Rapid spatial learning in cooperative and non-cooperative cichlids
Introduction
Many animals are “asocial” and live with or interact with other individuals for only a short period of their lives (e.g. while mating or caring for young), while others spend their entire lives with the same relatively stable group of individuals (Kutsukake, 2009, Ward and Webster, 2016). The Social Intelligence Hypothesis posits that animals living in conditions with greater social complexity have evolved enhanced cognitive abilities to cope with the difficulties of social life (Byrne, 1994, Byrne and Whiten, 1988, Holekamp, 2007). Some versions of the Social Intelligence Hypothesis suggest that social living favours social cognition specifically, while other versions argue that social living has a broader impact on cognition (Byrne and Whiten, 1988, Whiten and Byrne, 1997, Whiten, 2000). Much work has been done to investigate the relationship between social complexity and brain size evolution; however, the critical cognitive mechanisms that co-evolved with group-living and enable a highly social lifestyle have not received the same attention (Ashton et al., 2018, Johnson-Ulrich, 2017, Kummer et al., 1997, Reader and Laland, 2002).
The most tightly knit and well-coordinated group-living species are cooperatively breeding animals. In cooperatively breeding social groups, subordinate group members aid dominant group members in the care of the dominant’s young (Solomon and French, 1997). Researchers have argued that cooperative breeding requires that individuals recognize their own group-members, remember past interactions, and use this information to inform future behaviour (Iwaniuk and Arnold, 2004, Reddon et al., 2016, Thornton and McAuliffe, 2015). Hence, social memory and cheater detection (the ability to discern whether group-members are performing tasks that benefit the group) are thought to be important cognitive traits in the evolution of cooperation (Burkart and van Schaik, 2010, Dugatkin, 2002, West et al., 2007). Cooperative breeding also requires the formation of strong and stable social bonds and the ability to resolve conflict within a group (Balshine et al., 2017, Hick et al., 2014, Reddon et al., 2019). A hypothesis known as the Cooperative Breeding Hypothesis suggests that social challenges are especially pronounced in cooperatively breeding groups, as these animals often live in strict hierarchies for which they must remember their relative rank, and monitor the rank and contributions of others (Burkart and van Schaik, 2010, Iwaniuk and Arnold, 2004, Thornton and McAuliffe, 2015). Thus, according to this second hypothesis, cooperative breeders are expected to have highly developed socio-cognitive abilities (Burkart and van Schaik, 2016). Our aim was to investigate whether the enhanced socio-cognitive ability expected to develop in highly social species according to the Cooperative Breeding Hypothesis, might also extend to other domains, giving these highly social species an advantage when performing other cognitive tasks.
Spatial navigation is one such cognitive task and is a key requirement for foraging, migration and predator avoidance; activities that are directly linked to fitness and can have a social component (e.g. local enhancement in foraging; Burns and Rodd, 2008; Fagan et al., 2013; Fukumori et al., 2010; Pravosudov and Roth, 2013). The degree of habitat complexity that an animal needs to contend with is known to shape the brain (Carbia and Brown, 2019). In small mammals, like North American sciurids (e.g. squirrels), arboreality is the best predictor of relative brain size and is highly correlated with habitat (Budeau and Verts, 1986). In fishes, examples linking brain and habitat come from cichlid fishes, where species inhabiting structurally complex underwater environments often have larger telencephalons (Ectodini clade; Pollen et al., 2007), and the same has been shown for sunfish (Lepomis gibbosus; Axelrod et al., 2018), and sticklebacks (Pungitius pungitius; Gonda et al., 2009). The telencephalon is part of the teleost forebrain, and hosts the lateral telencephalic pallium, an area thought to represent the fish homologue of a hippocampus—the brain structure implicated in spatial learning and memory of mammals and birds (Durán et al., 2010, Rodrıguez et al., 2002, Vargas et al., 2009). Studies that lesion or ablate part of the telencephalon in fishes confirm its prominent role in spatial learning (Broglio et al., 2003, Riedel, 1998), but the telencephalon is also involved in the regulation and expression of social behaviour (Flood and Overmier, 1981, Scace et al., 2006). Although the role of habitat complexity in sculpting the fish brain is well established (Gonda et al., 2009, Kotrschal and Taborsky, 2010, Salvanes et al., 2013, Strand et al., 2010, White and Brown, 2015), little is known about how the demands of complex group-living (i.e. social complexity) might shape the fish brain. Also, cognitive differences or advantages might not manifest as visible changes in brain morphology or brain size, which makes understanding the connection between social complexity, and other cognitive abilities like spatial cognition, particularly difficult.
