Strap line

It started when two canine scientists decide to become pen pals in an era of digital media...

Tuesday, 21 August 2012

It's all about objective multiples...

sciseekclaimtoken-5036b68d5c7f8
First thoughts...
Hey Julie!

I hope you had a great weekend. Can I tell you more about assessing welfare? Hell to the yeah I can! 

(By the way, who gave you that photo of me working at 3:17am?!) 

I could probably blog every day until the next Canine Science Forum about measuring animal welfare - but there are so many other things I want to discuss with you, that simply won’t do.

In thinking how to respond to your question, I popped out to my bookcase (technically, it’s more of a book wall) and grabbed the first few books I thought of when I think of ‘understanding and assessing canine welfare’. There were 23 of them. Hmmm. And that’s before I even considered pulling up relevant scientific journal papers. So...

Interpreting and measuring animal welfare isn’t easy. There’s no litmus test for animal welfare, no smartphone app' that you can use to scan an animal and get a quick result as to whether its welfare is good, neutral or poor (but how cool would that be!).
It’s not easy, but it’s by no means impossible. 

Saliva collecting is easy with slobbery dogs
As scientists, we love an objective measure (one that requires minimal judgement by the person making the assessment, which reduces the chance of human bias and/or error) so we really love something that gives us numbers to analyse. This is where measurements of physical functions, like heart rate, weight, stress hormones and immune system function are great. No matter what mood I’m in, how much coffee I’ve drunk or what time of the night I’m reviewing my PhD data, I can’t alter the fact a salivary cortisol measure came in at 0.29µg/dL. Because that’s what the value is. It just is.

This dog obviously has 'good drive' (source)
Behavioural information is awesome because we can generally collect it without interfering with the animals, reducing any experimenter effect. It’s important that its measurement is still done objectively and this is where an ethological background (oh hey, like mine!) can be really handy. There are many industry-based terms which may mean different things to different people, or in different contexts. For example, in the working dog industry, terms like ‘drive’, ‘willingness’, ‘suspicion’ and ‘instinct’ are behavioural traits commonly referred to, but what is good drive? Has this boxer got ‘good drive’? How do we measure it? Is a livestock herding dog’s drive the same as a guide dog, police dog or racing greyhound’s drive? 

When we measure behaviour objectively using an ethogram, and look at frequencies (how often within a timeframe) or durations (length of time) of certain actions (like barking, sitting, walking, interaction with another dog, person or toy) we remove the human interpretation (and potential error in interpretation) from the data collection stage, which is SUPER important. So rather than rate the subjective term drive, a scientist would tend to measure objective behaviours or actions, like command-response time, running speed or directional tracking of eye movements.

(source)
Environment is also important as it gives context to the physiological and behavioural data. For example, if a dog’s cortisol reading in my PhD work was unusually high (usually associated with stress) and the video recording showed that dog barking incessantly in its kennel yard for an extended period of time, I might interpret that as a dog with a significant stress response to being kennelled. However, if I know that at that particular time of behavioural sampling and prior to the saliva collection there was a thunderstorm taking place, it might change how I interpret the data in front of me. It’s critical not to overlook even the simplest of environmental factors (like temperature) that have been shown to affect both animal behaviour and physiology.

This is truly just the tip of the ‘assessing animal welfare’ iceberg and there's always more to learn. I haven’t touched on the hot topics of ‘quality of life’, ‘affective states’, individual vs. population welfare, the interplay of public attitudes/perceptions and ethics with animals welfare or why there is currently an emphasis about avoiding bad welfare rather than promoting the good – lucky we have time on our side for all of that! 

Hopefully this has helped give you an idea of why multiple (physiological and behavioural) measures are so important in assessing canine welfare – the more information you have to help with the interpretation of an animal’s experience, the better – until that smartphone app' comes along anyway!

So how was your weekend, anyway?

Mia

Dreschel, N.A. & Granger, D.A. (2009). Methods of collection for salivary cortisol measurement in dogs, Hormones and Behavior, 55 (1) 168. DOI: 10.1016/j.yhbeh.2008.09.010

Martin, P. & Bateson, P. (2007) Measuring Behaviour: An Introductory Guide. (3rd Ed.) Cambridge University Press. ISBN: 9780521535632


© Mia Cobb 2012