Typologies of Dairy Farms with Automatic Milking System in Northwest Spain and Farmers’ Satisfaction

Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=tjas20 Italian Journal of Animal Science ISSN: (Print) 1828-051X (Online) Journal homepage: https://www.tandfonline.com/loi/tjas20 Typologies of Dairy Farms with Automatic Milking System in Northwest Spain and Farmers’ Satisfaction Ángel Castro, José M. Pereira, Carlos Amiama & Javier Bueno To cite this article: Ángel Castro, José M. Pereira, Carlos Amiama & Javier Bueno (2015) Typologies of Dairy Farms with Automatic Milking System in Northwest Spain and Farmers’ Satisfaction, Italian Journal of Animal Science, 14:2, 3559, DOI: 10.4081/ijas.2015.3559 To link to this article: https://doi.org/10.4081/ijas.2015.3559 ©Copyright Á. Castro et al. Published online: 17 Feb 2016. Submit your article to this journal Article views: 773 View related articles View Crossmark data Citing articles: 2 View citing articles [Ital J Anim Sci vol.14:2015] [page 207] Typologies of dairy farms with automatic milking system in northwest Spain and farmers’ satisfaction Ángel Castro, José M. Pereira, Carlos Amiama, Javier Bueno Departamento de Ingeniería Agroforestal, Universidad de Santiago de Compostela, Spain Abstract The aim of this study was to determine the characteristics of the dairy farms that installed an automatic milking system (AMS). A survey of 38 dairy farms with AMS, in Galicia (Spain), collected information on quantitative and qualitative variables. Following elimination of redundant variables, categorical principal component analysis identified 4 factors accounting for 43.7% of the total variance. Using these factors, the farms studied were subjected to hierarchical cluster analysis which differentiated 4 types of farms: (A) farms with more leisure and quality of life where the AMS covered the expectations of farmers (29%); (B) farms that removed cows more often due to AMS and farmers with more stress (34%); (C) farms with little leisure and farmers with no successor (21%); (D) large farms with many fulltime employees (FTE) where the AMS had covered farmer’s expectations the least (11%). Generally the farms were based on a family structure with a high percentage of FTE. With the adoption of AMS these farms sought to increase milk production, save labour and have more flexibility. With 87% of farms with free cow traffic the activity that took the most of the farmer’s time was fetching cows for milking (1 h/day). Nearly 58% of farmers were completely satisfied with their AMS, although this value reached 91% in farms with herd sizes below the average which were better adapted to the use of one AMS. Introduction Over the last 20 years, the European dairy sector has suffered a profound adjustment in its production structure, having lost two thirds of the dairy farms in the European Union of 12 member states. The major adjustment corresponded to Spain, where only 17% of the farms that existed in 1988 are now operative (Sineiro et al., 2009) and these have tripled the number of dairy cows. In 2010, cow milk production in Spain amounted to 6357 million Tm, the greatest contribution being from Galicia (Spanish autonomous community located in the North West of the country). Over the last decades, both labour and land prices have increased, whereas the price of milk has decreased, making it necessary to increase productivity per hour and per hectare (de Koning, 2004). Furthermore, the elimination of the quotas planned for 2015 creates a new scenario, so that in the near future it will be the market that will regulate milk production, which will result in greater price instability (Sineiro et al., 2009). New developments and technologies in the different fields of dairy production have come into play, facilitating the work in the farms and increasing their productivity. One of them is the automatic milking system (AMS), a system that not only reduces the amount of work required, but it also makes it more flexible, which is the main reason for installing this system (Hogeveen et al., 2004). Dairy farming is a very demanding activity, the number of hours required being determined by the dynamics of the dairy farm, especially the milking routine. The milking is a process that has to be done at least twice a day, every day of the year, and it demands long hours and specialized labour. Dairy farming can be an activity that ties down to the point that no holidays or free weekends can be taken. The potential successors notice the hard, demanding labour that their parents carry out, a labour that sometimes does not get enough return for all the efforts involved, which discourages them from following their parents’ footsteps. The AMS could be part of the solution, as farms with an AMS use 29% less labour than farms with conventional milking systems (Bijl