- Email: [email protected]
Biosecurity on Finnish cattle, pig and sheep farms – results from a questionnaire
Preventive Veterinary Medicine 117 (2014) 59–67
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
Biosecurity on Finnish cattle, pig and sheep farms – results from a questionnaire Leena Sahlström ∗ , Terhi Virtanen, Jonna Kyyrö, Tapani Lyytikäinen Risk Assessment Research Unit, Finnish Food Safety Authority Evira, Mustialankatu 3, 00790 Helsinki, Finland
a r t i c l e
i n f o
Article history: Received 25 November 2013 Received in revised form 27 June 2014 Accepted 7 July 2014 Keywords: Biosecurity Questionnaire Cattle Pig Sheep GLM
a b s t r a c t Biosecurity is important in order to prevent disease transmission between animals on farms as well as from farm to farm. Personal biosecurity routines such as hand washing and the use of protective clothing and footwear are measures that should be used at all farms. Other measures are for example related to purchasing new animals to the farm. A questionnaire-based survey was undertaken to study the frequency of use of different biosecurity measures on cattle, pig and sheep farms in Finland. Information about which biosecurity measures are in use is needed for contingency planning of emerging diseases or when combating endemic diseases. Knowledge about the level of biosecurity of a farm is also needed in order to assess if and where improvement is needed. Information regarding biosecurity levels may benefit future animal disease risk assessments. A total of 2242 farmers responded to the questionnaire resulting in a response rate of 45%. The implementation frequencies of different biosecurity measures are reported. The results revealed differences between species: large pig farms had a better biosecurity level than small cattle farms. There were also differences between production types such as dairy farming versus beef cattle farming, but these were not as remarkable. Sheep farming in Finland is sparse and the large number of hobby farmers keeps the biosecurity level low on sheep farms. This might represent a risk for the entire sheep farming industry. The Finnish farmers were satisfied with their on-farm biosecurity. Eighty percent of the farmers report that they were satisfied even though the biosecurity level was not particularly high. The implementation of biosecurity measures could be further improved. Even though the disease situation in Finland is good today, one must be prepared for possible epidemics of threatening diseases. © 2014 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Biosecurity The animal health status in Finland is very good regarding epizootic diseases. However, the situation can change very rapidly as we have seen in other countries
∗ Corresponding author. Tel.: +358 404 893375. E-mail address: leena.sahlstrom@evira.fi (L. Sahlström). http://dx.doi.org/10.1016/j.prevetmed.2014.07.004 0167-5877/© 2014 Elsevier B.V. All rights reserved.
such as in the foot-and-mouth-disease epidemic in Great Britain and bluetongue epidemics in many northern European countries. Some infectious agents are transmitted via direct contact between animals, whereas others can also be transmitted indirectly through contaminated equipment, vehicles, people and vector animals (Amass and Clark, 1999). FAO defines biosecurity as: “The implementation of measures that reduce the risk of the introduction and spread of disease agents” (FAO, 2010). The importance of biosecurity is underlined in the European Union health strategy for 2007–2013 “Prevention is better than
60
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
cure” (European Commission, 2007; Dekker, 2011). Biosecurity measures prevent both direct disease transmission between animals and indirect transmission between farms (Ellis-Iversen et al., 2011). Contacts such as transports and visits between farms and especially trade of animals facilitate disease spread. All precautions in animal trade, like quarantine, buying animals from a limited number of farms, etc. decrease the risk of disease transmission. Also biosecurity actions, including hygienic measures such as washing hands and using boots and coveralls (Amass et al., 2003, 2004), as well as transport logistics and “allin-all-out” systems, to mention a few, decrease the risk of spreading diseases (Madec, 2001). On-farm biosecurity measures are implemented differently depending on the farm. Biosecurity routines on pig farms have earlier been examined in different countries; e.g. Denmark (Boklund et al., 2003/2004; Boklund et al., 2004), Belgium (Ribbens et al., 2008), Chile (Pinto and Urcelay, 2003) and recently in Canada (Bottoms et al., 2012). In Sweden biosecurity measures have also been investigated on cattle and sheep farms (Nöremark et al., 2010). However, Finland differs from many other countries regarding the animal population which is sparsely distributed throughout the relatively large country. Even the areas with the highest farming densities in Finland have less than 0.5 farms/km2 including all cattle, pig sheep and goat farms (Tike, 2009). Information about the biosecurity level on farms is important for contingency planning of emerging diseases, when combating endemic diseases in a country, or to see if and where the biosecurity needs to be improved. In addition, information on biosecurity level may benefit models for animal disease spread as well as risk assessments. 1.2. Aim The aim of this survey was to describe the on-farm biosecurity routines used on cattle, pig and sheep farms in Finland and the possible differences in biosecurity depending on animal species, farm size and production type. 2. Material and methods 2.1. Administration of the questionnaire A written questionnaire was designed to collect information about the level of biosecurity and hygiene practices on Finnish cattle, pig and sheep farms. The questionnaire was sent in March 2011 to each of 3000 cattle and 1000 pig farmers. Another slightly modified questionnaire was sent in June 2011 to 866 sheep farmers in Finland. An invitation letter was enclosed, in which the farmers were informed that their response will be treated confidentially and the results reported so that no individual answers could be recognized. A reminder was sent by mail to the farmers who did not respond, 4 weeks after the original questionnaire was mailed. No compensation was paid for the response. The questionnaires were sent by mail but could also be answered online. The online link was provided in the mailed questionnaire and on the reminder card but also in the main agricultural newspapers in March 2011. A link to
the pig and cattle questionnaire was published in two agricultural newspapers and the websites of the Association for Animal Disease Prevention (ETT), the National Health Classification Registry (Sikava and Naseva) and the Finnish Food Safety Authority Evira. The link to the sheep questionnaire was published in June 2011 in two agricultural newspapers, an electronic newsletter (Saparo) and the website of the Finnish Food Safety Authority Evira. 2.2. Questionnaire design The 10-page questionnaire consisted of questions on general demography (age, sex, education and experience in animal husbandry), the farm (production type and type of buildings) and questions about several different on-farm biosecurity measures and hygienic precautions. There were 17 (14 for the sheep farmers) different biosecurity measures from which the farmers were requested to choose and mark the ones that were always in use at the farm. Biosecurity measures that were given as separate options were for example: the use of boots, or the use of protective clothing by the farmers and visitors, respectively, hand washing, use of a separate loading area, cleaning of stables and pest control. In addition, the farmers were also asked in eight separate questions about routines when purchasing animals as there are no animal markets in Finland; did they purchase animals from more than one other farm, did they enquire about the health status of the selling farm, the use of a quarantine, and the use of farm-specific transport vehicles for animals. One of the eight questions regarding trade of animals was if the farmer followed the guidelines of ETT. ETT (the Association for Animal Disease Prevention) is a farmers’ association in Finland which, among other things, prepares guidelines for animal trade. Another question in the questionnaire dealt with satisfaction. The respondents were asked to mark if they were satisfied, not satisfied or did not know if they were satisfied or not with the biosecurity on their farm. In both the cattle/pig and sheep questionnaires there were additional questions, the results of which are not presented in this paper. The questionnaire was pretested by an expert group of veterinarians working with biosecurity and it was edited according to their comments. The complete questionnaire (in Finnish or Swedish) is available upon request from the authors. 2.3. Selection of farmers There were 16,714 cattle farms, 2343 pig farms and 2576 sheep farms in Finland in 2009 (Finnish farm registry, Tike, 2009). The sample of 3000 cattle, 1000 pig and 866 sheep farmers represented 18%, 43% and 33% of the farms in Finland in 2009, respectively. The sample size was determined based on earlier experience of response rate (approximately 20%) among Finnish farmers and the aim was to get as big and representative sample as possible. Sampling was based on farm size which is defined here as the number of animals (cows and heifers >6 months, sows and finishers > 3 months) on the farm. The selection of large farms was motivated because there is a tendency towards larger farm size whereas smaller farms tend to discontinue
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
their production (Pyykkönen et al., 2010). The largest 10% of the farms were sampled more often than other farms. Among pig and sheep farmers, the largest 10% of all pig and sheep farms received the questionnaire, and the rest of the farms were randomly selected based on randomly generated numbers (SPSS software) from the Finnish farm registry 2009 (Tike, 2009). While there are so many more cattle farms in Finland, only 45% of the largest 10% was randomly chosen, and the rest was a random sample (randomly generated numbers, SPSS software) from the Finnish farm registry 2009 (Tike, 2009). 2.4. Statistical methods used Data were entered into Webropol 1.0 survey software and stored in a database using Microsoft Office Excel 2010 (Microsoft Co., Redmond, Washington, DC, USA). Statistical analyses were performed using SPSS Statistics 19 software (SPSS, Chicago, IL) and R 2.14.0 software (R core team, 2013). 2.5. Multivariable models 2.5.1. Implementation of biosecurity measures and practices The implementation frequencies of biosecurity measures and practices were estimated with multivariable models for a farm of typical size (Table 2) of each farm type in Finland (Tables 4 and 5). Farm types in these models were sow farm, farrow-to-finish farm, fattening pig farm, dairy farm, beef cattle farm, suckler cow farm and sheep farm. Analyses were performed using logistic regression (generalized linear models, GLM, family = binomial, link = logistic). There were 17 models for biosecurity measures in use at the farm (Table 4) and 8 models for biosecurity practices concerning purchase of live animals (Table 5). The equation for the models is as follows: Implementation of a biosecurity measure/practice = farm type + farm type × standardized farm size. Standardizing the farm size was performed as follows: the farm size for each farm was transformed with natural logarithm and then adjusted by the mean and standard deviation (z-score, SPSS) of the farm size of that production type in Finland. Thus, the standardized typical size of a farm type (calculated with the mean and the standard deviation of the ln transformed farm sizes for a farm type) was equal to 0. A negative value for farm size represents a smaller and a positive value a larger farm than the typical one for that farm type. The model did not have a common intercept, so every farm type got a separate intercept from the analysis representing the implementation frequency on a typical sized farm in that farm type. Farm type × farm size interactions were included into the model to allow separate slopes of farm size-related influence on implementation frequency of a measure within each production type. Thus, the statistical significance of the interaction term reflects a significant effect of the farm size on the implementation frequency within the farm type. The coefficients for the farm types were transformed into implementation probabilities of a typical sized
61
farm (frequencies) according to the following equation (McCullagh and Nelder, 1989): P(implementation of a measure) = exp(coefficient of a farm type)/[1 + exp(coefficient of a farm type)]. Confidence intervals for implementation frequencies were calculated by applying the standard error (SE) of the coefficient of a farm type: 95% CI for implementation of a measure = exp (coefficient ± 1.96 × SE)/[1 + exp(coefficient ± 1.96 × SE)]. 2.5.2. Association of the satisfaction of the farmer with the implementation of the biosecurity measures Generalized linear models (GLM, family = binomial, link = logistic) were also built for 17 biosecurity measures to find a possible association of satisfaction with the implementation of the measure. The model included farm type (pig or cattle), standardized farm size and whether or not the farmer was satisfied as explanatory variables. Also, the interaction term farm type × satisfaction was included when it was statistically significant. The equation for the models is as follows: Implementation of a biosecurity measure = type + standardized farm satisfaction + farm size + satisfaction × farm type. Statistical significance of parameters in GLM models was determined by dividing the parameter estimate by the standard error and comparing the result with the z-distribution (standard normal distribution) using R software. Statistical significance cut off level was defined as ***<0.001, **<0.01, *<0.05. 3. Results 3.1. Response rate The overall response rate was 42% (2036 replies); cattle farmers 41% (1236), pig farmers 42% (420) and sheep farmers 44% (380). The division into different production types is presented in Table 1. The reminder after 4 weeks increased the response rate substantially; 750 (34%) replies Table 1 Respondents from a biosecurity questionnaire to cattle, pig and sheep farmers in Finland were divided into three different production types among each species (except sheep farmers) as described in the table. The types are based on registration data from 2009 (Tike, 2009). The type correlates well with the production type given by the farmers in the questionnaire; the kappa value is 0.85. Production type
Number of respondents
Proportion of respondents (%)
Dairy Suckler cow Beef cattle Cattle (tot) Sow Farrow-to-finish Fattening pigs Pig (tot) Sheep Missing
893 140 181 1214 146 151 123 420 380 22a
74 12 15 100 35 36 30 100 na na
na = not applicable. a Known to be cattle farmers but missing information about production type.
