No CrossRef data available.
Article contents
Mechanistic models must link the field and the lab
Published online by Cambridge University Press: 19 March 2019
Abstract
In the theory outlined in the target article, an animal forages continuously, making sequential decisions in a world where the amount of food and its uncertainty are fixed, but delays are variable. These assumptions contrast with the risk-sensitive foraging theory and create a problem for comparing the predictions of this model with many laboratory experiments that do not make these assumptions.
Information
- Type
- Open Peer Commentary
- Information
- Copyright
- Copyright © Cambridge University Press 2019
References
Abreu, B. F. & Kacelnik, A. (1999) Energy budgets and risk-sensitive foraging in starlings. Behavioral Ecology 10:338–45.Google Scholar
Bateson, M. & Kacelnik, A. (1997) Starlings’ preference for predictable and unpredictable delays to food. Animal Behaviour 53(6):1129–42. https://doi.org/10.1006/anbe.1996.0388.Google Scholar
Bednekoff, P. A. & Houston, A. I. (1994) Dynamic models of mass-dependent predation, risk-sensitive foraging, and premigratory fattening in birds. Ecology 75(4):1131–40.Google Scholar
Fawcett, T. W., Fallenstein, B., Higginson, A. D., Houston, A. I., Mallpress, D. E., Trimmer, P. C. & McNamara, J. M. (2014) The evolution of decision rules in complex environments. Trends in Cognitive Sciences 18(3):153–61.Google Scholar
Higginson, A. D., Fawcett, T. W., Trimmer, P. C., McNamara, J. M. & Houston, A. I. (2012) Generalized optimal risk allocation: Foraging and antipredator behavior in a fluctuating environment. The American Naturalist 180(5):589–603.Google Scholar
Houston, A. & McNamara, J. (1985) The choice of two prey types that minimises the probability of starvation. Behavioral Ecology and Sociobiology 17(2):135–41.Google Scholar
Houston, A. I. (1991) Risk-sensitive foraging theory and operant psychology. Journal of the Experimental Analysis of Behavior 56(3):585–89.Google Scholar
Houston, A. I. (2009) Flying in the face of nature. Behavioural Processes 80(3):295–305.Google Scholar
Houston, A. I. & McNamara, J. M. (1989) The value of food: Effects of open and closed economies. Animal Behaviour 37:546–62.Google Scholar
Houston, A. I. & McNamara, J. M. (1999) Models of adaptive behaviour: An approach based on state. Cambridge University Press.Google Scholar
Houston, A. I., Welton, N. J. & McNamara, J. M. (1997) Acquisition and maintenance costs in the long-term regulation of avian fat reserves. Oikos 78:331–40.Google Scholar
Kacelnik, A. & Bateson, M. (1996) Risky theories: The effects of variance on foraging decisions. American Zoologist 36:402–34. https://doi.org/10.1093/icb/36.4.402.Google Scholar
Kacelnik, A. & El Mouden, C. (2013) Triumphs and trials of the risk paradigm. Animal Behaviour 86(6):1117–29.Google Scholar
Mallpress, D. E., Fawcett, T. W., Houston, A. I. & McNamara, J. M. (2015) Risk attitudes in a changing environment: An evolutionary model of the fourfold pattern of risk preferences. Psychological Review 122(2):364.Google Scholar
McNamara, J. M. (1996) Risk-prone behaviour under rules which have evolved in a changing environment. American Zoologist 36(4):484–95.Google Scholar
McNamara, J. M. & Houston, A. I. (1992) Risk-sensitive foraging: A review of the theory. Bulletin of Mathematical Biology 54(2–3):355–78.Google Scholar
Orduna, V. & Bouzas, A. (2004) Energy budget versus temporal discounting as determinants of preference in risky choice. Behavioural Processes 67:147–56. http://dx.doi.org/10.1016/j.beproc.2004.03.019.Google Scholar
Stephens, D. W. (1981) The logic of risk-sensitive foraging preferences. Animal Behaviour 29(2):628–29.Google Scholar
Target article
Related commentaries (22)
Beyond “incentive hope”: Information sampling and learning under reward uncertainty
Random isn't real: How the patchy distribution of ecological rewards may generate “incentive hope”
Anselme & Güntürkün (A&G) present a mechanistic model of foraging that makes functional sense. All models are simplifications of the complex world inhabited by an organism. Here, we highlight some of these simplifying assumptions made by A&G and point out how these assumptions contrast with previous attempts at understanding the effect of uncertainty and why exploring these assumptions is important.
Firstly, they A&G focus on variation in delays until food is obtained rather than amounts of food. Secondly, they focus on sequential as opposed to simultaneous choice. Thirdly, they assume that uncertainty is fixed over time.
