http://d.doi.org/10.3168/jds.2015-9979 Supplemental Table S1. Distributions 1 of herd, housing and management variables considered as potential predictors in logistic regression analyses. Variable Routine herd data Continuous variables Median (range) herd size (n lactating) 70 (22-211) change in herd size (%) 0 (-26.3-36.5) average age cows >2 y (mo) 54.4 (42.2-70.9) cows older than 5 y (%) 33.8 (2.9-62.8) cows < 60 DIM (%) 17.9 (0-35.0) replacement (%) 27.5 (13.7-58.3) slaughter (%) 3.4 (0.0-29.8) Discrete variables Level (number of herds) slaughter of cows < 210 DIM yes (61) no (118) on-farm mortality of cattle: aged 0-3 d (%) 2.3 (0-18.2) aged > 2 y (%) 0.5 (0-7.2) herd biosecurity status open (82) closed (97) certified organic 2 yes (3) no (176) certified BVD free status yes (67) no (112) certified IBR free status yes (69) no (110) certified salmonella free status yes (53) no (126) average milk yield per cow/d (kg) 26.4 (13.5-34.5) net result 3 ( ) 2356.5 (1526.0-2984.0) change in net result (%) 1.6 (0-9.3) average DIM (d) 180 (132.0-276.0) average fat (%) 3.0 (2.4-5.4) average protein (%) 5.0 (3.3-6.2) average urea (%) 45.6 (13.6-57.5) average ratio fat/protein cows <60 1.3 (0.9-1.8) DIM FFA (mmol/100 g) 29.3 (20.0-89.0) butryric acid bacteria yes (152) no (27) bulk tank milk SCC (10 3 cells/ml) 196.5 (57.7-436.3) cows with udder infection 4 (%) 19.6 (3.9-43.4) cows with new udder infection 4 (%) 8.1 (2.1-25.5) cows with SCC > 400,000 cells/ml 10.7 (0-30.4) (%) heifers with udder infection 4 (%) 20.0 (0-100) average SCC (10 3 cells/ml) of cows: 0-60 DIM 173.9 (28.3-1107.9) 60-120 DIM 162.3 (30.4-1170.7) 120-210 DIM 198.5 (44.4-795.3) >210 DIM 244.5 (68.3-1230.0) non-return 56 d (%) 63.2 (0-100) average services per cow 1.8 (1.0-4.0) cows with more than 2 services (%) 21.7 (0-100) abortion (%) 7.1 (0-59.3)
http://d.doi.org/10.3168/jds.2015-9979 average epected calving int. (d) 418.6 (375.4-511.7) average realized calving int. (d) 415.0 (346.7-570.0) average int. calving-first service (d) 94.7 (63.8-229.7) Housing and Management air inlet side walls yes (75) no (104) air inlet roof top yes (137) no (42) light intensity 5 light (108) dark (68) brushes no (46) fied (51) rotating (80) average width of alleys (cm) 228.7 (125.0-365.0) average width of passages (cm) 201.6 (95.0-366.7) width of alley behind the feeding rack 300.0 (210.0-475.0) alleys with dead ends yes (100) no (79) type of flooring system 2 slatted (171) solid (8) slippery floor 6 yes (117) no (62) floor scraping frequency (times/d) 2.0 (0-24.0) rims or pits in the floor yes (45) no (134) cow to stall ratio 1.0 (0.6-1.8) predominant surface of lying area mattress/ deep other (81) bedding (65) stall divisions cantilever (85) mushroom (57) other (37) average stall length (cm) 225.0 (112.0-280.0) average stall width (cm) 110.5 (95.8-126.7) average height of stall neck rail (cm) 111.0 (93.0-151.0) stalls with head lunge impediments all (28) some (127) none (24) bedding height (cm) 1.0 (0-20.0) stalls with fecal contamination 7 < 50% average feed space per cow (cm) 64.2 (16.3-116.4) type of feeding rack average feeding rack height (cm) 143.0 (78.0-165.0) smallest feeding rack height (cm) 140.0 (78.0-165.0) yes (76) no (103) headlocks (95) safety headlocks (53) continuous availability of roughage 8 yes (152) no (27) roughage contaminated with manure yes (25) no (154) and/or moulds, or heat coming out concentrate dispensers in the stable yes (132) no (47) group drinkers yes (156) no (22) sufficient number and length of yes (148) no (31) drinkers 9 dry cow groups (far-off and closeup) yes (75) no (104) heifer housing same building (36) other building (50) automatic milking yes (42) no (137) lactation groups yes (15) no (164) daily (versus non-daily) deposition of yes (158) no (21) roughage on the feed bunk different types of roughage no (29) yes, fed mied (93) other (31) both (60) yes, fed separately
http://d.doi.org/10.3168/jds.2015-9979 frequency of pushing feed (times/d) 3 (0-30) concentrates fed as TMR yes (22) no (157) maimum amount of concentrates 9.7 (0-15.0) (kg DM/cow/d) DIM when ma. amount of 21 (0-90) concentrates is fed (d) maimum waiting time for milking (min) routine herd claw trimming yes no individual cow claw trimming yes no between herd trimming events who trims claws professional stockperson both (97) (35) (47) frequency of footbaths per month 1 (58) >1 (68) none (53) length of transition period > 4 weeks yes (126) no (53) predominant place of calving in yes (145) no (32) calving pen or on pasture heifers are introduced in lactating yes (87) no (92) group before calving heifers are introduced in lactating yes (153) no (26) group individually Holstein-Fr. breed for 10% of cows yes (157) no (22) summer pasturing yes (138) no (41) outdoor loafing area in winter for dry yes (16) no (163) cows and/or young stock temporary fiing of cows after yes (36) no (143) milking cows in heat are fied or separated yes (59) no (114) sick pen for ill cows only yes (24) no (155) stalls cleaning frequency (times/d) 2.0 (0.3-6.0) stalls littering frequency (times/d) 2.0 (0-15.0) 1 Because herds in the present study were not selected randomly, they were not representative of the population of Dutch dairy farms as a whole. 2 Variable ecluded from analysis due to observed prevalence < 5% 3 Economic returns per average kg milk, fat, and protein, based on 305 d milk yield, fat contents, and protein contents, corrected for calving interval, and age and season of calving. 