Science is advanced by identifying novel connections among disparate findings and by developing new theories (Jack et al. Reference Jack, Crivelli and Wheatley2018). Anselme & Güntürkün (A&G) do both, integrating proximate neural mechanisms with evolutionary theories and extending the integration via mathematical modeling. Their thesis provides an impressive and timely connection between two disparate fields: the psychology of motivation and animal foraging. When bringing together separate bodies of literature, even the most scholarly integration can overlook important details (Fitzpatrick et al. Reference Fitzpatrick, Hobson, Mendelson, Rodríguez, Safran, Scordato, Servedio, Stern, Symes and Kopp2018). We believe that the authors underappreciate a critical feature of natural environments – the spatial distribution of resources – and how profoundly the spatial distribution of food will affect selection on neural mechanisms and animal behavior, as well as the outcome of foraging models (Arditi & Dacorogna Reference Arditi and Dacorogna1988; McIntyre & Wiens Reference McIntyre and Wiens1999; Pyke Reference Pyke2015).
Most laboratory trials are based on random reward schedules. In natural environments, food may be rare, but it is seldom distributed randomly in space or in time (Iwasa et al. Reference Iwasa, Higashi and Yamamura1981; Racey & Swift Reference Racey and Swift1985). Given that a single plant produces many seeds, an animal finding one seed is predictive of finding another nearby. River valleys, tree fall gaps, and low-lying wet areas support a local abundance of plants that are otherwise rare across the landscape. Nests, burrows, and swarms often contain multiple edible prey. For most animals, then, the probability of finding food items is not independent of recent experience. As resources predict other resources nearby, local foraging is incentivized after finding an initial reward (Krebs et al. Reference Krebs, Ryan and Charnov1974). Expectation of enhanced rewards can be seen even in humans when finding one good restaurant causes people to anticipate other good restaurants in the area. By assuming an environment where the probability of reward is independent of previous events, A&G miss key dynamics inherent in natural contexts, particularly a likely explanation for the apparent overperformance of animals under conditions where reward is uncertain. Although the experimental paradigm for sign-tracking studies may be one of random reward, the neural circuits and behaviors that are tapped likely evolved under conditions where rewards were not truly random but changed in probability based on recent foraging experience. This nonrandom distribution of rewards in natural environments can explain increased attention when reward is uncertain, a phenomenon that has been previously explained using hypotheses including frustration and simple attention to unexpected outcomes.
The statistical distribution of food resources has important implications for proximate neural mechanisms and emergent foraging behaviors. A&G connect stress to elevated dopamine levels, which are in turn connected to increased exploration, risk taking, and gambling. If food is discovered after a stressful period of starvation, there is good reason for the animal to expend additional energy and incur additional risk if the probability of imminent success is higher than it has been. The fact that exploration and risk taking are particularly enhanced when a stressed animal is feeding or drinking further suggests that these pathways are well suited to helping organisms exploit local resource patches. Furthermore, dopamine is highest when uncertainty is greatest, increasing the salience of immediate rewards and promoting learning, potentially reinforcing a recently successful foraging strategy. Rather than elevated performance under uncertainty being something of an odd anomaly in need of explanation, it is likely to be incredibly adaptive in natural environments, allowing organisms to adjust their motivation and investment in response to the local probability of finding additional resources in the same place or time or using a similar foraging strategy. Even the use of the word hope in the concept of “incentive hope” underscores the way that one successful foraging event appropriately changes the expectation of success for additional attempts in a similar time and space.
If the phenomenon of incentive hope arises from the spatial and temporal distribution of food in evolutionary environments, the hypothesis predicts that some species should express incentive hope more strongly than others and that specific behavioral contexts can elicit incentive hope. For example, one prediction is that incentive hope will be more strongly expressed by species that experience patchy resources than by species that inhabit environments where resources are comparably rare, but are more evenly distributed across the landscape. Indeed, some of the species that have been most studied for sign-tracking and foraging behavior (such as rats and small passerines) have highly clumped food resources. During laboratory training, the expression of incentive hope may be exacerbated by supplying non-independent rewards (where animals experience runs of success and failure and thereby learn to respond more strongly after encountering a first successful response). The importance of these recent experiences could be tested in a sign-tracking paradigm where the reward is still delivered in 50% of the trials, but the probability of reward is contingent on the previous probability of reward. The evolved tendency to expect strings of rewards might also be revealed in people and other animals through slot machine scenarios that include near misses and “payouts” that are less than the cost of play (Harrigan et al. Reference Harrigan, Collins, Dixon and Fugelsang2010; Scarf et al. Reference Scarf, Miles, Sloan, Goulter, Hegan, Seid-Fatemi, Harper and Colombo2011). Finally, the simulation developed by A&G is well suited for determining whether clumped resources increase the tendency and ability of organisms to gather and store resources over the scenario of random resource distribution.
A&G use the term incentive hope to indicate that the animal “is in a state of motivational excitement for possible good news (rewards) when bad news (non-rewards) is likely” (sect. 4.2.1, end of para. 1). A critical thing to note about incentive hope and foraging behavior is that, although bad news may be likely, once one food item is found, bad news is less likely than it was before.
