Does novelty influence the foraging decisions of a scavenger?

Introduction

Knowledge about the immediate environment is essential for survival. Animals navigate their surroundings and gather crucial information on food sources, mating partners, shelter, and predators (Mettke-Hofmann, Winkler & Leisler, 2002; Dall et al., 2005; Moretti et al., 2015; Sarkar & Bhadra, 2022), often driven by their exploratory behaviour. An enhanced exploratory behaviour may further lead to innovative problem-solving (Wat, Banks & McArthur, 2020; Klump et al., 2022). Thus, exploratory behaviour directly or indirectly influences the survival and reproduction of animals. However, neophobia, or avoiding objects or other environmental aspects, such as space, food, etc., solely because they have never been experienced, can constrain exploratory behaviour (Stöwe et al., 2006a). While a suppressed tendency to explore can decrease the risk of encountering predators, it can also substantially limit opportunities to discover novel resources, like food (Stöwe et al., 2006b). Animals inhabiting urban and semi-urban (i.e., human-dominated) environments experience a lower predation pressure than those in rural and wild habitats (Fischer et al., 2012; Eötvös, Magura & Lövei, 2018). Therefore, reduced object-neophobia in urban-dwelling animals can enhance exploratory behaviour, providing benefits (but see, Meddock & Osborn, 1968). A long-standing bias of scientists, especially behavioural ecologists, has been to ignore wildlife in urban environments (Magle et al., 2012), particularly ‘subsidised’ animals that ‘exploit’ anthropogenic food resources. Consequently, our knowledge of these animals’ neophobic behaviour is obscured. In recent years, with the formalisation of urban ecology, scientific research has developed more interest in understanding the behavioural and cognitive aspects of decision-making of animals living close to humans.

Human-dominated environments impose novel challenges (Ditchkoff, Saalfeld & Gibson, 2006), which can be perceived differently by animals that are domesticated, non-domesticated synanthropes, and non-domesticated but coerced to live around humans (Beckman, Richey & Rosenthal, 2022). While the extent of damage by the Anthropocene is irrevocable (Wagner et al., 2021), a wide range of animals exhibit behavioural adaptations and plasticity in traits to benefit from human-induced rapid environmental changes (Sih et al., 2010). Reduced object-neophobia and enhanced exploratory behaviour are such behavioural adaptations (Griffin, Netto & Peneaux, 2017). However, it is still unresolved to what extent neophobia is repeatable (i.e., a personality trait) or plastic, due to inconsistency in definition and objective assessments (Takola et al., 2021; Kimball & Lattin, 2023), but see, (Day et al., 2003; Quinn et al., 2009; Grunst et al., 2019). The problem increases further when neophilia, or the preference for novelty over familiarity (Day et al., 2003), is considered the opposite of neophobia. Although neophilia can promote exploratory behaviour (Day et al., 2003), object-neophobia and neophilia are thought to be shaped by different selective forces, and they do not necessarily represent two extremes of a continuum (Greenberg & Mettke-Hofmann, 2001). Most importantly, neophobic and neophilic responses are highly context-specific, with studies typically testing neophilia in neutral and neophobia in non-neutral, such as foraging contexts (Takola et al., 2021). Empirical evidence suggests that urban-living animals have reduced object-neophobia (Tryjanowski et al., 2016; Greggor et al., 2016; Jarjour et al., 2019; Biondi et al., 2020; Miller et al., 2022, but see Mazza et al., 2021), which, with or without the presence of neophilia, can drive their enhanced exploratory behaviour (Griffin, Netto & Peneaux, 2017; Thompson et al., 2018; Breck et al., 2019; Dammhahn et al., 2020). Despite a wealth of knowledge gathered by empirical studies on these traits, our understanding of the ecologically relevant contexts under which such traits can be expressed and/or be plastic is lacking (see Gordon, 2011). Consequently, testing more species—that live in close proximity to humans—in ecologically relevant contexts, such as foraging, is imperative.