Sex also impacts the brain through differential expression of key hormones such as testosterone and estrogen, which have implications for things like motivation (Becker and Taylor, 2008) and metabolism (Hewitt et al., 2003), and consequently behaviour and/or cognition. In humans, for example, males often outperform females on dynamic 3-dimensional spatial tasks and these sex differences are attributable to proximate biological mechanisms (e.g. sex hormones) and developmental mechanisms (e.g. play patterns; Geary, 1995). Recent research suggests that differences in wayfinding and navigational strategies in humans are further influenced by environment and experience (Livingstone-Lee et al., 2014), and that females are better able to use mapping strategies (allocentric methods i.e. remembering elements in the environment, allowing the formation of real-world representations) in order to effectively navigate their environment. That said, multiple strategies are used in tandem and are frequently switched, integrated, and combined (Fernandez-Baizan et al., 2019). In fishes, a number of previous studies have found sex differences in performance during spatial challenges (in the guppy Poecilia reticulata, Reader and Laland, 2000, in the cichlid Astatotilapia burtoni, Wallace and Hofmann, 2021, as well as other fish species, Costa et al., 2011). However, sex differences in non-spatial tasks are seldom observed (e.g. discrimination of food quantities, object recognition memory; Lucon-Xiccato and Bisazza, 2017a, Lucon-Xiccato and Bisazza, 2017b). Similar to humans, sex differences in the use of navigational strategies have also been described in a number of fish species (Salena et al., 2021). As such, we might expect to find sex-specific differences in spatial abilities for other fishes.
Here, we describe a comparative study that assessed whether spatial learning and memory performance differed between three cooperatively breeding cichlid fish species and three of their non-cooperative relatives, and further investigated for sex differences. All six species were Lamprologini cichlids, a tribe of fish from Lake Tanganyika in Africa (Day et al., 2007) and a clade that has evolved group-living and cooperative breeding on five separate occasions (Dey et al., 2017, Reddon et al., 2017). While many Lamprologini species live in social groups, rely on conspecific group members for protection and cooperate to raise young, other closely related Lamprologini species rarely interact with conspecifics (apart from their mated partners or with a neighbour during a territorial standoff). These less social species do not cooperate, nor do they form permanent groups (Balshine et al., 2017, Hick et al., 2014). Using a maze learning paradigm with repeated trials to assess spatial learning and memory, we tested the following six territorial Lamprologini cichlids: the three cooperatively breeding cichlid species were Neolamprologus pulcher, Neolamprologus multifasciatus and Julidochromis ornatus, and the three non-cooperative species were Telmatochromis temporalis, Altolamprologus compressiceps and Neolamprologus tretocephalus (see Supplementary Fig. 1 for additional details on the evolutionary relations of these species and the morphological measures of the individuals used). These fishes have fairly comparable habitats, and can all be found in shallow rocky areas of Lake Tanganyika (Barlow, 2008, Brichard, 1989, Konings, 1998). Hence, these closely related but socially diverse species offer a powerful model system to explore how social living molds the brain and cognitive abilities.
We hypothesized that the cooperative species would initially outperform the non-cooperative species, and improve more or get faster over repeated trials. We reasoned that because cooperatively breeding species must cope with the cognitive demands of social life, this lifestyle might make them better problem solvers in other aspects of cognition, and lead to enhanced spatial performance in the maze. Furthermore, because animals that live in groups often subdivide their territories to avoid conflict over space-use (Effenberger and Mouton, 2007, Schradin and Lamprecht, 2002, Werner et al., 2003), we posited that group-living and cooperatively breeding species may require a more detailed delineation of their territorial space—to adhere to their individual sub-territories. Thus, we expected that this would lead to better spatial performance in the cooperative species. Additionally, in many cichlids including N. pulcher, males have larger home ranges and disperse sooner and faster than females (Desjardins et al., 2008, Stiver et al., 2007, Wong et al., 2012) and we therefore predicted that males across all six species would complete the maze faster than females.
