Sunday, 7 October 2012

Can sheep feel hunger?

I think it is time I blog about some of my own research. I did a few interesting studies on hunger in sheep during my PhD in New Zealand, of which three have now been published (see list of publications if you’re interested)!

Sheep farming systems are mostly extensive in New Zealand, which means that the amount of land used is relatively large while the input of resources is relatively low. As opposed to the more intensive farming systems (the kind of farming most common in Europe, with limited land, and high input of labour, nutrition and technology), extensively farmed animals are generally reared outside all year around and receive minimal nutritional supplementation. This looks like to ideal picture for most people; happy animals in a paddock with plenty of space! Of course it’s great for sheep to be grazed all year around on lush green paddocks. However, what happens when those paddocks are not so lush and green? For example, during summer droughts or very cold winters?

There was actually a major summer drought in the year I conducted the studies (2008), and many farmers were not prepared for this. Pasture was insufficient to feed all animals and nutritional supplementation was not readily available and extremely extensive. Unfortunately, this meant poor welfare for both sheep and farmers! However, this was not the first time, and most likely not the last time this happened. Similar scenarios also occur(ed) in Australia! One of the problems with such low nutrition in summer is that animals are thin when they are mated and have to cope with a subsequent pregnancy, possibly without the opportunity to gain sufficient weight later on. Similar problems occur during cold winters, when pregnant sheep may lose a lot of weight and body condition score.

Body condition score (BCS) is a measure of how fat sheep are, and is assessed by manually feeling the fat and muscle cover on the sheep’s back and giving it a score between 1 (extremely thin) and 5 (severely obese). During my research I changed the BCS of pregnant sheep (by changing their diet) to simulate drought conditions in order to get a good indication of ewe welfare. This sounds pretty mean, but it is very important research because many industry people and policy makers still believe that it is ok for sheep to be extremely skinny because this is “normal and natural”. Although it seems logic that poorly-fed skinny sheep are hungry and have poor welfare, this has not been scientifically proven. Such evidence is really needed in order to change things for the better!

The first problem I ran into is how to know whether sheep are hungry or not. You can’t really tell just by looking at them (although the fact that they’re skinny may give some indication, but again, there is no real scientific evidence that skinny sheep are hungry) and sheep can’t tell you how they feel! However, there are some indirect ways in which we can “ask” the sheep how it feels. So the first study I did was aimed at developing a new methodology to measure hunger in sheep.

For this study, we build 2 identical races (each 3 m wide and 20 m long). Each race had an automated feeding station (equipped with an ultrasound sensor) that delivered about 5 g of a food reward every time the sheep approached it. There was also a moving gate that pushed the sheep to a pre-set distance along the race after the sheep had eaten the reward. The sheep could eat as many rewards in a day as it liked, as long as it was willing to walk the distance for each reward. We also changed the distance that sheep had to walk for each reward. We call this changing the “cost” or “price” of the reward. It’s kind of like a sheep in a supermarket: if food is cheap they will consume a lot, but the food becomes less attractive when it gets more expensive. Therefore, we expect sheep to “buy” a lot when the price is low, but gradually “buy” less as food becomes more expensive. Just like humans really.
Sheep getting a reward in the race. 
The beeps signal that the gate is going to move.

To test whether we can measure hunger in sheep, we divided the sheep into two groups; one group was fed as much food as they wanted before the test, while the other half did not receive any food for 24 h (fasting). We then tested all sheep at 5 different costs. The hypothesis was that a fasted sheep would be willing to pay a higher price for its food and also consume more food compared to the satiated sheep. Indeed, this is what we found. We successfully developed a methodology that allows sheep to indicate how much they are willing to pay for their food.

In the next study, we wanted to simulate more chronic food restriction conditions similar to what a sheep would experience on a farm. We chose to use pregnant sheep (Coopworth x Texel crosses), as they are likely to be the most vulnerable to undernutiriton. We divided 22 pregnant sheep with a good BCS (BCS of 3) into three different treatments: low BCS aimed at BCS 2, medium BCS aimed to stay at BCS 3 and high BCS aimed at BCS 4.  We adjusted the sheep’s diets to make sure they gained/lost the appropriate amount of weight/BCS for a period of 6 weeks. Note that a BCS of 2 is not an extremely low BCS, the animal is still healthy, alert and energetic. During droughts, farmed sheep may have a BCS of less than 2.

