Innovative problem solving refers to the acquisition of a novel behavior, which may allow an individual to exploit a new resource or environment (Ramsey et al., 2007; see also Reader and Laland, 2003). Variation in innovative problem solving both within and across species has been linked to cognitive abilities (Reader, 2003; Lefebvre et al., 2004; Griffin et al., 2014). However, motivational factors, in particular persistence in exploration and neophobia, are also related to success in problem solving tasks (Reader, 2003; Reader and Laland, 2003; Stöwe et al., 2006a,b). Explorative behavior, measured mostly in terms of the time spent (persistence in) investigating and interacting with an object/task, has been consistently shown to relate to problem solving success across a wide range of species (great tits: Morand-Ferron et al., 2011; blue tits: Morand-Ferron and Quinn, 2011; hyenas: Benson-Amram and Holekamp, 2012; meerkats: Thornton and Samson, 2012). Furthermore, in a number of bird species it was found that individuals who were least reluctant to approach novel objects (i.e., showed less neophobia) were also the quickest to solve novel foraging tasks (Webster and Lefebvre, 2001; Seferta et al., 2001; Auersperg et al., 2011). Similarly, in wild hyenas less neophobic individuals were significantly more successful than more neophobic ones in accessing a food puzzle box (Cole et al., 2011 for no links between neophobia and problem solving success; Benson-Amram and Holekamp, 2012, but see also Biondi et al., 2010).

A number of hypotheses posit that domestication may have negatively affected dogs' independent problem solving abilities (Frank, 1980; Frank and Frank, 1982, 1985; Frank et al., 1989; Frank, 2011) supported by studies showing that wolves are more successful than dogs in different tasks both as puppies (Frank and Frank, 1982, 1985; Frank et al., 1989) and adults (Hiestand, 2011). These differences in problem-solving skills have been related to differences in wolves' and dogs' cognitive complexity (e.g., Frank and Frank, 1985) and to dogs' greater sensitivity to social inhibition when tested with humans present (Topál et al., 1997; Udell, 2015).

Indeed when humans are present during problem solving tasks, dogs consistently behave differently compared to wolves. In the first studies comparing equally raised wolves and dogs, Frank and Frank (1985) presented pups with puzzle boxes of increasing complexity, noting that wolf pups were overall significantly more successful in obtaining the reward. Interestingly however, authors describe wolves “attacking each puzzle immediately” (pp. 271) and persisting “until the problem was solved or time ran out” (pp. 271) in contrast to dogs quickly reverting to seeking human attention upon discovering that the food was not immediately available, and then laying down until time elapsed. Similarly, when testing 4-month-old wolves and pet dogs in a manipulation task that suddenly became unsolvable, dogs quickly looked back toward the human handler whereas wolves did not (Miklósi et al., 2003). More recently, Udell (2015) presented adult wolves and dogs with a problem solving task and found that wolves were more persistent (and consequently more successful) than both pet and shelter dogs, regardless of human encouragement received during testing, whereas dogs consistently looked longer toward the human. Hence, taken together results suggest that when a human is present in the room, dogs show less persistent behavior than wolves in object manipulation tasks.

Interestingly however, differences between wolves and dogs in such tasks also emerge when humans are not present during testing. Udell (2015) tested adult dogs and wolves also when alone, and found wolves to be more persistent (and successful) in their efforts to obtain the reward. Furthermore, in a study comparing adult and 6-month-old wolves' and dogs' performance in a string-pulling task, differently from dogs, wolves required no prior training to solve the task, showed a greater variety of behaviors exhibited on the rope and in general were more persistent (Hiestand, 2011).

Hence, it appears that motivational factors and the explorative tendencies of wolves and dogs per-se (independently from human presence) differ, which may play a major role in the observed differences in problem solving abilities, as was indeed hinted at already in Frank and Frank's writing (“higher curiosity and explorative drive” (pp. 272, Frank and Frank, 1985) in wolves than in dogs).