Let us examine the assumption of variability in the delay until food is obtained. In this context, A&G see a functional account as a general explanation of preference for variability. This is at odds with the conclusion reached by Bateson and Kacelnik (Reference Bateson and Kacelnik1997), who argue that a mechanistic account is required. In general, food rewards can vary in delay, amount, or both. Variation in amount is analysed by the theory of risk-sensitive foraging (RSF). As we now discuss, there are places where A&G would benefit from closer links to the RSF theory.
Under RSF, if one assumes that the organism must reach a critical level of reserves at the end of a foraging period (i.e., foraging is interrupted), it can be shown that the organism should be risk averse if food is plentiful and risk taking if food is scarce. This is the famous daily energy-budget rule (Stephens Reference Stephens1981; tests reviewed by Kacelnik & El Mouden Reference Kacelnik and El Mouden2013). The rule is relevant to this quote from A&G: “food unpredictability does not constitute an immediate danger (see Abreu & Kacelnik Reference Abreu and Kacelnik1999; Orduna & Bouzas Reference Orduna and Bouzas2004)” (sect. 2.1, para 1). It is not clear that the cited articles show what is claimed. These lab studies are not about the danger of starvation. They are arguing against the energy-budget rule as an explanation of behavior.
Existing RSF theory is also relevant to this statement: “So, exploiting the immediately available resources is often an optimal strategy, even if those resources are in small amounts” (sect. 3.4, para. 3). Unlike most of the target article, this point is about variation in amount. Assume that an animal sequentially encounters prey items that differ in energy and handling time. Houston and McNamara (Reference Houston and McNamara1985) report that if there are no interruptions to foraging, then survival is maximised by eating everything that results in a gain of energy (i.e., the energy content of a prey item is greater than the energy spent handling). If there is a day-night cycle, then the optimal decision depends on energy reserves and time until nightfall (Houston & McNamara Reference Houston and McNamara1985). This shows that the optimal strategy depends on the ecology of the forager, a point emphasised by Houston (Reference Houston1991), McNamara and Houston (Reference McNamara and Houston1992), and Houston and McNamara (Reference Houston and McNamara1999). How will the approach of A&G perform when faced with simultaneous choice between options that vary in amount?
The model of A&G also assumes that the level of uncertainty remains fixed over time (an organism is either in an S-800, U-200, or U-60 environment). In an ecologically realistic situation, the availability of food will vary over time. A model that assumes variation in uncertainty might make very different predictions about the psychological factors that affect the consumption of food (cf. McNamara Reference McNamara1996). For example, when the uncertainty of the environment is autocorrelated, the current state becomes informative (e.g., Fawcett et al. Reference Fawcett, Fallenstein, Higginson, Houston, Mallpress, Trimmer and McNamara2014; Mallpress et al. Reference Mallpress, Fawcett, Houston and McNamara2015) and might be used to regulate the level of consumption. Similarly, if the animal can choose how much effort to devote to foraging, the optimal decisions depend on the degree to which the environment fluctuates (Higginson et al. Reference Higginson, Fawcett, Trimmer, McNamara and Houston2012).
Examining these assumptions and their role in model predictions is crucial from another perspective. A functional model is based on a particular ecological context and consequently does not necessarily need to explain an animal's behavior outside this context (Houston Reference Houston2009; Houston & McNamara Reference Houston and McNamara1989; McNamara Reference McNamara1996). However, a mechanistic model, such as that proposed by A&G, must also explain behavior during laboratory experiments. This makes two of the three assumptions that we have highlighted above – sequential rather than simultaneous choice and variability in delays rather than amounts – problematic for A&G's account. Many lab experiments use simultaneous choice between options that vary in amounts (see the articles discussed by Kacelnik & Bateson Reference Kacelnik and Bateson1996; Kacelnik & El Mouden Reference Kacelnik and El Mouden2013). A mechanistic account, such as that of A&G, must be general enough to explain behavior in these contexts. A related point is that A&G have produced a spatially explicit model. This is a complication that is not required to evaluate the model using data from laboratory experiments, and a simpler model that looks at choices and energy reserves will have a smaller set of assumptions and be more tractable for examining predictions in the laboratory. If the spatial component is retained, it could be used to investigate behavior when there is a spatial correlation between items (see Fawcett et al. Reference Fawcett, Fallenstein, Higginson, Houston, Mallpress, Trimmer and McNamara2014).
To conclude, it would be interesting to develop a synthesis of optimal levels of energy and RSF. Fat regulation is based on mass-dependent costs (Houston et al. Reference Houston, Welton and McNamara1997). Most models of RSF do not include these costs (for an exception, see Bednekoff & Houston Reference Bednekoff and Houston1994). A general model would predict both fat levels and choices between variable options.