Blog Author: Crystal Uminski
Citations: MacArthur, R.H., and Pianka, E.R., “On optimal use of a patchy environment.” The American Naturalist, Vol. 100 No. 916 (Nov-Dec 1966), p. 603-609.
Rodriguez, J. et al., “Does optimal foraging theory explain the behavior of the oldest human cannibals?” Journal of Human EvolutionVol 131 (2019), p. 228-239.
Authors:
Robert MacArthur (1930 – 1972) was an American ecologist who is noted for his work in niche partitioning. (We’ll be reading his paper on niche partitioning in warblers later in the semester.) MacArthur is credited with establishing the use of the hypothetico-deductive (H-D) method (a process of testing predictions based on existing theory) in the field of ecology. MacArthur was considered a pioneering leader in establishing hypothesis-driven science and worked to provide other ecologists with examples of how H-D science should be conducted. MacArthur first used the H-D method in his 1957 paper, “On the relative abundance of bird species.” In the six years prior to MacArthur’s paper, Fretwell (1975) estimated that only 5% of papers published in Ecologyinvolved testing a hypothesis. MacArthur helped to steer ecology away from observation-based work (such as the style used by Clements) and towards experimentation. Eric Pianka (b. 1939) is an American evolutionary ecologist who completed post-doctoral work with Robert MacArthur at Princeton University. Pianka has published over 200 scientific papers and is noted for his work on desert lizards. Pianka has been a faculty member at the University of Texas at Austin since 1968.
Jesus Rodriguez is a Spanish biologist who specializes in the paleoecology of European Pleistocene mammals. He has researched the evolution of mammalian community structure and the influence of climate change on the extinction of megafauna, but now studies human-carnivore interactions using mathematical models. Rodriguez is currently a curator at the National Research Center on Human Evolution in Spain.
MacArthur and Pianka (1966):
MacArthur and Pianka created mathematical models to determine the optimal utilization of time and energy budgets for a species. Their work is the foundation for optimal foraging theory. In “fine-grained” environments (where the prey species are located in the proportion in which they occur), the time predators spend eating each item is divided into search time (TNS) and pursuit time (TNP). When additional prey items are added to the diet (N+1), the change in search and pursuit time can be calculated. When predators increase their diet to include a larger number of prey options, search time can be reduced but pursuit time can increase (some of the additional prey items may be more difficult to find or catch). Figure 1 shows that in a hypothetical predator in a fine-grained environment, the optimal diet contains 4 prey species. Increasing the number of prey species would cause a greater increase in pursuit time than a decrease in search time (which is not beneficial to the predator). The S’ curve in Figure 1 represents a scenario where the density of the prey species is reduced by half. The inclusion of S’ shows that the number of prey species in the optimal diet is not static – in the S’ scenario, the optimal diet includes 5 prey species.
In patchy environments, the hunting time (H) increases when the predators are in less-suitable patches. Delta H represents the increase in hunting time per item when an additional patch of environment is added to the predator’s itinerary. The optimal diet in patchy environments is determined by comparing the change in hunting time to the change in travelling time (T) per prey item caught. Figure 2 shows that in a hypothetical predator in a patchy environment, the optimal number of patches in the itinerary is 3. Increasing the number of patches would decrease the travelling time but increase the hunting time.
MacArthur and Pianka concluded that in productive, fine-grained environments, predators with low search/pursuit ratios should be more restricted in their diet, but this conclusion does not hold in patchy environments where predators spend most of their time searching. Predators with large pursuit/search ratios should show restricted patch utilization in productive environments where food is dense. Optimal predators that have competition for food sources will shrink patch utilization but not the number of prey items in the diet.
Rodriguez (2019):
Rodriguez uses the framework of optimal foraging theory, as well as the lens of human behavioral ecology, to explain cannibalism in an ancestral human species. Rodriguez’s paper relies on the assumptions in MacArthur and Pianka’s paper about prey choice and prey availability and aimed to answer whether humans were a high- or low-ranked prey type in the diet of ancient hominins. If humans were a high-ranked prey type, this would support the theory of nutritional cannibalism, but if the humans were a low-ranked prey type, this indicates a scarcity of the higher-ranked prey types and that cannibalism was a last-resort mechanism for survival.
The different prey species in the ancestral human diet were accounted for in the remains identified at an archaeological site. The seven sets of human bones at the site that had “anthropogenic modification” were considered the cannibalized humans. The remains of 9 other mammal species that were larger than 5 kilograms and had evidence of anthropogenic modification were classified as the other prey types. The caloric content of human and non-human prey types was calculated. MacArthur and Pianka’s model assumes that the predator encounters the resources at random in proportion to the resource abundance in the environment, so the amount of each prey type at the archeological site is assumed to be in proportion to the amount of that prey type in the environment. Rodriguez used a random encounter model to predict theoretical prey densities.