et al., 2007). For a dairy farm, the implementation of an AMS means an important innovation that provides advantages; however, it is not free of difficulties. Besides adaptation of the cows, the farmer will need to acquire higher level management skills for cows, business and technology to optimize the investment in automation (Reinemann, 2008). In some countries in Northern Europe, a large proportion of the farmers who introduced an AMS changed their grazing strategy after AMS adoption, using stable feeding instead (Mathijs, 2004). In spite of the higher cost of AMS technology compared with the conventional systems for harvesting milk, many dairy farmers are willing to pay a premium for the improved quality of life offered by AMS (Reinemann, 2008). The goal for the future is to reduce production costs, improving productivity per employee and labour conditions (Sineiro et al., 2009) to make it more attractive and similar to other sectors. Studies that look into socio-economic aspects of AMS and into motivations for adopting this system instead of a conventional milking parlour were carried out in North America (de Jong et al., 2003), where dairy farms have different dynamics of expansion, and in Northern Europe, where AMS was already developed and adopted many years ago (Hogeveen et al., 2004; Mathijs, 2004). Furthermore, some different socio-economic aspects were expected, see for example data about farms with AMS described by a Dutch accounting agency (Bijl et al., 2007) which is different from Galician data (total land use, pasture, milk quota, total labour, etc.). Considering the importance of the dairy sector in northwestern Spain and the recent introduction of AMS in this area, the objective was to determine the structural characteristics and operations of the dairy farms that installed an AMS, as well as to find out the main reasons farmers had for deciding to invest in an AMS, and knowing the implications that its adoption had. Corresponding author: Dr. Ángel Castro Ramos, Escuela Politécnica Superior, Departamento de Ingeniería Agroforestal, Universidad de Santiago de Compostela, R/Benigno Ledo, 27002 Lugo, Spain. Tel. +34.9828.23200 - Fax: +34.9828.23001. E-mail: [email protected] Key words: Automatic milking system; Farm structure; Farm typology; Farmer satisfaction. Acknoledgements: the authors are grateful for the financial support granted by the Autonomous Government of Galicia through the Directorate General for Research & Development (PGIDT/PGIDIT Project, Ref: 07MRU013291PR), as well as to the farmers who facilitated access to their data. Received for publication: 7 July 2014. Accepted for publication: 6 March 2015. This work is licensed under a Creative Commons Attribution NonCommercial 3.0 License (CC BYNC 3.0). ©Copyright Á. Castro et al., 2015 Licensee PAGEPress, Italy Italian Journal of Animal Science 2015; 14:3559 doi:10.4081/ijas.2015.3559 Italian Journal of Animal Science 2015; volume 14:3559 PAPER [page 208] [Ital J Anim Sci vol.14:2015] Materials and methods Area of study Continuing with the study initiated by our research group concerning the efficiency of AMS in dairy farms in northwestern Spain (Castro et al., 2012), data was collected by conducting an in-person survey that covered all the farm owners within the Autonomous Community of Galicia that had AMS installed in their farms. Galician dairy farms contributed with 37.9% of the total Spanish cow milk production in 2010. Galicia occupies an area of 29,343 km2, with an average population density of 93.6 inhabitants/km2. In May 2011, there were 328.821 dairy cows in Galicia, representing two fifths of the country’s total (MARM, 2011). In 2009, livestock unit per hectare was 2.89 and milk production per hectare was 17,253 kg milk per year (Barbeyto and López, 2012). Milk yield data published from 2010 from Galician herds showed an average milk yield per cow in 305 d of 8971 kg with 3.76% fat and 3.17% protein (AFRICOR, 2010). Data collection Firstly, a questionnaire was designed. For that purpose, other studies were consulted on the implementation of AMS in other regions, management practices, social aspects and motivations of farmers for installing AMS (Alibés et al., 2002; de Jong et al., 2003; Hogeveen et al., 2004; Mathijs, 2004; Kristensen and Noe, 2004). These studies colCastro et al. Table 1. Active variables considered divided into main topics. Topic Variable Characteristics of farms Size of farms and number of cows Milk quota, kg/year Contracted agricultural labours out to professionals Total full time employed in the farm Contracted full time employed in the farm Characteristics of farmers Age of the farmer interviewed, years Education level of farmer Contracted farm service for management of farm Existence of successor in the farm Who cooperates with labour on the farm Statements of