62
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Table 2 Number of animals on a farm in different production types in 2009 in Finland (Tike, 2009). The minimum number of animals was one in all production types. A typical farm referred to in the text is the mean of ln transformed farm sizes of all Finnish farms within a farm type. Production type
Sow Farrow-to-finish Fattening pigs Dairy Beef cattle Suckler cow Sheep
Number of animals Mean
90th percentile
Max
107 259 448 43 32 51 12
502 955 1410 97 139 130 76
4227 6455 4586 574 1895 853 728
were received after the reminder. The majority of the respondents used the paper form, but 17% answered online. 3.2. Production type The cattle farm respondents represented either dairy farms, beef cattle herds or farms with suckler cows. The largest group of respondents was dairy farmers. Among pig farmers the respondents were divided into three groups representing sow farms, farrow-to-finish farms, and farms with only fattening pigs/finishers (Table 1). If a farmer had more than one species on the farm he was categorized according to the main production type. Sheep farmers were not divided into separate groups for the analysis of the questionnaire. However, the sheep farmers are quite a heterogeneous group, mainly consisting of hobby farmers with a small number of animals (mean number of ewes is 2) and only a few professional farmers that make sheep farming a living (only 18% earns more than 70% of their income from sheep farming (Table 3)).
production types. The farmers’ use of biosecurity measures is presented in Table 4. The use of the different biosecurity measures separated between species and production type is presented in Table 4. The results indicate that farm size usually has a significant positive effect on the use of different biosecurity measures as shown in Table 4. In a few exceptions farm size had a negative effect on the use of a biosecurity measure. This was true for “The farmer and his family and the visitors wash their hands after working in the stable” on sow farms, “control of rodents and birds in the animal stable or shelter at the feeding table” and “control of rodents and birds in the feed storage” on dairy farms. In the questionnaire to the sheep farmers, the questions about protective clothing, boots and hand washing were not separated between visitors and the owners because there are so few visitors on sheep farms. In the comparison between species, the cattle and pig owners’ habits (not including the visitors) are compared to the sheep farmers. Among the sheep farmers 5.7% reported (ticked the box) that they did not use any of the biosecurity measures mentioned. The corresponding answer among the cattle and pig farmers was 1.1% and 0.2%, respectively. 3.5. The farmers’ routines in trading animals Table 5 shows results only for those respondents that reported that they buy live animals (303 (77%) of the pig farmers, 614 (54%) of the cattle farmers and 182 (48%) of the sheep farmers). A significant positive effect of farm size is reported in connection to Table 5. In most cases there is a positive effect of farm size on the use of a biosecurity measure. The only negative effect of farm size is noted for “I tend to buy animals from as few farms as possible” on beef cattle farms (Table 5).
3.3. Description of the respondents
3.6. Satisfaction with biosecurity
A description of the respondents including average age, sex, education, farming experience, and main income is summarized in Table 3. Education is here defined as all education in the agricultural field after comprehensive school which includes nine grades, from the age of 7 to 16. The degree of education correlates with the farm size (p < 0.001) and the largest farms (10%) have a positive correlation with the main income (p < 0.001). The characteristics of the sheep farmers differ significantly from those of cattle and pig farmers. Less than 20% of the sheep farmers responded that they got their main income from sheep farming. Also the degree of education in the agricultural field is sparser among the sheep farmers than among the other types of farmers. The group of non-respondents was similar to respondents regarding the production types and the fraction belonging to the largest 10% in farm size.
Of the respondents, depending on production type, 80–88% were satisfied with their biosecurity, 3–6% were not satisfied and 8–13% did not know whether they were satisfied or not. Pig and sheep farmers were satisfied slightly more often than cattle farmers. When satisfaction was related to biosecurity measures in use at the farms, it was revealed that satisfied farmers significantly (p < 0.05) more often used several biosecurity measures than those who were not satisfied or did not know whether they were satisfied or not, i.e. out of 17 measures analysed 10 were significantly more often marked as in use by satisfied farmers (Table 6).
3.4. The farmers’ use of biosecurity measures Biosecurity was better implemented on pig farms than on cattle and sheep farms. There are big differences in some biosecurity measures between species but also between
4. Discussion Pig farmers implement more biosecurity measures on their farms than cattle and sheep farmers do. This was expected due to differences in management practices such as the frequent use of the all-in-all-out system on pig farms and the small number of pigs kept outdoors. Free range farming complicates the use of many traditionally used biosecurity measures for example a barrier between clean
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
63
Table 3 Presentation of age, gender, education, years of experience, and earning the main income from the farm among respondents in a Finnish questionnaire about biosecurity. Farmers of
Average age of farmers (min–max)
Women, % of respondents (95% CI)
Agricultural educationa , % of respondents (95% CI)
Farming experience, mean number of years (90% PI)
Main income >70%, % of respondents (95% CI)
Cattle Pig Sheep
47 (19–76) 47 (19–76) 52 (18–78)
28 (26–31) 19 (15–22) 47 (42–52)
68 (65–70) 75 (71–79) 42 (37–47)
22 (4–40) 22 (4–40) 16 (3–32)
75 (72–77) 59 (54–63) 18 (14–22)
CI = confidence interval; PI = percentile interval. a Here education includes additional or higher degree of education in the field of agriculture gained by the respondent after comprehensive school that includes 9 grades.