Despite the potential significance explorative behaviors have for problem solving, few studies have directly investigated these tendencies in wolves and dogs. Early studies comparing wolves and dogs led most authors to agree that despite socialization, wolves are more neophobic than dogs (Fentress, 1967; Klinghammer and Goodmann, 1987; Zimen, 1987), but these descriptions contrast with those of wolf pups immediately “attacking” new puzzle boxes in experimental settings (Frank and Frank, 1985). This contradiction might be explained by focusing on different aspects of whether and how animals approach novelty. A recent study testing human-raised pack-living adult wolves and dogs kept in a game park setting under identical conditions, compared their reactions to novel objects and found that dogs were less likely to approach a novel object than wolves (i.e., only 7 of the 13 dogs but all 11 wolves approached the object), and wolves were more persistent in their investigation of the objects than dogs. At the same time however, wolves were slower in making contact with the object and fled from it more often than dogs (Moretti et al., 2015). Hence, it appears that adult wolves approach novel objects in their environment slower than dogs and they do so in a more cautious and fearful manner, but they then spend more time investigating it. Unequivocally however, wolves appear more interested in novel objects in their environment than dogs. Confirming this difference, another study found that wolves, at the age of 4 months, spent more time exploring an unfamiliar outdoor kennel than dogs (Topál et al., 2005). As outlined above, such an elevated motivation to approach and explore novel objects and places is likely to increase success also in problem solving. Interestingly, however Gácsi et al. (2005) found no differences in wolves and dogs' approach behavior to novel objects at 3 and 4 weeks: the animals moved in an unfamiliar room and approached different objects and persons in a similar way. This may suggest that at early ages dogs and wolves explore their environment similarly and only later (possibly between the ages of 1 and 4 months) dogs lose their interest in novel objects as compared to wolves.

Hence, to further elucidate the potential effects of domestication on dogs' explorative and neophobic behaviors at an early age, in the current study, we presented identically-raised and kept wolf and dog puppies at 5, 6, and 8 weeks of age, with a novel environment and a novel object test. In order to address neophobia in different ways, we recorded not only (1) likelihood and (2) latency to contact the novel objects but also (3) time spent exhibiting fear-related behaviors such as freezing and running around with a lowered posture (i.e., “escape” see Table 2); (4) the frequency of behaviors commonly associated with stress in these species (e.g., panting, yawning, scratching etc.) and the (5) time spent in proximity to the exit. Explorative behaviors we measured in terms of the (1) time the animals spent moving around in a novel environment maintaining a relaxed body posture and (2) actively investigating and playing with a novel object.

Based on results of the adult wolf-dog comparison, we hypothesized that already at this young age, wolves would show greater interest than dogs in exploring a new environment and a novel object, but that they would also show higher levels of stress and fear-related behaviors.

In line with our predictions wolves explored a new environment more than dogs at the age of 5 weeks and interacted longer with the novel object than dogs both at 6 and 8 weeks. Furthermore, in line with previous results with adults (Moretti et al., 2015), there was a tendency for wolves to be more likely to contact the novel objects than dogs.

A potential reason for these differences is that wolves and dogs have a different developmental trajectory in terms of their motor abilities. Frank and Frank (1982) compared wolves and Malamute dog puppies on a number of problem solving tasks and concluded that wolf pups developed motor skills significantly faster than dogs. For example, wolf pups could climb over their 45 cm pen at 3 weeks of age but Malamute dog pups at 32 days of age were not yet able to climb over their 15 cm barrier. Similarly, at 6 weeks of age wolf pups had much better motor performance than dogs. However, in the present novel object task (both at 6 and 8 weeks of age), there was no difference between wolves and mixed-breed dogs in the latency to contact the object nor the person for those animals that did so. It is therefore possible that middle-sized mixed-breed dogs' motor skills develop differently than that of Malamutes or that the requirements of the current tasks were equally within the animals' motor abilities. It is still possible that especially at 6 weeks, dog pups tired more easily due to their less well-developed motor abilities, thereby influencing the duration of exploration. However, considering also the likelihood of contacting the object tended to differ between wolves and dogs, we suggest that differences in motor abilities are not sufficient to explain the different explorative tendencies of wolf and dog pups. Moreover, the fact that the explorative differences between wolves and dogs we observe in the current study have been found also in older puppies and adults (Topál et al., 2005; Moretti et al., 2015) supports that this behavioral difference is genuine rather than due to differing mobility skills.

