Evolutionary Ecology

Variability and change in subsistence strategies have been key topics in archaeological research. Although many different theoretical frameworks have been utilized in such studies, an evolutionary ecology approach has many distinct advantages for studying subsistence. This annotated bibliography therefore reviews recent developments and critiques in evolutionary ecology and how these theoretical inquiries pertain to archaeological studies of subsistence. Theoretical statements are presented first to address the basic subject matter, the history of the approach, problems within evolutionary ecology, and the application of evolutionary ecology models to archaeology. Although understanding the theoretical framework evolutionary ecology operates under is vital, it is equally if not more important to understand how these concepts are played out in particular case studies. I therefore offer case studies that variably deal with archaeological and ethnographic data. My primary purpose in this presentation is to explore the application of evolutionary ecology to studies of subsistence in the Pacific. However, much of the important work done in this area has come from outside Oceania, primarily in the Americas. Consequently I devote the section following theoretical statements to case studies based in North and South America, while my final section focuses on the few but increasingly consummate studies done on Pacific Islands.

This section reviews current and past literature that focuses on the validity of evolutionary ecology and its applicability to certain problems in anthropology and archaeology. As Broughton and O'Connell (1999) discuss, evolutionary ecology uses natural selection theory to understand behavior and adaptation in an ecological framework, and employs optimality models to understand fitness related behavior. Thus, a primary focus for evolutionary ecology has been optimality models, which are especially useful in understanding foraging and other subsistence activities. Models include optimal foraging, prey choice, patch choice, and central place foraging. Charnov (1976) is one of the original articles proposing a patch choice model, and is important because it is a primary source for all subsequent variations of the model. Likewise, Krebs and Davies' (1987) article is included specifically because it is from the source field of behavioral ecology rather than an anthropological derivation. Articles on human behavioral ecology and risk-sensitive adaptive tactics are incorporated as examples of wider applications than simplistic optimization models.

There are several distinct advantages of an evolutionary ecology approach, including its comprehensive and integrative treatment of behavior, but the primary benefit is testability. Using modeling to build hypotheses, these can then be compared empirically to ethnographic or archaeological cases. If derivations from the model are found, this can be explained within the broader theoretical framework the models were built from, namely evolutionary theory. Problems in applying evolutionary ecology models to human societies are also discussed by several authors. Smith (1987) mentions that perceptual or cognitive mechanisms are not incorporated, but also stresses that optimal foraging models have been largely successful in accounting for the empirical data. Kaplan and Hill (1992) highlight the need for choosing appropriate empirical units in study, and mention that data requirements for some models are not regularly attainable in archaeology. In this regard, the use of qualitative assessments of data, such as a change from high ranked resources to low ranked resources over time, may be more useful in archaeology than the quantitative tests more feasible in ethnographic studies. Winterhalder and Smith (2000) outline a variety of new directions that the field is moving in, including a more recent focus on horticulturalists and agriculturalists in a discipline that has been primarily concerned with foragers. Articles discussing ecosystems theory, historical ecology, and human behavioral ecology are included to widely review the various source fields and parallel work that case studies might utilize in applying evolutionary ecology.

The case studies presented use ethnographic data, archaeological data, or a combination of both in their application of evolutionary ecology to subsistence behavior. Ethnographic data has been the primary source for testing optimization theory, and is useful for archaeologists because it tests and expands upon various models. Many of the questions presented in these case studies, such as group size, sharing, competition and cooperation, probably cannot be analyzed using archaeological data because the data requirements are simply not feasible. However, archaeology offers time depth in understanding change, and there have certainly been successes in adjusting optimality models to archaeological data (see especially Butler 2001). Future research will undoubtedly be concerned with linking archaeological evidence (likely involving subsistence data) to wider prehistoric social practices. Ethnographic studies are also crucial in understanding the taphonomic factors that affect archaeological site formation. Ethnographic case studies in the Pacific, such as Thomas (2002) and Bird and Bird (1997), highlight differential transport and processing as factors that can heavily influence the archaeological record, and the subsequent application of models to that record. Several articles offer potential solutions to these dilemmas based on ethnographic studies.