and dirty areas, rodent and wild bird control, and the control of access for visitors and pets. This is true also for cattle and sheep farms during the pasture season. There is large variation in farm size within all farm types, ranging from hobby farms with only a few animals to large professional farms with hundreds of dairy cows or thousands of pigs. A correction for farm size was done in order to show the biosecurity standard on a typically sized Finnish farm. The standard sized sheep farm in Finland is relatively small because of the large fraction of hobby farmers. The differences between sheep farmers compared to cattle and pig farmers regarding education, age and income also reflect the large group of hobby farmers. The hobby farmers are not necessarily as aware of disease risks as the professional farmers are. This might be a health risk for the whole sheep population. In case of an animal disease that also affects other species, such as foot and mouth disease, the disease could also spread to other production sectors. On the other hand, hobby farmers are likely less connected with professional farms. It is also evident from the results of the questionnaire that the size of the farm affects the use of a biosecurity measure on the farm (Tables 4–6). Large farms have more biosecurity measures in use than smaller farms, which is consistent with studies e.g. in Sweden (Nöremark et al., 2010) and Belgium (Ribbens et al., 2008). However, there are a few exceptions revealed in the survey; smaller dairy farms have a higher probability of using rodent control, and on smaller sow farms there is a higher probability that farmers and visitors wash their hands after working in the stable. These are important biosecurity measures and should not be neglected on larger farms. Due to their challenging realization, there are rather few published studies regarding the effect of biosecurity measures, such as the changing of boots and coveralls, on disease prevention. However, even though the effect of individual measures has not been studied in detail, it can be concluded that these cheap and easy means of protecting farm animals are warranted. On the other hand, there are more expensive or demanding types of biosecurity measures that are more frequently in use on larger farms or in some production types than others. For example a separate loading space for animals and a quarantine facility are more expensive or difficult to implement. If these are not planned when the stables are built it may be a too costly a task to implement them later. Bottoms et al. (2012) speculate that large farms might see the risk of a pathogen introduction as more disastrous than small farms or they have the financial potential
to establish even the more expensive biosecurity measures. In addition, the expenses of biosecurity per animal will be lower with increasing farm size (Niemi et al., 2013). There are differences between production types and their use of biosecurity measures as seen in Tables 4 and 5. However, the differences seem logical and are mostly easily explained based on different management practices. For example a separate loading space might not be needed in a finishing pig stable if the all-in-all-out system is applied. A leak proof container for dead animals might seem unnecessary for a dairy farm because of the small number of dead animals. To improve the overall biosecurity on a national level, one needs to increase the use of biosecurity measures on farms that are not as well managed in terms of biosecurity, especially on smaller cattle and sheep farms. Many biosecurity measures do not have to be expensive, e.g. the use of farms-specific protective clothing and boots. The most important factor is to convince the farmers to change their habits in order to improve biosecurity on their farms. This is not an easy task (Racicot et al., 2012). However, it is an important factor in enhancing the biosecurity in the country, motivating the farmers to improve biosecurity by implementing more biosecurity measures on their farms. Vermin control is quite well applied on most farm types as seen in this survey, which is in accordance with Ribbens et al. (2008). On the other hand, a hygiene barrier between the clean and dirty areas is missing on the majority of farms, and the use of protective clothing and boots could be improved on many farms. Hand washing is a documented measure to prevent spread of diseases, but there are still farmers who do not wash their hands after they finish their work. In our survey sow farms had the highest frequency of quarantines in use (57%; 44–69, 95% CI) (Table 5). This is approximately the same frequency as on large sow farms in Chile (Pinto and Urcelay, 2003), where the herd size ranged between 50 and 3000 sows; Md 420. Quarantine can be understood in different ways; however, the numbers seem to be quite high in Finland taking into account that the sow farms sampled in Chile were larger than in Finland (Table 2). Nöremark et al. (2010) reported that 74% of Swedish pig farmers and 23% of Swedish cattle farmers required visitors to use protective clothing when entering the stables. We on the other hand found that visitors on Finnish sow-, farrow-to finish- and fattening pig farms used protective clothing on 86, 85 and 75% of the farms, respectively. On Finnish cattle farms the corresponding frequencies were
64 Table 4 Results from multivariable models for implementation of biosecurity measures. Coefficients from the models are transformed into probabilities for implementation on typically sized pig, cattle and sheep farms in Finland based on a questionnaire study to farmers [% (95% CI)]. Statistically significant effect of farm size is marked ***<0.001, **<0.01 and *<0.05 accordingly. The farm size effect was generally positive, but negative in the following activities and types of production: “The farmer and his family and the visitors wash their hands after working in the stable” in sow farms, “control of rodents and birds in the animal stable or shelter at the feeding table” and “control of rodents and birds in the feed storage” in dairy farms. na = not applicable. Sow farm
Farrow-to-finish
Fattening pigs
Dairy farm
Beef cattle
Suckler cow
Sheep farm
The farmer and his family use protective clothing in the stables The farmer and his family use boots or protective shoes in the stables Visitors use protective clothing (coveralls) Visitors use boots The farmer and his family wash their hands after working in the stable Visitors wash their hands after the visit The use of a barrier that separates the clean area from the dirty area and that is not passed without changing protective clothing and shoes The use of a separate loading area Washing the loading area after use Outside the animal stables there is a leak-proof container for dead animals The animal stables are cleaned between each “batch” Doors are kept locked Animals are divided into compartments The traffic on the farm is organized so that biosecurity aspects are taken into account Control of rodents and birds Control of rodents and birds in the animal stable or shelter at the feeding table Control of rodents and birds in the feed storage
89 (83–94)*
82 (75–88)
85 (77–91)
62 (59–65)*
69 (60–76)
59 (50–68)*
24 (19–30)***
79 (71–85)
73 (64–79)
86 (78–91)
63 (60–66)***
59 (50–68)
66 (57–74)
36 (30–42)***
86 (78–91) 81 (73–87) 98 (93–99)**
85 (77–90)*** 75 (67–82)*** 91 (85–95)
75 (65–83)*** 73 (63–81)*** 85 (78–91)
51 (47–54)*** 68 (64–71)*** 88 (85–90)**
34 (25–43)*** 39 (30–49)*** 84 (77–89)
27 (19–36)* 50 (40–0)*** 81 (73–87)
na na 67 (61–72)***
89 (82–94)* 35 (27–44)*
75 (67–81) 27 (20–35)*
66 (56–74) 26 (19–36)
63 (59–66) 11 (9–13)***
43 (34–52) 8 (5–15)
42 (34–51) 9 (5–15)
na 3 (1–6)***
59 (50–67)*** 24 (18–33) 44 (35–53)***
74 (67–81) 36 (28–45) 54 (45–63)***
25 (17–36)*** 18 (11–27) 59 (49–69)***
3 (2–4)*** 2 (1–4) 1 (1–2)
2 (1–7) 1 (0–6) 0 (0–4)*
7 (3–13) 3 (1–8) 2 (0–6)
6 (4–11)*** na 4 (2–7)***
32 (24–41)
39 (31–48)
79 (71–86)
17 (15–20)
43 (34–52)**
26 (19–34)
42 (36–49)*
49 (40–58)** 51 (41–60)*** 46 (38–55)
47 (39–56)* 53 (45–62)*** 42 (34–51)*
56 (46–66)*** 47 (37–58)*** 29 (20–39)**
12 (10–14) 10 (8–12)*** 26 (23–28)***
15 (10–22) 15 (9–24)*** 11 (6–20)***
8 (4–14) 30 (22–39) 23 (16–3)***
15 (11–20)*** 24 (18–30)*** 11 (7–15)***
89 (83–94) 59 (50–68)
90 (83–94) 64 (54–72)
97 (91–99) 65 (55–73)
75 (72–78)** 54 (50–57)***
67 (58–75)*** 37 (29–46)
69 (60–76) 17 (11–25)
52 (46–58) 38 (32–44)***
56 (47–64)
60 (51–68)*
65 (55–73)
42 (38–45)*
34 (26–43)
30 (22–39)
41 (35–48)**
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Variable description
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
65
Table 5 Results from multivariable models for implementation of biosecurity practices concerning purchase of live animals. Coefficients from the models are transformed into probabilities for implementation on typically sized pig, cattle and sheep farms in Finland based on a questionnaire study to farmers [% (95% CI)]. The figures in the table refer to 303 (77%) pig farmers, 614 (54%) cattle farmers and 182 (48%) sheep farmers that returned the questionnaire and reported that they buy live animals. Statistically significant effect of farm size is marked ***<0.001, **<0.01 and *<0.05, accordingly. The farm size effect was generally positive, but negative in the following activity and type of production: “I tend to buy animals from as few farms as possible” in beef cattle farms. Variable description
Sow farm
Farrow-to-finish
Fattening pigs
Dairy farm
Beef cattle
Suckler cow
Sheep farm
I use a farm-specific transport vehicle when moving animals I use a quarantine, which is a building or compartment that keeps newly purchased animals separate from the other animals on the farm I check the disease status of the selling farm before purchasing I buy animals preferably from the same farms I follow the guidelines of ETT (Association for Animal Disease Prevention) when purchasing animals I buy animals via slaughterhouses that coordinate the trade between farms I tend to buy animals from as few farms as possible I demand a health certificate before purchase (Sheep farmers check the Maedi-Visna and scrapie status of the sheep before buying)
10 (4–22)*
27 (18–39)
13 (7–22)
37 (31–43)
46 (35–58)*
28 (17–41)*
91 (82–95)***
57 (44–69)
37 (27–49)
17 (27–49)
15 (11–21)**
36 (26–48)
51 (38–65)
10 (5–19)***
61 (48–72)*
67 (55–74)
37 (26–49)
90 (85–93)
41 (31–53)
85 (74–92)
81 (70–89)***
84 (73–91)
84 (74–91)
57 (46–68)
55 (49–62)
54 (42–65)
62 (48–75)
na
94 (86–98)
89 (79–95)
93 (85–97)
92 (88–95)**
86 (76–93)
88 (78–94)
na
78 (67–86)
84 (74–91)
92 (85–96)
21 (16–27)
80 (71–87)**
54 (40–68)*
na
95 (86–98)
93 (84–97)
86 (77–92)
92 (88–95)
69 (57–79)*
92 (81–97)
93 (83–97)***
33 (22–47)
44 (32–56)
16 (9–27)
71 (65–77)
14 (8–24)
82 (69–90)
70 (56–80)**
na = not applicable.
51, 34 and 27% on dairy, beef and suckler cow farms, respectively (Table 4). The farm size in Sweden in 2006 was slightly larger with an average cattle herd size of 64 animals compared to Finland (Table 2). The average pig herd size in Sweden with 116 sows and 495 piglets and fattening pigs was also bigger compared to Finland (Nöremark et al., 2010). In Sweden there is a separate loading area on 22% of the cattle farms and 86% of the pig farms (Nöremark et al., 2010). In Finland loading areas are very uncommon on cattle farms; 3, 2 and 7% have a loading area on dairy, beef and suckler cow farms, respectively. On pig farms in Finland loading areas are more common as 74% of the farrow to finish farms, 59% of the sow farms and 25% of the fattening pig farms report using one (Table 4). The Finnish farmers are generally satisfied with their biosecurity; more than 80% of the respondents reported that they were satisfied. This is in accordance with results from other studies, e.g. in Sweden 72% of the farmers were satisfied, even though they did not apply any or only few routines related to biosecurity (Nöremark et al., 2010). Spanish pig farmers also seem generally satisfied with the biosecurity applied on their farms with a self-scoring mean of 6.72 (min 0, max 10; SD 1.57) (Casal et al., 2007). The Spanish farm size in the study of Casal et al. (2007) ranged from 70 to 7000 sows (25, 50 and 75% quartiles were 203, 328 and 532 sows, respectively) and thus were much larger than the Finnish sow farms (Table 2). The Finnish farmers
that reported satisfaction with their biosecurity more often implemented biosecurity measures than the unsatisfied farmers. However, farmers were still quite satisfied even though there was improvement potential in the level of biosecurity as a whole. When farmers were satisfied with a lower biosecurity level it might reflect a lack of knowledge of biosecurity or an underestimation of the risks threatening their production. One might also assume that satisfied farmers are less likely to improve their biosecurity. The favourable animal disease situation in Finland might lull farmers into a false feeling of security. Nevertheless, it is important to be prepared for a rapid change in the disease situation. However, as long as the farmer does not experience any problems with animal diseases, he will most likely be satisfied and not feel the need to improve the biosecurity level (Alarcon et al., 2013). In other words, motivating farmers is the key to improving biosecurity on farms and in the country as a whole. One might suspect a bias among the respondents towards a higher response likelihood among farmers more interested in biosecurity. Furthermore, one can assume that farmers with larger farms would be more interested in biosecurity, because of the higher economic risk involved. However, a comparison between the respondents and non-respondents revealed no difference regarding farm size or production type. Farmers can also report that they use more biosecurity measures than they really do because they might experience that they are expected to
66
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Table 6 Results from multivariable models for the association of satisfaction with implementation of biosecurity measures. Generalized linear models included farm type in question (pig or cattle), standardized farm size and whether the farmer was satisfied with his biosecurity measures in use. Estimates are given for basal level (intercepts) and predictors. Statistically significance is marked ***<0.001, **<0.01 and *<0.05, accordingly. Also, an estimate for the interaction term between farm type and satisfaction is included when it is statistically significant. Typical size means here the mean of ln transformed and standardized farm sizes of all Finnish farms within that farm type. Non-satisfied farmers include here also the ones that were unsure. Variable Farmer and his family use protective clothing in the stables Farmer and his family use boots or protective shoes Visitors use protective clothing Visitors use boots Farmer and his family wash their hands after working Visitors wash their hands after the visit Use of a barrier that separates the clean area from the dirty area and that is not passed without applying protective clothing and shoes Use of a separate loading area Washing the loading area after use Animal stables are cleaned between each “batch” Outside animal stables is a leak-proof container for dead animals Doors are kept locked Traffic is organized so that biosecurity aspects are taken into account Animals are divided in compartments Control of rodents and birds on the feeding table Control of rodents and birds in the feed storage General control of rodents and birds
Basal level (not satisfied, cattle farm, typical size)
Satisfied
Pig farm
Standardized farm size
Interaction term satisfied × pig farm
0.35*
0.21
0.65
0.20***
0.77*
0.50***
0.05
0.76***
0.20***
n.s.
