(7) animal behaviour RM

things to remember about behaviour

• behaviour is all observable responses that an individual / animal gives to internal or external stimuli - can see and record

  • stimuli = events that cause an organism to perform an activity or start a reaction

• can manipulate the stimuli and see how the behaviour changes

• this allows us to infer the reasons why behaviours occur to be made and tells us about the individuals physiology, psychology and wellbeing

behaviour:

• normally involves movement

• categorised by particular types (they tell us the underlying physiological and psychological state of the animal)

  • social behaviours - interaction (directed behaviours) to another individual

    • may be associative, performed by individuals in the same space and time

    • or interactive, animals doing something to another

  • foraging behaviours - set pattern that a species uses to locate and identify food

    • tells us a bout internal motivation and energetic requirement

behaviour can be what?

1. pure

2. applied function of animal behaviour

what is the pure function

• evolutionary biology - where have we come from, who are our most common ancestors and how does this impact behaviour

• behavioural ecology - how has the animal evolved to the habitat it lives in

• compare cognition and consciousness - where do our brain abilities come from

• human and animal comparisons - sheds light onto the reasons and mechanisms for behaviour

  • e.g. how do evolutionary trends affect the mental health crisis

pure behaviour studies at UoE

• evolutionary components to human behaviour

• cross species comparisons (aggression, sociality)

  • social networks analysis can measure a behaviour and how its similar across species

• why do behaviour occur in the form that they do (e.g. problem solving or learning)

  • squirrels and how their behaviour has evolved

what is the applied function

1. provides data that can be used for

   • improving how we keep and manage animals

   • improving how we interact with animals that we use (improving quality of life for cows improves their milk yield)

   • adding to our understadning of internal motivation and the reasons for behavioural performance

     • need to understand motivations for performance

     • e.g. allow giraffes feeding behaviours or they will show abnormal patterns of behaviour

what is a key aspect of the applied function

animal welfare

• the state of an individual as it attempts to cope with its environment (Broom, 1986)

• can measuring coping and look at individual state by observing behaviour

• humans interacting with animals influences their quality of life

5 domains fro animal welfare assessment

physiological domains - what we provide the animal with

1. nutrition

2. living environment

3. health and fitness

behavioural domain - things we can measure

4. interactions with humans

5. interactions with other animals

6. interactions with the environment

example of observations

• when we group participants must make sure there are specific descriptions of what we are examining so data can be repeated

• observational with experimental approach

types of animal behaviour (more on this later)

• states - long duration / anything you can time

• events - short duration / anything that can’t be timed (vocalisation, bout of aggression)

• bouts - the grouping of events into a set pattern (courtship ritual)

  • sequence of events that can be stitched together (e.g. mirror, signal, manoeuvre)

what do these behaviours tell us?

• ethogram - list of all behaviours the individual will perform built from the literature (define and describe)

  • e.g. sleeping - lion is unconscious but breathing (must be very specific)

• time-activity budget → energy → motivation

  • lots of energy means a high motivational state

  • inferences of motivation come from what we have recorded

• animal welfare = state of the individual… infer via behavioural observation

ss is a population of monkeys that moved from one enclosure to another, the behavioural states change which are helpful for understadning choice, preferences and quality of life (grooming, dozing and being awake linked to environmental changes)

planning a behavioural study, what needs to be considered

it is complex and need to take into account many variables

• independent

• dependent

• extraneous (weather, time, other people)

hypotheses

• null

• alternative / experimental

sampling and recording

• sampling

  • who to follow and watch, and what to measure (e.g. how two individuals form a social bond and whether the presence of others influences the social dynamic)

• recording

  • how and when to observe / record information (e.g. when the giraffes are alone and when another individual is present)

  • focal individual - know the animal you are following

  • scan sample - individuals in a group, hard to identify

  • ethical review - stringent process, write your methods and submit for peer review

    • have a duty of care to look after animals welfare

the species observed and question being asked will dictate the recording methods utilised in the experimental design

different recording techniques

recording techniques

• instantaneous (e.g. instant, scan sampling of a group or individual), good if you know the individual but can’t do continuous due to setup/environment