Rodriguez found that rhinos and bison were high-energy prey and that human individuals provided a “moderate amount of energy” (about 13% of the accounted calories). Compared to other prey species, the humans represented a high-ranked resource. Following the Prey Choice Model, high-ranked prey are harvested on encounter. The lower-ranked non-human prey were consumed in proportion to their abundance in the environment, but the high-ranked human prey were harvested in a proportion higher that expected based on estimated abundance. Humans were positively selected as a prey option. Rodriguez proposes that even though humans were a lower-calorie food source, there was a low cost of acquisition, which made humans a higher-ranked prey option.
The nature of human behavior may have violated the assumptions of random encounter, which contributed to the higher-than-expected number of human prey at the archaeological site. The human victims may have been individuals of the same group as the cannibals and died of natural causes (and if this is the case, the energetic cost of “hunting” this prey is zero). Rodriguez concluded that ancestral humans followed the principles of optimal foraging theory and that cannibalism can be thought of as an adaptive strategy.
My thoughts:
After reading a biography of MacArthur which emphasized his importance in the movement of hypothesis-driven science, I was a surprised to see that this paper was so heavily grounded in theory. MacArthur did hint at a need for experimentation, indicating that the delta P value “must be empirically determined” for each species, but other than a brief mention that there is “evidence for this from herons,” MacArthur uses mathematical models to explain optimal diets of predators rather than empirical data collected from experimentation.
I am not very familiar with the field of archaeology, but I did question whether Rodriguez’s results could have been affected by a preservation bias in the fossil record? Were human fossils more prevalent because human fossils were more likely to be preserved than other species?
In most of the papers that we have looked at in class so far, humans are considered exceptions to the ecological frameworks. It is interesting to see how human behavior (even if it is an ancestral species of human) can fit into the principles of optimal foraging theory. Although modern humans don’t necessarily “hunt” for food, can optimal foraging theory still explain some of our food-seeking behaviors today? There’s a whole host of socioeconomic factors that contribute to the “food desert” effect in urban areas, but can some of the concepts of optimal foraging theory be applied to explain why low-cost, high-calorie, processed foods are so prevalent in certain areas?
Both papers talked about diet and feeding efficiency. MacArthur developed a very straightforward model to estimate the relationships between the time for hunting & persuing and the number of available food types or available places for resources (patches). Ideally the shortest time cost should the most preferable way in terms of energy budget. Overall it is easy to follow and the theory behind it is reasonable. However, many of the assumptions are too ideal to fit into the real system. The first thing that comes into my mind is about food preference. The author assumed that the predator would search and chase preys equally, and the food preference is only about the time difference in Ts and Tp. But what about the energy that the food provides? The optimal prey may not always be the most time-efficient one, new prey species may be easier to catch but provide much less energy. Also, he thought that the specialists were those who have high slope of delta p, which means they can still eat something else, it just took a long time for the specialists to capture & eat the new prey. This is not that intuitive for me because I think the food preference of specialists are limited by their specific structures or organs instead of by their ability to capture the food. Nevertheless, I think this is a simple model that can be modified into more complex one in order to take other factors into consideration.
ReplyDeleteThe companion paper studied the reason for cannibalism in ancient humans in the perspective of energy efficiency, and the authors found that human cannibalism can be predicted by maximizing nutrient return and minimizing cost. I am not familiar with the methods they use, but I do think this paper considered the energy cost which is more reasonable than time cost.
I agree that MacArthur and Pianka's paper was oversimplified, but I do not think this detracted from the usefulness of their ideas. Intuitively, it makes sense that a diet which maximizes benefits and minimizes costs will be selected for over time. And, as Crystal mentioned, this may very well still play a role in modern day human diets, whether unconsciously or not. However, like Yuguo, I couldn't help but think about the other types of costs that may play a role in diet composition. Specifically, health costs came to mind. Does cannibalism increase your risk of contracting a disease from your food? Is it dangerous to subdue conspecifics? If you underestimate your potential human prey, what are the odds that you will get eaten instead? Although cannibalism may be justifiable from an optimal foraging theory perspective, it seems that there is more to consider than when predator and prey belong to different species. I also wondered about the relatedness between individuals in a cannibalistic interaction. How closely related were the eater and the eaten? Relatedness between individuals in a cannibalistic interaction is an important consideration in terms of optimizing diet to maximize fitness.
ReplyDeleteI think the MacArthur paper was a great starting point for optimal foraging theory, which came alone ten years later.
ReplyDeleteI have major problem with the Rodriguez paper. They did not really, in my mind, pay close enough attention to what I think is a very valid alternative explanation. The eaten humans were members of the same clan that died of natural causes and were then consumed. They do mention this and don't give a lot of evidence as to why this hypothesis was not explored more. I think the possibility of scavenging should have been much more prevent and presented sooner. The fact that all the human remains showed signs of consumption, and the age of the individuals, leds strong evidence to scavenging, not active predation.
