The objectives of this current study were to characterize the types of farms investigated by DAFF for animal welfare incidents, using a case study approach and secondly to use these data to enhance EWS. The EWS is a partnership between interested parties who wish to identify ways in which farm animal welfare can be further improved and to identify and resolve animal welfare problems before they become overpowering. It is hoped that by better understanding these case farms we will develop a better understanding as to why, despite the EWS already being in existence, welfare problems still occur on some Irish farms. By studying KPIs we hope to be able to ascertain ways in which to enhance the EWS to make it more effective. This study will form part of a broader study which will include comparisons with control herds and a study of the human factors that influence farm animal welfare incidents.
The welfare of farm animals is well provided for under Irish legislation and on-farm welfare incidents are not common. Traditionally, in Ireland, there has been a perception that welfare problems arise more commonly on farms run by older bachelors living in rural areas, where there is a lack of family support and poor access to health and social facilities. However, in this study the herd keepers ranged in age from young to old (31 to 84 years) and in sex (14 males and 5 females). It is likely therefore that where welfare problems arise, that the underlying cause is not a singular trigger, but may possibly relate to sociological, health, psychological, economic and other circumstances. In five of the case herds the responding SVI/VI believed that problems started to occur after the death of a parent. At the time of the incident two herd keepers were suffering from ill health, one of whom also had reported problems related to alcohol, and another who was reported to be suffering from depression (Table 1). The SVIs/VIs dealing with each welfare incident reported problems on the case farms that included high mortalities, poor management skills, registration issues, carcases unburied on the farm and a prior history of welfare problems. This tallies with the earlier unpublished DAFF study where on the 494 farms investigated 1552 dead animals were found, 78 animals were in extremis and were euthanased on-farm and 17% had issues with tagging and registration (Pat Flanagan: An investigation into On-Farm Welfare Incidents, 2007) and with a study by Collins et al  on equine welfare on an Irish farm where abandoned carcases were found, injuries were untreated and there was a history of prior welfare problems.
Our findings indicate that the late registration of calves; the use of on-farm burial as a method of carcase disposal; an increasing number of moves to knackeries over time and records of animals moved to 'herd unknown' were notable on the case farms and warrant further investigation. In contrast, no pattern was established between changes in herd size and the number of calves registered per cow per year between the case study herds (Tables 2 and 3).
Under Regulation EC 1760/2000, all calves must be registered within 27 days of birth. A DAFF communication (DAFF, unpublished data) on all calf registrations in Ireland in 2010 (including these case herds) showed that 2.22% of all 116,815 Irish herd keepers registered calves in excess of the 27 day limit. This figure is similar to the 2.26% of Irish farmers who despite being entitled to receive payments of greater than €100 from the Single Farm Payment Scheme did not apply for these payments (DAFF, unpublished data), possibly indicating problems with completing the paperwork involved. In this study the majority of those herds with a defined animal welfare incident date registered calves late in the incident year and in the years leading up to the incident (Figure 1). All of the long-term herds have registered calves late at some time between 2001 and 2009 (Figure 2). In recent years there has been an increase in the amount of paperwork that herd keepers are required to complete. Studies by Lobley et al. , Raine , Simkin et al. , Booth et al. , McGregor et al.  and Deary et al.  have identified problems with paperwork, legislation and finances as a cause of stress in farmers. Stress can affect decision making and the ability to cope, and may in the current context manifest as failure to manage the herd appropriately. The increased administrative burden may be greater for those farmers with jobs off-farm or with no formal training in agriculture. We do not know if the herd keepers in this study had off farm employment, what level of formal education they received, what their economic status was or what family and community supports were in place.
On-farm burial is prohibited in Ireland, except by licence under EU regulation 1774/2002. In the case study herds with defined animal welfare incident dates, seven herds (58.3%) buried animals on farm during the study period (Figure 3). Only one of the six herds, where the incident was determined to be long-term, did not bury animals on-farm after 2002 (Figure 4). In April 2009, DAFF ceased to make financial contributions to the Fallen Animals Scheme, except for animals aged 48 months or older. This scheme helped with the cost of disposing of animals through knackeries. As the most recent welfare incident date from the case herds was April 2009, it is unlikely that the abolition of the subsidy was a reason for the on-farm burials recorded in this study. On-farm burials may indicate that herd keeper is trying to conceal a welfare problem or that he/she is ignorant of, or showing a disregard for, the legal requirements of animal disposal. A further study will compare on-farm burial in these herds with control herds.
