Thread by @pec_exeter

"Why are pheasants so dumb?"

We’ve been asked this repeatedly over the past 5 years as we study how selection acts on cognition, using the ultimate bird-brain as our study species

Our @ERC_Research funded project finished this week😪, so here's a 🧵 outlining our findings

@ERC_Research It’s a lot of work (39 papers/completed manuscripts); the product of a fantastic team of researchers & collaborators.

In this thread we’ve set these papers (with links) within the context of our overarching central question.

This figure summarises the thread structure

@ERC_Research This 🧵 might be an indulgence, but we think stepping back & showing how components of a large-scale project fit together can help give a clearer picture than just talking about 1 paper at a time

WARNING: If you don’t like pheasants or cognition, now may be a good time to leave

@ERC_Research Our Question:

How does natural selection sift individual differences in basic cognitive abilities (learning, memory, executive function) and thus how may such abilities evolve?

Differs from comparative approach usual in cognition studies.

Uses a behavioural ecology approach

@ERC_Research So what?
Unclear what constrains cognitive exaggeration. Why don’t all individs have rapid learning/perfect memory/faultless executive function?

Why no Einstein pheasants?

Use model system to explore general principles for what maintains variation & drives change in basic cogn

@ERC_Research Why pheasants:

🐥🐥Can be reared in large numbers, under controlled conditions, w/o parents

Can utilise cognitive tasks devised for 🐓🐓

Can be released into wild and fate followed

No evidence of strong selection having acted on specific abilities (e.g. caching or homing)

@ERC_Research We asked:

a) How did individuals differ in cognitive performance across a battery of different tasks?

b) What caused these differences?

c) What were the consequences of these differences?

(1 sub-thread per question)

We also explored some natural history along the way

@ERC_Research Sub-thread 1
How do individuals differ in cognitive performance across a battery of different tasks?

This involved testing 200 chicks per year, resulting in around 240,000 separate trials

EVERY ONE OF WHICH WAS ADMINISTERED BY HAND🤪😫🤬😱

@ERC_Research We wrote a couple of reviews considering:
1) how one should measure and interpret individual diffs royalsocietypublishing.org/doi/full/10.10…

2) and how these could be linked to fitness academic.oup.com/beheco/article…

@ERC_Research Q: How repeatable are such measures of cognition? This is essential to understand whether our assays provide informative measures
A: Broadly, not very (r = 0.19-0.28), but at levels similar to other behavioural traits commonly studied
royalsocietypublishing.org/doi/full/10.10…

@ERC_Research Q: Do we assay individuals without bias? If not, then estimates of variation within a population may be skewed
A: With care, we can, but sex, body condition and prior experience can all affect participation in cognitive tasks. Ensure we account for this.
sciencedirect.com/science/articl…

@ERC_Research Q: Does prior experience matter? If so, need to ensure full knowledge of individual’s history before testing
A: Very much! Both subtle experience of physical properties & broader early rearing conditions
royalsocietypublishing.org/doi/full/10.10…
royalsocietypublishing.org/doi/full/10.10…
link.springer.com/article/10.100…

@ERC_Research How did we measure basic cognitive abilities?

Using some fairly crazy contraptions, a dedicated crew of researchers and a LOT of patience

@ERC_Research We measured learning and memory using discrimination and spatial tasks. For details see:
royalsocietypublishing.org/doi/full/10.10…
royalsocietypublishing.org/doi/10.1098/rs…
sciencedirect.com/science/articl…
animalbehaviorandcognition.org/uploads/journa…
peerj.com/articles/5738/

@ERC_Research We measured executive functions, typically behavioural inhibition, using a range of detour, stop/change and reversal tasks. For details see:
link.springer.com/article/10.100…
royalsocietypublishing.org/doi/full/10.10…
royalsocietypublishing.org/doi/full/10.10…
link.springer.com/article/10.100…

@ERC_Research Q: Can an individual’s performance be summarised by 1 measure? Do pheasants have 'g'?

A: Not really. Within tasks assaying similar associative processes, we found some consistency sciencedirect.com/science/articl…
But no single ‘g’ factor emerged across domains royalsocietypublishing.org/doi/full/10.10…

@ERC_Research Sub-thread 2
What causes these individual differences in cognitive performance?

Involved: Rearing birds under different conditions; Collaborating w @NERCBAF to build pedigree of >900 birds; Observing social interactions to construct social networks and dominance relationships.

@ERC_Research @NERCBAF Q: How does the physical environment shape cognitive performance?

A: Increasing spatial complexity improves spatial working memory royalsocietypublishing.org/doi/full/10.10…

A: Increasing spatial unpredictability improves behavioural inhibition (enrichment?)
link.springer.com/article/10.100…

@ERC_Research @NERCBAF Q: How does the physical environment shape cognitive performance?

A: Increasing diet complexity improves motor skills (prey capture)
besjournals.onlinelibrary.wiley.com/doi/10.1111/13…
A: Increasing unpredictability of information in the environment impedes behavioural inhibition
MS in revision Anim Cogn

@ERC_Research @NERCBAF Q: How does the social environment shape cognitive performance?

Review of the relationship between cognition and social factors
royalsocietypublishing.org/doi/full/10.10…

Social factors affect cognitive performance BUT cogn perf also affects social relationships. Complex interplay

@ERC_Research @NERCBAF Q: How does the social environment shape cognitive performance?

A: More dominant individuals show faster learning
royalsocietypublishing.org/doi/10.1098/rs…

A: Unclear whether dominance is cause or consequence of ‘better’ cognition
peerj.com/articles/5738/

@ERC_Research @NERCBAF Q: How does the social environment shape cognitive performance?