farmers It is important to have new technologies at an early stage in the farm It is important to have some free time and take a holiday every year Reasons to install an AMS As the first option, why have you installed an AMS? Reasons to adopt an AMS AMS adopted Number of AMS installed Implications of AMS in barn Did you make any change in the barn to install the AMS? Position of the AMS in the barn Area of corridors per cow, m2 Stalls per cow Implications of AMS in strategies before and after adoption of the AMS Milking labour time before installation of the AMS, min Milking labour time after installation of the AMS, min Start milking labour in the morning before installation, a.m. Start milking labour in the morning after installation, a.m. Start milking labour in the afternoon before installation, p.m. Start milking labour in the afternoon after installation, p.m. Milking in parlour after installation of the AMS Implications of AMS on health My physical health has improved My mental health has improved My sleeping quality has improved Implications of AMS on leisure and quality of life I have more time for my family I have more time for hobbies The quality of life of our family has improved Adaptation of cows Selected traffic of cows Time until adaptation of the cows to AMS milking, when cows were milked voluntarily, days Any cows were removed due to AMS problems Adaptation of farmers The AMS covered your expectations Previous experience with computerized management systems Farmer were satisfied with data of AMS software Hours worked Checking of alarm lists and problems with AMS and computer, min/d Cows had to be fetched, min/d AMS maintenance; changed teat cup liners; cleaning the robot etc., min/d Checking information cow data, writing reports, min/d Other labours, min/d Other implications of AMS Contracted maintenance service Periodically review AMS, automatic milking system. [Ital J Anim Sci vol.14:2015] [page 209] lected interesting variables that we took into account when developing our questionnaire. This would also allow us to compare the results of the other regions with our own. The final questionnaire included both quantitative and qualitative variables belonging to the following topics: owner and owner family profile, farm structure, statements by farmers, reasons for installing an AMS, implications of AMS (on health, quality of life, strategies before and after the AMS introduction, barn design), adaptation of the cows, adaptation of the farmers, hours worked and AMS adopted. Firstly, a telephone survey was conducted to assess the willingness of farmers to participate in the study. If confirmed, an in-person interview was organized. During each visit, in addition to the interview, a layout of the barn with the AMS was drawn. Finally, all the owners who had an AMS in Galicia by September 2009 (n=38; 46 AMS) accepted the in-person interview. The census of farms with AMS was maintained until 2012. Data analysis The survey contained 78 variables to characterize the farms that invested in AMS. The information provided by the variables was both quantitative and qualitative. Sometimes a quantitative variable does not provide more information than a qualitative variable (Grande and Abascal, 2005). Also, in order to be able to analyze many variables simultaneously, they must all be of the same typology. Therefore, quantitative variables were transformed into 3 classes, using its quantile position with respect to the mean. This provided the frequencies of observations that were within the quantiles: <25%, between 25 and 75%, and >75% of the mean value for a chosen variable. It provided the frequencies of observations within each of the classes for each qualitative variable. Then a categorical principal component analysis (CATPCA) was performed to identify the factors that best accounted for the variations between the farms considered. This procedure quantifies categorical variables simultaneously, thereby reducing the dimensionality of the data (Meulman and Heiser, 2010) and allowing us to detect the factors that best characterized the farms. The criterion followed to select the number of dimensions consisted in taking a number that was small enough for the interpretations to make sense. The analysis was carried out using IBM SPSS Statistics 19 (SPSS, 2010). Information provided by the analytical variables and considered in the present study was redefined and coded into nominal, ordinal or binary variables that summarized the data from the survey. The 45 active variables finally considered in the CATPCA – divided into main topics – are defined in Table 1. Simultaneous analysis of all the variables is difficult to interpret so it is important to choose an active topic, i.e., those variables that form a homogeneous unit. In this study the attention was focused on determining the