As discussed in the introduction, similarly to former studies (e.g., Frank and Frank, 1985, 1987; Miklósi et al., 2003), dogs' reduced exploration of the environment and objects also in the current study may have been a side effect of dogs spending more time in the vicinity/contact with the people present during testing. However, no difference between wolves and dogs emerged either in the likelihood or the latency to contact the people in the room in either test. As for proximity there was a tendency for wolf pups at 6 weeks to spend more time close to the person, although no such difference emerged at 8 weeks. Considering the experimenter was a stranger, it may be that wolves' tendency to spend more time in proximity with her may be in fact an expression of the greater overall explorative tendency observed in wolves compared to dogs (in line with results also from other studies Gácsi et al., 2005; Topál et al., 2005). Although it is possible that the presence of the human affected the explorative behavior of the animals, it is interesting to note that the same pattern of results, with wolves being more explorative than dogs, emerges both in the novel environment test, in which the cameraperson was present but out of sight, and in the novel object test, where the experimenter was quietly standing against the wall. This consistency in the pattern of result would suggest that it is not directly linked to the effect of the human being present.

Another possibility is that dogs explored the object and the environment less because they were more distressed due to separation from the human caregiver and/or their peer group. However, no differences between wolves and dogs emerged in fear-related behaviors and time spent in proximity to the exit. Furthermore, no wolf-dog difference emerged in stress-related behaviors at 5 and 6 weeks and an opposite pattern emerged at 8 weeks, with wolves more likely to exhibit stress-related behaviors and doing so more frequently than dogs. It would therefore seem highly unlikely that dogs' reduced explorative behavior was due to heightened separation anxiety during testing. Our results at week 8, however, are in line with findings of the study conducted with adults in which wolves, although approaching novel objects more readily than dogs, exhibited more fearful behaviors (Moretti et al., 2015). Fear-responses in this former study with adults were specifically related to the object (i.e., walking, running, or jumping away from the object; Moretti et al., 2015). Although in the current study, it is unclear whether the differences in frequency of stress behaviors were due to the presence of the object or other elements of the test, e.g., separation from their peers and/or caregiver, the fact that these behaviors became more frequent as the animals got older (and hence less susceptible to separation anxiety) seems to support the former rather than latter explanation.

Overall, based on the current study and in line with the study comparing adult wolves and dogs (Moretti et al., 2015), it appears that domestication has reduced dogs' environmental exploration tendencies and persistence in investigating new objects (as reflected in the shorter duration of interaction time with the novel objects). However, whereas as adults, wolves showed a more intense fear reaction to the novel object compared to dogs (Moretti et al., 2015), this pattern did not emerge as clearly in puppies, although the higher stress-related behaviors in wolves at 8 weeks may be an indication of their neophobic reaction.

The question which remains to be addressed is what factor/s may have changed during domestication, which reduced dogs' persistence in exploring novel objects and also their fear of such objects?

Frank proposed that a consequence of domestication is dogs' reliance on man “as an intermediary between animal and environment” (pp. 272; Frank and Frank, 1985).

In line with this, a number of studies have found that dogs preferentially seek out humans in problem solving tasks, when given such an opportunity (Frank and Frank, 1985; Miklósi et al., 2003; Udell, 2015). Hence one possibility is that during the course of domestication dogs have been selected for a greater dependence/reliance on humans, which has resulted in a reduced motivation for independent explorative behavior.

A second important aspect is that domestication is thought to have affected the timing of an animal's “critical period of development,” i.e., the window of opportunity during which exposure to social and environment stimuli will facilitate acceptance of these as adults (Freedman et al., 1960; Fox and Stelzner, 1966; Lord, 2013 for a review of the dog-wolf literature in particular). This “critical period” is thought to coincide with the onset of mobility (walking and exploration of the environment) and close when animals show avoidance rather than an approach response to novel objects. For example, when comparing foxes selected for tameness to control foxes at 45 days, Trut et al. (2004) found that whereas in the control population there was a decrease in explorative behavior in a new environment and a corresponding increase in both the fearful behaviors and glucocorticoids in the blood, no such changes occurred in the “tame” pups. The “critical period” of the animals selected for tameness, was significantly longer, in that there was no decrease in explorative activity and increase in fear and glucocorticoid levels even at 3 months of age. This extended window of opportunity for exploration with no increase in fear is thought to contribute to the ease with which these animals are socialized to different elements of their environment, including humans.