A: Birds in bigger groups show faster learning
sciencedirect.com/science/articl…

A: Early-life learning predicts later life assortment and this, but not learning speed itself, predicts survival
MS in revision for J Anim Ecol

@ERC_Research @NERCBAF Sub-thread 3
What are the consequences of individ diffs in cognitive performance?

Involved: tracking released birds across Devon; fighting technology👩‍🔧; building an entire reverse GPS system👨‍💻; crawling through mud, ditches & hedges to find corpses🥶; scanning >1,000,000 pics🤓

@ERC_Research @NERCBAF Q: Do ‘smarter’ birds survive better?
A: Birds w better working spatial memory survive longer
royalsocietypublishing.org/doi/full/10.10…

A: Birds slow to reverse learnt association survive longer
royalsocietypublishing.org/doi/full/10.10…

A: Birds w medium lateralization survive longer
nature.com/articles/s4159…

@ERC_Research @NERCBAF Q: Could cognitive ability bring social benefits?

A: Dominant birds = faster learners but unclear if cause or consequence of dominance
royalsocietypublishing.org/doi/10.1098/rs…
peerj.com/articles/5738/

A: Early-life learning predicts later assortment and this⬆️survival
MS in revision J Anim Ecol

@ERC_Research @NERCBAF Q: Does cognitive ability influence movement ecology?

A: Birds with faster spatial learning: use more diverse foraging locations
MS in review w J Anim Ecol

AND show more rapid improvements in movement path efficiency
MS in prep

@ERC_Research @NERCBAF Sub-thread 4
What about pheasant biology?

Despite being most common bird in UK, little is known about their natural history, biology and ecology. Understanding this is critical to behavioural ecology

We also explored issues relating to their rearing & release for shooting

@ERC_Research @NERCBAF Death (the serious research)

Most released pheasants die in <1 year

We reviewed what kills them link.springer.com/article/10.100…
and explored how rearing conditions can reduce mortality
royalsocietypublishing.org/doi/full/10.10…
besjournals.onlinelibrary.wiley.com/doi/10.1111/13…

@ERC_Research @NERCBAF Death (the slightly odd research)

We looked at patterns of pheasant roadkill over time and space
royalsocietypublishing.org/doi/full/10.10…

& how personality predicts when they are shot(!!)
sciencedirect.com/science/articl…

@ERC_Research @NERCBAF Welfare

Almost no studies consider the welfare of gamebird 🐥🐥 during rearing. We reviewed the limited work
MS in press at Animal Welfare

We showed that thermography can be used to reveal rapid changes in arousal following frustration
MS in revision at Roy Soc Open Sci

@ERC_Research @NERCBAF Sexual Segregation

Why do pheasants segregate by sex?
We described changes in segregation over time in the wild
onlinelibrary.wiley.com/doi/abs/10.111…
We tested 3 prior hypotheses
peerj.com/articles/5674/
none applied

BUT early social preferences drove later segregation link.springer.com/article/10.100…

@ERC_Research @NERCBAF Vigilance & Group Size

Sexes differ in adjusting vigilance as harem sizes ⬆️
We calculated optimal harem size to maximise predator detection. It ~matches observed harem size sciencedirect.com/science/articl…

(May be our favourite paper – described as ‘old fashioned behavioural ecology’)

@ERC_Research @NERCBAF Sub-thread 5
Thanks & Acknowledgements

Field work is HARD
Measuring cognition is HARD
Studying fitness in the wild TAKES A LONG TIME

All this has been a team effort👨‍👩‍👧‍👦👨‍👩‍👧‍👦

Project was only possible b/c of generous 💰💰 for risky research by @ERC_Research and grant from @NERCBAF

@ERC_Research @NERCBAF Fieldwork was conducted at North Wyke Farm, home of @Rothamstead

The staff at North Wyke provided phenomenal support👩‍🍳👨‍💻👩‍🔧🤾‍♀️

Access to neighbouring land was facilitated very generous Devon farmers👨‍🌾

We loved working in Mid Devon – beautiful in (almost) all conditions

@ERC_Research @NERCBAF Assistants, Co-authors & Others who had to wear some bizarre costumes include:
@lucapstick90 @heatherwarrendr @CamilleTroisi @ChristinaHMeier @JayCee_161 @CornishJackdaws @CetaceanMike @NeeltjeBoogert @JMorandFerron @ali__wilson & many others sensible enough not to be on Twitter

@ERC_Research @NERCBAF @lucapstick90 @heatherwarrendr @CamilleTroisi @ChristinaHMeier @JayCee_161 @CornishJackdaws @CetaceanMike @NeeltjeBoogert @JMorandFerron @ali__wilson The @pec_exeter team displayed constant good humour, enthusiasm & professional approach, despite some utterly miserable physical, environmental, intellectual & personal circs

@Jayden_vanHorik @MarkAWhiteside @EllisJGLangley @MsChristineB @PipLaker @rjpheathcote

👏Thanks👏

@ERC_Research @NERCBAF @lucapstick90 @heatherwarrendr @CamilleTroisi @ChristinaHMeier @JayCee_161 @CornishJackdaws @CetaceanMike @NeeltjeBoogert @JMorandFerron @ali__wilson @Jayden_vanHorik @MarkAWhiteside @EllisJGLangley @MsChristineB @PipLaker @rjpheathcote Thank YOU for getting to the end of this 🧵

We hope you can see how we spent past 5 years & now understand & empathize with the cognitive world of the pheasant (and cognitive evolution more generally)

So, if anyone asks “Why are pheasants so dumb?”, maybe point them this way