degree of satisfaction of farmers that had installed an AMS, taking into account the variables which in our view may affect this satisfaction. All other variables are illustrative; that is, those variables that are not used for obtaining the analytical results, but which are related to these and facilitate their interpretation. There are many variables that can be considered to contribute to each of the dimensions; but only the variables with factorial loadings greater than 0.55 were selected in each dimension. Once the CATPCA has been carried out with the active variables, the first factors found will be used for classifying the sample of farms in homogeneous classes. This is done by means of a hierarchical cluster analysis. A cluster analysis allowed grouping the farms that were similar. Other classification techniques were used by researchers to study the different typologies of dairy farms, but the discriminate analysis may be used to validate this methodology (Riveiro-Valiño et al., 2009). The characteristics of each group in the active topic as well as the rest of the illustrative variables collected in the survey were then statistically described. The differences between the groups of farms obtained were contrasted with regards to the quantitative variables by a oneway ANOVA with Student-Newman-Keuls mean comparison and, with regards to the qualitative variables, by a Pearson chi-square test of contingency tables. Results Automatic milking system farms typology Table 2 shows the results of the CATPCA. From the beginning we chose 4 dimensions: 43.7% of the total variance is accounted for by these 4 dimensions with eigenvalues greater than 3, a fairly high proportion for an analysis involving such a large number of variables. Cronbach’s alpha for all of the dimensions is greater than 0.7, which means that the test for these samples of farms has a good reliability. A fifth dimension would have a Cronbach’s alpha close to 0.7 which would decrease the reliability. The first dimension, or principal component, accounts for almost 14.8% of the total variance. With the 7 variables that are considered to contribute to this dimension, it can be referred to as the amount of labour with AMS. The second principal component accounts for 11.5% of the total variance with 4 contributing variables and can be referred to as the implications of installation of AMS. The 4 variables that can be regarded as associated with the third dimension helped in referring to it as Typologies of farms with AMS Table 2. Results of the categorical principal component analysis. Dimension (combination of variables grouped) 1 2 3 4 Total (milking labour time (position of the AMS in the barn; (contracted (successor; before AMS; size of farm; area of corridors per cow; agricultural I have more milk quota; stalls per cow; labours out; my mental possibilities contracted FTE; checking of alarm lists and health has improved; to spend time cows that had to be fetched; problems with AMS any cows were removed due with my my physical health has improved; and computer) to AMS problems; family; age of number of robots installed) education level of farmer) the farmer) Score dimensions 0.70; 0.68; 0.66; 0.66; 0.87; 0.85; 0.85; -0.56 0.80; -0.61; 0.61; -0.57 0.62; 0.58; 0.56 0.61; -0.58; 0.56 Cronbach’s alpha 0.869 0.824 0.782 0.739 0.971 Eigenvalues 6.66 5.15 4.25 3.61 19.67 Variance, % 14.81 11.45 9.44 8.01 43.70 AMS, automatic milking system; FTE, full time employee. [page 210] [Ital J Anim Sci vol.14:2015] farmer’s level of professionalism. Of the many variables that contribute to the fourth dimension, we only considered three and these variables suggest that this dimension may be appropriately referred to as future of the farm. After the CATPCA, an aggregative hierarchical cluster was carried out from the farms in the sample on the basis of the four emerging dimensions, which allowed us to classify them into four classes or groups of farms. Figure 1 shows the dendrogram obtained from the farms studied. We considered it appropriate to make a cut at reference level 10 so that four classes were obtained, because with fewer groups it is easier to interpret. The four types of farms consisted of 11, 13, 8 and 4 dairy farms, and they were classified as type A, B, C, and D, respectively. However, this cut produces a loss of 2 farms, which we believe necessary to be able to separate the farms into well differentiated groups. As a result, mean values of main qualitative and quantitative variables for each typology group of farms are showed in Tables 3, 4, 5, 6 and 7. The main active characteristics of all of the farms that allow more discrimination between groups (P<0.05) were: contracted FTE; I have more time for my family; any cows were removed due to AMS problems; milk quota; I have more time for hobbies; farm size; previous