Although the function of birdsong has been well studied in male songbirds, the function of female song is less well understood. This is partly due to a historical view of females occupying a passive role compared with males, which led to the neglect of female song. We examined the female song in Galápagos yellow warblers, Setophaga petechia aureola, and show for the first time that it is widespread in the form of solo songs as well as duets. We interrogated the seasonal patterns and functions of female song by carrying out simulated territory intrusions using playbacks of male, female or duet songs during breeding and nonbreeding seasons and conducted a territory retention survey for over a year. We measured the association between aggressive response and singing behaviour, sex-specific patterns of response and territory retention across years. Females sang mostly during the nonbreeding season and predominantly in male-led duets. Although females were strongly aggressive towards female song playback, they gave the weakest singing response to it. There was no association between female aggressive behaviour and song output in response to a simulated intruder. Moreover, the probability of territory retention across years was not explained by song output or aggression in response to intruders, though evidence for this was weak due to the small sample size. We suggest that female song in this year-round resident island system does not function for signalling aggression in territory defence or intrasexual competition but may have other functions.

The impact of feral cats on Australian wildlife contributes to species decline across the country. Reptiles play a significant role in feral cat diet, especially in the arid zone where reptile diversity and density are high. In this study, predator–prey interactions between feral cats and two arid-zone gecko species, the common knob-tail gecko, Nephrurus levis , and the beaded gecko, Lucasium damaeum , were investigated. Typically, N. levis responds to threats by bluffing, whereas L. damaeum flees. We hypothesized that these behavioural differences may affect their likelihood of avoiding predation. Using Perspex reptile enclosures in a large outdoor cat pen, the behaviour of gecko and feral cat was observed to assess how cats and geckos responded during and after interactions. Cats visited the Perspex enclosures with N. levis 11 times more than the control enclosures, and they had on average 15 s longer interactions. Reptile behaviour was compared before, during and after cat interactions. N. levis spent significantly more time in a stationary head-down hiding response during and after cat interactions, whereas L. damaeum showed no behavioural change. The longer length of interactions between cats and N. levis indicates that N. levis may be more susceptible to cat predation than L. damaeum despite their recognition of cats as a threat. Reptile body size showed no effect on the number of interactions with cats and interaction length, which indicates that other factors such as prey preference or posture may have been responsible for the increased interest towards N. levis . Bluffing behaviour was not observed possibly because of the enclosure, which prevents cats from approaching closer than 30 cm to geckos. However, in the wild, this behaviour could reduce predation impacts. Further research on reptile responses to feral predators is necessary to understand the interspecific variation in susceptibility of reptiles to cat predation. • Two arid geckos exhibited poor antipredator response to cats. • Gecko size had no influence on cat interactions. • Lucasium damaeum did not change behaviour in response to cats. • Cryptic predatory response for Nephrurus levis did not prevent detection by cats.

Extensive laboratory studies have shown that animals strongly devalue rewards with associated costs, such as time or effort. However, these laboratory discounting tasks have limited ecological validity, whereas in natural decision-making scenarios, animals may wait or expend more effort for a substantially better outcome. In our experiment, a discounting task in which wild grey squirrels, Sciurus carolinensis , in their natural habitat chose between climbing a short distance to receive a less preferred food or a longer distance to receive a preferred food was used. The squirrels exhibited relatively shallow discounting, slightly devaluing their preferred food with the increase of distance, compared with the steep declines observed in laboratory-tested animals. Social effects were also observed where less dominant squirrels showed steeper devaluation of the preferred food, which may be due to an increased risk of losing the reward as distance increased. Separately, squirrels were more likely to climb a longer distance in the presence of conspecifics, perhaps choosing the more time-consuming and effortful option to avoid increased competition for the easier, more immediate reward. The findings of this study indicate that understanding animals’ decision-making behaviour requires more ecologically valid approaches, such as considering the social context in which choices are made. • Grey squirrels show relatively shallow discounting in an ecologically relevant task. • Subordinate squirrels showed steeper discounting. • Squirrels exhibited shallower discounting in the presence of conspecifics.

Sentience is an increasingly popular topic in animal welfare and behavioural research. However, studying sentience can be difficult and contentious. Sentience has frequently been associated with ‘feelings’ and ‘subjective states’ not directly accessible to human observers; thus, investigating awareness in animals could be an alternative approach. Frameworks of awareness have been developed for the experimental examination of cognitive capacities that might underpin ‘feelings’ as affective states and subjective experiences. Using a framework of four hierarchically arranged levels of awareness (that is, perceptual, cognitive, assessment and executive), the awareness of an intertidal crustacean, the common hermit crab, Pagurus bernhardus , was assessed. Inspired by recent work on cognitive abilities and task solving in terrestrial hermit crabs, a series of four experiments was designed to test the ability of P. bernhardus to solve a simple conditional task: if the crab changes its shell, then it can escape confinement. Crabs were exposed to different levels of confinement, including complete confinement, confinement requiring shell changes for escape, confinement requiring no shell changes for escape and no confinement. Apart from testing crabs in the absence of an external motivation (experiment 1), crab responses were tested in the presence of food (experiment 2), an additional shell (experiment 3) and under hypoxic conditions (experiment 4) to ensure that individuals were motivated to escape. Crabs were more likely to escape confinement if they did not have to change shells. Sequence analysis of the behavioural patterns of crabs revealed that they are not tactically assessing different components of their confinement to aid in task solving. Collectively, these experiments indicated that P. bernhardus displays perceptual awareness without evidence of insight or forward planning. • Task-solving ability of hermit crabs was used to examine awareness and sentience. • Crabs were given an escape task, that is, changing shells to escape confinement. • Crabs were more likely to escape if a shell change was not required for escape. • Behavioural sequence analysis showed no evidence of forward planning. • Hermit crabs may be only perceptually aware.