Science is advanced by identifying novel connections among disparate findings and by developing new theories (Jack et al. Reference Jack, Crivelli and Wheatley2018). Anselme & Güntürkün (A&G) do both, integrating proximate neural mechanisms with evolutionary theories and extending the integration via mathematical modeling. Their thesis provides an impressive and timely connection between two disparate fields: the psychology of motivation and animal foraging. When bringing together separate bodies of literature, even the most scholarly integration can overlook important details (Fitzpatrick et al. Reference Fitzpatrick, Hobson, Mendelson, Rodríguez, Safran, Scordato, Servedio, Stern, Symes and Kopp2018). We believe that the authors underappreciate a critical feature of natural environments – the spatial distribution of resources – and how profoundly the spatial distribution of food will affect selection on neural mechanisms and animal behavior, as well as the outcome of foraging models (Arditi & Dacorogna Reference Arditi and Dacorogna1988; McIntyre & Wiens Reference McIntyre and Wiens1999; Pyke Reference Pyke2015).
Most laboratory trials are based on random reward schedules. In natural environments, food may be rare, but it is seldom distributed randomly in space or in time (Iwasa et al. Reference Iwasa, Higashi and Yamamura1981; Racey & Swift Reference Racey and Swift1985). Given that a single plant produces many seeds, an animal finding one seed is predictive of finding another nearby. River valleys, tree fall gaps, and low-lying wet areas support a local abundance of plants that are otherwise rare across the landscape. Nests, burrows, and swarms often contain multiple edible prey. For most animals, then, the probability of finding food items is not independent of recent experience. As resources predict other resources nearby, local foraging is incentivized after finding an initial reward (Krebs et al. Reference Krebs, Ryan and Charnov1974). Expectation of enhanced rewards can be seen even in humans when finding one good restaurant causes people to anticipate other good restaurants in the area. By assuming an environment where the probability of reward is independent of previous events, A&G miss key dynamics inherent in natural contexts, particularly a likely explanation for the apparent overperformance of animals under conditions where reward is uncertain. Although the experimental paradigm for sign-tracking studies may be one of random reward, the neural circuits and behaviors that are tapped likely evolved under conditions where rewards were not truly random but changed in probability based on recent foraging experience. This nonrandom distribution of rewards in natural environments can explain increased attention when reward is uncertain, a phenomenon that has been previously explained using hypotheses including frustration and simple attention to unexpected outcomes.
The statistical distribution of food resources has important implications for proximate neural mechanisms and emergent foraging behaviors. A&G connect stress to elevated dopamine levels, which are in turn connected to increased exploration, risk taking, and gambling. If food is discovered after a stressful period of starvation, there is good reason for the animal to expend additional energy and incur additional risk if the probability of imminent success is higher than it has been. The fact that exploration and risk taking are particularly enhanced when a stressed animal is feeding or drinking further suggests that these pathways are well suited to helping organisms exploit local resource patches. Furthermore, dopamine is highest when uncertainty is greatest, increasing the salience of immediate rewards and promoting learning, potentially reinforcing a recently successful foraging strategy. Rather than elevated performance under uncertainty being something of an odd anomaly in need of explanation, it is likely to be incredibly adaptive in natural environments, allowing organisms to adjust their motivation and investment in response to the local probability of finding additional resources in the same place or time or using a similar foraging strategy. Even the use of the word hope in the concept of “incentive hope” underscores the way that one successful foraging event appropriately changes the expectation of success for additional attempts in a similar time and space.
If the phenomenon of incentive hope arises from the spatial and temporal distribution of food in evolutionary environments, the hypothesis predicts that some species should express incentive hope more strongly than others and that specific behavioral contexts can elicit incentive hope. For example, one prediction is that incentive hope will be more strongly expressed by species that experience patchy resources than by species that inhabit environments where resources are comparably rare, but are more evenly distributed across the landscape. Indeed, some of the species that have been most studied for sign-tracking and foraging behavior (such as rats and small passerines) have highly clumped food resources. During laboratory training, the expression of incentive hope may be exacerbated by supplying non-independent rewards (where animals experience runs of success and failure and thereby learn to respond more strongly after encountering a first successful response). The importance of these recent experiences could be tested in a sign-tracking paradigm where the reward is still delivered in 50% of the trials, but the probability of reward is contingent on the previous probability of reward. The evolved tendency to expect strings of rewards might also be revealed in people and other animals through slot machine scenarios that include near misses and “payouts” that are less than the cost of play (Harrigan et al. Reference Harrigan, Collins, Dixon and Fugelsang2010; Scarf et al. Reference Scarf, Miles, Sloan, Goulter, Hegan, Seid-Fatemi, Harper and Colombo2011). Finally, the simulation developed by A&G is well suited for determining whether clumped resources increase the tendency and ability of organisms to gather and store resources over the scenario of random resource distribution.
A&G use the term incentive hope to indicate that the animal “is in a state of motivational excitement for possible good news (rewards) when bad news (non-rewards) is likely” (sect. 4.2.1, end of para. 1). A critical thing to note about incentive hope and foraging behavior is that, although bad news may be likely, once one food item is found, bad news is less likely than it was before.