1.65*** 0.58*** −0.60
0.72*** 0.67*** 0.23**
n.s. n.s. 1.07*
−0.65*** 0.38** 1.90***
0.60*** 0.21 0.02
0.07 −2.69***
0.34* 0.63**
0.78*** 1.23***
0.03 0.47***
n.s. n.s.
−3.83*** −4.55*** −1.75***
0.65** 0.90** 0.58***
3.45*** 2.73*** 1.16***
0.72*** 0.38*** 0.23***
n.s. n.s. n.s.
5.15***
0.89***
n.s. −0.79* n.s.
−4.83***
−0.26
−2.67*** −1.75***
0.67** 0.80***
2.87*** 0.52***
0.31*** 0.75***
−2.14*** −0.67***
0.29 0.74***
1.92*** 0.54***
0.93*** −0.24***
n.s. n.s.
−1.03***
0.69***
0.89***
−0.03
n.s.
0.28
1.35***
0.83***
0.33***
n.s.
n.s. = non significant and thus not included in the model.
do so. This results in a too positive picture of the on-farm biosecurity level in Finland. If one hypothesizes that the respondents would be more interested in biosecurity than non-respondents, and that there is a positive bias in their responses, one could say that the need for improvement of biosecurity is even bigger than shown in this study. However, we might still expect that the positive bias is somewhat the same in all production types and the differences shown between the production types would still be approximately the same. A questionnaire is proven to be a good tool for the collection of data (Ribbens et al., 2008). Advantages with a postal questionnaire are that it is cheap to administer and you easily reach a large group of people. The response rate, however, is not as high as it might be in a telephone interview. On the other hand, in reality a telephone interview approach limits the number of farmers that can be contacted. A weakness of the questionnaire at hand is that while the same questionnaire was used for both cattle and pig farmers for the sake of comparability, some of the farmers commented that not all questions were suitable for their particular farm type. This must be kept in mind e.g. when interpreting answers regarding different compartments in the stable or the loading space, which might mean completely different things, depending on the species of animal kept on the farm. The evaluation of the answers of a questionnaire may be challenging because even though the questions are made as clear as possible, the questions may
be comprehended differently depending on the respondent’s current situation. Finland differs from many European countries by its sparsely distributed animal population. There might also be large differences in on-farm-biosecurity between European countries. These differences should be explored further, considering that EU regulations on animal diseases are the same for all Member States and trade between European countries is common and occurs in large volumes. 5. Conclusion Further improvement of the biosecurity level is possible on farms in Finland. Pig farms generally have a higher level of biosecurity than cattle and especially sheep farms. The large proportion of hobby farmers among sheep farms and smaller farms in cattle and pig production, have a negative influence on the level of biosecurity. Counselling regarding potential risks should be emphasized in order to increase the level of biosecurity especially on smaller farms, so as to decrease the risks to production as a whole. Acknowledgements The authors would like to thank the Ministry of Agriculture and Forestry (Makera Dnr 2803/502/2008) for funding this project. The authors would also like to thank all the farmers that have contributed by responding to the
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
questionnaire and Kitty Schulman for critically reviewing the paper. References Alarcon, P., Wiedland, B., Mateus, A., Dewberry, C., 2013. Farmers’ Perceptions, Attitudes, Influences and Management of Information in the Decision-Making Process for Disease Control. Society for Veterinary Epidemiology and Preventive Medicine, Madrid, Spain, 20-22.3 2013. ISBN 978-0-948073-20-5. Amass, S.F., Mason, P.W., Pacheco, J.M., Miller, C.A., Ramirez, A., Clark, L.K., Ragland, D., Schneider, J.L., Kenyon, S.J., 2004. Procedures for preventing transmission of foot-and-mouth disease virus (O/TAW/97) by people. Vet. Microbiol. 103, 143–149. Amass, S.F., Pacheco, J.M., Mason, P.W., Schneider, J.L., Alvarez, R.M., Clark, L.K., Ragland, D., 2003. Procedures for preventing the transmission of foot-and-mouth disease virus to pigs and sheep by personnel in contact with infected pigs. Vet. Rec. 153, 137–140. Amass, S.F., Clark, L.K., 1999. Biosecurity considerations for pork production units. Swine Health Prod. 7, 217–228. Boklund, A., Alban, L., Mortensen, S., Houe, H., 2004. Biosecurity in 116 Danish fattening swineherds: descriptive results and factor analysis. Prev. Vet. Med. 66, 49–62. Boklund, A., Mortensen, S., Houe, H., 2003/2004. Biosecurity in 121 Danish sow herds. Acta. Vet. Scand. Suppl. 100, 5–14. Bottoms, K., Poljak, Z., Dewey, C., Deardon, R., Holtkamp, D., Friendship, R., 2012. Evaluation of external biosecurity practices on southern Ontario sow farms. Prev. Vet. Med. 109, 58–68. Casal, J., De Manuel, A., Mateu, E., Martin, M., 2007. Biosecurity measures on swine farms in Spain: perception by farmers and their relationship to current on-farm measures. Prev. Vet. Med. 82, 138–150. Dekker, A., 2011. Biosecurity and FMD transmission. Vet. Rec. 168, 126–127. Ellis-Iversen, J., Smith, R.P., Gibbens, J.C., Sharpe, C.E., Dominiguez, M., Cook, A.J.C., 2011. Risk factors for transmission of foot-and mouth disease during an outbreak in southern England in 2007. Vet. Rec. 128, http://dx.doi.org/10.1136/vr.c6364. European Commission, 2007. A New Animal Health Strategy for the European Union (2007–2013) Where Prevention Is Better Than Cure,
67
http://ec.europa.eu/food/animal/diseases/strategy/docs/animal health strategy en.pdf (accessed 12.08.13). Food and Agriculture Organization of the United Nations/World Organisation for Animal Health/World Bank, 2010. Good practices for biosecurity in the pig sector – issues and options in developing and transition countries. In: FAO Animal Production and Health Paper No. 169. FAO, Rome, http://www.fao.org/docrep/012/i1435e/ i1435e00.pdf (accessed 06.12.12). Madec, F., 2001. Biosecurity on Pig Units: A Major Issue for Herd Health Maintenance, http://www.cnpsa.embrapa.br/abravessc/pdf/Palestras2001/Francois Madec1.pdf (accessed 13.08.13). McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models, second ed. Chapman & Hall, pp. 511. Niemi, J.K., Sahlström, L., Virtanen, T., Kyyrö, J., Lyytikäinen, T., Sinisalo, A., Lehtonen, H., 2013. The value of biosecurity and the role of economic incentives in promoting the adoption of biosecurity. In: Paper Prepared for Presentation at the V Workshop on Valuation Methods in Agro-food and Environmental Economics: Methodological and Empirical Challenges in Valuation Methods. 137th EAAE Seminar, Castelldefels, Spain, 27–28 June 2013. Nöremark, M., Frössling, J., Lewerin, S., 2010. Application of routine that contribute to on-farm biosecurity as reported by Swedish Livestock farmers. Transbound. Emerg. Dis. 57, 225–236. Pinto, C.J., Urcelay, V.S., 2003. Biosecurity practices on intensive pig production systems in Chile. Prev. Vet. Med. 59, 139–145. Pyykkönen, P., Lehtonen, H., Koivisto, A., 2010. Structural Change and Investments in Finnish Agriculture 2010–2020. PTT Working Papers 125, pp. 24, ISBN 978-952-224-061-3 (pdf), ISSN 1796-4784 (pdf) (In Finnish, English summary). Available from: http://www.ptt.fi/ dokumentit/tp125 1111100930.pdf (accessed 15.08.13). R Core Team, 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org/ Racicot, M., Venne, D., Durivage, A., Vaillancourt, J.-P., 2012. Evaluation of strategies to enhance biosecurity compliance on poultry farms in Quebec: effect of audits and cameras. Prev. Vet. Med. 103, 208–218. Ribbens, S., Dewulf, J., Koenen, F., Mintiens, K., De Sadeleer, L., de Kruif, A., Maes, D., 2008. A survey on biosecurity and management practices in Belgian pig herds. Prev. Vet. Med. 83, 228–241. Tike Information Centre of the Ministry of Agriculture and Forestry, 2009. Finnish Farm Registry.
