Weak evidence that asset protection underlies temporal or contextual consistency in boldness of a terrestrial crustacean

Abstract

In recent years, many studies have investigated the potential state dependence of individual differences in behaviour, with the aim to understand the proximate and ultimate causes and consequences of animal personality. Among the potential state variables that could affect behavioural expression is size and mass, but few studies have found associations at the among-individual levels. Insufficient sampling and incorrect analysis of data are cited as impediments to detecting correlations, if they exist. Here, we conducted a study using 100 pillbugs (Armadillidium vulgare) and assayed their defensive behaviour 24 times each over time and across familiarity contexts, to test the asset protection hypothesis that predicts a negative correlation between boldness and mass, and with increases in mass over time. Multivariate mixed models revealed that despite mostly consistent individual behavioural differences over time (modest slope variance) and across contexts (near-parallel reaction norms), and 18-fold range in starting mass, there was no correlation between individual mean mass and boldness. However, individuals that gained more mass over time may have been more ‘shy’ compared to those gaining less mass, but the correlation was weak and observed variation in mass gain was small. There was also a mean level trend of increasing shyness over time that was coincident with mean level mass increases over time. Together, our study provides weak evidence for the asset protection hypothesis, whereby individuals that accumulate more resources are thought to protect them through risk averse behaviour. Significance statement Individual variation in ‘state’, such as mass or energy reserves, is thought to be a predictor of individual differences in behaviour that are consistent over time. However, few studies reveal such links, and several studies suggest insufficient sampling may explain null results in most studies. We studied 100 animals sampled 24 times each in a controlled setting to reveal stable individual differences in mean behaviour over time and across contexts; however, individual behaviour was unrelated to large differences in individual mass but weakly related to increases in mass through time whereby individuals became more shy and those growing faster were somewhat more shy. Our results provide little evidence for the asset protection hypothesis.

Time‐specific convergence and divergence in individual differences in behavior: Theory, protocols and analyzes

“…Similarly, investigators who have tested for temporal changes in individual differences using random regression models typically estimate the correlation between elevation and slope (CORRe t s) at just one point in time (e.g., Beveridge et al, 2022 ; Biro et al, 2014 ; Class & Brommer, 2016 ; Dingemanse, Bouwman, et al, 2012 ; Martin & Reale, 2008 ; Mathot et al, 2012 ; Thys et al, 2021 ). This practice might be adequate for detecting convergence or divergence if either of those patterns was sustained from the beginning to the end of the study (e.g., patterns indicated in Figure 1 ).…”

Phylogenetic meta-analysis reveals system-specific behavioural type–behavioural predictability correlations

“…Hence, the two traits are often (in 66% of assessed cases) not independent, but their relationship is species-specific. Empirical studies using cut-edge methodology to reveal covariation between components of between- and within-individual behavioural variation started to accumulate lately [ 9 , 10 , 22 – 27 , 47 , 77 , 117 ], but the evolutionary analysis of behavioural predictability is still in the pioneering stage. We hope that our synthesis highlights the need for additional empirical research applying advanced experimental design (>6 number of individual repeated assays coupled with robust statistical methods).…”

“…[ 46 ] Taeniopygia guttata Sunda zebra finch Aves Beveridge et al . [ 47 ] Armadillidium vulgare common pill bug Isopoda Beyts et al . [ 48 ] Engystomops pustulosus tungara frog Amphibia Biro and Adriaenssens [ 10 ] Gambusia holbrooki eastern mosquitofish Actinopterygii Brand et al .…”

Predation cues amplify the effects of parasites on the personality of a keystone grazer

“…The 95% credible interval (95% CI) for the posterior probability distribution of the effect is reported. Fixed effects and correlation parameters with credible intervals not overlapping zero are considered significant (Beveridge et al, 2022; Bürkner, 2017; Roth & Sterck, 2020).…”

“…The aim was to determine which effects were significant and supported for inclusion into the MV model (c.f. Beveridge et al, 2022). Morphological measurements were strongly collinear and were analysed separately.…”