Ethnographic and archaeological cases are presented in both the Pacific and the Americas, in order to thoroughly survey both a regional and temporal view of evolutionary ecology studies. In both these areas resource intensification, depression, and conservation have been extensively examined (e.g. Broughton 1994, 1997; Butler 2001; Aswani 1997). Relating these trends to more complex social and political processes is a future but increasingly practical goal. As mentioned, studies of horticulturalists and agriculturalists using evolutionary ecology is becoming more widely practiced. Ladefoged and Graves (2000), Stevenson et al. (2002), and Field (2002) are all case studies that use varying aspects of evolutionary ecology to understand agricultural societies. On the opposite end of the spectrum, a major topic lacking in this review is a discussion of pre-Holocene applications for evolutionary ecology. As mentioned, evolutionary ecology has largely focused on hunter-gatherers and thus has had wide applications in early hominid evolution. However, my primary concern is subsistence in Oceania, and therefore a more recent temporal focus (within the last 10,000 years) is appropriate. Subsistence has been a widely discussed topic in the Pacific; the application of a particular approach such as evolutionary ecology affects the types of questions we ask, and ultimately can influence the answers we come up with for the “big” questions such as migration or human diversity. A review of case studies both within and outside of Oceania should highlight the successes, failures, and potential for understanding subsistence through human evolutionary ecology in unique island settings.

1999 On evolutionary ecology, selectionist archaeology, and behavioral archaeology. American Antiquity 64(1): 153-65.

The goal of this article is to call attention to evolutionary ecology as a theoretical framework. Broughton and O’Connell fault Schiffer (1996) as promoting a dialogue between behavioral and selectionist archaeologies, while ignoring the place of evolutionary ecology in the discussion. In response, Broughton and O’Connell first outline the major tenets of evolutionary ecology, Dunnellian selectionism, and behavioral archaeology, using examples and case studies to demonstrate each approach. After a critical review, they proceed to discuss why they find evolutionary ecology the most compelling approach for studying human behavioral evolution, and at what points they believe evolutionary ecology overlaps with the other two frameworks.

Evolutionary ecology uses natural selection theory to understand behavior and adaptation in an ecological framework, and employs optimality models to understand fitness related behavior. Broughton and O’Connell state three primary strengths of the approach: it is comprehensive (capable of making predictions about a wide range of behaviors related to fitness); it is integrative (has the potential to link various aspects of behavior); and it is testable (models can be falsified). Applications of models like the Fine-grained Prey Model are clearly useful in subsistence studies, and Broughton and O’Connell use California as a case study. They show that evolutionary ecology can explain changes in diet breadth, resource depression, and technology.

Broughton and O’Connell give a brief but fair description of Dunnellian selectionism and behavioral archaeology. Although they find similarities in all three theoretical frameworks, they specifically point to faults in the competing two approaches. For Dunnellian selectionism, similarities include the use of Darwinian evolution, an interest in evaluating data empirically, and treating humans in the same analytic terms as other organisms (i.e. both avoid intention as an explanatory device). Broughton and O’Connell claim, however, that Dunnellian selectionism ignores ecological factors, uses evolutionary theory in such a way as to exclude behavior, and rejects predictive modeling. For behavioral ecology, Broughton and O’Connell agree that understanding the relationship between behaviors and the archaeological consequences is important, and applaud the use of ethnographic testing. However, they disagree that a “new social theory” needs to be constructed; evolutionary ecology has already demonstrated an approach successfully applied for decades. As an example the authors discuss early hominid studies, presenting a case study for behavioral archaeology that stops short of any explanatory power, and comparing this to an evolutionary ecology model that has wide explanatory abilities.

This paper is part of an important theoretical debate between three competing approaches. All have positive aspects to contribute, and it is crucial to understand what those positive (and negative) aspects are. This will allow a more critical look at each approach, and how the frameworks are being applied within archaeology. In the discussion of subsistence, evolutionary ecology has contributed immense amounts of work in both ethnographic and archaeological contexts, and, more importantly, in a large percentage of cases an evolutionary ecology approach demonstratively works well. There are, however, several issues that Broughton and O’Connell, and this three-way debate in general, do not address. Although apparently very fruitful for hunter-gatherer societies and even hominid evolution studies, evolutionary ecology is more rarely applied to complex societies. The sheer volume of case studies that have tested models seem to well support evolutionary ecology as a viable framework; I think the next step is what the Broughton and O’Connell article (and Schiffer 1996) hints at – we need to find points at which these differing theories articulate. Only then will we be able to begin understanding how to widen applications of the model to complex societies, and other areas not yet breached.