Contents lists available at ScienceDirect
Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed
Biosecurity on Finnish cattle, pig and sheep farms – results from a questionnaire Leena Sahlström ∗ , Terhi Virtanen, Jonna Kyyrö, Tapani Lyytikäinen Risk Assessment Research Unit, Finnish Food Safety Authority Evira, Mustialankatu 3, 00790 Helsinki, Finland
a r t i c l e
i n f o
Article history: Received 25 November 2013 Received in revised form 27 June 2014 Accepted 7 July 2014 Keywords: Biosecurity Questionnaire Cattle Pig Sheep GLM
a b s t r a c t Biosecurity is important in order to prevent disease transmission between animals on farms as well as from farm to farm. Personal biosecurity routines such as hand washing and the use of protective clothing and footwear are measures that should be used at all farms. Other measures are for example related to purchasing new animals to the farm. A questionnaire-based survey was undertaken to study the frequency of use of different biosecurity measures on cattle, pig and sheep farms in Finland. Information about which biosecurity measures are in use is needed for contingency planning of emerging diseases or when combating endemic diseases. Knowledge about the level of biosecurity of a farm is also needed in order to assess if and where improvement is needed. Information regarding biosecurity levels may benefit future animal disease risk assessments. A total of 2242 farmers responded to the questionnaire resulting in a response rate of 45%. The implementation frequencies of different biosecurity measures are reported. The results revealed differences between species: large pig farms had a better biosecurity level than small cattle farms. There were also differences between production types such as dairy farming versus beef cattle farming, but these were not as remarkable. Sheep farming in Finland is sparse and the large number of hobby farmers keeps the biosecurity level low on sheep farms. This might represent a risk for the entire sheep farming industry. The Finnish farmers were satisfied with their on-farm biosecurity. Eighty percent of the farmers report that they were satisfied even though the biosecurity level was not particularly high. The implementation of biosecurity measures could be further improved. Even though the disease situation in Finland is good today, one must be prepared for possible epidemics of threatening diseases. © 2014 Elsevier B.V. All rights reserved.
1. Introduction 1.1. Biosecurity The animal health status in Finland is very good regarding epizootic diseases. However, the situation can change very rapidly as we have seen in other countries
∗ Corresponding author. Tel.: +358 404 893375. E-mail address: leena.sahlstrom@evira.fi (L. Sahlström). http://dx.doi.org/10.1016/j.prevetmed.2014.07.004 0167-5877/© 2014 Elsevier B.V. All rights reserved.
such as in the foot-and-mouth-disease epidemic in Great Britain and bluetongue epidemics in many northern European countries. Some infectious agents are transmitted via direct contact between animals, whereas others can also be transmitted indirectly through contaminated equipment, vehicles, people and vector animals (Amass and Clark, 1999). FAO defines biosecurity as: “The implementation of measures that reduce the risk of the introduction and spread of disease agents” (FAO, 2010). The importance of biosecurity is underlined in the European Union health strategy for 2007–2013 “Prevention is better than
60
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
cure” (European Commission, 2007; Dekker, 2011). Biosecurity measures prevent both direct disease transmission between animals and indirect transmission between farms (Ellis-Iversen et al., 2011). Contacts such as transports and visits between farms and especially trade of animals facilitate disease spread. All precautions in animal trade, like quarantine, buying animals from a limited number of farms, etc. decrease the risk of disease transmission. Also biosecurity actions, including hygienic measures such as washing hands and using boots and coveralls (Amass et al., 2003, 2004), as well as transport logistics and “allin-all-out” systems, to mention a few, decrease the risk of spreading diseases (Madec, 2001). On-farm biosecurity measures are implemented differently depending on the farm. Biosecurity routines on pig farms have earlier been examined in different countries; e.g. Denmark (Boklund et al., 2003/2004; Boklund et al., 2004), Belgium (Ribbens et al., 2008), Chile (Pinto and Urcelay, 2003) and recently in Canada (Bottoms et al., 2012). In Sweden biosecurity measures have also been investigated on cattle and sheep farms (Nöremark et al., 2010). However, Finland differs from many other countries regarding the animal population which is sparsely distributed throughout the relatively large country. Even the areas with the highest farming densities in Finland have less than 0.5 farms/km2 including all cattle, pig sheep and goat farms (Tike, 2009). Information about the biosecurity level on farms is important for contingency planning of emerging diseases, when combating endemic diseases in a country, or to see if and where the biosecurity needs to be improved. In addition, information on biosecurity level may benefit models for animal disease spread as well as risk assessments. 1.2. Aim The aim of this survey was to describe the on-farm biosecurity routines used on cattle, pig and sheep farms in Finland and the possible differences in biosecurity depending on animal species, farm size and production type. 2. Material and methods 2.1. Administration of the questionnaire A written questionnaire was designed to collect information about the level of biosecurity and hygiene practices on Finnish cattle, pig and sheep farms. The questionnaire was sent in March 2011 to each of 3000 cattle and 1000 pig farmers. Another slightly modified questionnaire was sent in June 2011 to 866 sheep farmers in Finland. An invitation letter was enclosed, in which the farmers were informed that their response will be treated confidentially and the results reported so that no individual answers could be recognized. A reminder was sent by mail to the farmers who did not respond, 4 weeks after the original questionnaire was mailed. No compensation was paid for the response. The questionnaires were sent by mail but could also be answered online. The online link was provided in the mailed questionnaire and on the reminder card but also in the main agricultural newspapers in March 2011. A link to
the pig and cattle questionnaire was published in two agricultural newspapers and the websites of the Association for Animal Disease Prevention (ETT), the National Health Classification Registry (Sikava and Naseva) and the Finnish Food Safety Authority Evira. The link to the sheep questionnaire was published in June 2011 in two agricultural newspapers, an electronic newsletter (Saparo) and the website of the Finnish Food Safety Authority Evira. 2.2. Questionnaire design The 10-page questionnaire consisted of questions on general demography (age, sex, education and experience in animal husbandry), the farm (production type and type of buildings) and questions about several different on-farm biosecurity measures and hygienic precautions. There were 17 (14 for the sheep farmers) different biosecurity measures from which the farmers were requested to choose and mark the ones that were always in use at the farm. Biosecurity measures that were given as separate options were for example: the use of boots, or the use of protective clothing by the farmers and visitors, respectively, hand washing, use of a separate loading area, cleaning of stables and pest control. In addition, the farmers were also asked in eight separate questions about routines when purchasing animals as there are no animal markets in Finland; did they purchase animals from more than one other farm, did they enquire about the health status of the selling farm, the use of a quarantine, and the use of farm-specific transport vehicles for animals. One of the eight questions regarding trade of animals was if the farmer followed the guidelines of ETT. ETT (the Association for Animal Disease Prevention) is a farmers’ association in Finland which, among other things, prepares guidelines for animal trade. Another question in the questionnaire dealt with satisfaction. The respondents were asked to mark if they were satisfied, not satisfied or did not know if they were satisfied or not with the biosecurity on their farm. In both the cattle/pig and sheep questionnaires there were additional questions, the results of which are not presented in this paper. The questionnaire was pretested by an expert group of veterinarians working with biosecurity and it was edited according to their comments. The complete questionnaire (in Finnish or Swedish) is available upon request from the authors. 2.3. Selection of farmers There were 16,714 cattle farms, 2343 pig farms and 2576 sheep farms in Finland in 2009 (Finnish farm registry, Tike, 2009). The sample of 3000 cattle, 1000 pig and 866 sheep farmers represented 18%, 43% and 33% of the farms in Finland in 2009, respectively. The sample size was determined based on earlier experience of response rate (approximately 20%) among Finnish farmers and the aim was to get as big and representative sample as possible. Sampling was based on farm size which is defined here as the number of animals (cows and heifers >6 months, sows and finishers > 3 months) on the farm. The selection of large farms was motivated because there is a tendency towards larger farm size whereas smaller farms tend to discontinue
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
their production (Pyykkönen et al., 2010). The largest 10% of the farms were sampled more often than other farms. Among pig and sheep farmers, the largest 10% of all pig and sheep farms received the questionnaire, and the rest of the farms were randomly selected based on randomly generated numbers (SPSS software) from the Finnish farm registry 2009 (Tike, 2009). While there are so many more cattle farms in Finland, only 45% of the largest 10% was randomly chosen, and the rest was a random sample (randomly generated numbers, SPSS software) from the Finnish farm registry 2009 (Tike, 2009). 2.4. Statistical methods used Data were entered into Webropol 1.0 survey software and stored in a database using Microsoft Office Excel 2010 (Microsoft Co., Redmond, Washington, DC, USA). Statistical analyses were performed using SPSS Statistics 19 software (SPSS, Chicago, IL) and R 2.14.0 software (R core team, 2013). 2.5. Multivariable models 2.5.1. Implementation of biosecurity measures and practices The implementation frequencies of biosecurity measures and practices were estimated with multivariable models for a farm of typical size (Table 2) of each farm type in Finland (Tables 4 and 5). Farm types in these models were sow farm, farrow-to-finish farm, fattening pig farm, dairy farm, beef cattle farm, suckler cow farm and sheep farm. Analyses were performed using logistic regression (generalized linear models, GLM, family = binomial, link = logistic). There were 17 models for biosecurity measures in use at the farm (Table 4) and 8 models for biosecurity practices concerning purchase of live animals (Table 5). The equation for the models is as follows: Implementation of a biosecurity measure/practice = farm type + farm type × standardized farm size. Standardizing the farm size was performed as follows: the farm size for each farm was transformed with natural logarithm and then adjusted by the mean and standard deviation (z-score, SPSS) of the farm size of that production type in Finland. Thus, the standardized typical size of a farm type (calculated with the mean and the standard deviation of the ln transformed farm sizes for a farm type) was equal to 0. A negative value for farm size represents a smaller and a positive value a larger farm than the typical one for that farm type. The model did not have a common intercept, so every farm type got a separate intercept from the analysis representing the implementation frequency on a typical sized farm in that farm type. Farm type × farm size interactions were included into the model to allow separate slopes of farm size-related influence on implementation frequency of a measure within each production type. Thus, the statistical significance of the interaction term reflects a significant effect of the farm size on the implementation frequency within the farm type. The coefficients for the farm types were transformed into implementation probabilities of a typical sized
61
farm (frequencies) according to the following equation (McCullagh and Nelder, 1989): P(implementation of a measure) = exp(coefficient of a farm type)/[1 + exp(coefficient of a farm type)]. Confidence intervals for implementation frequencies were calculated by applying the standard error (SE) of the coefficient of a farm type: 95% CI for implementation of a measure = exp (coefficient ± 1.96 × SE)/[1 + exp(coefficient ± 1.96 × SE)]. 2.5.2. Association of the satisfaction of the farmer with the implementation of the biosecurity measures Generalized linear models (GLM, family = binomial, link = logistic) were also built for 17 biosecurity measures to find a possible association of satisfaction with the implementation of the measure. The model included farm type (pig or cattle), standardized farm size and whether or not the farmer was satisfied as explanatory variables. Also, the interaction term farm type × satisfaction was included when it was statistically significant. The equation for the models is as follows: Implementation of a biosecurity measure = type + standardized farm satisfaction + farm size + satisfaction × farm type. Statistical significance of parameters in GLM models was determined by dividing the parameter estimate by the standard error and comparing the result with the z-distribution (standard normal distribution) using R software. Statistical significance cut off level was defined as ***<0.001, **<0.01, *<0.05. 3. Results 3.1. Response rate The overall response rate was 42% (2036 replies); cattle farmers 41% (1236), pig farmers 42% (420) and sheep farmers 44% (380). The division into different production types is presented in Table 1. The reminder after 4 weeks increased the response rate substantially; 750 (34%) replies Table 1 Respondents from a biosecurity questionnaire to cattle, pig and sheep farmers in Finland were divided into three different production types among each species (except sheep farmers) as described in the table. The types are based on registration data from 2009 (Tike, 2009). The type correlates well with the production type given by the farmers in the questionnaire; the kappa value is 0.85. Production type
Number of respondents
Proportion of respondents (%)
Dairy Suckler cow Beef cattle Cattle (tot) Sow Farrow-to-finish Fattening pigs Pig (tot) Sheep Missing
893 140 181 1214 146 151 123 420 380 22a
74 12 15 100 35 36 30 100 na na
na = not applicable. a Known to be cattle farmers but missing information about production type.
