourteen experimental pens were established with a flock size of five and different space allowance in the deep bedding area: 1.8, 2.7, and 3.6 m(2) resting area per heifer. The total space allowance inclusive of an unstrawed area at the feed manger was 3.6, 4.5, or 5.5 m(2) per heifer in herd 1, 3.6 or 4.5 m(2) per heifer in herd 3, and 3.0 or 4.9 m(2) per heifer in herd 2. Eating space was standardized within pens to 60 cm per animal. Feed was available ad libitum. At 1.8 m(2) resting area per heifer, the heifers were unable to synchronize their resting behaviour, and the variance of the daily lying time was higher compared with heifers at 2.7 or 3.6 m(2) resting area. None of the heifers were lying on the unstrawed area. Total lying time was not affected by the space allowance of resting area. There was a tendency towards more aggressive interactions at the lowest resting area of 1.8 m(2) per heifer. Frequency of abnormal behaviour was very low and was not affected by the space allowance. Daily gain and feed intake were increased in heifers with a resting area of 2.7 or 3.6 m(2) compared with 1.8 m(2), but the feed conversion ratio was unaffected. The prevalence of heel horn erosion was higher at 1.8 m(2) resting area compared with 2.7 and 3.6 m(2). For heifers with access to 2.7 or 3.6 m(2) compared with 1.8 m(2) resting area per heifer both welfare (behaviour and health) and production were improved.

Studies of calf suckling systems were carried out in the dairy herd at Research Centre Foulum and in two private organic dairy herds. The objectives of the presented study was to describe the confrontation between the late lactation cow (the 'aunt') and calf in order to develop one cow-calf-couple. In the research herd, special focus was directed against the behaviour of the calves. In this herd, 20 late lactating cows were confronted with one calf each. The calves were divided into four groups of 5 calves each: mothered versus bucket fed (with a suckler bucket) during the colostrum period, and housed together with the late lactating cow for one versus two days before being included into a group of 4 cow-calf suckler pairs. In the private herds, special focus was directed against advantages and disadvantages of the whole system in a broader sense. In one private herd, 10 late lactating cows were housed with 2 calves each, and all calves were fed colostrum from suckler buckets. In the other private herd, 11 late lactating cows were housed with 3 calves each and all calves had been mothered during the colostrum period.A successful establishment of a calf-aunt couple seemed highly dependent on the calf's previous suckling experience when confronted with late lactating dairy cows (the 'aunts'). The calves having been mothered were significantly better adjusting to the suckling systems. After 8 days in the shared box, the bucket fed calves suckled the cows approximately as much as the mothered calves, but there was significantly more cross suckling among these calves. Observations from the video recordings showed that interactions among the cows caused severe disturbance in the group seen as a whole.

The hypothesis that daily gain is associated with resting behaviour in heifers is tested in this paper. The test is based on two experiments conducted on Danish commercial dairy farms with Danish Friesian heifers (300–400 kg) in the winter periods. Group size was 5–6 heifers per pen. Experiment A, with 80 heifers, investigated the effect of different space allowances (1.5 or 3.0 m2 per heifer on fully slatted floor) and different floor types (fully slatted or with access to bedding). Experiment B, with 70 heifers, investigated the effect of space allowance in the deep bedding area (1.8; 2.7 and 3.6 m2 resting area per heifer). Data from experiment A showed a tendency towards a positive correlation between lying periods and daily gain for heifers housed on fully slatted floor but not when housed with access to bedding. There tended to be a positive correlation between number of lying periods and daily gain at a low but not at a high space allowance. The data from experiment B showed a tendency towards a positive correlation between lying time and daily gain at a low space allowance in the deep bedding area and not at a high space allowance. Among heifers in the pens with access to bedding in experiment A, it was found that one particular heifer in each of the four pens lay significantly more on the slatted floor and this heifer had an average daily gain which was less than the others.