  • divide the sample period into short intervals and record behaviour on the time point, need to be close enough so nothing is missed but not so close that we get pseudo-replication (do not give subject time to change behaviour naturally)

  • instantaneous focal sample - note the behaviour occurring at every time point (e.g. record every 2 mins for an hour) - might be walking, feeding, standing

  • or can do an instantaneous scan sample of all animals but give the number of individuals performing each behaviour at each time point

    • overall score expressed as a proportion of the observation time

• continuous - recording the total time spent by an individual on each behaviour, good if you know the individual

  • exact record of behaviour recorded at each occurrence (mins or seconds on each), keep timing behaviour until animal does something else

  • necessary for true frequencies / duration

• one-zero sampling (yes I saw it, no I didn’t, good for rare activity)

  • record whether the behaviour has occurred in sample period

  • have same observation time of an hour, broken down into 2 mins and say whether you see the particular rare behaviour - can then work our the rate the behaviour occurs for

  • often goes in conjunction with behavioural sampling

different sampling techniques

sampling techniques

• ad lib sampling (e.g. write down and see what occurs, use for an ethogram) - all behaviours are recorded

  • good for creating an ethogram

• focal sampling - focus on one individual

• scan sampling - scan group at regular intervals (different behaviours in the group being performed - good if you can’t tell individuals apart)

• behavioural / event sampling - record occurrence of a particular behaviour (good if that behaviour is the focus)

  • tally chart for a frequency or rate (all-occurrence)

about sample sizes

• want as many subjects as possible (but case studies are useful too)

• as much data as possible

  • this will allow us to determine if the data follows a normal distribution

• stats and significance will depend on what you have watched / observed / counted

  • t-tests are really good for behavioural data - allow us to decide on other tests to do to figure out where particular differences lie

  • can ignore the idea they only work on parametric data as long as you have a fairly large sample (25+), t tests will still provide valid inferential analysis

t-tests for comparing animals

• one sample t-test good for comparing a treatment against a baseline

  • know the average amount of time giraffes should chew for, then compare how long our sample spend chewing with the published mean

• paired t-tests are good for comparing data from a population measured twice

  • differences between feeding and foraging in the morning and afternoon and compare within same population

• can also do post hoc testing to see where differences caused by variables like age category lie

• repeated measures testing is really good to tell us what is effecting behaviours being observed

example of continuous recording

• focal sample as we know who they are

• may be interested in motivation using forced swim test or using a wheel

  • start stop watch when its on the wheel and stop it when the mouse gets off

  • can see whether depression effects how the animal engages with the environment

example of instantaneous recording

• scan sample as cannot tell who is who

• so have to do scan instantaneous recording

  • take 20 and count behavioural data at each time point

time intervals need to be based on the ecology of the species

example of one-zero recording

• rare behaviour of throwing their backs in the air and belching

• every 2 mins watching ducks and looking at how long they spend swimming and not swimming

  • every 2 mins ask have i seen this display, yes or no (coded 0 or 1)

more on the different measures of behaviour

• frequency

  • the number of occurrences of the behaviour per unit of time (e.g. how many times see the courtship display within an hour)

• duration

  • the length of time for which a single occurrence of the behaviour lasts (how long does the courtship behaviour last)

• latency

  • the time from a specific stimulus to the first occurrence of the behaviour - good for lab studies (e.g. provide animal with enrichment, then how long until the first important behaviour)

• bouts

  • a short period of a specific activity, normally intense, that can be timed (good for vocalisations)

  • will have a specific start and end point and a latency between bouts is required for measurement

  • e.g. courtship - all behaviours that make up the dance is the bout and the latency is the time for the group to get organised then to start performing the courtship display

  • some behaviours allow us to split the activity into bouts (latency is before foot stamping, bout is the foot stamping, then once its finished how logn until it start again)

but what about quick stuff that is hard to time

many event behaviours that occur instantaneously (biting, pecking, licking, scratching are events, socialising is the state)

