INTRODUCTION
Within populations, individual animals often exhibit behavioural differences that are consistent across time and context . Although this phenomenon, widely referred to as animal personality , has been demonstrated across very diverse animal taxa , the empirical literature is dominated by vertebrate studies . In comparison, personality variation in invertebrates generally, and aquatic invertebrates in particular, has been less well studied . This omission matters because understanding personality variation is not only central to fundamental research on animal behaviour, but also increasingly relevant across more applied fields such as welfare , ecotoxicology and invasion biology . Here we investigate the presence and structure of among-individual variation in the red cherry shrimp (Neocaridina heteropoda ). This is a popular ornamental species that, being easy to maintain and breed under laboratory conditions, is an emerging model for decapod research with relevance to ecosystem stability, ecotoxicology, and aquaculture .
From a fundamental perspective, among-individual variation in behaviour is widely assumed to have functional importance, impacting fitness via effects on survival and/or reproduction . While quantitative genetic studies have also shown that among-individual differences are heritable , understanding the evolutionary causes and consequences of this variation remains a challenge. Why does variation persist? Are among-individual differences adaptive, or do they reflect evolutionary constraints (e.g. trade-offs)? Why can’t all individuals adjust behaviour plastically to be optimal for the conditions they find themselves in? Hypothesised answers to these questions often invoke state-dependence, predicting that personality will covary with other traits (e.g. metabolic rate, body size) or intrinsic variables (e.g. age, sex) as a consequence of feedback between behaviour and state . For example, if small individuals face greater starvation risk, they may need to be ‘bolder’ (less risk averse) and more explorative to find resources faster, even if this behaviour increases predation risk . State-dependence means behaviour is likely to be integrated within broader life history strategies , and differ systematically between sexes . For example, in some systems males, on average, need to adopt riskier behaviour than females in order to obtain mating opportunities . Conversely, female behaviour may be selected for to reduce costs from male harassment .
Although evolutionarily motivated studies dominate the literature on animal personality , this phenomenon is increasingly recognised as having wider implications and applications. For example, Prentice et al. (2022) argues how the integration of personality traits with stress physiology means artificial selection on behavioural biomarkers could be used to improve welfare in fish aquaculture (see also e.g. . Ecotoxicology is another applied field in which the potential importance of among-individual differences in behaviour has been recently highlighted . For example, great tits (Parus major ) with high levels of lead in their blood and high levels of multiple metals in their feathers, showed lower explorative behaviours on average , while insecticide exposure lowered behavioural repeatabilities in spiders , reducing the relative importance of among-individual differences. Furthermore, found long-term fluoxetine exposure in guppies (Poecilia reticulata ) erodes variation in activity levels between individuals. Ubiquitous contamination may therefore impair behaviour and future adaptive potential of phenotypic variation to anthropogenic-induced alterations within both terrestrial and aquatic landscapes.
In decapod crustaceans, such as the species we investigate here, several applications of personality variation have been suggested. First, just as in fishes, personality traits may be relevant to welfare outcomes in captivity, which are under increasing scrutiny following recognition of sentience . Second, decapod behavioural change following sub-lethal exposure to environmental pollutants could contribute to bioassays relevant for monitoring ecosystem health and susceptibility of benthic and/or sediment dwelling invertebrates to pollutants . Third, since decapods demonstrate trait-biased dispersal, with bolder individuals outcompeting conspecifics, variation in boldness and activity may link to invasive success . Finally, many billions of decapod crustaceans are harvested from wild fisheries and raised in aquaculture systems for human consumption annually . Personality differences have been shown to predict trappability in decapods just as in vertebrates , and are also expected to be integrated with life-history traits important for production in aquaculture .
Here we characterise personality variation in the red cherry shrimpNeocaridina heteropoda (syn. N. davidi ), a small (<30mm) caridean species that is emerging as a model for pharmaceutical and ecotoxicological research . A short generation time and fast development also make it amenable to genetic studies and potentially a convenient model system for decapod aquaculture . Cherry shrimp are of commercial aquaculture importance themselves as an ornamental species . Unfortunately, release by aquarists combined with a wide tolerance of water and temperature parameters means they have become invasive outside their native range . Despite this, basic biological information on this species is scarce (but see e.g. . In particular there is a lack of baseline behavioural data that may, for example, impede use of ‘behavioural endpoints’ in ecotoxicology . Very little is currently known about the amount or structure of behavioural variation among-individuals. Nor is it known whether state-dependent behavioural variation, if present, is linked to intrinsic variables such as size or sex.
We focus specifically on ‘shy-bold’ variation , an aspect of personality that describes differences in behavioural response to (perceived) risk . We use two simple testing paradigms, Open Field Trials (OFT) and Food and Shelter trials (FST), coupled with multivariate behavioural phenotyping and a repeated measures design. Our specific aims are to (i) test for repeatable among-individual differences of the behavioural measures of boldness across the OFT and FST, (ii) determine whether the structure of multivariate behavioural variation observed is consistent with expectations given an underlying shy-bold among-individual axis, and (iii) determine whether size and/sex explains behavioural variation among-individuals.