Badger visits to farm yards – new paper published
Several research studies have shown that badgers will enter farm yards where they may contaminate the farm environment (feed, water troughs etc) or come into close contact with cattle (Garnett, Delahay & Roper 2002; Tolhurst et al. 2009; Judge et al. 2011; O’Mahony 2015). Feeding opportunities seem to be a key driving factor behind this behaviour, as yards and buildings may contain high energy animal feeds which are attractive to badgers (click here to see a list of feed sources attractive to badgers).
Work by APHA has shown that practical measures such as sheet gates or electric fences can be used to keep badgers out of farm yards (Judge et al. 2011), and it is recommended that farmers use such measures where practical in their farms to reduce the risk of disease spread (see the TBhub). However, previous studies on badger activity in farm yards have shown that levels of activity can vary hugely (Mullen et al. 2015; Woodroffe et al. 2017). This raises the question as to how widespread this behaviour is and whether particularly farms are more likely to experience badger visits than others.
Recently we published a paper in the Journal PLOS One on the subject of badgers in farm buildings…
This study aimed to answer several questions:
- How widespread are badgers visits to farm yards?
- Which farm characteristics are associated with visits by badgers?
- Can we predict this behaviour and identify farms where visits occur?
The paper is open access (free to view), but I thought it would be useful to produce a short summary for those who want the key information.
What did the study involve?
The study involved monitoring badger activity at 155 farms across south-west England and Wales. Each farm was monitored for one month using trail cameras (click here for advice on using cameras to monitor badger activity).
Each farm was also surveyed for badger field signs (setts, latrines etc) and a range of other farm characteristics were recorded (farm type, number of cattle, numbers of different buildings etc).
A range of statistical analyses were then conducted to look at how common badgers visits were, and also determine which farm characteristics were associated with badger visits.
How widespread were badger visits to farm yards?
Overall 41% of farms had evidence of farm visits (badgers recorded on camera), although the level of activity (number of nights badgers were seen) was quite variable.
Which farm characteristics were associated with badger visits?
The presence/absence of badger visits was significantly related to several farm characteristics……
- Badger visits were more likely at farms with higher numbers of feed stores and cattle sheds (suggesting more feeding opportunities?)
- Badger visits were more likely at farms with higher numbers of badger setts (within 500m), and where the nearest active sett was located closer to the farm.
- Badger visits were less likely at farms which housed large numbers of cattle and where there was a farm house on dwelling on site. This could indicate that the farm was larger or had higher levels of human disturbance that may deter badgers?
The frequency of visits (ie the number of nights badgers were recorded) was related to only two factors:
- Number of cattle housed (higher number = fewer visits)
- Distance to the nearest active badger sett (closer sett = more visits
Can we predict this behaviour and identify farms where visits occur?
Statistical models (which identified the different relationships above) were used to predict badger visits at 40 farms with known levels of badger visitation which were intensively monitored as part of an earlier study (Judge et al. 2011). Much like when developing a new diagnostic test, different cut off values can be chosen (ie above this cut off farms are classed as being likely to experience badger visits) which affect the accuracy of the models predictions, along with the sensitivity and specificity.
- Sensitivity – the % of farms with badger visits that were correctly classified
- Specificity – the % of farms without badger visits that were correctly classified
Overall the models could predict the presence /absence of badger visits with a 72% accuracy (sensitivity = 62%, specificity= 81%). However these number could potentially be changed by using a different cut off value.
The models were less accurate at predicting the frequency of badger visits (ie not only identifying farms where badgers were present, but those which had lots of visits rather than just one or two). Although, farms predicted as having badgers present tended to be those with higher overall levels of activity.
We also created an interactive tool where users can enter the characteristics of a given farm and generate predictions.
What can we conclude from this study?
This study further confirms that badger visits to farm yards may be widespread, affecting a significant proportion of farms within the area studied (south west England and wales). Although it should be noted that the farms were not randomly selected, so the sample of farms could be slightly biased. The study also highlights that certain farms are more likely to be of visited than others, which could help to focus biosecurity measures in a more targeted way.
The results from the study differ from several recent studies, which suggest that badger visits to farm yards are uncommon or that badgers avoid farm yards. It is important to stress there are some important differences with this study and these earlier studies, both in their scale and in the way that badger activity was recorded. This study also demonstrates that certain types of farms are more/less likely to be visited, so it is possible that farms monitored in these earlier studies fall into a low risk/less likely to be visited category.
For further details on what this study involved and for a more thorough discussion of the results – read the full paper here.