Lauriane Bégué

Reliable individual identification is fundamental to ecological and behavioural studies, yet many marking techniques are invasive and may affect animal behaviour or survival. Biometric identification approaches offer a non-invasive alternative for individual recognition using phenotypic traits. In amphibians, body patterns, especially colouration, have been used to identify individuals, but species lacking permanent body patterns require alternative traits. This study evaluated whether iris patterns could serve as a marker for individual recognition in the treefrog Osteocephalus oophagus . A capture-recapture study was conducted over three breeding seasons (2021–2023) in French Guiana. At each encounter, eye photographs were compared with an image library to identify individuals without handling. Manual identifications were independently verified using Wild-ID software. Across 392 capture events involving 103 individuals, iris patterns were highly distinctive among individuals and enabled repeated identification across encounters. Within-season recapture rates ranged from 56% to 80%, indicating that iris patterns remained sufficiently stable for short-term monitoring. However, only three individuals were re-identified across years, preventing a robust assessment of long-term pattern stability. As these individuals showed no substantial alteration in iris patterns, the low interannual recapture rate was more likely explained by population turnover or emigration than by changes in iris morphology. The findings suggest that iris recognition provides an effective, non-invasive method for short-term individual identification in O. oophagus , reducing the need for repeated handling and supporting ethical wildlife research. Further studies are needed to assess the long-term stability of iris patterns and evaluate the applicability of this method to other amphibian species.

Environmental conditions during early development can considerably shape individual behavior, and laboratory studies suggest that predator presence is an important factor influencing both within- and among-individual behavioral variation. To test this under natural conditions, we studied neotropical poison frog (Allobates femoralis) tadpoles from pools with and without dragonfly larvae (tadpole predators) under naturally varying environmental conditions. We measured 2 behavioral traits, distance moved in a novel environment and time to emerge from a shelter, across 4 repeated trials in predator-exposed (n = 48) and predator-naïve (n = 169) tadpoles. Predator-exposed tadpoles moved less, emerged later in their first trial and were more unpredictable in emergence time than predator-naïve individuals. Moreover, among-individual variation in both behaviors was higher in predator-exposed tadpoles, although behavioral repeatability did not differ significantly between predator-exposed and predator-naïve groups. Predator-naïve individuals exhibited behavioral and predictability syndromes, individuals that moved more also emerged sooner, and individuals more predictable in 1 behavior were more predictable in the other. Despite only being significant in predator-naïve individuals, these correlations did not differ between predator-exposed and predator-naïve groups. A personality–plasticity association occurred only in predator-exposed tadpoles: individuals that initially moved less showed less behavioral change across trials. These findings demonstrate that even in naturally variable environments, predation risk shapes behavioral variation at multiple levels, thereby contributing to the emergence and maintenance of “animal personality”.

Background: Parental care is costly for the caregiver. Therefore, parents should be able to discriminate between their own and conspecific offspring to avoid costly misdirected care. Infanticide, the intentional killing of conspecific young by adult individuals, occurs in many animal taxa. It has been shown to have several benefits for the perpetrator, such as increasing mating opportunities, ensuring dominance, or reducing a competitor’s fitness; but infanticide may also minimise risks of misdirected parental care. Previous studies in Allobates femoralis, a poison frog with parental care, have shown that males transport all tadpoles present in their territory to water bodies, regardless of whether they have sired the clutch or not. However, when taking over a new territory, males cannibalise clutches from the previous territory holder. These findings raise the question as to which factors actually shape parental care and infanticidal behaviour in male A. femoralis after territory takeover. To answer this question, we designed a laboratory experiment, in which we tested males with different territorial status and recent mating activity. We recorded tadpole transport and cannibalism and compared the occurrence of these two behaviours across our different experimental conditions. Results: We found that territory ownership, relatedness to clutches, and possibly also recent mating activity influenced parental behaviours. However, we were unable to clearly disentangle the factors influencing cannibalistic behaviours. Our results also confirmed that males use territorial recognition to discriminate between their own and unrelated offspring, and that they commit infanticide likely to avoid misdirected parental care. Conclusions: Transport and cannibalism appear to be impacted by several factors in different ways. We found that the territorial status and relatedness to the clutch both influence parental behaviours in male poison frogs, whereas the factors influencing infanticidal behaviours remain unclear. Therefore, transport and cannibalism appear to be two independent processes, and factors influencing one behaviour do not necessarily affect the other. Further studies should investigate associated neuroendocrine changes, to better understand the mechanisms underlying parental and infanticidal behaviour in poison frogs. Our findings suggest that the decision-making processes involved in tadpole transport and clutch cannibalism appear to be more complex than previously thought.

Animals need to recognize different individuals, both con- and heterospecifics, to make appropriate decisions. In the wild, responses to familiar individuals may vary depending on the context, which can be beneficial. However, differing responses towards human experimenters can influence experimental outcomes. Such effects might be particularly overlooked in reptiles which are frequently viewed as cognitively less advanced. We tested Tokay geckos’ (Gekko gecko) ability to distinguish between familiar and unfamiliar handlers in two situations: in a novel situation (exerting physical constraint) and a routine situation (feeding from forceps as during regular husbandry). Geckos showed sex-specific differences towards familiar and unfamiliar handlers in a routine situation, but not in a novel situation, in which they showed individual repeatability. Our results further advance our understanding of reptile cognition revealing important insights into context specific responses in relation to handler identity with implications for experimental animal studies that are rarely considered.