experience with computerized management systems; number of AMS installed; my mental health has improved; AMS maintenance, change teat cup liners, cleaning the robot. So with these variables, the 4 groups found, can be defined. Type A: farmers with more leisure and better quality of life where the automatic milking system covered their expectations This group corresponded to 11 AMS dairy farms (29%) with the smallest average herds (61.6 cows) (Table 3) and with 1 AMS/farm. These are dairy farms with an average owned area of 29.4 ha. As we can see the rented area did not show significant differences between groups but this fact may be interpreted as a difference if we consider the farm size in each group. So the area available per cow could indicate differences. On these farms grass crop (32.2 ha) is more common for silage than corn. All of the farms contracted out some agricultural labours to professionals, the preparation of the total mixed ration (TMR) being the labour least contracted out (9%). Around 46% of farms are general partnerships with no difference in total FTE compared with other groups but with 0.4 contracted FTE. The farmers in this group are young representing the overall average age. It is a group of farmers with the same percentage of primary, university and agricultural education levels (27%). A successor is not guaranteed in 45% of these farmers. All of these farmers have dairy farming as their main profession. These are the AMS users who agreed the most that their mental health had improved (64%). All of them considered that their physical health had improved and more than half of them said that their sleeping quality had not changed. All of these farmers agreed that they had more time for their family and more time for hobbies. Almost all the farmers (91%) had previous experience with computerized management systems and in the same percentage the AMS covered their expectations, and is almost significantly different to the other groups (P=0.054). Type B: farms that removed cows more often due to automatic milking system and farmers with more stress Thirteen farms (34%) formed this AMS group with a herd size of 71.5 cows. They are the group that contracted out the most agricultural TMR labour (62%), as well as the total agricultural labours. More than half the farmCastro et al. Figure 1. Dendrogram for 38 farms with automatic milking system showing the results of hierarchical clustering and the four classes (A, B, C and D). [Ital J Anim Sci vol.14:2015] [page 211] Typologies of farms with AMS Table 3. Structure and characteristics of farmers for each group of dairy farms with automatic milking system obtained in the cluster analysis. Variables Overall Type A Type B Type C Type D P (n=11) (n=13) (n=8) (n=4) Characteristics of farms Farm size, cows 75.4 61.6b 71.5b 72.9b 140.0a 0.005 Type of enterprise, % 0.246 Family farm 31.6 36.4 15.4 50.0 25.0 Cooperative 5.3 9.1 7.7 0.0 0.0 Agricultural transformation society 28.9 9.1 46.2 25.0 25.0 General partnership 31.6 45.5 30.8 25.0 25.0 Limited liability company 2.6 0.0 0.0 0.0 25.0 Own area, ha 34.9 29.4b 34.3b 27.0b 72.8a 0.021 Rent area, ha 11.0 12.8 10.3 13.0 7.3 0.729 Corn area, ha 25.3 17.9b 24.7b 22.9b 56.0a 0.011 Grass area, ha 25.4 32.2 22.0 24.1 23.0 0.431 Others forages area, ha 3.4 0.0b 4.2b 1.5b 16.3a 0.029 Forages bought, kg 62,579 45,272 50,846 66,125 172,500 0.309 Milk quota, kg 720,847 588,363b 651,815b 713,125b 1,481,250a 0.001 Milk quota is enough, % 55.3 36.4 53.8 37.5 75.0 0.514 Contracted agricultural labours out to professionals, % Nothing 2.6 0.0 0.0 0.0 25.0 0.042 Harvesting silage 94.7 100 92.3 100 75.0 0.250 Tilling the land 55.3 72.2 46.2 37.5 50.0 0.432 Slurry 21.1 18.2 38.5 12.5 0.0 0.300 TMR 36.8 9.1 61.5 50.0 0.0 0.018 Others 10.5 0.0 23.1 12.5 0.0 0.287 Total FTE, n 2.8 2.7 2.9 2.5 3.4 0.441 Contracted FTE, n 0.6 0.4b 0.4b 0.4b 2.8a 0.000 Tractors per farm, n 2.5 2.6 2.4 2.6 2.8 0.853 Other machinery, n 5.6 5.4 4.9 6.9 6.8 0.208 Characteristics of farmers Sex farmer owner interviewed, % 0.298 Man 89.5 100 84.6 75.0 100 Woman 10.5 0.0 15.4 25.0 0.0 Age of farmers, years 38.9 36.5ab 33.8ab 46.1b 28.8b 0.044 Education level, % 0.028 Primary school 36.8 27.3 23.1 75.0 25.0 Secondary school 13.2 18.2 15.4 12.5 0.0 Agricultural education 31.6 27.3 53.8 12.5 0.0 University 18.4 27.3 7.7 0.0 75.0 Contracted farm services, % Reproduction 81.6 81.8 76.9 87.5 100 0.728 Milk quality 60.5 72.7 53.8 50.0 75.0 0.647 Feeding 92.1 90.9 92.3 100 100 0.782 Others 18.4 9.1 23.1 12.5 25.0 0.768 Successor, % 44.7 54.5 61.5 0.0 50.0 0.038 Labour in co-operation in the farm, % 0.201 Wife or husband 18.4 27.3 7.7 37.5 0.0 Single person 5.3 0.0 0.0 12.5 0.0 Person employed 39.5 36.4 30.8 37.5 100 With child, laws or parents 28.9 27.3 46.2 12.5 0.0 Partner 7.9 9.1 15.4 0.0 0.0 Dairy farming is the main profession, % 