Movement numbers, timing, and holding types
The foot and mouth disease epidemics in 2001 and 2007 are both noticeable as a drop in movement volume in Figure 1. Even at the height of the 2001 epidemic, however, there was still a certain amount of movement going on; licenses were granted for movements within the infected area, from the uninfected area to the infected area, and within the uninfected area.
Figure 1 shows a clear seasonal pattern to movement volumes, with peaks in April and October of each year. Previous work has looked at seasonal patterns in cattle movements in more detail, and shown both that most movements occur during the working week, with a peak on Wednesdays, and that there is a seasonal peak in the number of births in spring, and a smaller one in September [15, 16]; the longer-term analysis presented here shows that this seasonal pattern has continued.
As would be expected, the vast majority of movements involve agricultural holdings, markets, and slaughterhouses. Table 1 shows that agricultural holdings are net exporters of animals, the numbers of animals entering and leaving markets are roughly the same, and that slaughterhouses are net importers of animals. Since animals are born on farms, pass through markets, and die at slaughterhouses, these figures are reassuringly predictable.
Comparing tables 2 and 3 shows that there were substantially fewer movements between animal holdings in 2008-2009 than would be expected by chance; the majority of this difference is explained by the greater number of movements from animal holdings to markets and slaughterhouses, and from markets to animal holdings. Similarly, there is very little movement of animals from market to market, animals instead moving to or from animal holdings. As well as being an interesting insight into the structure of the cattle industry in the UK, these figures would be valuable for constructing an economic model of livestock movements, which in turn might be a useful technique for predicting future patterns of livestock movement in the UK.
Livestock numbers and ages
Figures 2, 3, and 4, and Table 4 provide some insight into the mortality of British cattle. Table 4 shows unremarkably that the majority of cattle deaths occur at red meat slaughterhouses; also that animal holdings and red meat slaughterhouses account for nearly all (99%) cattle deaths between them. In the light of concerns about the risk BSE posed to human health, The Fresh Meat (Beef Controls) Regulations 1996 were introduced on 29 March 1996. They banned cattle that were over thirty months old from entering the human food chain; instead the animals were slaughtered, and farmers paid compensation under the over thirty months slaughter scheme. This ban was relaxed on 7 November 2005, when older cattle were again eligible to enter the human food chain, provided they tested negative for BSE. The effect of this so-called "over thirty month rule" (OTM) is clear to see in Figures 2 and 3 -- there is a substantial spike in the number of cattle dying at thirty months old. Figure 3 shows the distribution of ages of animals dying at red meat slaughterhouses. There is a substantial peak at around a week of age, particularly among dairy cattle; male dairy calves are worth very little, so some are slaughtered at a young age to save the cost of rearing them; rennet may also be extracted from the abomasums of calves. Animals are typically slaughtered for veal at around 6 months of age; it is clear from Figure 3 that this remains a insignificant beef product in the UK. Intensively reared beef is produced from beef and dairy animals of around 18 months of age; these animals are fed cereals and concentrates and so come to slaughter weight faster than more extensively-reared animals, and the peaks in Figure 3 at around 500 days are due to this type of beef production. Finally, extensive beef suckler systems where beef cattle are reared more slowly on grass result in animals reaching slaughter weight at around 24 months; they result in the step in the number of beef cattle dying at around 700 days old. While Figure 3 shows the relative importance of different beef rearing regimes, Figure 4 shows the ages at which animals die on farms, generally representing a loss to the farmer. As would be expected, the majority of losses occur in young animals, succumbing to disease early in their life, although there is a small peak at 30 months, again probably due to the OTM scheme.
Frequency and distance of movements
Figures 5 and 6 show the distributions of the number of times an animal moves in its life, and the distances over which animals are moved, respectively. The extreme x-values in these figures should be treated with some caution -- it seems unlikely that an animal would travel 4,838 km (roughly four times the road distance between Land's End and John o'Groats) in its lifetime, for example, although pedigree animals may be taken to many showgrounds during their lives. Figure 5 shows that most animals move only a few times during their lifetimes; a single move (from birth location to slaughterhouse) is most common. Dairy animals are more likely to make two moves during their lifetimes than beef animals; this is most likely due to male dairy calves moving once to a fattening unit, and thence to slaughter. Figure 6 shows that while around 18% of animals move less than a kilometre during their life, there is then a very broad spread of distances travelled, with dairy cattle moving less far than beef cattle. This pattern of most animals moving only a few times, and for a short distance has been described before in the UK , though the longer-term analysis presented here shows a higher proportion of animals moving only once in their lifetime. Research on cattle movements in Italy, Portugal and Sweden has shown that most movements are short-range (with a few longer distance movements), suggesting that this may be a common pattern across at least Western Europe [19–21]. This is the first analysis to consider the total distance animals move in their lifetimes and its relationship to the number of times animals move.