A knackery is a plant in which unprocessed Category 2 material (material unfit for human consumption) is handled and/or temporarily stored for the purpose of further transportation to its final destination. Between 2002 and 2009 the total number of cattle on-farm deaths in Ireland ranged from 192,437 to 234,088, which equates to 9.1% to 10.9% of the total number of cattle disposals, with a peak in 2008 at 12.1% and a low in 2003 of 8.6% (DAFF, unpublished data). For the 12 case herds with a defined animal welfare incident date, the number of animals disposed of at knackeries in relation to all exits from the farm increased annually during the study period, from a median of 5.1% three years prior to the incident, 14.6% two years prior to the incident, 16.8% in the year prior to the incident and 21.8% in the year that the incident occurred (Figure 5). The high median in the year the incident occurred is likely to be due in part to DAFF intervention in welfare cases where burial on farm was not allowed as a method of disposal of carcases already present on farm and to the euthanasia of animals too ill to move off the farm. The increase in the number of on-farm deaths of animals in the study in the two years preceding the incident year suggests that there was an increase in the numbers of animals that were not fit to enter the human food chain that either died on-farm or were euthanased on farm. This is likely due to disease, poor condition and unfitness to travel. All of the long-term herds recorded moves to knackeries in excess of the national average at various stages between 2001 and 2009 (Figure 6).
Another of our findings that was notable was moves to 'herd unknown' location. A 'herd unknown' location is used to record situations where the final end of life point of the animal cannot be determined with absolute certainty. In 2010, there were several cases of cattle rustling reported in Ireland . It is believed that some animals that die on-farm from welfare and disease issues are buried on-farm and subsequently reported as stolen or lost by the herd keeper. Cattle herds are subject to an annual Tuberculin test in Ireland. At the time of test, herd profile data including the location and registration details of the animals due to be presented at the TB test in each herd, as recorded on AIM, can be downloaded by the testing veterinarian from AHCS. Any discrepancies found are flagged on the system and subjected to intensive systematic investigations by DAFF. Whilst most of the discrepancies are resolved through this process, if the discrepancy cannot be resolved with certainty, the tag number is listed in 'herd unknown' location. In this study, the location of various animals in seven of the 12 herds where the incident date was known was unable to be resolved (Figure 7). Only one of the long-term herds did not record any moves to this location in the years 2001 to 2009 (Figure 8). The high percentage of the case study herds with moves to 'unknown location' during the three years leading up to a welfare incident indicates that moves to this location are of concern, however it is likely that these animals were assigned retrospectively to 'herd unknown' location in the years preceding the incident after an investigation was completed. Attention should be paid to discrepancies that cannot be resolved within a timely manner.
In 2009 (the year that this study commenced), there were 117,287 herds in Ireland. The average age of farmers was 55.8 years . The most current information on average farm size is for 2007, where it was 32.3 hectares . We were surprised at the diversity of the herd keepers in this study with respect to age and sex, and with the range in sizes of the farms (Table 1). Some factors such as age tallied with the national average, but others, e.g. average farm size, did not.
Socio-economic factors are likely to play a key role in animal welfare incidents e.g. family support; changes in personal circumstances etc., as welfare problems are not always due to intentional neglect or cruelty by herd keepers. A study by Sanne et al.  identified that male agricultural workers had the highest level of depression of all occupational groups and that the level of anxiety in male farmers was significantly higher than the average level among all working male participants. Due to long working hours and distance from mental health care facilities many herd keepers experiencing mental health issues may not get help. During the Celtic Tiger years, off-farm employment of either the farmer and/or spouse rose year on year from 45% of farms in 2001 to a peak of 58% in 2006 and 2007 . This should be of concern given the extent to which off-farm income supports the viability of many farms, leading to many herd keepers not having sufficient time to devote to their livestock. This can give rise to welfare problems.
There are limitations with this study, including the small number of case study herds. The data for KPIs were all taken from national databases, which rely on herd keepers providing the correct information. There is a potential for bias in the case studies chosen as they were nominated by SVIs and not selected randomly. In addition, this study did not evaluate other factors such rural isolation, availability of family and community support, accessibility of health care and off-farm employment, which may also be important key indicators. This will be addressed in further studies, as will comparisons with control herds.