97.4 100 100 100 75.0 0.042 Nature of other enterprises, % Products made of milk 5.3 0.0 0.0 0.0 50.0 0.001 Vegetables, forages, grains and fruit 7.9 9.1 0.0 12.5 25.0 0.419 Other livestock 10.5 9.1 15.4 12.5 0.0 0.849 Contract work for others farmers 5.3 0.0 15.4 0.0 0.0 0.290 Veterinary and advisory services 5.3 9.1 0.0 0.0 25.0 0.223 Other enterprises 13.2 9.1 15.4 12.5 25.0 0.882 Statements of farmers It is important to have new technologies at an early stage in the farm, % 76.3 63.6 84.6 87.5 50.0 0.334 It is important to have some free time and to go on holiday every year, % 94.7 100 100 100 75.0 0.042 It is important what other farmers think of me, % 10.5 18.2 0.0 25.0 0.0 0.236 TMR, total mixed ration; FTE, full time employee. a,bDifferent letters in the same row denote significant (P≤0.05) differences among means. [page 212] [Ital J Anim Sci vol.14:2015] ers in this group (54%) have agricultural training and it was this group that resorted the least to contracting an economic management service for their farms before and after of the installation (8%). 77% of them had to remove some cows due to problems associated with the AMS. This can influence the fact that these farmers are who least agreed that their mental health had improved (8%). Also these farms are a minority regarding the fact that their sleeping quality had not changed (15%). However, they are the group whose succession is the most guaranteed (62%). All of them believe that the AMS is the future and a challenge for their farms and for this reason they adopted this machine. They were the farms that spent the most time on milking with AMS, including the time spent on fetching the cows to be milked (72 min/day). Type C: farmers with little leisure and without successor This group consisted of eight farms (21%) with a herd size similar to group type B (72.9 cows). Half of them are family farms. Only 13% of these farmers have more time for their families and for hobbies. They are the group of farmers with the lowest level of education with 75% of them having studied primary education. None of these farmers has a successor and they are also the group with the highest average age (46.1 years). They felt that they did not have more time for their family and hobbies (87%). Also, 13% of them had to remove cows with problems when they installed the AMS. Furthermore, the AMS covered the expectations of about 50% of the farmers. Only a quarter of the farmers had previous experience with computerized management systems. They were the farmers that spent the least time on AMS maintenance, e.g., changing teat cup liners or cleaning the milking robot (7.5 min/day). Moreover, the time spent on fetching cows was about 75 min per day. Type D: large farms with many contracted full-time employees and where the automatic milking system covered the expectations of the farmers the least This group includes 11% of farms and is composed of large farms with a herd size of 140 cows. This size can influence other characteristics such as milk quota (1,481,250 kg), number of AMS installed (2), area of corn crop (56 ha). They are farms that had more contracted FTE (2.8). In 25% of these dairy farms, dairy farming is not the main profession. Also 50% of them make other dairy products. These are farms that contracted out the least amount of agricultural services, for example none of these farms contracted out the preparation of the TMR. They are the group of farmers with the highest level of education (75% of them have a university education). Despite being large dairy farms, only 50% of these farms are sure to have a successor, maybe because they are the youngest farmers (28.8 years), them being the actual successor. In contrast to other groups (P<0.05) 25% of these farms do not believe that it is important to have some free time and to go on holiday each year and also, the AMS only covered the expectations of a quarter of them. Once categorized by typology, the farms being studied were described based on different sections. Castro et al. Table 4. Reasons to install an automatic milking system for each group of dairy farms obtained in the cluster analysis. Variables Overall Type A Type B Type C Type D P (n=11) (n=13) (n=8) (n=4) Reasons for installing an AMS Already knew some farmer with AMS, or personal contact with an AMS farmer, % 57.9 63.6 53.8 50.0 50.0 0.928 Farmer considered buying a milking parlour before by an AMS, % 36.8 54.5 38.5 12.5 50.0 0.297 As the first option, why have you installed an AMS?, % 0.734 It’s a saver labour 2.6 0.0 0.0 12.5 0.0 To expand the farm 2.6 9.1 0.0 0.0 0.0 To increase milk production 57.9 63.6 46.2 62.5 50.0 To reduce production costs 5.3 0.0 7.7 12.5 0.0 For the future 2.6 0.0 7.7 0.0 0.0 Other 28.9 27.3 38.5 12.5 50.0 Reason to adopt an AMS, % Labour reduction 81.6 63.6 92.3 87.5 100 0.193 Labour