Dogs (Canis familiaris) inhabit a wide range of habitats, from living as pets in human households to roaming freely in human-dominated environments. Unlike pets, free-ranging dogs are primarily under natural and sexual selection pressures (Range & Marshall-Pescini, 2022); thus, acquiring information about their surroundings is key to survival. Several recent studies have provided evidence that free-ranging dogs’ socio-cognitive abilities, from understanding human cues, attentional states and facial expressions to forming trust with and learning socially from humans, have made them successful in human-dominated environments (Bhattacharjee et al., 2017b; Bhattacharjee et al., 2017c; Bhattacharjee et al., 2019; Bhattacharjee, Sau & Bhadra, 2018; Bhattacharjee, Sau & Bhadra, 2020; Brubaker et al., 2019; Bhattacharjee & Bhadra, 2020; Bhattacharjee & Bhadra, 2022; Lazzaroni et al., 2023; Cimarelli et al., 2023). However, in contrast to socio-cognitive skills, little is known about individual traits, such as exploratory behaviour and their contribution to free-ranging dogs’ success in human-dominated environments. These dogs spend considerable time and energy walking and foraging solitarily (Sen Majumder, Chatterjee & Bhadra, 2014), which could potentially include encountering unknown human artefacts, and if that explored, can be rewarding (such as packaged containers, garbage bins, etc., D Bhattacharjee, A Bhadra, pers. obs., 2010–2024). Therefore, reduced object-neophobia can be beneficial for free-ranging dogs. Pet and captive pack-living dogs exhibit little to no neophobia and enhanced neophilia (Kaulfuß& Mills, 2008; Moretti et al., 2015) in choice tasks when provided with familiar and unfamiliar objects. The presence of group members in captive pack-living dogs further induces exploratory behaviour, indicating risk-sharing (Moretti et al., 2015). However, considering early age classes (five, six, and eight weeks of age), a dog-wolf comparative study suggests that wolves are more persistent in exploring a novel environment and novel objects when compared with dogs (Marshall-Pescini et al., 2017a). While various scavenging strategies of free-ranging dogs have been reported in the past, from following simple rules of thumb to extractive foraging techniques (Mangalam & Singh, 2013; Bhadra et al., 2015), how these dogs respond to novelty during foraging and whether their exploratory behaviour and foraging decisions are constrained or influenced by object-neophobia is unknown.

We conducted an ecologically relevant object-choice test with free-ranging dogs to investigate whether their foraging decisions depend on (or are constrained by) object-neophobia. Free-ranging dogs encounter plastic garbage bags and pouches and extract food leftovers from them (Bhattacharjee et al., 2017a). On the contrary, plastic balls are unlikely to be encountered during scavenging (D Bhattacharjee, A Bhadra, pers. obs., 2010–2024). Accordingly, we used plastic balls as novel and plastic pouches as familiar objects. A between-subject study design was employed, and dogs were randomly provided with one of the following conditions with specific combinations of objects—an opaque ball and an opaque pouch (test), an opaque ball and a translucent white pouch (test), an opaque pouch and a translucent white pouch (control), and two opaque balls (validation). All the objects contained equal-sized food items, except for the last condition (i.e., two opaque balls), in which of one the balls was false-baited. Since free-ranging dogs benefit from reduced intraspecific competition during scavenging and experience little to no predation pressure in human-dominated environments (Sarkar, Sau & Bhadra, 2019; Sarkar et al., 2023), we hypothesised that they would not exhibit object-neophobia. In particular, we expected dogs to inspect the novel objects first rather than the familiar ones in the test conditions but show comparable choices in the control and validation conditions. Since object-neophobia might have been selected against during domestication (cf. Moretti et al., 2015), we expected adults and juveniles to respond similarly, i.e., no ontogenetic differences in their behavioural responses. Additionally, as humans exert significant anthropogenic stress on free-ranging dogs (Bhattacharjee & Bhadra, 2020; Bhattacharjee & Bhadra, 2021) and even cause mortality (Paul et al., 2016), we hypothesised that dogs would be cautious in their approach to the objects. Thus, we expected them to show comparable latencies to approach the familiar and novel objects first under different conditions.

Results

(i) First inspection (object choice)

(a) Ball and opaque pouch—62% (23 out of 37) of the adults and 69% (18 out of 26) of the juveniles inspected the ball first. Regardless of age classes, dogs first inspected the ball significantly more over the opaque pouch (Binomial test: p = 0.02, 95% Confidence Interval/CI [0.52–0.76], Fig. 1). However, we found no significant effect of age class (GLM: z = 0.6, p = 0.54) and sex (GLM: z = 0.3, p = 0.76) on object choice. These results suggest that both adults and juveniles comparably preferred the novel object over the familiar one.

(b) Ball and translucent white pouch—73% (25 out of 34) of the adults and 65% (17 out of 26) of the juveniles inspected the ball first. Overall, dogs first inspected the ball significantly more than the translucent white pouch (Binomial test: p = 0.002, 95% CI [0.57–0.81], Fig. 1). Age class (GLM: z = −0.68, p = 0.49) and sex (GLM: z = 1.12, p = 0.26) did not influence object choice. Thus, similar to the ball and opaque pouch condition, adults and juveniles first inspected the novel object more than the familiar one, even when the familiar object provided visual cues of the food item.

