Random sampling is the method of choice to obtain a representative sample of the population [
31]. In the present study herds were conveniently selected based upon motivation, as was done in previous studies to guarantee an adequate follow up and to minimize data loss [
3,
8,
10,
13,
32-
35]. Because the sample size included 5% of the population with more than 90% of the active veterinarians and integrators represented, and because housing and feeding are highly standardized in the Belgian veal industry, the possible selection bias, caused by this selection procedure, is believed to be limited. Because of the convenience selection the sample can only be assumed indicative but not representative for the complete Belgian veal industry at present. The estimation of morbidity was based upon individual treatment by the producer, assuming that treatment rates accurately reflected illness and that increased treatment rates indicated a higher degree of morbidity at that time [
10,
33]. Antimicrobial use is however influenced by socioeconomic factors and also the personal attitude of the producer might have influenced the difference in treatment rates between the farms [
36]. Since easily administrable oral group treatments are frequently used throughout the production cycle, farmers only tended to individually treat calves when severely ill or when only few calves require treatment [
21]. In that respect, the individual calf treatments do reflect severe individual calf illness as perceived by the producers.
In the present study, the mortality risk (5,3%) was higher than previously reported for white veal calves in Canada (3,7%), the United States (2,5% and 4,2%) and Switzerland (3,0%) [
10,
11,
13,
37]. Including beef cohorts in the present study might explain the higher losses compared to studies on dairy veal calves only, since beef calves are more likely to die [
38,
39]. However, also the mortality risk within the dairy calves was relatively high (4,9%) compared to previous studies. The most likely explanation is probably the longer production cycle (28 weeks) compared to previously studied systems (16-21 weeks), which increased the days at risk. A second explanatory factor might be the housing system. In the older studies, calves were housed in individual stalls during the complete production cycle. It was postulated that this creates a higher opportunity for individual monitoring and care compared to contemporary group housing [
11]. Also the possibility to control feed uptake in individually housed calves might have been a protective factor, since the mortality risk of digestive diseases was far smaller (19,2%) in individual housing, compared to group housing in Belgium and Switzerland (41,9% and 52,0% respectively) [
13]. The exact influence of the housing type on mortality remains unclear, since individual housing is nowadays forbidden and previous comparative studies did not report mortality data [
40]. Nevertheless, the Swiss study shows that low mortality risks can be achieved in group housing in large pens. However, the fact that in that study, calves were purchased within a day, at a minimum age of three weeks and were housed at low stocking density (> 3,5 m
2/calf) most likely also contributed to the lower mortality risk.
High mortality risks (8.2%) have been reported in farms which purchase young calves from different origin [
41]. Surprisingly, the mortality risk in veal calves was similar to live born calves in dairy replacement herds in Great Britain (5.0%), Norway (4,6%), Sweden (4.0%) and crossbred cow-calf farms in Switzerland (5.0%) and even smaller than reported in large scale dairy calf rearing in Northern America (7,6% and 13,3%) [
2,
3,
7,
9,
35,
42]. Veal producers in Belgium appear to be reasonably able to manage and care for the young, highly stressed calves from multiple origin. However, compared to conventional calf rearing, preventive and metaphylactic antimicrobial drug use plays an important role in this management [
21]. Whether the current mortality risk can be maintained with less antimicrobial use is an important question for future research.
Compared to North American (5,5%) or Australian bobby calves (0,6%), transport related mortality was low in Belgian veal calves (0.3%), most likely due to shorter transportation times [
43,
44]. The finding of hydranencephalia in several dummy calves, was associated with the 2007 bluetongue outbreak in Northern Europe and illustrates how close monitoring of veal calves can assist in the detection of calf diseases of global interest [
45]. Diarrhea and related mortality was mainly an issue in the first weeks after arrival, consistent with the risk period in conventional calf rearing [
2]. The incidence rate of diarrhea (0,30 cases per 1000 calf days at risk) was smaller than in Swedish (1.17) and North American (1.50) dairy calves, most likely because calves were also monitored in the neonatal period in the latter studies [
2,
7,
8,
46]. All major pathogens of the neonatal enteritis complex were found and surprisingly also
E. Coli F5, suggesting that certain calves were much younger than two weeks old. Although
Salmonella spp. are historically reported as one of the major causes of mortality in veal calves in Belgium and recent studies still confirmed its presence on Danish veal herds, the bacteria could not be isolated from any of the suspicious cases [
47-
49]. In contrast to conventional dairy calves, diarrhea and respiratory disease occurred simultaneously in the first three weeks after arrival, which is most likely a consequence of commingling [
8,
13,
50].