flexibility 92.1 90.9 92.3 100 75.0 0.533 Get rid of hired labour 28.9 27.3 15.4 12.5 75.0 0.083 Improving technical parameters 65.8 54.5 84.6 50.0 75.0 0.291 It’s a future challenge 78.9 81.8 100 50.0 75.0 0.046 Other activities 78.9 72.7 92.3 75.0 75.0 0.612 AMS adopted Number of robots installed 1.3 1b 1.3b 1.3b 2.1a 0.010 Why have you installed this brad of AMS?, % Finance 5.3 0.0 7.7 12.5 0.0 0.630 Operation washing, techniques arm 55.3 63.6 53.8 62.5 25.0 0.578 Good and bad publicity of other brands 7.9 18.2 0.0 0.0 25.0 0.190 After sales service technical assistance 31.6 27.3 38.5 25.0 50.0 0.781 Considered as the best brand 36.8 27.3 30.8 50.0 50.0 0.674 Other 15.8 9.1 23.1 12.5 25.0 0.768 AMS, automatic milking system. a,bDifferent letters in the same row denote significant (P≤0.05) differences among means. [Ital J Anim Sci vol.14:2015] [page 213] Structure of dairy farms with automatic milking system: characteristics of farmers The average herd size for these farms was 75.4 cows (Table 3). Today the farms are enterprises and, as such, they have different mercantile structures. Basically, they can be categorized as family farms (32%), general partnerships (32%) and agrarian transformation societies (29%). The farm area amounts to 45.9 ha, 11.0 of which is rented land. Most farms grow corn for silage (25.3 ha/farm) as well as grass (25.4 ha/farm) or other winter forage. The forage cultivated was not sufficient in the farms as the forage bought exceeded 62 tonnes per farm and year. A large proportion of the farms with AMS had a milk quota of between 500,000 and 700,000 kg, and the group with a quota of more than 1,000,000 kg is also an important one. Less than half of the farmers consider that their milk quota is not sufficient. They contract many labours out to professionals, such as the harvesting of silage Typologies of farms with AMS Table 5. Implications of automatic milking system for each group of dairy farms obtained in the cluster analysis. Variables Overall Type A Type B Type C Type D P (n=11) (n=13) (n=8) (n=4) Implications of AMS in barn Did you make any change in the barn to install the AMS?, % 0.400 New barn 15.8 0.0 0.0 12.5 0.0 Nothing 2.6 72.7 76.9 37.5 75.0 A lot 13.2 9.1 7.7 37.5 0.0 A few 68.4 18.2 15.4 12.5 25.0 The internal distribution of the barn was changed 5.3 0.0 7.7 12.5 0.0 0.630 Position of the AMS in the barn°, % 0.290 POU 44.7 63.6 46.2 37.5 25.0 PIC 10.5 0.0 15.4 12.5 25.0 POC 2.6 0.0 7.7 0.0 0.0 PIU 10.5 9.1 7.7 25.0 0.0 LIU 2.6 0.0 7.7 0.0 0.0 LOC 5.3 27.3 7.7 0.0 25.0 LOU 15.8 0.0 7.7 25.0 0.0 Unknown 7.9 0.0 0.0 0.0 25.0 Area of corridors, m2/cow 5.4 5.4 5.5 5.6 3.8 0.360 Stalls per cow 1.1 1.1 1.1 1.1 0.8 0.281 Implications of AMS in strategies before and after adoption of the AMS Grazing before, % 18.4 27.3 15.4 12.5 25.0 0.827 Grazing after, % 0.0 0.0 0.0 0.0 0.0 TMR before, % 92.1 90.9 92.3 87.5 100 0.905 TMR after, % 97.4 100 92.3 100 100 0.611 Free barn before, % 94.7 90.9 100 87.5 100 0.573 Free barn after, % 100.0 100 100 100 100 Milking labour time before, h 4.0 3.8 3.9 4.1 5.3 0.110 Milking labour time after, h 2.1 1.6 2.7 1.8 2.8 0.134 Start milking labour on the morning before, a.m. 7.6 8.1 7.3 7.5 7.3 0.429 Start milking labour on the morning after, a.m. 8.5 8.5 8.4 8.4 8.1 0.901 Start milking labour on the afternoon before, p.m. 21.5 21.9 21.3 21.5 21.9 0.493 Start milking labour on the afternoon after, p.m. 20.8 21.0 20.4 20.8 21.3 0.238 Size herd before 71.2 58.6b 68.4b 68.5b 127.5a 0.005 Economic management before, % 36.8 45.5 7.7 50.0 75.0 0.041 Economic management after, % 36.8 36.4 15.4 50.0 75.0 0.126 Milking parlour before, % 92.1 90.9 100 87.5 100 0.573 Milking parlour after, % 26.3 27.3 23.1 25.0 50.0 0.763 Implications of AMS on health My physical health has improved, % 81.6 100 84.6 62.5 50.0 0.077 My mental health has improved, % 34.2 63.6 7.7 12.5 50.0 0.013 My sleeping quality has improved, % 39.5 63.6 15.4 25.0 50.0 0.079 Implications of AMS on leisure and quality of life I have more time for my family, % 68.4 100 84.6 12.5 25.0 0.000 I have more time for hobbies, % 63.2 100 69.2 12.5 25.0 0.001 The quality of life of our family has improved, % 71.1 81.8 76.9 62.5 25.0 0.169 AMS, automatic milking system. °Locations of AMS in the barn: P vs. L, orientation of AMS longitudinal axis perpendicular (P) or longitudinal (L) to the longitudinal axis of the barn or hallways; I vs. O, farmer access to the AMS located inside (I) or outside (O) of the cubicle area; C vs. U, centered (C) position, with barn surface area almost equally available to both sides, or uneven (U) position, with less or no available barn surface area to one side. a,bDifferent letters in the same row denote significant (P≤0.05) differences among means. [page 214] [Ital J Anim Sci vol.14:2015] (95%), tilling the land (55%) or preparing TMR on 