Final Report Summary - EVOLECOG (The evolutionary ecology of individual variation in cognitive performance)
Research summary
The aim of this project was to characterize individual behavioral variation in innovativeness, learning and personality, and to explore a variety of causes and consequences of this variation for individuals in a population of great tits across a fragmented landscape of conifer and deciduous planted woodlands. The first major step was to setup a new study population in the Cork area by deploying 250 nestboxes across 9 woodland fragments and to establish a breeding population, which has been achieved. The second was to establish assays for measuring associative learning, innovativeness and personality to be conducted in captivity. The second two of these traits were straightforward but despite several attempts over several years, in the wild and in captivity, both in Cork and using a captive population at Heteren in the Netherlands, we could not achieve the first. Birds would not learn to associate specific colored keys with food rewards. This was surprising because we had already demonstrated that wild great tits in a UK population could do so, but we think reflects poor motivation in the population, and the fact that birds were too quick to give up trying, or that they did not mind being punished when they made mistakes and were happy to wait for awards at random. Although too late for this grant, in fact the experience gained over several years of trying has led to the development of other cognitive traits that are being used to good effect now under a different grant. The main results arising from this grant are based around individual measures of innovative lever pulling and the reactive proactive personality axis, which are summarized as follows.
One of the first main results showed that both personality and innovative problem solving in captivity were repeatable over lifetime but that consistency was context dependent. Specifically, innovativeness under standardized conditions did not predict innovativeness in a social competitive environment during an experiment in captivity. The study also showed that a range of environmental factors did not influence performance during the assay, including time in captivity, colorific intake prior to or during assays. A second significant result done in collaboration with colleagues at the Netherlands was that the personality measure ‘exploration behaviour’ (EB) did not predict whether individuals could innovate during social foraging trials but that the tendency to innovate over successful trials increased over time and was higher for slow explorers. In fact most birds could solve the lever pulling devices under these social circumstances, suggesting that the ability to solve these tasks is more widespread than originally thought, but that the use of this new foraging techniques is constrained by social context (social learning and shared vigilance). This is a significant result because it distinguishes between performance to innovate in isolation from the ability to do so a social context.
In another objective, we found no links between either innovative problem solving or personality and whether birds were trapped in deciduous or coniferous woodlands. After two years we already expected this would not be the case because gene flow was so high in the population. Although we found no variation between the habitats in personality or innovativeness, we did find remarkably different patterns of life history variation; for example, the ubiquitous decline in clutch size with lay date is absent in our deciduous sites, where clutch sizes and breeding success are lower because of low food availability as shown using cameras at the nest. This paper alone is a critically important description of the basic ecology of the species in this new study system, which is not only a unique description of life history variation across a fragmented landscape, but will also inform future research in my study population.
Small sample sizes meant we could not do a robust test of the hypothesis that innovativeness was linked to life history variation or indeed to fitness. However, we found negative reproductive selection on female personality, a relationship that was observed across both habitat types, and seemed to be caused by faster exploring females provisioning at a slower rate than slow explorers. Male exploration behaviour was negatively related to fledgling condition in deciduous fragments, a relationship that was reversed in coniferous fragments. Our results suggest that although the effects observed are context and sex specific, individuals at the slow end of this personality axis are favoured across our study system. Lack of progress on a cognitive trait has been disappointing but the link between personality, provisioning behavior and fitness is a significant advance in the field of personality, which was an integrative part of this project. Also recruitment was so low that it was impossible to generate a pedigree from the dataset. However, I conducted an animal model analysis on problem solving data collected by a former PhD student in another population, the first such analysis of its kind. There was little evidence for an additive genetic basis to the trait but instead we found evidence for the developmental stress hypothesis, because adults who were poor at solving problems had been born in oak-tree poor habitats and in high local population density areas. This is a significant advance because it is the largest scale study of its kind on a wild population.
For another objective, we experimentally showed that ‘tight sitting’ during incubation when faced with a predator correlated with both clutch size and reproductive success. Furthermore, amongst females that fled, nest failures increased with lay date. Our results suggest that personality is not linked to pace of life syndrome during this stage of the breeding season in females. Instead, a more immediate measure of risk-taking behaviour, tight sitting, predicted female reproductive investment and subsequent success in line with the pace of life theory. These results suggest that simple measures of personality and innovativeness do not necessarily predict important functional behavioural variation directly linked to life history variation, and provides an important cautionary note when expecting to find links between standard measures of personality or cognition, and functional behaviour.