(c) Opaque pouch and translucent white pouch—52% (15 out of 29) of the adults and 63% (19 out of 30) of the juveniles inspected the opaque pouch first. In general, dogs did not differ in their first inspection of the familiar objects (Binomial test: p = 0.29, 95% CI [0.44–0.70], Fig. 1), even when they differed in visual obscurity levels. No significant effect of age class (GLM: z = 0.78, p = 0.43) and sex (GLM: z = −0.5, p = 0.61) on object choice was found. (d) Ball and scent ball –56% (22 out of 39) of the adults and 54% (21 out of 39) of the juveniles first inspected the ball. Overall, dogs did not discriminate between the novel objects during the first inspection, even when the strengths of olfactory cues differed (Binomial test: p = 0.49, 95% CI [0.42–0.65], Fig. 1). We did not find any significant effect of age class (GLM: z = −0.33, p = 0.74) and sex (GLM: z = −0.5, p = 0.61) on object choice.

Discussion

Using an ecologically relevant experimental design, we investigated whether free-ranging dogs exhibit object-neophobia and whether their foraging decision-making is dependent on, and in particular, constrained by, neophobic behaviour. We tested dogs from two age classes to evaluate if such behavioural responses have any ontogenetic developmental basis. As hypothesised, dogs, irrespective of age classes, first inspected the ball significantly more than the opaque pouch. Similarly, adults and juveniles inspected the ball significantly more than the translucent white pouch. We further investigated dogs with balls of varying strengths of olfactory cues and found no significant difference in their first inspection. Within the four conditions, we found no significant differences in latencies between dogs who first inspected the two objects (i.e., dogs who inspected the ball and dogs who inspected the opaque pouch, dogs who inspected the ball and dogs who inspected the translucent white pouch, dogs who inspected the opaque pouch and dogs who inspected the translucent white pouch, dogs who inspected the ball and dogs who inspected the scent ball). We discuss these findings and their implications, emphasising object-neophobia, exploratory behaviour and foraging decision-making in free-ranging dogs.

Our results indicate that free-ranging dogs do not exhibit object-neophobia in a scavenging context, as demonstrated by their first inspection of the ball over familiar plastic pouches. This implies that object-neophobia has no direct constraint on their potential exploratory behaviour. While we did not investigate the exploration of objects after the first inspection, object choice in the form of first inspection highlights a crucial step of exploratory behaviour and foraging decision-making (Mettke-Hofmann et al., 2006; Takola et al., 2021). These findings align with the conclusions from previous studies, which suggest a reduced neophobia in dogs, albeit in non-foraging contexts (Kaulfuß& Mills, 2008; Moretti et al., 2015). Human-dominated environments offer food subsidies, and these predictable resources can have broad ecological and evolutionary implications for animals (Oro et al., 2013). A wide range of species exploit these resources (Biswas et al., 2023), resulting in substantial competition both within and between species. Furthermore, anthropogenic activities can alter the spatiotemporal availability of such resources and directly or indirectly determine how animals eventually utilise them (Markus & Hall, 2004; Murray & St. Clair, 2017; Ramírez et al., 2020; Bhattacharjee & Bhadra, 2021; Egert-Berg et al., 2021). To maximise their food intake while avoiding potential anthropogenic stressors, animals may engage in opportunistic scavenging and explore novel resources, including objects. A reduction in object-neophobia in free-ranging dogs can, therefore, be an adaptive strategy to promote exploratory behaviour, especially during foraging.

A comparable response of adults and juveniles can be attributed to selection against neophobia during domestication, where reduced object-neophobia can provide adaptive benefits (Moretti et al., 2015) in human-dominated environments. In free-ranging dogs, different ontogenetic phases hold varying degrees of significance. The juvenile phase of development (3–6 months) involves complete independence from the mothers and venturing into the immediate environment (Paul et al., 2016); developmentally, this is a crucial phase as dogs start to forage on their own and may experience novel environmental aspects, like potentially rewarding novel objects. It has been shown that juvenile free-ranging dogs are reluctant to approach and follow the communicative intents of unfamiliar humans (Bhattacharjee et al., 2017b). Therefore, it can be assumed that scavenging, but not begging for food from humans, is the predominant feeding strategy in juveniles. A reduced object-neophobia can thus provide considerable foraging benefits to juveniles. Conversely, adults can rely on scavenging and begging (Sen Majumder et al., 2014; Bhadra et al., 2015; Boitani et al., 2016), and depending on the energy requirements and other external factors (such as human disturbance), they may flexibly switch between those strategies, where reduced object-neophobia would still be beneficial. Nonetheless, object-neophobia in dogs appears to be a trait that was selected against during domestication (but see Marshall-Pescini et al., 2017a). The ‘social-ecology’ hypothesis suggests that feeding ecology and social organisation may act together as mechanisms to drive dogs’ interactions with environmental, particularly novel stimuli, including objects (Marshall-Pescini et al., 2017b). Free-ranging dogs’ scavenging behaviour in human-dominated environments, thus, can contribute to the reduction in object-neophobia.