BCS 2 sheep
BCS 4 sheep

After 6 weeks, we tested them at 5 different costs in the races. We found that the low BCS ewes were willing to pay the highest price for their food. When cost was high (50 m per reward), the low BCS ewes were willing to walk 14km on a day just to get their daily requirements! It took them about 22 hours to walk this distance (they took little breaks in between), and they hardly got any sleep. This really shows how hungry they were! The high BCS ewes, on the contrary, did not want to walk for their food at all. Even when cost was really low (only 1 m per reward) they choose to eat much less then their daily requirements. At the highest cost (50m) they only ate 10% of the daily requirements. They actually slept the most part of the day, and did not bother to walk for their food! On the days that they were not in the races and food was freely available, however, they ate about 1.6 times the daily requirements. This is a really interesting result, because it shows that sheep will overeat and get fat if tasty food is easily available. If they have to make a little bit of effort, however, they reduce food intake to very minimal levels!

We also looked at the energy balance of the sheep and at some of the physiological parameters involved in regulating food intake and body weight. As was to be expected, the low BCS ewes were in a negative energy balance and struggled to get enough energy. The high BCS ewes had plenty of energy, leading to significant weight and fat gain! The medium BCS were intermediate, and seemed the most energetically balanced. Interestingly, we also found correlations between glucose (blood sugar), leptin (a hormone secreted from the fat cells that regulates food intake) and hunger; sheep with lower glucose and leptin levels pay a higher price for food.

We have therefore shown that we can measure hunger in sheep! And we have provided evidence that skinny sheep experience hunger! In fact, we can recommend to anyone keeping sheep that a BCS of 3 is optimal. A BCS of 2 is really too skinny and will make sheep feel hungry! There is no more excuse for skinny sheep!

Tuesday, 14 August 2012

Dalmatian dog adopts spotted lamb

A beautiful example of two different species caring for each other! This Dalmatian dog adopted a cute spotted lamb. Take a look at the video:

Friday, 6 July 2012

Do factory-farmed pigs know what they are missing?

You have probably heard this argument many times in defence of the factory farming industry:  if animals have never experienced any better conditions, then that means they don’t know what they are missing and therefore they are not suffering. But is this really true?

Does the fact that animals are always kept in barren conditions mean that they really “don’t know any better” and therefore it is OK?

I don’t think so… and a recent study has provided evidence that animals kept in barren conditions, without ever having experienced good conditions, are in a bad emotional state and are suffering.

In this experiment, two groups of young pigs were randomly assigned to housing in either an enriched or barren environment. The barren environment was pretty typical for factory farming and met the minimal requirements for intensive pig housing according to EC directives. The pigs had 1.2m2 space each, the pens had a partially slatted concrete floor and there was a wood log to play with. Pigs in the enriched environment were kept on a solid concrete floor (1.9m2 space per pig) covered with fresh straw and had metal chains and logs, sticks and cardboard boxes to play with.

Pigs were then trained on a judgement bias task. The judgement bias methodology comes from human psychology and has recently been adapted to measure emotional states in animals. It is a very promising methodology. It basically measures whether the animal perceives its environment as “optimistic” or “pessimistic”. It is kind of a glass half full or half empty type of test.

Pigs were trained to associate a note played on glockenspiel with a tasty food reward (apple): when they heard the note they had to approach a hatch in a training arena were they received an apple.  They were also trained to associate the sound of a dog clicker trainer with an aversive event (a plastic bag waved in the face). When pigs approached the hatch, the bag was waved in their face, which then taught them NOT to approach the hatch when hearing the dog clicker trainer.  Once the pigs had learned to always approach the hatch after hearing the glockenspiel but never to approach when hearing the clicker trainer they were ready for the judgement bias test.

In the judgement bias test, a third unfamiliar sound was played (a squeaky dog toy) and it was observed whether the pigs approached the hatch or not. A pig expecting to be rewarded with the apple approached the hatch after hearing the ambiguous sound, and therefore made an optimistic choice. While pigs expecting to be punished by the plastic bag did not approach the hatch, and made a pessimistic choice.

The results of the study were very interesting. The first time judgement bias was tested, the pigs kept in the barren environment approached the hatch significantly less often compared to the pigs kept in the enriched environment.

This shows that even though the barren-housed pigs had never experienced any enriched conditions, they interpreted the ambiguous sound as more pessimistic and were expecting to be punished more often! In humans, such pessimistic interpretation of events is often seen in depressed people!