In another experiment where birds had a choice of 3 kinds of food, we found that fast exploring great tits consumed a higher proportion of sunflower seeds, while slower individuals preferred peanuts. Problem solving performance was positively correlated with energy intake but not food preference. Peanuts accounted for a larger proportion of total daily energy intake for coniferous birds, which also lost more weight on average. Our results illustrate that a complex array of factors can determine foraging behaviour and success, including personality, innovativeness, state variables, time, and habitat origin, highlighting the challenge of explaining how selection acts on foraging performance over time. Finally, in another experiment, although we found no significant relationship between exploration behavior or innovativeness in isolation and competitiveness, individuals that were less competitive were more likely to spontaneously perform string-pulling behaviour during the dyadic trials, the first direct experimental demonstration of competitive exclusion leading to innovation, and therefore a significant advance. Our results support the hypothesis that innovations provide a means for less competitive individuals to access resources in line with the “necessity drives innovation hypothesis.
To date 6 papers linked to the project have been or are in the process of being published (under review, in revision etc) and at least another two are planned for the coming year. Together these have all helped advance our understanding of the causes and consequences of behavioural variation in a wild population, especially in the context of personality, and innovativeness. An additional 6 papers have been published in the related field during the course of the project and would have been facilitated indirectly by knowledge gained because of the CIG.
Career development
From a career development perspective, the CIG has had a major impact. First it has allowed me to establish a long-term and relatively unique study population. Second it provided me with the proof of concept for two other major grants, an ERC Consolidator grant for €2 million, and a Science Foundation Ireland ERC Support grant for €300,000. The first of these ultimately led to my promotion to Professor and established my permanent faculty position at UCC, while the second allowed me to build a state of the art aviary for conducting behavioural experiments on great tits. The CIG also allowed me to indirectly develop a range of other projects on avian ecology, especially on seabirds, such that my group now numbers 19, including 5 postdocs, 2 research assistants, 9 PhD students, 2 MRes students and a Research Support Officer. In truth, this is larger than I had intended but arose because grant success rates were higher than I had anticipated. Building up this group has taken a lot of effort whilst delivering 140 hours teaching per year, but from now on I am expecting the reward from this effort to increase considerably in terms of publication outputs. Thank you Marie Curie!
http://ornithology.ucc.ie/current-projects/evolutionary-ecology-individual-variation-cognitive-performance-marie-curie-grant-334383/
The aim of this project was to characterize individual behavioral variation in innovativeness, learning and personality, and to explore a variety of causes and consequences of this variation for individuals in a population of great tits across a fragmented landscape of conifer and deciduous planted woodlands. The first major step was to setup a new study population in the Cork area by deploying 250 nestboxes across 9 woodland fragments and to establish a breeding population, which has been achieved. The second was to establish assays for measuring associative learning, innovativeness and personality to be conducted in captivity. The second two of these traits were straightforward but despite several attempts over several years, in the wild and in captivity, both in Cork and using a captive population at Heteren in the Netherlands, we could not achieve the first. Birds would not learn to associate specific colored keys with food rewards. This was surprising because we had already demonstrated that wild great tits in a UK population could do so, but we think reflects poor motivation in the population, and the fact that birds were too quick to give up trying, or that they did not mind being punished when they made mistakes and were happy to wait for awards at random. Although too late for this grant, in fact the experience gained over several years of trying has led to the development of other cognitive traits that are being used to good effect now under a different grant. The main results arising from this grant are based around individual measures of innovative lever pulling and the reactive proactive personality axis, which are summarized as follows.
One of the first main results showed that both personality and innovative problem solving in captivity were repeatable over lifetime but that consistency was context dependent. Specifically, innovativeness under standardized conditions did not predict innovativeness in a social competitive environment during an experiment in captivity. The study also showed that a range of environmental factors did not influence performance during the assay, including time in captivity, colorific intake prior to or during assays. A second significant result done in collaboration with colleagues at the Netherlands was that the personality measure ‘exploration behaviour’ (EB) did not predict whether individuals could innovate during social foraging trials but that the tendency to innovate over successful trials increased over time and was higher for slow explorers. In fact most birds could solve the lever pulling devices under these social circumstances, suggesting that the ability to solve these tasks is more widespread than originally thought, but that the use of this new foraging techniques is constrained by social context (social learning and shared vigilance). This is a significant result because it distinguishes between performance to innovate in isolation from the ability to do so a social context.