• walking, feeding etc are state behaviours that can be in a time-activity budget

• but if you can’t time it, it cannot be in a time-activity budget

• these are event behaviours and must be counted and displayed as a frequency or rate (of occurrence)

  • divide total number of times seen event by observation time and this gives us rate

  • ad lib sampling helps with this as you can do states and events at the same time

what are time-activity budgets

• proportion of behaviours displayed during the study period

• e.g. a group of pigs watched for 5 hours

  • 15% of the time was spent foraging

  • 20% was spent asleep

    • would need to show error (using SE or SD bars for an average time-activity budget)

• this would be expressed graphically to show how the individuals expend energy on behaviour (and therefore shows us the importance of each activity)

  • time = energy = motivation

recording sheets

• record as much of the ‘noise’ as possible to explain any behaviours you may observe

• scan is the count of number of animals performing that behaviour rather than a cross

  • can then calculate mean and SD

data entry sheet

this tells us what happens at what time, at what date and all the other information around it

• this allows for analysis of predictor variables and easy application into Jamovi

developing work

• seek out precedent and use the literature

• base your methods on those that are published (remember to cite!) - this is not plagiarism!!

• record states and events at the same time - maximise the amount of data you can record

• think about the question you want to answer and the most appropriate species to use to answer it - is it a causation, development, an evolution, function etc question

ethics and peer review

• Animal Scientific Procedures Act regulates use of all animals that are mammals, bird, fish, reptiles, amphibians and 2 invertebrates (octopi and crustaceans)

  • need home office licence if keeping animals

• comments and feedback are required to ensure valid methods are provided

• check the quality of evidence used in support of your aims and research question

• always comply with legislation and policy surrounding animal welfare and ethical practices

ethics form includes the 3 R’s

1. Reduction - have you reduced the number of animals to the most statistically relevant population (best possible for statistical significance)

2. Refinement - have you considered this is the best species to use (why do you need a primate, not a mouse, why can’t it be done on humans)

3. Replacement - what is it that you can change or adapt that means we don’t have to use animals or can you use technology to reduce your presence and improve animal welfare, don’t wear same colours as the keepers

   • good welfare = better results

what makes animal behaviour good compared to human research

• animal subjects won’t decide not to participate

• they can be managed in one place

• provide real world insight into the human world from applied and pure perspective

• animal research is covered by the same ethical review as human research

example study

Rose & Robert - activity patterns and enclosure usage of a little-studied zoo species, the sitatunga (antelope)

• time-activity budget on an individual level - can see where behavioural differences lie

calculating daily activity patterns

these can be calculated quite easily using animal behaviour data

• create a behavioural profile for the animal over a 24 hour period

• best if behavioural observations have occurred during night and day

  • use of technology allow for this

• or can do over a 12 hour daylight period if cannot do 24 hours to see how behaviour changes on a temporal scale

  • need to apply relevant methods to each species (e.g. no good collecting data on owls during the day)

example of fluctuation in activity overtime

• % of time on each behaviour (feeding and rumination)

• time on day on x axis

measuring this gives us a good understudying of welfare state to allow for us to positive impact animals quality of life in zoo’s

do’s for animal behaviour research

• run pilot study - know animals, the population, what you are going to do and the constraints

• read papers detailing similar species or similar research question before starting

• con struct ethogram before you begin

• base methods on precedent

don’ts

• do not continue using a sampling protocol that is not generating useful or valid data

• do not assume that you need hours and hours of data to collect accurate information on specific aspects of animal behaviour

• however, don’t think behaviour is an easy research option - quality data is better than quantity

evaluating data

• think question and not species

  • introduce the bigger picture, what element of behaviour are you studying

  • then introduce species because that is what you are testing question on

  • analyse data using appropriate inferential stats

  • expand discussion and conclusions to show application to the wider world (why does the research matter)

checklist for behavioural recording and summary

sampling

• who to watch

• what to measure

  • focal

  • scan

  • behavioural

  • event

recording

• how to watch

• when to measure

  • continuous

  • instantaneous

  • one-zero