The first inspection of the novel object, overriding the partial visual cue of food item from the familiar object, can be attributed to free-ranging dogs’ reliance on a complex multimodal sensory information system of vision and olfaction during foraging. Free-ranging dogs are primarily known to rely on olfactory cues to make their foraging decisions (Bhadra et al., 2015; Banerjee & Bhadra, 2019; Sarkar et al., 2023). Our results indicate that they might also use visual cues and rely more on them than olfaction when encountering potentially novel objects. In non-scavenging conditions, dogs have been shown to tolerate human artefacts (Kaulfuß& Mills, 2008). Thus, it is safe to assume that dogs paid particular attention to the visual appearance of the novel object. At the same time, they did not do so for the familiar objects with different visual obscurities. On the other hand, the ball and scent ball condition validates our results by demonstrating that novelty indeed played a role, and dogs did not fully rely on (the strength of) olfactory cues. These findings indicate that dogs can use their multimodal sensory information system rather flexibly, depending on the context, to guide their foraging decisions.

The appearances of novel objects, from simple to relatively ‘complex’ sizes and shapes, can be perceived differently by animals, eliciting varying responses (Mettke-Hofmann et al., 2006). In the current study, we could not measure the construct validity due to the use of only one type of novel object (Greggor, Thornton & Clayton, 2015; Kimball & Lattin, 2023), and to what extent the plastic ball was novel could not be assessed. We were cautious about calling it a truly novel object and hence considered it ‘potentially novel’. It is still possible that some of the dogs might have seen plastic balls of different sizes or colours previously if they did not interact with them. Furthermore, a recent study has shown a strong preference for the colour yellow in free-ranging dogs in the context of foraging. This preference could be strong enough to override the attraction towards food rewards (Roy et al., 2024). Since our plastic balls were green in colour, they would have appeared a light shade of yellow to the dogs, thereby potentially adding an additional cue to the object. Although non-significant, it is worth mentioning that a substantial percentage of adults and juveniles first inspected the familiar plastic pouches. A few possibilities can explain the finding: first, these dogs did not perceive balls as novel objects (see Mettke-Hofmann et al., 2006); second, the state of hunger might have forced dogs to choose the plastic pouches as reliable food sources over the balls (see Alley, 2018); third, these individuals had shy or avoidant personalities (see Sloan Wilson et al., 1994). From our current study with one-off experiments, it is impossible to pinpoint which mechanism played the most significant role. Nevertheless, we believe that it would be an oversimplification to assert that free-ranging dogs have reduced object-neophobia, and thus, we recommend careful assessments of the highlighted mechanisms to reach a firm conclusion.

Studies with novel object tests often consider investigating the time animals take to approach or explore the objects. Individuals with reduced neophobia are expected to approach the novelty quicker than their neophobic counterparts (Takola et al., 2021; Kimball & Lattin, 2023). One may argue that dogs in our study did not differ in their latencies to first inspect the novel and familiar objects. This could be explained by the study design, where we provided two-way object choice conditions instead of separately presenting novel or familiar objects. The findings are also in line with our prediction that although free-ranging dogs may show reduced object-neophobia, the negative impact of anthropogenic stressors (Paul et al., 2016; Bhattacharjee & Bhadra, 2020; Bhattacharjee & Bhadra, 2021) may force them to explore their environments with caution (Greenberg, 2003). Besides, solitary foraging is prevalent in free-ranging dogs; consequently, a lack of intraspecific competition may aid in that process. However, this does not disregard the potential influence of competition and subsequent shorter latencies to approach novel objects during group foraging events.

Conclusions

We conclude that free-ranging dogs exhibit reduced object-neophobia in a scavenging context by significantly inspecting a relatively simple object with potential novelty more over familiar objects. Due to the use of only one type of novel object, we could not test its construct validity and subsequently failed to measure the extent of novelty. However, with adequate control and validation phases, we found that dogs indeed paid attention to the appearance of the novel object and made their foraging decisions. In future, it would be useful to include novel objects with different shapes and sizes to examine whether our results hold. Additionally, with our one-off tests (due to the between-subject study design), we could not check for repeatability and inter-individual differences in dogs’ behavioural responses to novel objects. Such an approach would be instrumental in resolving whether object-neophobia is a personality (see Sih, Bell & Johnson, 2004) or a plastic (see Vincze et al., 2016; Greggor et al., 2016) trait. Nevertheless, we provide the first experimental evidence of free-ranging dogs’ behavioural responses to novelty during scavenging.

Supplemental Information