Wolves' and dogs' critical period has also been reported to differ. A number of authors have reported an increase in avoidance of novelty in wolves at 6 weeks of age (Scott and Marston, 1950; Fentress, 1967; Woolpy and Ginsburg, 1967; Zimen, 1987), whereas in dogs an increase in neophobic behaviors appears to start at around 8 weeks of age, when it is reported that pups will avoid a novel object unless they have been exposed to it for several days (Scott and Marston, 1950; Scott, 1958; Freedman et al., 1960). Our current results are not wholly in agreement with the outlined differential timeline in wolves and dogs. Indeed wolves were more explorative at both 6 and 8 weeks of age, whereas based on the reported information they should have been less explorative at both these times since their critical period should have closed at 6 weeks. Nevertheless, assessment of critical periods can also be based on the expression of stress-related behaviors in general. Considering wolves' showed a higher frequency of stress related behaviors than dogs at 8 weeks, these results are in line with studies suggesting that wild canids' critical period may be shorter than that of dogs. So the increased stress-related behaviors in wolves at 8 weeks is in agreement with the outlined critical period timeline, but the increased explorative behaviors are not. Compared to wolves, dogs showed both less stress behaviors and less exploration of the novel object at 8 weeks. Considering results from the present study with young puppies largely show a similar pattern to those observed with older individuals and adults (Topál et al., 2005; Moretti et al., 2015), it would appear that the wolf-dog differences in explorative behaviors are not limited to a specific developmental phase.

Finally, an often somewhat overlooked aspect which has changed during domestication and has likely affected explorative behaviors, is dogs' ecological niche (Coppinger and Coppinger, 2001, 2016). Indeed, explorative tendencies and neophobia have both been linked to ecological variables in a species environment (Clarke and Lindburg, 1993; Greenberg and Mettke-Hofmann, 2001; Mettke-Hofmann et al., 2002, 2005; Reader and Laland, 2003; Martin and Fitzgerald, 2005). Wolves rely predominantly on group hunting requiring extraordinary persistence considering success rates are between 15 and 50% (Mech et al., 2015). In contrast, free-ranging dogs (i.e., 70–80% of the world's dog population- Lord et al., 2013; Hughes and MacDonald, 2013), live in proximity to human settlements and rely predominantly on solitary scavenging of human waste (Butler and Du Toit, 2002; Butler, 2004; Vanak and Gompper, 2009a,b; Hughes and MacDonald, 2013; Newsome et al., 2014). Human generated food can be considered a more reliable food source compared to live prey, hence in line with past research on other species, the difference in the foraging ecology of wolves and dogs may explain, at least in part the difference in their explorative patterns. While wolves are dependent on greater persistence to obtain their elusive preys (reflected in their greater duration of exploration of novel objects both as pups and adults), dogs as adults at least (Moretti et al., 2015) show less fearful behaviors than wolves when confronted with novel objects, which would be in line with their greater reliance on scavenging in more “humanized” environments.

Whatever the selective pressures affecting wolves' and dogs' differing explorative behaviors (and it is possible that all three aspects combined have played a role in these changes), considering the growing literature highlighting the link between persistence and neophobia on problem solving skills (e.g., Benson-Amram and Holekamp, 2012; Thornton and Samson, 2012), these aspects require further investigation to assess their potential role in domestication and need to be taken into account when comparing the cognitive abilities of wolves and dogs.

ZV, EK, and FR: Contributed to the conception and design of the study and critical revision of the manuscript; SM: Contributed to the analyses and interpretation of the data as well as the drafting and revising the manuscript. Approval of the final version was given by all authors.

SM and FR were supported by funding from the European Research Council under the European Union's Seventh Framework Programme (FP/2007–2013)/ERC Grant Agreement No. [311870]. ZV was supported by the Vienna Science and Technology Fund (WWTF project CS11-026) and by the Austrian Science Fund (FWF project P21244-B17). EK was supported by the Hungarian Academy of Sciences (MTA 01 031, János Bolyai Research Scholarship). The authors further thank many private sponsors including Royal Canin for financial support and the Game Park Ernstbrunn for hosting the Wolf Science Center. The funders had no role in study design, data collection, and analysis, decision to publish, or preparation of the manuscript.