Introduction
Animal personality, defined as inter-individual consistent differences in behaviour across time and context has received growing interest in the past two decades (e.g. Roche et al., 2016). Personality studies have important ecological and evolutionary implications determining such aspects as: space use, species geographic distributions, tendencies to be invasive, response to environmental change, speciation rates, social interactions and fitness consequences (e.g. Briffa and Weiss, 2010; Lichtenstein et al., 2017, 2017; Sih et al., 2004). Repeatability is a highly informative metric that provides a standardized estimate of consistency of individuality, that is, personality (Roche et al., 2016). However, several aspects of animal personality research currently face intense criticism. For instance, there is a underrepresentation of field studies compared to lab-based tests (Archard and Braithwaite, 2010). Also, the amount of studies on invertebrate personality is drastically disproportionate to the number of species and behavioural traits (e.g. Kralj-Fišer and Schuett, 2014). Further, most studies concerns the “Big Five” of animal personality (boldness, aggressiveness, sociability, exploration and activity), ignoring other traits that may bring in-depth understanding of the phenomena and the possible associations between commonly and rarely measured traits (Koski, 2014). Finally, metrics of the personality traits should be chosen with caution in order to be applicable for a given study organism and to represent ecologically relevant information (Carter et al., 2013).
Compared with the large number of laboratory experiments on captive and captive-bred animals, few studies have focused on personality in the wild (Archard and Braithwaite, 2010; Carere and Maestripieri, 2013; Fisher et al., 2015b; Hertel et al., 2020). This skew is unfortunate since lab-based experiments often are affected by a number of constraints: e.g. captivity stress, selective trapping, learning, homogeneity of the laboratory environment, artificial and relaxed selection and reduced pool of potential mates (reviewed by Archard and Braithwaite, 2010). As a consequence, individuals in laboratory conditions may behave in ways not representative for the natural environment and hence showing ecologically irrelevant behavioural patterns (Niemelä and Dingemanse, 2014).
Because of this potential for ecologically irrelevant behavioural patterns in the lab it is important to validate the relevance of findings in the lab with corresponding field studies. Studies designed to compare lab-based and field-based assessments of personality show different results, some fail to find any correlations between lab and field behaviour whereas others provide similar personality estimates in the two environments. For example, studies on crickets showed repeatability of exploration and activity both in laboratory and natural conditions. However, shyness was repeatable in artificial conditions only (Fisher et al., 2015b). Study on zebra finches showed personality both in laboratory and field conditions, but there was no correlation between the two situations (McCowan et al., 2015). In a recent example on sea anemones, Osborn and Briffa (2017) showed that the transition from field to laboratory environment might influence personality assessments. Results from this kind of transplant experiments can therefore be biased by the translocation itself (Niemelä and Dingemanse, 2017). On the other hand, studies on great tits (Cole and Quinn, 2012) or striped mice (Yuen et al., 2016) showed that individuals behaved consistently both in the laboratory and field. The above examples denote that well-designed assessment of personality in a model organism measured in natural field conditions may be of high applicability for understanding the ecological and evolutionary consequences of animal personality (Wolf and Weissing, 2012) in natural populations (Archard and Braithwaite, 2010; Osborn and Briffa, 2017).
Despite the fact that invertebrates represent the most numerous group of animals on Earth (Larsen et al., 2017; Stork, 2018), personality studies on this taxa are still scarce when compared to studies on vertebrates (Gosling, 2001; Kralj-Fišer and Schuett, 2014; Mather and Logue, 2013). However, in recent years insects started to play an important role in animal behavioural research (Keiser et al., 2018). This is because insects’ sexual and social behaviours represent a great variety of phenomena that are rare or absent in vertebrates, providing new possibilities for addressing ecological or evolutionary questions on personality causes and consequences (Carere and Maestripieri, 2013). Also studies on insects in many instances are less ethically controversial and less time consuming because of a shorter lifespan when compared to vertebrates (Córdoba-Aguilar et al., 2018). However, studies on insect personality in natural conditions without handling and captivity trauma are rare (e.g. Fisher et al., 2015b).
Beyond ‘ The big five’ (boldness, aggressiveness, exploration, activity and sociability) which became the blueprint for animal personality studies (Mather and Logue, 2013; Réale et al., 2007; Van Oers and Naguib, 2013), we have limited understanding of other personality aspects. For example, behaviours related to mating have an extraordinary role in species ecology and evolution, but these behavioural traits have received relatively little attention in animal personality studies (Koski, 2014). For instance, The term ‘sociability‘ is being used as a proxy for a whole range of behaviours. These include: hiding in presence of a conspecifics’ smell (Cote et al., 2008), grooming in chimpanzees (Koski, 2011), aggregation at food sources in fruit flies (Scott et al., 2018), tendency to shoal in mosquitofish (Brodin et al., 2019) and mating behaviours (Sih et al., 2014). It is possible that the same test/metric may actually measure different traits in two different species (Koski, 2014). For instance, testing response to a predator (as a proxy of boldness) in open areas, which is used for e.g. kangaroos (Blumstein and Daniel, 2003), may not be adequate for a passerine, which usually inhabits, and are preyed upon, in more closed habitats (Whittingham et al., 2004). Also, multiple tests of one personality trait, could be of higher validity in describing a given trait (Carter et al., 2013), as has been done on guppies (Poecilia reciculata ) where boldness was measured in three experiments (Burns, 2008).
Already established model organisms (i.e., non-human species representing a larger group of organisms used for comparative and integrative research on specific scientific problems, Leonelli and Ankeny, 2013), intensively bred and studied under lab-conditions over several decades, have their limitations and may not be very useful as models for some lines of research. For instance, one of the most significant model organisms, the fruitfly (Drosophila melanogaster ), intensively used for testing molecular mechanisms of behaviour (Kain et al., 2012; Roberts, 2006; Sokolowski, 2001), has been reared for many generations in homogeneous, non-natural, environments of molecular biology laboratories. This have more than likely resulted in the species adaptation to stable environments and a change in its behavioural reaction to novel conditions when compared to natural populations (Archard and Braithwaite, 2010). For example, a recent study of how anxiolytic pharmaceuticals can affect zebrafish behaviour showed that the behaviour of wild zebrafish was changed by the exposure to the anxiolytic, while the lab-reared zebrafish was unaffected (Vossen et al., 2020). Hence, to increase the ecological relevance of studies including behavioural traits we need to both expand the number and taxonomic breadth of model organisms, as well as re-stock or replace existing lab-populations (Behringer et al., 2009; Leonelli and Ankeny, 2013).
Here, we report a set of behavioural traits tested for their repeatability over time and contexts in the damselfly Calopteryx splendens measured in natural field conditions. We measured traits related to three behavioural axes: (i) courtship behaviour, (ii) aggressiveness and (iii) boldness. Since this is the first study onC. splendens personality in the wild, we test three traits within each behavioural axis to make sure they are applicable to this study system (Carter et al., 2013). The repeatability was assessed in two different contexts: on undisturbed original patches (males’ territories) and after partial deterioration of a territory. Our results indicate that C. splendens is an excellent model for studies on animal personality and behavioural syndromes in nature.