62
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Table 2 Number of animals on a farm in different production types in 2009 in Finland (Tike, 2009). The minimum number of animals was one in all production types. A typical farm referred to in the text is the mean of ln transformed farm sizes of all Finnish farms within a farm type. Production type
Sow Farrow-to-finish Fattening pigs Dairy Beef cattle Suckler cow Sheep
Number of animals Mean
90th percentile
Max
107 259 448 43 32 51 12
502 955 1410 97 139 130 76
4227 6455 4586 574 1895 853 728
were received after the reminder. The majority of the respondents used the paper form, but 17% answered online. 3.2. Production type The cattle farm respondents represented either dairy farms, beef cattle herds or farms with suckler cows. The largest group of respondents was dairy farmers. Among pig farmers the respondents were divided into three groups representing sow farms, farrow-to-finish farms, and farms with only fattening pigs/finishers (Table 1). If a farmer had more than one species on the farm he was categorized according to the main production type. Sheep farmers were not divided into separate groups for the analysis of the questionnaire. However, the sheep farmers are quite a heterogeneous group, mainly consisting of hobby farmers with a small number of animals (mean number of ewes is 2) and only a few professional farmers that make sheep farming a living (only 18% earns more than 70% of their income from sheep farming (Table 3)).
production types. The farmers’ use of biosecurity measures is presented in Table 4. The use of the different biosecurity measures separated between species and production type is presented in Table 4. The results indicate that farm size usually has a significant positive effect on the use of different biosecurity measures as shown in Table 4. In a few exceptions farm size had a negative effect on the use of a biosecurity measure. This was true for “The farmer and his family and the visitors wash their hands after working in the stable” on sow farms, “control of rodents and birds in the animal stable or shelter at the feeding table” and “control of rodents and birds in the feed storage” on dairy farms. In the questionnaire to the sheep farmers, the questions about protective clothing, boots and hand washing were not separated between visitors and the owners because there are so few visitors on sheep farms. In the comparison between species, the cattle and pig owners’ habits (not including the visitors) are compared to the sheep farmers. Among the sheep farmers 5.7% reported (ticked the box) that they did not use any of the biosecurity measures mentioned. The corresponding answer among the cattle and pig farmers was 1.1% and 0.2%, respectively. 3.5. The farmers’ routines in trading animals Table 5 shows results only for those respondents that reported that they buy live animals (303 (77%) of the pig farmers, 614 (54%) of the cattle farmers and 182 (48%) of the sheep farmers). A significant positive effect of farm size is reported in connection to Table 5. In most cases there is a positive effect of farm size on the use of a biosecurity measure. The only negative effect of farm size is noted for “I tend to buy animals from as few farms as possible” on beef cattle farms (Table 5).
3.3. Description of the respondents
3.6. Satisfaction with biosecurity
A description of the respondents including average age, sex, education, farming experience, and main income is summarized in Table 3. Education is here defined as all education in the agricultural field after comprehensive school which includes nine grades, from the age of 7 to 16. The degree of education correlates with the farm size (p < 0.001) and the largest farms (10%) have a positive correlation with the main income (p < 0.001). The characteristics of the sheep farmers differ significantly from those of cattle and pig farmers. Less than 20% of the sheep farmers responded that they got their main income from sheep farming. Also the degree of education in the agricultural field is sparser among the sheep farmers than among the other types of farmers. The group of non-respondents was similar to respondents regarding the production types and the fraction belonging to the largest 10% in farm size.
Of the respondents, depending on production type, 80–88% were satisfied with their biosecurity, 3–6% were not satisfied and 8–13% did not know whether they were satisfied or not. Pig and sheep farmers were satisfied slightly more often than cattle farmers. When satisfaction was related to biosecurity measures in use at the farms, it was revealed that satisfied farmers significantly (p < 0.05) more often used several biosecurity measures than those who were not satisfied or did not know whether they were satisfied or not, i.e. out of 17 measures analysed 10 were significantly more often marked as in use by satisfied farmers (Table 6).
3.4. The farmers’ use of biosecurity measures Biosecurity was better implemented on pig farms than on cattle and sheep farms. There are big differences in some biosecurity measures between species but also between
4. Discussion Pig farmers implement more biosecurity measures on their farms than cattle and sheep farmers do. This was expected due to differences in management practices such as the frequent use of the all-in-all-out system on pig farms and the small number of pigs kept outdoors. Free range farming complicates the use of many traditionally used biosecurity measures for example a barrier between clean
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
63
Table 3 Presentation of age, gender, education, years of experience, and earning the main income from the farm among respondents in a Finnish questionnaire about biosecurity. Farmers of
Average age of farmers (min–max)
Women, % of respondents (95% CI)
Agricultural educationa , % of respondents (95% CI)
Farming experience, mean number of years (90% PI)
Main income >70%, % of respondents (95% CI)
Cattle Pig Sheep
47 (19–76) 47 (19–76) 52 (18–78)
28 (26–31) 19 (15–22) 47 (42–52)
68 (65–70) 75 (71–79) 42 (37–47)
22 (4–40) 22 (4–40) 16 (3–32)
75 (72–77) 59 (54–63) 18 (14–22)
CI = confidence interval; PI = percentile interval. a Here education includes additional or higher degree of education in the field of agriculture gained by the respondent after comprehensive school that includes 9 grades.