In another objective, we found no links between either innovative problem solving or personality and whether birds were trapped in deciduous or coniferous woodlands. After two years we already expected this would not be the case because gene flow was so high in the population. Although we found no variation between the habitats in personality or innovativeness, we did find remarkably different patterns of life history variation; for example, the ubiquitous decline in clutch size with lay date is absent in our deciduous sites, where clutch sizes and breeding success are lower because of low food availability as shown using cameras at the nest. This paper alone is a critically important description of the basic ecology of the species in this new study system, which is not only a unique description of life history variation across a fragmented landscape, but will also inform future research in my study population.
Small sample sizes meant we could not do a robust test of the hypothesis that innovativeness was linked to life history variation or indeed to fitness. However, we found negative reproductive selection on female personality, a relationship that was observed across both habitat types, and seemed to be caused by faster exploring females provisioning at a slower rate than slow explorers. Male exploration behaviour was negatively related to fledgling condition in deciduous fragments, a relationship that was reversed in coniferous fragments. Our results suggest that although the effects observed are context and sex specific, individuals at the slow end of this personality axis are favoured across our study system. Lack of progress on a cognitive trait has been disappointing but the link between personality, provisioning behavior and fitness is a significant advance in the field of personality, which was an integrative part of this project. Also recruitment was so low that it was impossible to generate a pedigree from the dataset. However, I conducted an animal model analysis on problem solving data collected by a former PhD student in another population, the first such analysis of its kind. There was little evidence for an additive genetic basis to the trait but instead we found evidence for the developmental stress hypothesis, because adults who were poor at solving problems had been born in oak-tree poor habitats and in high local population density areas. This is a significant advance because it is the largest scale study of its kind on a wild population.
For another objective, we experimentally showed that ‘tight sitting’ during incubation when faced with a predator correlated with both clutch size and reproductive success. Furthermore, amongst females that fled, nest failures increased with lay date. Our results suggest that personality is not linked to pace of life syndrome during this stage of the breeding season in females. Instead, a more immediate measure of risk-taking behaviour, tight sitting, predicted female reproductive investment and subsequent success in line with the pace of life theory. These results suggest that simple measures of personality and innovativeness do not necessarily predict important functional behavioural variation directly linked to life history variation, and provides an important cautionary note when expecting to find links between standard measures of personality or cognition, and functional behaviour.
In another experiment where birds had a choice of 3 kinds of food, we found that fast exploring great tits consumed a higher proportion of sunflower seeds, while slower individuals preferred peanuts. Problem solving performance was positively correlated with energy intake but not food preference. Peanuts accounted for a larger proportion of total daily energy intake for coniferous birds, which also lost more weight on average. Our results illustrate that a complex array of factors can determine foraging behaviour and success, including personality, innovativeness, state variables, time, and habitat origin, highlighting the challenge of explaining how selection acts on foraging performance over time. Finally, in another experiment, although we found no significant relationship between exploration behavior or innovativeness in isolation and competitiveness, individuals that were less competitive were more likely to spontaneously perform string-pulling behaviour during the dyadic trials, the first direct experimental demonstration of competitive exclusion leading to innovation, and therefore a significant advance. Our results support the hypothesis that innovations provide a means for less competitive individuals to access resources in line with the “necessity drives innovation hypothesis.
To date 6 papers linked to the project have been or are in the process of being published (under review, in revision etc) and at least another two are planned for the coming year. Together these have all helped advance our understanding of the causes and consequences of behavioural variation in a wild population, especially in the context of personality, and innovativeness. An additional 6 papers have been published in the related field during the course of the project and would have been facilitated indirectly by knowledge gained because of the CIG.
Career development
From a career development perspective, the CIG has had a major impact. First it has allowed me to establish a long-term and relatively unique study population. Second it provided me with the proof of concept for two other major grants, an ERC Consolidator grant for €2 million, and a Science Foundation Ireland ERC Support grant for €300,000. The first of these ultimately led to my promotion to Professor and established my permanent faculty position at UCC, while the second allowed me to build a state of the art aviary for conducting behavioural experiments on great tits. The CIG also allowed me to indirectly develop a range of other projects on avian ecology, especially on seabirds, such that my group now numbers 19, including 5 postdocs, 2 research assistants, 9 PhD students, 2 MRes students and a Research Support Officer. In truth, this is larger than I had intended but arose because grant success rates were higher than I had anticipated. Building up this group has taken a lot of effort whilst delivering 140 hours teaching per year, but from now on I am expecting the reward from this effort to increase considerably in terms of publication outputs. Thank you Marie Curie!
http://ornithology.ucc.ie/current-projects/evolutionary-ecology-individual-variation-cognitive-performance-marie-curie-grant-334383/