37% of farms. Work on the farm was carried out by 2.8 full time employees (FTE). Total FTE can be divided into family members and contracted labourers. In our study the contracted FTE per farm was 0.6. Only 11% of the interviewed farmers, whom we considered to be the farm managers, were women. The average age of these farmers is 38.9 years. Most farmers (37%) had primary education, agricultural specific training (AT) was completed by 32%, and a group of 18% had university degrees. A dairy farm usually contracts out various important services for the correct performance of the farm, carried out by professionals such as agricultural engineers or veterinarians, feeding services (92% of farms), reproduction services (82%) or such as milk quality in 61% of farms. Most of the farmers performed the normal farm labours helped by employed personnel (40%), by parents, in-laws or children (29%), or by their wives or husbands (18%). The succession is certain in less than half of the farmers (45%). For almost all the farmers (97%) dairy farming is their main activity, however, 37% of the farmers had other businesses of a very diverse nature. They seek to diversify with other productions in order to support the dairy farming activity. The main activity was producing other livestock (10.5%). The farmers were asked about statements on new technologies, free time and what other farmers think of them, and we wanted to know their opinions about these. About 76% of farmers agreed with the statement it is important to have new technologies at an early stage on the farm. Almost all of them (95%) agreed with the statement it is important to have some free time and go away on holiday every year. However, 11% of the interviewed farmers said they were sensitive to what others thought of them. Reasons for installing an automatic milking system In general, 58% of the farmers already knew another farmer with an AMS, or had personal contact with one before installing the AMS (Table 4). Of the farmers that bought an AMS, just over one third (37%) had considered buying a new milking parlor. The farmers were asked about their main motivation for investing in an AMS. They had to choose the answer that they considered to be the most important from a number of open answers. More than half (58%) of these farmers sought, by installing the AMS, to increase milk production. In order to know in more detail and more specifically their reasons for adopting an AMS, the farmers had to indicate their particular reasons from a closed list of options. A farmer could choose one or several answers, the result being the percentage of farmers that chose each option. The listed reasons were: labour reduction, labour flexibility, dismissing labour, improving technical parameters, facing the future, challenge, and other activities. Most farmers (92%) chose labour flexibility and Castro et al. Table 6. Adaptation of farmers and cows for each group of dairy farms with automatic milking system obtained in the cluster analysis. Variables Overall Type A Type B Type C Type D P (n=11) (n=13) (n=8) (n=4) Adaptation of cows Selected traffic, % 0.476 Free 86.8 90.9 69.2 100 100 Forced 10.5 9.1 23.1 0.0 0.0 Guided 2.6 0.0 7.7 0.0 0.0 Time until adaptation of the cows to AMS milking, when cows were milked voluntarily, days 188.4 183.3 177.0 251.6 188.8 0.854 Any cows were removed due to AMS problems, % 31.6 0.0 76.9 12.5 0.0 0.000 Adaptation of farmers The AMS covered farmer’s expectations, % 57.9 90.9 46.2 50.0 25.0 0.054 Previous experience with computerized management systems, % 47.4 90.9 30.8 25.0 25.0 0.007 It was easy to understand the AMS software, % 89.5 90.9 92.3 75.0 100 0.520 Farmer was satisfied with data of AMS software, % 73.7 72.7 92.3 50.0 50.0 0.135 AMS, automatic milking system. Table 7. Labours for each group of dairy farms with automatic milking system obtained in the cluster analysis. Variables Overall Type A Type B Type C Type D P (n=11) (n=13) (n=8) (n=4) Hours worked in AMS Checking of alarm lists and problems with AMS and computer, min/d 10.5 4.4 8.4 16.3 8.0 0.275 Cows had to be fetched, min/day 69.1 45.0 72.3 75.0 127.5 0.064 AMS maintenance, changed teat cup liners, cleaning the robot, etc., min/d 15.8 10.0b 18.9b 7.5b 39.3a 0.012 Checking results of information, data cows, to make reports, etc., min/d 25.7 28.6 31.9 15.6 21.3 0.387 Other labours, min/d 1.2 0.0 2.7 0.0 0.0 0.518 Other implications of AMS Changes in genetic selection, % 0.507 None 78.9 72.7 84.6 87.5 75.0 Udders 15.8 9.1 15.4 12.5 25.0 Milking speed 5.3 18.2 0.0 0.0 0.0 Contracted secure, % 86.8 100 69.2 87.5 100 0.137 Contracted maintenance service, % 26.3 9.1 30.8 25.0 25.0 0.637 Periodically review, % 65.8 45.5 69.2 75.0 100 0.211 AMS, automatic milking system. a,bDifferent letters in the same row denote significant (P≤0.05) differences among means.