and dirty areas, rodent and wild bird control, and the control of access for visitors and pets. This is true also for cattle and sheep farms during the pasture season. There is large variation in farm size within all farm types, ranging from hobby farms with only a few animals to large professional farms with hundreds of dairy cows or thousands of pigs. A correction for farm size was done in order to show the biosecurity standard on a typically sized Finnish farm. The standard sized sheep farm in Finland is relatively small because of the large fraction of hobby farmers. The differences between sheep farmers compared to cattle and pig farmers regarding education, age and income also reflect the large group of hobby farmers. The hobby farmers are not necessarily as aware of disease risks as the professional farmers are. This might be a health risk for the whole sheep population. In case of an animal disease that also affects other species, such as foot and mouth disease, the disease could also spread to other production sectors. On the other hand, hobby farmers are likely less connected with professional farms. It is also evident from the results of the questionnaire that the size of the farm affects the use of a biosecurity measure on the farm (Tables 4–6). Large farms have more biosecurity measures in use than smaller farms, which is consistent with studies e.g. in Sweden (Nöremark et al., 2010) and Belgium (Ribbens et al., 2008). However, there are a few exceptions revealed in the survey; smaller dairy farms have a higher probability of using rodent control, and on smaller sow farms there is a higher probability that farmers and visitors wash their hands after working in the stable. These are important biosecurity measures and should not be neglected on larger farms. Due to their challenging realization, there are rather few published studies regarding the effect of biosecurity measures, such as the changing of boots and coveralls, on disease prevention. However, even though the effect of individual measures has not been studied in detail, it can be concluded that these cheap and easy means of protecting farm animals are warranted. On the other hand, there are more expensive or demanding types of biosecurity measures that are more frequently in use on larger farms or in some production types than others. For example a separate loading space for animals and a quarantine facility are more expensive or difficult to implement. If these are not planned when the stables are built it may be a too costly a task to implement them later. Bottoms et al. (2012) speculate that large farms might see the risk of a pathogen introduction as more disastrous than small farms or they have the financial potential
to establish even the more expensive biosecurity measures. In addition, the expenses of biosecurity per animal will be lower with increasing farm size (Niemi et al., 2013). There are differences between production types and their use of biosecurity measures as seen in Tables 4 and 5. However, the differences seem logical and are mostly easily explained based on different management practices. For example a separate loading space might not be needed in a finishing pig stable if the all-in-all-out system is applied. A leak proof container for dead animals might seem unnecessary for a dairy farm because of the small number of dead animals. To improve the overall biosecurity on a national level, one needs to increase the use of biosecurity measures on farms that are not as well managed in terms of biosecurity, especially on smaller cattle and sheep farms. Many biosecurity measures do not have to be expensive, e.g. the use of farms-specific protective clothing and boots. The most important factor is to convince the farmers to change their habits in order to improve biosecurity on their farms. This is not an easy task (Racicot et al., 2012). However, it is an important factor in enhancing the biosecurity in the country, motivating the farmers to improve biosecurity by implementing more biosecurity measures on their farms. Vermin control is quite well applied on most farm types as seen in this survey, which is in accordance with Ribbens et al. (2008). On the other hand, a hygiene barrier between the clean and dirty areas is missing on the majority of farms, and the use of protective clothing and boots could be improved on many farms. Hand washing is a documented measure to prevent spread of diseases, but there are still farmers who do not wash their hands after they finish their work. In our survey sow farms had the highest frequency of quarantines in use (57%; 44–69, 95% CI) (Table 5). This is approximately the same frequency as on large sow farms in Chile (Pinto and Urcelay, 2003), where the herd size ranged between 50 and 3000 sows; Md 420. Quarantine can be understood in different ways; however, the numbers seem to be quite high in Finland taking into account that the sow farms sampled in Chile were larger than in Finland (Table 2). Nöremark et al. (2010) reported that 74% of Swedish pig farmers and 23% of Swedish cattle farmers required visitors to use protective clothing when entering the stables. We on the other hand found that visitors on Finnish sow-, farrow-to finish- and fattening pig farms used protective clothing on 86, 85 and 75% of the farms, respectively. On Finnish cattle farms the corresponding frequencies were
64 Table 4 Results from multivariable models for implementation of biosecurity measures. Coefficients from the models are transformed into probabilities for implementation on typically sized pig, cattle and sheep farms in Finland based on a questionnaire study to farmers [% (95% CI)]. Statistically significant effect of farm size is marked ***<0.001, **<0.01 and *<0.05 accordingly. The farm size effect was generally positive, but negative in the following activities and types of production: “The farmer and his family and the visitors wash their hands after working in the stable” in sow farms, “control of rodents and birds in the animal stable or shelter at the feeding table” and “control of rodents and birds in the feed storage” in dairy farms. na = not applicable. Sow farm
Farrow-to-finish
Fattening pigs
Dairy farm
Beef cattle
Suckler cow
Sheep farm
The farmer and his family use protective clothing in the stables The farmer and his family use boots or protective shoes in the stables Visitors use protective clothing (coveralls) Visitors use boots The farmer and his family wash their hands after working in the stable Visitors wash their hands after the visit The use of a barrier that separates the clean area from the dirty area and that is not passed without changing protective clothing and shoes The use of a separate loading area Washing the loading area after use Outside the animal stables there is a leak-proof container for dead animals The animal stables are cleaned between each “batch” Doors are kept locked Animals are divided into compartments The traffic on the farm is organized so that biosecurity aspects are taken into account Control of rodents and birds Control of rodents and birds in the animal stable or shelter at the feeding table Control of rodents and birds in the feed storage
89 (83–94)*
82 (75–88)
85 (77–91)
62 (59–65)*
69 (60–76)
59 (50–68)*
24 (19–30)***
79 (71–85)
73 (64–79)
86 (78–91)
63 (60–66)***
59 (50–68)
66 (57–74)
36 (30–42)***
86 (78–91) 81 (73–87) 98 (93–99)**
85 (77–90)*** 75 (67–82)*** 91 (85–95)
75 (65–83)*** 73 (63–81)*** 85 (78–91)
51 (47–54)*** 68 (64–71)*** 88 (85–90)**
34 (25–43)*** 39 (30–49)*** 84 (77–89)
27 (19–36)* 50 (40–0)*** 81 (73–87)
na na 67 (61–72)***
89 (82–94)* 35 (27–44)*
75 (67–81) 27 (20–35)*
66 (56–74) 26 (19–36)
63 (59–66) 11 (9–13)***
43 (34–52) 8 (5–15)
42 (34–51) 9 (5–15)
na 3 (1–6)***
59 (50–67)*** 24 (18–33) 44 (35–53)***
74 (67–81) 36 (28–45) 54 (45–63)***
25 (17–36)*** 18 (11–27) 59 (49–69)***
3 (2–4)*** 2 (1–4) 1 (1–2)
2 (1–7) 1 (0–6) 0 (0–4)*
7 (3–13) 3 (1–8) 2 (0–6)
6 (4–11)*** na 4 (2–7)***
32 (24–41)
39 (31–48)
79 (71–86)
17 (15–20)
43 (34–52)**
26 (19–34)
42 (36–49)*
49 (40–58)** 51 (41–60)*** 46 (38–55)
47 (39–56)* 53 (45–62)*** 42 (34–51)*
56 (46–66)*** 47 (37–58)*** 29 (20–39)**
12 (10–14) 10 (8–12)*** 26 (23–28)***
15 (10–22) 15 (9–24)*** 11 (6–20)***
8 (4–14) 30 (22–39) 23 (16–3)***
15 (11–20)*** 24 (18–30)*** 11 (7–15)***
89 (83–94) 59 (50–68)
90 (83–94) 64 (54–72)
97 (91–99) 65 (55–73)
75 (72–78)** 54 (50–57)***
67 (58–75)*** 37 (29–46)
69 (60–76) 17 (11–25)
52 (46–58) 38 (32–44)***
56 (47–64)
60 (51–68)*
65 (55–73)
42 (38–45)*
34 (26–43)
30 (22–39)
41 (35–48)**
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Variable description
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
65
Table 5 Results from multivariable models for implementation of biosecurity practices concerning purchase of live animals. Coefficients from the models are transformed into probabilities for implementation on typically sized pig, cattle and sheep farms in Finland based on a questionnaire study to farmers [% (95% CI)]. The figures in the table refer to 303 (77%) pig farmers, 614 (54%) cattle farmers and 182 (48%) sheep farmers that returned the questionnaire and reported that they buy live animals. Statistically significant effect of farm size is marked ***<0.001, **<0.01 and *<0.05, accordingly. The farm size effect was generally positive, but negative in the following activity and type of production: “I tend to buy animals from as few farms as possible” in beef cattle farms. Variable description
Sow farm
Farrow-to-finish
Fattening pigs
Dairy farm
Beef cattle
Suckler cow
Sheep farm
I use a farm-specific transport vehicle when moving animals I use a quarantine, which is a building or compartment that keeps newly purchased animals separate from the other animals on the farm I check the disease status of the selling farm before purchasing I buy animals preferably from the same farms I follow the guidelines of ETT (Association for Animal Disease Prevention) when purchasing animals I buy animals via slaughterhouses that coordinate the trade between farms I tend to buy animals from as few farms as possible I demand a health certificate before purchase (Sheep farmers check the Maedi-Visna and scrapie status of the sheep before buying)
10 (4–22)*
27 (18–39)
13 (7–22)
37 (31–43)
46 (35–58)*
28 (17–41)*
91 (82–95)***
57 (44–69)
37 (27–49)
17 (27–49)
15 (11–21)**
36 (26–48)
51 (38–65)
10 (5–19)***
61 (48–72)*
67 (55–74)
37 (26–49)
90 (85–93)
41 (31–53)
85 (74–92)
81 (70–89)***
84 (73–91)
84 (74–91)
57 (46–68)
55 (49–62)
54 (42–65)
62 (48–75)
na
94 (86–98)
89 (79–95)
93 (85–97)
92 (88–95)**
86 (76–93)
88 (78–94)
na
78 (67–86)
84 (74–91)
92 (85–96)
21 (16–27)
80 (71–87)**
54 (40–68)*
na
95 (86–98)
93 (84–97)
86 (77–92)
92 (88–95)
69 (57–79)*
92 (81–97)
93 (83–97)***
33 (22–47)
44 (32–56)
16 (9–27)
71 (65–77)
14 (8–24)
82 (69–90)
70 (56–80)**
na = not applicable.
51, 34 and 27% on dairy, beef and suckler cow farms, respectively (Table 4). The farm size in Sweden in 2006 was slightly larger with an average cattle herd size of 64 animals compared to Finland (Table 2). The average pig herd size in Sweden with 116 sows and 495 piglets and fattening pigs was also bigger compared to Finland (Nöremark et al., 2010). In Sweden there is a separate loading area on 22% of the cattle farms and 86% of the pig farms (Nöremark et al., 2010). In Finland loading areas are very uncommon on cattle farms; 3, 2 and 7% have a loading area on dairy, beef and suckler cow farms, respectively. On pig farms in Finland loading areas are more common as 74% of the farrow to finish farms, 59% of the sow farms and 25% of the fattening pig farms report using one (Table 4). The Finnish farmers are generally satisfied with their biosecurity; more than 80% of the respondents reported that they were satisfied. This is in accordance with results from other studies, e.g. in Sweden 72% of the farmers were satisfied, even though they did not apply any or only few routines related to biosecurity (Nöremark et al., 2010). Spanish pig farmers also seem generally satisfied with the biosecurity applied on their farms with a self-scoring mean of 6.72 (min 0, max 10; SD 1.57) (Casal et al., 2007). The Spanish farm size in the study of Casal et al. (2007) ranged from 70 to 7000 sows (25, 50 and 75% quartiles were 203, 328 and 532 sows, respectively) and thus were much larger than the Finnish sow farms (Table 2). The Finnish farmers
that reported satisfaction with their biosecurity more often implemented biosecurity measures than the unsatisfied farmers. However, farmers were still quite satisfied even though there was improvement potential in the level of biosecurity as a whole. When farmers were satisfied with a lower biosecurity level it might reflect a lack of knowledge of biosecurity or an underestimation of the risks threatening their production. One might also assume that satisfied farmers are less likely to improve their biosecurity. The favourable animal disease situation in Finland might lull farmers into a false feeling of security. Nevertheless, it is important to be prepared for a rapid change in the disease situation. However, as long as the farmer does not experience any problems with animal diseases, he will most likely be satisfied and not feel the need to improve the biosecurity level (Alarcon et al., 2013). In other words, motivating farmers is the key to improving biosecurity on farms and in the country as a whole. One might suspect a bias among the respondents towards a higher response likelihood among farmers more interested in biosecurity. Furthermore, one can assume that farmers with larger farms would be more interested in biosecurity, because of the higher economic risk involved. However, a comparison between the respondents and non-respondents revealed no difference regarding farm size or production type. Farmers can also report that they use more biosecurity measures than they really do because they might experience that they are expected to
66
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
Table 6 Results from multivariable models for the association of satisfaction with implementation of biosecurity measures. Generalized linear models included farm type in question (pig or cattle), standardized farm size and whether the farmer was satisfied with his biosecurity measures in use. Estimates are given for basal level (intercepts) and predictors. Statistically significance is marked ***<0.001, **<0.01 and *<0.05, accordingly. Also, an estimate for the interaction term between farm type and satisfaction is included when it is statistically significant. Typical size means here the mean of ln transformed and standardized farm sizes of all Finnish farms within that farm type. Non-satisfied farmers include here also the ones that were unsure. Variable Farmer and his family use protective clothing in the stables Farmer and his family use boots or protective shoes Visitors use protective clothing Visitors use boots Farmer and his family wash their hands after working Visitors wash their hands after the visit Use of a barrier that separates the clean area from the dirty area and that is not passed without applying protective clothing and shoes Use of a separate loading area Washing the loading area after use Animal stables are cleaned between each “batch” Outside animal stables is a leak-proof container for dead animals Doors are kept locked Traffic is organized so that biosecurity aspects are taken into account Animals are divided in compartments Control of rodents and birds on the feeding table Control of rodents and birds in the feed storage General control of rodents and birds
Basal level (not satisfied, cattle farm, typical size)
Satisfied
Pig farm
Standardized farm size
Interaction term satisfied × pig farm
0.35*
0.21
0.65
0.20***
0.77*
0.50***
0.05
0.76***
0.20***
n.s.