Section snippets
Animals and housing conditions
All fish were housed in the Aquatic Behavioural Ecology Laboratory at McMaster University, Hamilton, Ontario, Canada. Morphological information regarding our study specimens can be found in the supplementary material (see Supplementary Tables 1 and 2). N. pulcher and T. temporalis were the descendants of wild caught fishes from Lake Tanganyika, Africa, while N. multifasciatus, A. compressiceps, J. ornatus and N. tretocephalus were purchased from a commercial aquarist supplier (Finatics,
Social system
The three cooperative species did not complete the maze any faster than the three non-cooperative species (Fig. 2a; log-LMM: 2 = 2.07, df = 1, p = 0.15). Both cooperative and non-cooperative species completed the maze faster across trials (log-LMM: 2 = 30.70, df = 2, p < 0.001; Supplementary Fig. 2a cooperative, 2d non-cooperative), and there was no significant difference in their rates of improvement i.e. in the reduction in time taken to complete the maze across trials (Social System*Trial
Discussion
Contrary to our predictions, we did not find compelling evidence that cooperative species outperformed their non-cooperative relatives in the spatial task. Both cooperative and non-cooperative cichlids took less time to complete the maze following the first trial. Cooperative species did not display better inhibitory control, nor did they make fewer mistakes than their non-cooperative relatives. Overall, our results suggest that the challenges of group-living and cooperation do not offer any
Ethical note
All protocols were approved by the Animal Research Ethics Boards of McMaster University (Animal Utilization Protocol No. 18–04–16) and complied with the guidelines established by the Canadian Council on Animal Care (CCAC) regarding the use of animals in research and teaching.
Funding
This work was supported by a Natural Sciences and Engineering Research Council of Canada (NSERC) Discovery grant to Sigal Balshine (RGPIN-2016–05772). Matthew G Salena was supported by an Ontario Graduate Fellowship.
CRediT authorship contribution statement
Matthew G. Salena: Conceptualization, Methodology, Formal analysis, Investigation Writing – original draft, Writing – review & editing, Visualization. Angad Singh: Methodology, Investigation, Writing – review & editing. Olivia Weller: Investigation, Writing – review & editing. Xiang Xiang Fang: Investigation, Writing – review & editing. Sigal Balshine: Conceptualization, Methodology, Writing – review & editing, Supervision, Funding acquisition.
Acknowledgements
We would like to thank Brett Culbert, Reuven Dukas, Jonathan Pruitt, Andy Turko, Hossein Mehdi, Caitlyn Synyshyn, and the reviewers for helpful comments on the manuscript. We also express our gratitude to Greaton Tan for his assistance in designing the experimental tank figures. Lastly, we would like to thank the attendees and organizers of the McMaster Data Lunch, especially Michael Li, Ben Bolker, Jonathan Dushoff and Ian Dworkin, for their help regarding the statistical analyses.
References (103)
- et al.
Social motivation and conflict resolution tactics as potential building blocks of sociality in cichlid fishes
Behavioural Processes
(2017) - et al.
Social cognition in fishes
Trends in Cognitive Sciences
(2014) - et al.
Hastiness, brain size and predation regime affect the performance of wild guppies in a spatial memory task
Animal Behaviour
(2008) - et al.
Phylogenetic relationships of the Lake Tanganyika cichlid tribe Lamprologini: the story from mitochondrial DNA
Molecular Phylogenetics and Evolution
(2007) - et al.
Costs and benefits of polygyny in the cichlid Neolamprologus pulcher
Animal Behaviour
(2008) - et al.
Lateral but not medial telencephalic pallium ablation impairs the use of goldfish spatial allocentric strategies in a “hole-board” task
Behavioural Brain Research
(2010) - et al.
Spatial memory in young adults: Gender differences in egocentric and allocentric performance
Behavioural Brain Research
(2019) - et al.
Factors affecting the measure of inhibitory control in a fish (Poecilia reticulata)
Behavioural Processes
(2018) Sexual selection and sex differences in spatial cognition
Learning and Individual Differences
(1995)- et al.
Social behavior functions and related anatomical characteristics of vasotocin/vasopressin systems in vertebrates
Brain Research Reviews
(2001)