Octopus vulgaris Exhibits Interindividual Differences in Behavioural and Problem-Solving Performance

Abstract

By presenting individual Octopus vulgaris with an extractive foraging problem with a puzzle box, we examined the possible correlation between behavioural performances (e.g., ease of adaptation to captive conditions, prevalence of neophobic and neophilic behaviours, and propensity to learn individually or by observing conspecifics), biotic (body and brain size, age, sex) and abiotic (seasonality and place of origin) factors. We found more neophilic animals showing shorter latencies to approach the puzzle box and higher probability of solving the task; also, shorter times to solve the task were correlated with better performance on the individual learning task. However, the most neophilic octopuses that approached the puzzle box more quickly did not reach the solution earlier than other individuals, suggesting that strong neophilic tendency may lead to suboptimal performance at some stages of the problem-solving process. In addition, seasonal and environmental characteristics of location of origin appear to influence the rate of expression of individual traits central to problem solving. Overall, our analysis provides new insights into the traits associated with problem solving in invertebrates and highlights the presence of adaptive mechanisms that promote population-level changes in octopuses' behavioural traits.

Keywords: behavioural plasticity; cephalopods; extractive foraging; innovation; neophilia; octopus; problem solving; seasonality; social learning.

Figure 1

Schematic representation of the experimental plan and behavioural assays utilized in this study; asterisks indicate the feeding days (Adapted from Borrelli et al. 2020 [46]).

Figure 2

Boxplots for the estimated age (a), brain size (b), neophilia (c), and social learning scores (d) for O. vulgaris captured in autumn–winter or spring–summer. Blue dots represent individuals’ scores.

Figure 3

Boxplots of the brain size (a) and the Individual Learning score (b) of O. vulgaris captured in the S. Lucia-Circolo Posillipo or the Donn’Anna-Nisida areas in the Bay of Naples (Mediterranean Sea, Italy; see Supplementary Materials for details about the sites of capture). Blue dots represent individuals’ scores.

Figure 4

Spider plots (a) representing the score (normalized within the 0–100 range) of each variable in the two clusters. Barplots (b) of the corresponding means and standard error of the means (see the Supplementary Materials for the descriptive statistics of the two clusters). Asterisks indicate significant differences between clusters (p < 0.050).