BRD was the leading cause of morbidity and mortality. The incidence rate (0,95 cases per 1000 calf days at risk) was similar to Swedish (0.83; measured between birth and 13 weeks of age) or Minnesota dairy calves (1.00; measured between birth and 16 weeks of age), but smaller than in non-weaned Charolais calves in cow-calf herds (1,89; measured between birth and 26 weeks maximum) [
2,
8,
28]. Given the large amount of oral group antimicrobial treatments administered for respiratory disease in veal calves, the incidence is probably severely underestimated and a lot more calves is expected to have suffered from respiratory disease than indicated by individual treatment [
51]. Peak incidences of BRD were reached 2 to 6 weeks after arrival, which is at younger age than conventionally housed dairy heifer calves (10 weeks) [
2]. Commingling of calves is a major risk factor for BRD, and the peak incidence of respiratory disease is expected immediately after arrival [
50,
52]. Metaphylactic treatment at arrival, gradual decline of maternal immunity, incomplete maturation of the immune system and the slowly progressive nature of the dominant pathogens in European veal production, namely
Mycoplasma bovis and BVDV, might have influenced the occurrence of the peak incidence at the age of 1-1.5 months instead of at arrival [
24,
53-
55]. In contrast to cow-calf herds where the BRD incidence remains at a higher level (1.0%), hardly any veal calves still require individual treatment after 3 months of age [
2,
28]. Most likely the similar age and the all-in all-out management of veal calves limit respiratory disease to the first two months after arrival, whereas in conventional herds pathogens can constantly be transferred from older to younger calves. The long tail of the BRD mortality and morbidity curve is explained by a large proportion of chronic BRD cases (reoccurent and relapse). In addition to previous work, the present study confirms the association of BVDV with chronic pneumonia lesions and pleuritis in white veal calves [
24]. As in feedlot calves, the synergy between
M. bovis and BVDV is the cause of chronic, unresponsive pneumonia, often in association with arthritis and otitis (
M. bovis associated disease) [
24,
53,
56,
57]. In this respect, the higher incidence of arthritis and otitis compared to conventional calves is most likely the consequence of the high prevalence of
M. bovis in white veal cohorts [
8]. In the present study crossbreds had marked lower mortality due to respiratory disease. This heterosis effect has also been observed in other production systems [
38,
39,
58].
As mentioned earlier, digestive diseases were an more important cause of mortality in the recent studies on group housed calves, compared to an older study on individually housed calves [
10,
13]. In group housed veal calves in Switzerland, much more calves died from perforating abomasal ulceration (0.53% vs. 0.11% in the present study) and intestinal torsion (0.4% vs. 0.02%) compared to group housed calves in Belgium [
13]. In contrast very few calves (0.14%) died from ruminal bloat in Switzerland, whereas ruminal bloat (0.7%) and enterotoxaemia (0.5%) were the most important digestive causes of mortality in Belgium [
13]. Although both diseases occurred throughout the production cycle, the main risk period was situated near the end of the production round, when feed uptake was at its highest. In contrast to ruminal disorders, enterotoxaemia almost exclusively occurred in Belgian Blue veal calves. The causative agent is
Clostridium perfringens, but the identity of the toxin and the exact pathogenesis are still unclear [
59,
60]. Also Belgian Blue suckler calves are highly susceptible for enterotoxaemia, and it is unclear whether there is a breed predisposition or whether dietary differences between the studied production systems are the cause [
61].
Finally, one of the most remarkable causes of mortality in the present study, was idiopathic peritonitis, especially in dairy veal calves. Idiopathic peritonitis emerged only recently in veal calves and the peak incidence at week 9, shortly after the respiratory problems, suggests septicemic spread of bacteria from the lungs to the peritoneum. In one outbreak in Belgium
P. multocida capsular type F has been isolated from peritoneal fluid in two cases [
62]. Also,
P. multocida capsular type B was isolated from outbreaks of pleuritis and peritonitis in intensive dairy calf rearing facilities in New Zealand [
63]. In the present study no significant association between pneumonia and peritonitis could be demonstrated at necropsy and only
M. haemolytica and
E. coli could be isolated from peritoneal fluid. Given these contradictory necropsy results, and the identification of a specific risk period in the present study, more research is necessary to identify the aetiology of idiopathic peritonitis.
1 Koen De Bleecker,2 Miel Hostens,3 Jozefien Callens,2 Jeroen Dewulf,3 and Piet Deprez1
This article has been 