1.65*** 0.58*** −0.60
0.72*** 0.67*** 0.23**
n.s. n.s. 1.07*
−0.65*** 0.38** 1.90***
0.60*** 0.21 0.02
0.07 −2.69***
0.34* 0.63**
0.78*** 1.23***
0.03 0.47***
n.s. n.s.
−3.83*** −4.55*** −1.75***
0.65** 0.90** 0.58***
3.45*** 2.73*** 1.16***
0.72*** 0.38*** 0.23***
n.s. n.s. n.s.
5.15***
0.89***
n.s. −0.79* n.s.
−4.83***
−0.26
−2.67*** −1.75***
0.67** 0.80***
2.87*** 0.52***
0.31*** 0.75***
−2.14*** −0.67***
0.29 0.74***
1.92*** 0.54***
0.93*** −0.24***
n.s. n.s.
−1.03***
0.69***
0.89***
−0.03
n.s.
0.28
1.35***
0.83***
0.33***
n.s.
n.s. = non significant and thus not included in the model.
do so. This results in a too positive picture of the on-farm biosecurity level in Finland. If one hypothesizes that the respondents would be more interested in biosecurity than non-respondents, and that there is a positive bias in their responses, one could say that the need for improvement of biosecurity is even bigger than shown in this study. However, we might still expect that the positive bias is somewhat the same in all production types and the differences shown between the production types would still be approximately the same. A questionnaire is proven to be a good tool for the collection of data (Ribbens et al., 2008). Advantages with a postal questionnaire are that it is cheap to administer and you easily reach a large group of people. The response rate, however, is not as high as it might be in a telephone interview. On the other hand, in reality a telephone interview approach limits the number of farmers that can be contacted. A weakness of the questionnaire at hand is that while the same questionnaire was used for both cattle and pig farmers for the sake of comparability, some of the farmers commented that not all questions were suitable for their particular farm type. This must be kept in mind e.g. when interpreting answers regarding different compartments in the stable or the loading space, which might mean completely different things, depending on the species of animal kept on the farm. The evaluation of the answers of a questionnaire may be challenging because even though the questions are made as clear as possible, the questions may
be comprehended differently depending on the respondent’s current situation. Finland differs from many European countries by its sparsely distributed animal population. There might also be large differences in on-farm-biosecurity between European countries. These differences should be explored further, considering that EU regulations on animal diseases are the same for all Member States and trade between European countries is common and occurs in large volumes. 5. Conclusion Further improvement of the biosecurity level is possible on farms in Finland. Pig farms generally have a higher level of biosecurity than cattle and especially sheep farms. The large proportion of hobby farmers among sheep farms and smaller farms in cattle and pig production, have a negative influence on the level of biosecurity. Counselling regarding potential risks should be emphasized in order to increase the level of biosecurity especially on smaller farms, so as to decrease the risks to production as a whole. Acknowledgements The authors would like to thank the Ministry of Agriculture and Forestry (Makera Dnr 2803/502/2008) for funding this project. The authors would also like to thank all the farmers that have contributed by responding to the
L. Sahlström et al. / Preventive Veterinary Medicine 117 (2014) 59–67
questionnaire and Kitty Schulman for critically reviewing the paper. References Alarcon, P., Wiedland, B., Mateus, A., Dewberry, C., 2013. Farmers’ Perceptions, Attitudes, Influences and Management of Information in the Decision-Making Process for Disease Control. Society for Veterinary Epidemiology and Preventive Medicine, Madrid, Spain, 20-22.3 2013. ISBN 978-0-948073-20-5. Amass, S.F., Mason, P.W., Pacheco, J.M., Miller, C.A., Ramirez, A., Clark, L.K., Ragland, D., Schneider, J.L., Kenyon, S.J., 2004. Procedures for preventing transmission of foot-and-mouth disease virus (O/TAW/97) by people. Vet. Microbiol. 103, 143–149. Amass, S.F., Pacheco, J.M., Mason, P.W., Schneider, J.L., Alvarez, R.M., Clark, L.K., Ragland, D., 2003. Procedures for preventing the transmission of foot-and-mouth disease virus to pigs and sheep by personnel in contact with infected pigs. Vet. Rec. 153, 137–140. Amass, S.F., Clark, L.K., 1999. Biosecurity considerations for pork production units. Swine Health Prod. 7, 217–228. Boklund, A., Alban, L., Mortensen, S., Houe, H., 2004. Biosecurity in 116 Danish fattening swineherds: descriptive results and factor analysis. Prev. Vet. Med. 66, 49–62. Boklund, A., Mortensen, S., Houe, H., 2003/2004. Biosecurity in 121 Danish sow herds. Acta. Vet. Scand. Suppl. 100, 5–14. Bottoms, K., Poljak, Z., Dewey, C., Deardon, R., Holtkamp, D., Friendship, R., 2012. Evaluation of external biosecurity practices on southern Ontario sow farms. Prev. Vet. Med. 109, 58–68. Casal, J., De Manuel, A., Mateu, E., Martin, M., 2007. Biosecurity measures on swine farms in Spain: perception by farmers and their relationship to current on-farm measures. Prev. Vet. Med. 82, 138–150. Dekker, A., 2011. Biosecurity and FMD transmission. Vet. Rec. 168, 126–127. Ellis-Iversen, J., Smith, R.P., Gibbens, J.C., Sharpe, C.E., Dominiguez, M., Cook, A.J.C., 2011. Risk factors for transmission of foot-and mouth disease during an outbreak in southern England in 2007. Vet. Rec. 128, http://dx.doi.org/10.1136/vr.c6364. European Commission, 2007. A New Animal Health Strategy for the European Union (2007–2013) Where Prevention Is Better Than Cure,
67
http://ec.europa.eu/food/animal/diseases/strategy/docs/animal health strategy en.pdf (accessed 12.08.13). Food and Agriculture Organization of the United Nations/World Organisation for Animal Health/World Bank, 2010. Good practices for biosecurity in the pig sector – issues and options in developing and transition countries. In: FAO Animal Production and Health Paper No. 169. FAO, Rome, http://www.fao.org/docrep/012/i1435e/ i1435e00.pdf (accessed 06.12.12). Madec, F., 2001. Biosecurity on Pig Units: A Major Issue for Herd Health Maintenance, http://www.cnpsa.embrapa.br/abravessc/pdf/Palestras2001/Francois Madec1.pdf (accessed 13.08.13). McCullagh, P., Nelder, J.A., 1989. Generalized Linear Models, second ed. Chapman & Hall, pp. 511. Niemi, J.K., Sahlström, L., Virtanen, T., Kyyrö, J., Lyytikäinen, T., Sinisalo, A., Lehtonen, H., 2013. The value of biosecurity and the role of economic incentives in promoting the adoption of biosecurity. In: Paper Prepared for Presentation at the V Workshop on Valuation Methods in Agro-food and Environmental Economics: Methodological and Empirical Challenges in Valuation Methods. 137th EAAE Seminar, Castelldefels, Spain, 27–28 June 2013. Nöremark, M., Frössling, J., Lewerin, S., 2010. Application of routine that contribute to on-farm biosecurity as reported by Swedish Livestock farmers. Transbound. Emerg. Dis. 57, 225–236. Pinto, C.J., Urcelay, V.S., 2003. Biosecurity practices on intensive pig production systems in Chile. Prev. Vet. Med. 59, 139–145. Pyykkönen, P., Lehtonen, H., Koivisto, A., 2010. Structural Change and Investments in Finnish Agriculture 2010–2020. PTT Working Papers 125, pp. 24, ISBN 978-952-224-061-3 (pdf), ISSN 1796-4784 (pdf) (In Finnish, English summary). Available from: http://www.ptt.fi/ dokumentit/tp125 1111100930.pdf (accessed 15.08.13). R Core Team, 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, http://www.R-project.org/ Racicot, M., Venne, D., Durivage, A., Vaillancourt, J.-P., 2012. Evaluation of strategies to enhance biosecurity compliance on poultry farms in Quebec: effect of audits and cameras. Prev. Vet. Med. 103, 208–218. Ribbens, S., Dewulf, J., Koenen, F., Mintiens, K., De Sadeleer, L., de Kruif, A., Maes, D., 2008. A survey on biosecurity and management practices in Belgian pig herds. Prev. Vet. Med. 83, 228–241. Tike Information Centre of the Ministry of Agriculture and Forestry, 2009. Finnish Farm Registry.