After the first test, the housing of the pigs was then changed: pigs in the enriched group were moved to barren housing and the pigs in the barren group were now housed in enriched pens.

The next judgement bias test again showed that pigs kept in the barren conditions interpreted the ambiguous sound as more pessimistic.

The pigs were then changed back to their original housing, and results were similar.

The study also found that the length of time that pigs experienced the housing conditions affected how they interpreted the ambiguous cues.  Pigs that had been housed in the enriched environment for 5 weeks made more pessimistic choices when moved to the barren environment than pigs that had experienced the enriched conditions for only 7 days. It seems that once the animals are used to the enriched conditions, it is worse to experience the barren conditions.

Therefore, pigs that never experienced enriched conditions appear to know what they are missing. But pigs that do know what they are missing are affected by barren housing even worse!

So there is really no excuse for keeping pigs in barren environments! 

I wonder how these free-range pigs would have
interpreted the ambiguous sounds? I guess they would
have been pretty optimistic!
Reference: Douglas et al., 2012. Environmental enrichment induces optimistic cognitive biases in pigs. Applied Animal Behaviour Science. Article In Press. 

Monday, 21 May 2012

Does your sports horse need a friend?

Many horses kept for sports or recreational use are housed in individual stables, some only for a part of the day and others permanently. Some of these stables allow some form of contact with a neighbouring horse, while others minimize social contact. The reasons for housing horses in single pens are various; cleanliness of the horse, ease of getting close to the horse, protection from the elements, lack of pastures and some owners even believe that their horse has no need for social contact. However, could horses have a strong need for social contact with other horses, and could restricting social contact between horses have a detrimental impact on their welfare?

Horses can tolerate cold very well.
Make them wear a rug if you're worried about cold.

Feral horses are very social, and their survival depends on strong social bonds between families or bachelors. For example, the (early) detection of a predator and the following flight are important defence mechanisms for a group of horses. There is strong evidence that our modern sports horses still have a very basic need for social contact (from now on, when I speak of social contact, I mean social contact with other horses, not with other species including humans): former domesticated horses that have become feral display many social interactions and a very structured social organization. However, a main difference between feral horses and domesticated horses is that our domesticated horses generally do not choose their own social groups (we decide what individuals share a paddock) and paddock space may be limited. In addition, in domesticated horse groups there are generally more males (geldings) than in feral groups.

So what kind of social behaviours do horses engage in? First of all, there are agonistic interactions that help determining the hierarchy in the group. All social groups form hierarchies to determine which individuals get first access to the best food, shelter and mates. Hierarchies also ensure that it is “clear” what individuals have the highest rank in the group, which then minimizes aggression. The main agonistic behaviours that horses show are 1) threat behaviours consisting of attack, bite, threat to bite, and approach with ears flattened and 2) submissive avoidance behaviour. Kicking with the hind legs is used both while attacking and while defending from an attack, and is therefore context dependent.

This is my 21 year old (now retired) horse.
He always had plenty of time at pasture with his friends,

even when competing at advanced level dressage
(a long time ago...). 

Another important social behaviour is allogrooming when two horses groom each other. Foals start allogrooming within the first two weeks of life, usually with their mother or brothers and sisters. Research has shown that allogrooming is rewarding for horses, meaning that it results in the release of “pleasure” hormones (endorphins). Horses that are housed in single pens with minimal social contact are very keen to have physical head and neck contact with an unfamiliar horse. However, when two grooming horses were separated after 5 min of the social contact, the horses showed physiological and behavioural signs of frustration, again showing that allogrooming is important for horses. Play is most likely an equally important and rewarding behaviour in foals and geldings, although it may be less important for mares.

So does your horse suffer when it cannot experience the joys of allogrooming and play, or perhaps any other social behaviour that is rewarding? Unfortunately, very few studies have investigated this question directly and little data is available. However, it is known that chronic stress in horses leads to stereotypic behaviours which are repetitive behaviours that seemingly have no function, such as weaving, wind sucking, crib-biting, box-walking and wood-chewing. Stereotypic behaviours are generally interpreted as a sign of poor welfare. Several studies have shown that individual housing, irregular social contact and limited access to pasture are among the highest risk factor for developing stereotypic behaviours. Therefore, it is likely that limited social contact leads to suffering in horses.

To ensure the welfare of your horse, give it plenty of access to pasture and allow it to play and interact with other horses!

Reference: Vandierendonck et al., 2012. Applied Animal Behaviour Science 138, 194-202.

Thursday, 10 May 2012