Wolves are more prone to risk. Dogs aren't
Both human and non-humans species face decisions in their daily lives which may entail taking risks. At the individual level, a propensity for risk-taking has been shown to be positively correlated with explorative tendencies, whereas, at the species level a more variable and less stable feeding ecology has been associated with a greater preference for risky choices. In the current study we compared two closely related species; wolves and dogs, which differ significantly in their feeding ecology and their explorative tendencies.
Wolves depend on hunting for survival with a success rate of between 15 and 50%, whereas free-ranging dogs (which make up 80% of the world dog population), are largely scavengers specialized on human produce (i.e., a more geographically and temporally stable resource). Here, we used a foraging paradigm, which allowed subjects to choose between a guaranteed less preferred food vs. a more preferred food, which was however, delivered only 50% of the time (a stone being delivered the rest of time).
We compared identically raised adult wolves and dogs and found that in line with the differing feeding ecologies of the two species and their explorative tendencies, wolves were more risk prone than dogs.
Human and non-human animals are often faced with decisions which may involve taking risks. Humans are mostly risk-averse, so for example when asked to choose between two containers, one with 10 Euros for certain vs. the other with a 50/50 chance of containing 20 Euros or being empty, subjects will mostly choose the safe option.
Studies on risk-taking in non-human species have focused mostly on primates, where results have been more controversial: cotton-top tamarins, lemurs, and bonobos have been shown to be risk averse, whereas chimpanzees, macaques, and capuchin monkeys have been shown to be risk prone and common marmosets have been shown to be risk neutral.
However, the tendency for being risk prone seems to depend on several factors, including how the choices are presented, energy budget, individual differences and the ecology of the species.
When faced with the risk of a potential loss, both humans and a number of non-human species have been shown to increase their risk-preference, whereas if the same choices are framed in a context of a potential gain, the preference for risk taking can be reversed. This tendency for risk-preferences to reverse depending on whether the choice is framed as a loss or a gain is known as the ‘reflection effect’ or ‘loss aversion’ and is one of the primary predictions of Prospect Theory.
Energy budget has also been shown to affect risk-taking, with individuals in a positive energy budget state being more risk averse than an individuals in a negative state. Furthermore, in several species, variation in risk-taking behavior has been associated with individual differences in boldness and aggression and a genetic predisposition for risk-taking in great tits was shown to be positively associated with a greater inclination for explorative behaviors. The ecology of a species has also been consistently associated with risk-taking preferences. A comparison between three sympatric species of tits revealed that risk-preference was directly affected by their feeding ecology. Indeed the more insectivorous species were shown to be risk prone whereas the more granivorous species was risk averse. Comparative work across primate species further supports the link between feeding ecology and risk preference. For example, chimpanzees, that depend more heavily on patchy fruit trees and invest time in hunting where outcomes are often uncertain, have been shown to be more risk-prone and more willing to wait for a delayed reward than bonobos, that rely more heavily on terrestrial vegetation, a more temporally and spatially reliable food source.
In the current study, we addressed whether a species’ ecological environment may shape its risk preference by comparing wolves and dogs in a foraging task. Wolves and dogs are interesting to study in this context because they diverged only between 15000 and 30000 years ago. Wolves and free-ranging dogs show similar behavioral patterns in terms of living in pack-like groups and forming stable hierarchical structures. Crucially for the present study, however, they differ conspicuously in their foraging strategies. Wolves rely predominantly on hunting with success rates ranging between 10 and 50%, whereas free-ranging domestic dogs, although capable of preying on wild species, rely mostly on human generated resources, leading most researchers to conclude that free-ranging dogs can be considered scavengers specialized in exploiting human refuse. This is supported by the occurrence of key mutations to specific genes associated with the digestion of starch during domestication, indicating a direct adaptation to the new dietary opportunities offered by dogs’ association with humans.
Considering wolves’ higher reliance on hunting behavior compared to dogs’ scavenging on distributed and more stable food sources, we hypothesized that, in the feeding context, wolves’ would show a higher preference for making risky choices than dogs. Furthermore, in a recent study comparing identically raised wolf and dog pups, we found that the former showed a higher inclination to explore both a new environment and a novel object (Marshall-Pescini et al., submitted). Considering that great tits’ risk-taking has been positively associated with a higher propensity for explorative behavior (van Oers et al., 2004), our finding regarding the higher explorative tendencies of wolf compared to dog pups would also support predictions suggesting that wolves will be more risk-prone than dogs.
To test the risk preference of wolves and dogs, we adapted a method previously used with apes (Heilbronner et al., 2008; Rosati and Hare, 2012, 2013) in which animals are presented with a two-way choice task. One option results in a ‘safe outcome’ where animals always obtain a less preferred food item (i.e., a dry food pellet). The alternative choice represents the ‘risky outcome’ where on 50% of trials subjects obtain a preferred food item (i.e., a piece of meat), but in the remaining 50% they obtain a non-edible item (a stone). Before each trial, subjects are shown the potential outcome of each choice, so they are aware of the ‘odds’ of the risky vs. safe outcome. The same task was presented to comparably raised and kept wolves and dogs.