About 80 % of all Swedish dairy cows are kept in tie-stalls, but loose housing is increasing, probably because it is more profitable for the farmer. The design of the floors is of cruicial importance to the cows well-being. Traditionally, the most common type of floor in the alleys of a cubicle system still is a slatted manure-draining floor made of concrete. The typical Swedish design has 125-mm slats and 40-mm slots, maximally. With too wide slots, there is an increased risk of injuries to the claws when the cow makes a hasty move. On top of that, there will be too little support for the claws, leading to injuries. A dirty floor can also increase the risk of hygiene-related diseases, mainly dermatitis and heel horn erosion, which in severe cases even may cause lameness. It has been shown earlier that by reducing the slots from 40 to 30 mm one can still achieve a good drainage. This increases the support of the claws and thus decreases the risk of traumatic injuries.
The aim of the present study was to test a modified slatted concrete floor (100-mm slats and 30-mm slots) in a full-scale experiment, by comparing the cows locomotion and the floor’s cleanliness to a conventional (125/40-mm) slatted floor. The trial was conducted at a commercial dairy farm with cubicle housing and two compartments of equal size and design. In the beginning of the grazing period, approx. 4 months before grouping and housing of cows, the modified floor was put in the alley of one of the compartments. Apart from the type of floor, the two compartments were comparable with respect to housing design and management routines. All cows were Swedish Holsteins and were assigned randomly to the two compartments at housing.
A test track was prepared in the alley in each of the compartments. The floor was covered with a mixture of lime and manure and then one cow at the time passed and had her passing time recorded. From the footprints, step length, stride length, step angle, overlap and step abduction were measured. The assymetry (absolute difference between left and right step length) and average speed were calculated. Each one of the seven traits was analysed using a mixed random-intercept model, specifying floor type as the only fixed effect and cow identity as random. Floor cleanliness was studied by collecting and weighing the manure on top of the slats in six 60 x 80-cm test squares in each compartment before cleaning and 1, 2, 4 and 8 days after cleaning (new cleaning for each observation). Three squares were placed in the centre of the alley and three at the edges, behind the cubicles. The amount of manure was analysed using multiple regression, including floor type, location in alley, interaction between floor type and location, and time (1, 2, =4 days) as categorical variables in the model.
The cows had longer steps (+2.7 cm, P=0.006) and strides (+4.5 cm, P=0.002) and the amount of manure was smaller at the edges of the alley (-464 g, P=0.001) on the modified floor than on the conventiona...