-Miranda
I thought that the MacArthur and Pianka paper was very concise and the clearest of the papers we have read thus far in terms of defining variables and explaining the dynamic relationships between/among organisms. I thought that their taking into account of heterogenous (patchy) environments was particularly useful/relevant, as foraging patterns are dependent on availability of non-uniformly distributed resources. However, it made me question how we would factor in additional fragmentation of habitats, specifically areas organisms reply upon for foraging/travelling to forage locations, driven by humans today – particularly disruptions to these environments that are less visible (e.g., noise pollution, contamination/pollutants).
ReplyDeleteThe Rodríguez et al. paper presents a nice (albeit strange) case study applying optimal forage theory in a context where there is a relatively high rate of encounter between predators and prey (which happen to be the same species). One of the study’s main findings, that hominins were a high-ranked prey type, actually surprised me somewhat. I felt that their assumptions that bone marrow weight/fat would have been similar to modern-day humans was somewhat tenuous, given that humans today have very different diets and lifestyles that could affect these metrics. I would have expected that some of the other mammal species, such as cervids or bison, would have had a higher overall caloric content than humans, making these larger mammals the preferred prey type. I also really agree with and am now wondering about Miranda’s point, that there is a likelihood that few/not all of the consumed hominids were actually hunted. If they have a detailed fossil record from the particular study site, a brief discussion of artefacts of weapons/more consideration of the proof that these individuals were in fact hunted would have been interesting (and given more credibility to their conclusion that hominids were in fact preferred solely based on nutritional value).
I am particularly impressed this week by the PoE chapter (MacArthur) and the companion paper. The older paper is very general and has a broad application which is demonstrated in the companion paper. The companion paper (Rodrigues et al., 2019) takes a fascinating approach to the optimal foraging theory by taking human behavior into account. Essentially the Rodriguez paper approached the question"is cannibalism worth it energetically-speaking?" which makes for some interesting reading. First, I was fascinated that the individuals cannibalized were under the age of 18. I am not sure if this was because life expectancy was so short, but it seemed striking to me that those eaten were very young. It also may be a contributing factor to the conclusions made by the authors that humans are a high-ranked resource based on ease of acquisition. The authors mentioned this near the end of the paper, though I would have liked this expanded upon. At the same time, I think Miranda has a point above. Scavenging was not taken into consideration and, though the idea of hunting humans for food is glamorous, there may be other explanations for the consumption of human flesh.
ReplyDeleteThe table of human body parts and their calories was probably my favorite part. This is not a topic I think about every day, so seeing someone lay this out was fascinating.
I don't have much to say about that MacAruthur and Pianka paper other than it is brilliant in its generality. Clearly the optimal foraging theory can be applied to a variety of studies.
I had a hard time following the MacArthur and Pianka paper. I think I understood their reasoning about patch choice, but I was confused about diet breadth. Specifically, I don’t understand why an optimal predator wouldn’t alter their diet in the presence of a competitor. I don’t see how this is consistent with their earlier reasoning about how the reduction is search time (deltaS) is impacted by different densities of prey. Since pursuit time is unaffected by density, it seems to me that the optimal diet breadth would be affected by a competitor.
ReplyDeleteI liked how clearly the Rodriguez et al. paper stated their hypotheses and the implications of each possible outcome. I’m a bit skeptical about how they calculated the payoffs for each of the prey types, particularly for H. antecessor prey. Since the values of handling time would be different depending on if the cannibalism was of deceased members of their group vs. if they were hunting outsiders, I think a sensitivity analysis of how varying handling time impacts prey type ranking would be useful and could potentially help discriminate among hypotheses about cannibalism.
- David
I was glad for the chance to read the MacArthur and Pianka paper because I have used some degree of optimal foraging theory in my research but had not yet gone back to this fundamental paper. I think this very basic model was a necessary stepping stone to modern foraging theory adapted by Houston, MacNamara, and Lima starting in the 90s and the development of the marginal value theorem only 10 years after MacArthur and Pianka published this optimal foraging paper. Like other models we have talked about, I would argue that this fundamental model is still very useful and relevant today because it is broad enough to be a good starting point in most systems.
ReplyDeleteI liked the Rodriguez paper because the methods set the study up really well to answer the question of whether optimal foraging theory or human culture drove early cannibalism. I think OFT was a good model for them to use because they probably did not have fine resolution data to inform more complicated models. I was glad they acknowledged that they cannot discount the social and cultural effects, but wish they had gone into depth more on this. Humans are extremely social species and I think it would be difficult to tease apart the social and cultural components from OFT.