We agree wholeheartedly with Anselme & Güntürkün (A&G) that it is important not only to understand the functional explanation for why animals should gain more weight (carry more fat reserves) when food is unpredictable, but also to understand the mechanisms that drive this phenomenon. In the target article, A&G try to bring together functional explanations from behavioural ecology with the mechanistic explanations from behavioural psychology and behavioural neuroscience. Specifically, they speculate about a novel mechanism that they call “incentive hope,” which works alongside (and interacts with) the well-established “incentive salience” mechanism, which they equate with the “wanting” system sensu Berridge (Fig. 2 in the target article; Berridge et al. Reference Berridge, Ho, Richard and DiFeliceantonio2010). This “wanting” mechanism is physiologically represented by the dopaminergic innervation of the nucleus accumbens (NAc) by the ventral tegmental area (VTA). The implication of this proposed mechanism is that the increased foraging that leads to increased eating, increased fat reserves, and, in some species, increased food hoarding (a “seeking” response), only occurs once the incentive hope system is activated but not following activation of the “wanting” system alone.

We believe that the wanting system is broader than A&G propose, with a complex interplay of factors that influence it. Berridge et al. (Reference Berridge, Ho, Richard and DiFeliceantonio2010, p. 47) stated that the wanting system is activated “when a food cue is encountered in a mesolimbically primed state (or if cues are vividly imagined then)” (emphasis added). Clearly, “wanting” can be activated to search for food, not just in the presence of food. This is also how the late Jaak Panksepp conceptualized the same system, which he called the SEEKING system (his capitalization), which drives intensive exploration and searching for resources upon release of dopamine in the NAc (Alcaro & Panksepp Reference Alcaro and Panksepp2011; Ikemoto & Panksepp Reference Ikemoto and Panksepp1999). This implies that the wanting system can be activated in the absence of direct food cues, which leads to active searching for said food (and, indeed, active searching for other resources, such as sexual partners, water, and even a way to escape danger). Therefore, the wanting system, as typically conceptualized, controls not only the consummatory, but also the appetitive, phase of feeding behaviour, in other words, foraging itself.

Indeed, situations other than unpredictable foraging success lead to an increase in foraging, eating, and hoarding. In particular, food restriction or food deprivation leads to increased foraging and hoarding in hamsters (Bartness & Clein Reference Bartness and Clein1994; Keen-Rhinehart et al. Reference Keen-Rhinehart, Dailey and Bartness2010). This response seems to be initiated by homeostatic regulatory hormones and neuropeptides. Experimental peripheral administration of leptin decreases foraging and hoarding in both hamsters (Buckley & Schneider Reference Buckley and Schneider2003) and titmice (Henderson et al. Reference Henderson, Cockcroft, Kaiya, Boswell and Smulders2018). Leptin and insulin also directly inhibit the VTA, while ghrelin, an appetite-stimulating hormone in mammals, activates it (Palmiter Reference Palmiter2007). In hamsters, at least, we also know that intracerebroventricular injection of the hypothalamic neuropeptides – agouti-related protein (AgRP) and neuropeptide Y (NPY) – increases food intake, foraging, and hoarding (Day & Bartness Reference Day and Bartness2004; Teubner et al. Reference Teubner, Keen-Rhinehart and Bartness2012), but at different timescales. Whereas NPY seems to respond to short-term food shortages and induce immediate, but relatively short-lived, increases in foraging and hoarding, AgRP is more responsive to chronic food restriction and leads to much longer-lasting changes in foraging and hoarding (Thomas & Xue Reference Thomas and Xue2017) of the kind one would expect to occur in small birds when confronted with ongoing winter conditions. Indeed, AgRP neurons reduce their activity the moment food is detected (Ferrario et al. Reference Ferrario, Labouebe, Liu, Nieh, Routh, Xu and O'Connor2016), while activation of AgRP neurons in the absence of food leads to clear foraging behaviours (Dietrich et al. Reference Dietrich, Zimmer, Bober and Horvath2015). Less is known about the role of AgRP and NPY in foraging in wild birds. However, in birds in general these neuropeptides show strong evolutionary conservation with mammals in terms of their amino acid sequences, the neuroanatomical arrangement of the neurons synthesising them in the hypothalamus, and the sensitivity of their gene expression to nutritional state (Boswell & Dunn Reference Boswell and Dunn2017). In red junglefowl, as in rodents, AgRP expression is more sensitive to chronic food restriction, whereas NPY is more responsive to acute food restriction (Lees et al. Reference Lees, Lindholm, Batakis, Busscher and Altimiras2017), suggesting that the same dichotomy may exist in birds as well.

Therefore, when we return to Figure 2 in the target article, “physiological deprivation” does not just lead to increased consumption. Chronic food deprivation, signalled in mammals through respective decreases and increases in circulating concentrations of the metabolic hormones leptin and ghrelin, leads to increased AgRP synthesis, secretion, and neuronal activity, which controls foraging behaviour. It is very likely that this increased foraging driven by AgRP is, in turn, at least in part regulated by the mesotelencephalic dopamine system (Dietrich et al. Reference Dietrich, Bober, Ferreira, Tellez, Mineur, Souza, Gao, Picciotto, Araújo, Liu and Horvath2012; Roseberry et al. Reference Roseberry, Stuhrman and Dunigan2015). The wanting system is therefore enough to mediate these effects on appetitive behaviours without the addition of a new motivational system to the model.

So what is going on under unpredictable foraging conditions? Firstly, we agree with A&G that unpredictability is likely to lead to an increase in glucocorticoids and that this can also activate the mesotelencephalic dopamine system (Piazza et al. Reference Piazza, RougePont, Deroche, Maccari, Simon and LeMoal1996). In that sense, we believe unpredictability is just one of a complex interplay of different inputs that regulate the dopaminergic wanting system. Secondly, it is difficult (at least under natural conditions) to separate unpredictable foraging from at least some form of food restriction, making it very likely that the AgRP-mediated mechanisms we have outlined above also play a role in the increase in foraging, eating, and hoarding under winter-like conditions.

Bringing the knowledge from behavioural psychology to ecological questions is therefore important, and we thank A&G for doing so. Even though we disagree with the details of how they think unpredictable foraging affects foraging motivation, the target article is thought-provoking and raises some very interesting questions that will allow us to better understand how physiological mechanisms have evolved to help animals survive periods of scarce and unpredictable food supply.