An experiment was conducted on seven Danish commercial dairy farms with 200 Danish Friesian heifers (315 kg) in 152 days from the month of November. Founded on existing slatted floor systems, 32 experimental pens were established with Low (L) or High (H) space allowance (1.5 vs. 3.0 m(2) per animal), different type of separate lying area (Slatted (S) vs. Bedded (B) floor) and different flock size (6 vs. 12 animals). Eating spaces were standardized to about 50 cm per animal. Ad libitum feeding was practised and rations contained a maximum 30% concentrates (dry matter).An increase in space allowance from 1.5 to 3.0 m(2) per animal in pens with fully slatted floors increased the live weight gain by 174 g day(-1) or 31% (P = .04) and net energy intake per kg live weight gain was reduced by 23% (P = .04). Heifers housed in small pens lay down less than heifers in large pens. Access to bedding affected number of lying periods. Production, however, was not influenced by access to a bedded lying area. Flock size is of minor importance for heifers' production, when fed ad libitum and one eating space per animal is ensured. The prevalence of heel horn erosion was reduced and claw length was increased in heifers with access to bedding. No occurrence of tail tip necrosis regardless of pen type was found. Reproduction expressed as ages at pregnancy was not significantly influenced by any of the treatments.

Twenty-four pairs of monozygotic twin dairy cows (Danish Friesian and Red Danish) were assigned to four groups of 12 in a balanced incomplete block design. Group E: loose housing, free access to feeding table, deep bedding, a yard and pasture, milking twice a day. Group N: tie stall, concrete floor plus 1 kg straw, milking twice a day, no exercise. Group I: tie stall, rubber mats plus 2 kg straw, milking four times a day, no exercise. Group IE: as group I, except for 1 h daily exercise. The aim of this experiment was to compare allo- and self-grooming, exploration and abnormal behaviour (bar-biting, tongue-rolling, leaning) in an extensive environment and in tie-stalls, and to compare the behaviour in tie-stalls with and without the possibility of daily exercise.The total frequencies of social sniffing and licking were lower in group E compared with the tethered groups. Within the tethered groups, the frequencies of sniffing and licking were significantly lower in group IE. The frequency of self-grooming was lower in group E than in the three tethered groups. In cows of all groups, licking of the back and sides of the body constituted 40-45% of all grooming behaviour. When tethered, the cows directed only 30-32% of all licking behaviour against the hindquarters, whereas this behaviour pattern increased to 56% in the yard (group IE).The frequencies of all types of exploratory behaviours (sniffing and licking the equipment or the ground) were 2-3 times higher in the tie-stall than in loose housing systems. In group IE, sniffing the equipment and the ground was twice as high per hour in the yard than in the stable, while the frequency of licking the equipment was more than five times higher in the yard. The frequency of leaning against equipment was higher in the tie-stall than in group E and performed by 57-64% of the cows in the tie-stall.Permanent tethering of dairy cows seems to change the normal activity pattern and increase leaning behaviour. Daily exercise in a yard increased the frequencies of normal social behaviour, self-grooming and investigative behaviour, and decreased bar-biting, and therefore indicates some deprivation in permanently tethered cows.

Twenty-four pairs of monozygotic twin dairy cows (Danish Friesian and Red Danish) were assigned to four groups of 12 in a balanced incomplete block design. Group E: loose housing, free access to feeding table, deep bedding, a yard and pasture, milking twice a day. Group N: tie stall, concrete floor plus 1 kg straw, milking twice a day, no exercise. Group I: tie stall, rubber mats plus 2 kg straw, milking four times a day, no exercise. Group IE: similar to Group I, except for 1 h daily exercise.In general the duration of the lying-down pattern was significantly longer in the tie stalls and interruptions of the lying-down movement occurred more frequently in the tie stalls than under loose housing conditions. The duration of the lying-down movement was significantly longer in Group N than in Group E.Cows in Group E had significantly shorter total lying times compared with cows in the tie-stall groups, while no significant differences among the groups in the tie stalls were found. In Croup E the mean duration of resting bouts (lying periods which included one or more short standing periods for a maximum of 15 min) was significantly higher in the pasture. Lying flat on the side and with the head back or on the ground occurred more often in the pasture than in the deep-bedding area. In the tie-stall groups 1 h of daily exercise had no significant effect on time spent lying or the duration of resting bouts. The kneeling duration was shorter in the groups with than without 1 h of exercise.There were no inflammations of knees and hocks in Group E, but there were from eight to ten hock inflammations in each of the tie-stall groups. The frequency of trampled teats was significantly higher in cows kept on a concrete floor with a small amount of straw (Group N) than in cows kept on rubber mats with a large amount of straw (Group I and IE).Altogether the results suggest that the act of lying down in the stalls was aversive to the cows and that the lying area and/or the tie-stall system increased the risk of injuries to tethered cows.