There is a weak (ρ = 0.496) but statistically significant correlation between the number of times an animal is moved in its life, and the total distance it moves, as illustrated in Figure 7. Intuitively, animals that are moved more frequently would be expected to move further in their lifetimes, so it is a little surprising that this correlation is not stronger.
The duration of livestocks' stays on holdings is shown in Figure 8. Around 34% of all recorded stays are transient, i.e. the animal leaves the holding on the same day as it arrived there; these will be stays on markets. The effects of the OTM scheme are evident again, with a noticeable rise in stays of around 30 months.
From the point of view of understanding how cattle are moved, and potentially predicting future movement patterns, an interesting question is how habitual farmers are; if they are very habitual in their movement patterns, then one could reasonably assume that a farm will send its cattle to the same market next year that it did this year. Figure 9 enables this question to be addressed; it shows the number of times a movement occurs as a cumulative frequency distribution. Nearly a third of movements occur between 2 and 10 times, so some repetition of movements should be incorporated into any model of the UK cattle industry, but only to a limited extent.
Changes in movement patterns over the past decade
Given that the regulatory regime regarding animal movements has changed substantially in the recent past, particularly since the 2001 FMD epidemic, it is worthwhile to try and assess what effect these changes have had on the movement of animals. Figures 10, 11, 12, and 13 do this, on a yearly basis.
What is striking about these figures is how little has changed since 2002 overall, in contrast to work by Robinson and Christley which considered movements in the period 2002 to early 2005 . The availability of data for a longer period of time shows that while there was an increase in cattle movement in the period they studied (see e.g. Figure 13), that increase has not continued.
Issues with the CTS
The CTS was not set up with the intention that it might be useful as a control system for epidemic diseases such as FMD; the 2001 FMD outbreak in the UK and subsequent enquiries have led to changes in the collection of data, and the scope of such data. Specifically, the UK government has attempted to increase reporting of cattle movements by electronic means, and has introduced schemes to collect details on batch movements (rather than individual-level data) of sheep, pigs, and goats .
Not all movements of cattle are required to be reported to RADAR. Specifically, movements to shared grazing lands are not required to be reported, and neither are movements between holdings that have been "linked". The latter process is meant to allow farmers to move livestock between nearby holdings without the administrative burden of having to report the movements, but it has been abused by some farmers, who have "linked" holdings which are far away from each other . Given the original purpose of CTS, it is perhaps unsurprising that such movements need not be reported, but they may represent a substantial epidemiological risk.
A National Audit Office report noted that some keepers may be tempted to avoid the extra work associated with reporting animal movements, and that furthermore there may be financial advantages to deliberately contravening the identification and tracking requirements (particularly given standstill periods); some examples of detected fraud were illustrated, although there is little idea as to the scale of the problem . DEFRA has conducted a review of the livestock movement controls. In addition to issues regarding abuse of "linked" holdings, the review concluded that the current regulations are overly complex and should therefore be simplified. It additionally recommends that abattoirs should report the premises of departure of animals arriving at them, and that markets and collection centres should report the source and destination of animals passing through them, by electronic means. Regarding shared grazing lands, it suggests that a single Land Management Unit should be formed consisting of the common land and any in-bye land to which cattle on the shared grazing have free access; movements into and out of this area would have to be reported, and would induce a standstill period. It also advocates greater regulation of dealers and traders, specifically that those which hold livestock for mixing and sorting purposes be treated as collection centres (and so be subject to a formal approval procedure), and that CTS investigate movements of animals where a few days have passed between an "off" movement and the subsequent "on" movement, to attempt to determine whether the animals concerned stayed at an intermediate premises . In 2010, DEFRA consulted on proposals to simplify the livestock movement rules and holding identifiers in England, although no changes have yet been proposed as a result of that consultation. Problems remain, however. The current regulations are complex, which leads to errors in reporting, and are somewhat open to abuse. Furthermore, the data are not collected nor stored in a manner ideally suited to contact-network-based studies (although this latter situation has improved significantly with the production of ordered movement tables for each animal). How important the delay between movements and their reporting to RADAR is in terms of intervention during an outbreak is an unanswered question; during the brief 2007 FMD outbreak, livestock movement data were not available to researchers until the outbreak was over.
The importance of movements that are not required to be reported to RADAR in contact networks is unknown, and difficult to quantify nationally; a study in the Outer Hebridies showed that shared grazing land was a significant source of unreported contact between different keepers' cattle .