4 Udder infection is defined as SCC > 150,000 cells / ml in first parity cows and SCC > 250,000 cells / ml in second or higher parity cows 5 Light (versus dark ) was described as the observer being able (versus not being able) to read a newspaper in a cubicle around midday 6 Slippery was described as the observer eperiencing slipping and having little grip during turning 7 Fecal contamination of a stall was described as cow droppings or >20% manure cover in the rear 1/3 part of the lying area 8 Continuous availability was described as at least 180 L (1 wheelbarrow) roughage per 25 cows anytime during the farm visit 9 Sufficient refers to at least 1 water bowl for 15 cows and/or 4 cm of through per cow; not sufficient refers to otherwise (adapted from Welfare Quality, 2009) (57)
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Table S2. Potential predictors associated (P<0.20) with an animal-based dairy cattle welfare indicator in the univariable analyses. Animal-based welfare indicator Potential predictor Routinely collected herd data Severely lame cows (%) Cows with lesions/swellings (%) Cows with dirty hindquarter (%) Very lean cows (%) Avoidance distance inde Frequency of displacements Herd size Change in herd size Average age of cows > 2y Average milk yield/cow/d Average DIM Net result Slaughter % Replacement % On-farm mortality of cattle: aged 0-3 d aged > 2 y Herd biosecurity status Certified IBR free status Certified BVD free status Bulk milk SCC Cows with SCC > 400 000 Udder infection New udder infection Heifers with udder infection Average SCC of cows: < 60 DIM 60-120 DIM 120-210 DIM > 210 DIM Average milk fat Proportion fat to protein of cows < 60 DIM Average urea
http://d.doi.org/10.3168/jds.2015-9979 Average services per cow Cows with > 2 services Non-return 56 d Average calving interval % abortion Average epected calving interval Data collected with housing checklist Cow to stall ratio Predominant surface of lying area Average height of stall neck rail Average stall length Average stall width Head lunge impediments Type of stall division Stalls with fecal contamination Average width of passages Average width of alleys Width of alley behind the feeding rack Rims or pits in the floor Daily deposition of roughage on the feed bunk Type of feeding rack Average feeding rack height Smallest feeding rack height Average feed space per cow Roughage contaminated Continuous availability of roughage Frequency of pushing feed Lactation groups Alleys with dead ends Sufficient number and length of drinkers Light intensity Automatic milking Data collected with management interview
http://d.doi.org/10.3168/jds.2015-9979 Pasturing Use of footbaths Routine herd claw trimming Individual claw trimming Daily deposition of roughage on feed bunk Different types of roughage Concentrates fed as TMR Frequency of pushing feed Stalls littering frequency Stalls cleaning frequency Dry cow groups Length of transition period Cows in heat are fied or separated Sick pen for ill cows only Ma. amount of concentrates per cow/d Holstein-Fr. breed
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Figure S1. Specificity (i.e. percentage of herds with better welfare correctly identified as having better welfare) and relative number of farm visits at different levels of sensitivity (i.e. percentage of herds with poorer welfare correctly identified as having poorer welfare) when using a screening test for identification of herds with more than 61.2% cows with lesions or swellings Page 7 of 11
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Figure S2. Specificity (i.e. percentage of herds with better welfare correctly identified as having better welfare) and relative number of farm visits at different levels of sensitivity (i.e. percentage of herds with poorer welfare correctly identified as having poorer welfare) when using a screening test for identification of herds with more than 60.4% cows with a dirty hindquarter. Page 8 of 11
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Figure S3. Specificity (i.e. percentage of herds with better welfare correctly identified as having better welfare) and relative number of farm visits at different levels of sensitivity (i.e. percentage of herds with poorer welfare correctly identified as having poorer welfare) when using a screening test for identification of herds with more than 7.0% very lean cows. Page 9 of 11
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Figure S4. Specificity (i.e. percentage of herds with better welfare correctly identified as having better welfare) and relative number of farm visits at different levels of sensitivity (i.e. percentage of herds with poorer welfare correctly identified as having poorer welfare) when using a screening test for identification of herds with an avoidance distance inde less than or equal to 57.7. Page 10 of 11
http://d.doi.org/10.3168/jds.2015-9979 Supplemental Figure S5. Specificity (i.e. percentage of herds with better welfare correctly identified as having better welfare) and relative number of farm visits at different levels of sensitivity (i.e. percentage of herds with poorer welfare correctly identified as having poorer welfare) when using a screening test for identification of herds with more than 0.58 displacements per cow/h. Page 11 of 11