1976 Optimal Foraging, the Marginal Value Theorem. Theoretical Population Biology 9:129-136.

The goal of this article is to develop a model for optimal foraging by a predator in a “patchy habitat”. Charnov recognizes the large amount of work that has previously been done on optimal foraging and diet breadth, and wishes to expand that database with a specific model for understanding the behavior of animals in a patchy environment. This is the basis for the patch choice models seen in evolutionary ecology and throughout anthropology. Charnov states the general problem: If a predator encounters food items in a patch or clumps rather than scattered evenly across the landscape, how much time should it spend at each patch? Time spent traveling between patches is critical, as is the rate of depression of the patch. The longer an animal feeds, the less available food there will be. An optimal forager will leave a patch when it is worth the travel time to move to an un-depleted patch. More specifically, the predator will leave a patch “when the marginal capture rate in the patch…drops to the average capture rate for the habitat” (Charnov 1976: 132).

Charnov discusses these concepts in terms of mathematical formulas. He factors in variables such as energy cost per unit time in traveling between patches and energy cost per unit time while searching in a certain patch, as well as assimilated energy from hunting and proportions of different types of patches. He uses a graph that plots energy intake versus time in patch to show a curve which reaches a peak before heading steadily down. It is possible to draw a line using his formulas that approximates when an animal should leave the patch. Krebs, Ryan, and Charnov (1974) tested these models with controlled experiments on chickadees. They defined time between the last capture at a patch and the animal leaving the patch as “giving up time” (GUT). They predicted that in a rich environment, GUT would be lower than in an environment with resources across patch types. Indeed, their hypotheses were confirmed. GUT would be lower in a rich environment because there are enough resources that it is worthwhile to move around. Charnov specifies that his theorem is useful where predators increasingly deplete prey the longer they stay. Applications of similar models are seen in the case studies below, especially for shellfish in the Pacific (e.g. Aswani 1997).

This article was written more than twenty-five years ago, but it still is worthwhile considering today. Firstly, it is the original article discussing patch choice, and therefore most work in the area probably builds on this article. Furthermore, anthropologists and archaeologists using patch choice are probably also using this as a primary source, and so any faults in the original work might be compounded by uncritical use of the model. Application of such models to anthropological studies (as has been widely done in evolutionary ecology) might be problematic because culture adds a variable not seen in most animals. For instance, technology such as nets may significantly change the time it takes to obtain food from a patch. Therefore the question becomes whether we can apply relatively simple models to complex human behavior. From the sample of case studies below, it seems that a majority of the work in evolutionary ecology for the last twenty years has focused on proving that you can use modified versions of patch choice and optimal foraging for human populations. The majority of that work has focused on hunter-gatherers, probably because of models like patch choice that specify a need for clumps of resources, travel time, resource depression, etc. Agriculturalists and even horticulturalists do not fit into the patch choice model because they have little need to forage. The importance of this article for evolutionary ecology should be clear; it lays the groundwork for many of their theories. It is also clear from the case studies that the last twenty years have been well spent adjusting such models to our unique species, using primarily ethnographic data, but also some archaeological looks at long term change.

2000 Two Decades of a New Paradigm. In Adaptation and Human Behavior, edited by L. Cronk, N. Chagnon, and W. Irons, pp. 3-26. Walter de Gruyter, Inc., New York.

The purpose of this paper is to discuss the origins of human behavioral ecology, its reception by cultural anthropologists and the public in general, and where it is today. Human behavioral ecology has its roots in work by Williams, Hamilton, MacArthur and Pianka, developed in the 1960’s. This theoretical framework studies organisms as products of natural selection, which favors phenotypes that increase representation of their genes in future generations. They are interested in behavior as a part of the individual that can be acted on by natural selection. Human behavioral ecology applies this framework to human populations to test hypotheses about human behavior derived from evolutionary theory. At its inception, this was a new and somewhat radical development, forging a link between the anthropological and animal sciences.

Early attempts to apply this approach to humans were met by some serious resistance. A good deal of the hostility was towards “sociobiology”; Irons and Cronk argue that all of human behavioral ecology got lumped under this heading. Additionally, most of the objections originated from cultural anthropology. Some of these protests included arguments that the approach was logically indefensible, racist, sexist, and imperialist. Sahlins was a major proponent of these critiques. Irons and Cronk detail this time period and examine the issues that were most prominently debated. One major point is that human behavioral ecologists consider themselves rigorous in methodology, using testable hypothesis; it was ironic to be critiqued as logically indefensible by cultural anthropologists, and many human behavioral ecologists have tested situations Sahlins said were untestable. “Sahlins’ Fallacy”, the idea that some societies don’t have words for fractions and therefore cannot calculate genealogical relatedness, is absolutely ridiculous, since clearly plants don’t know the formula for chlorophyll but they’re still green. These types of arguments show an ignorance of biology rather than a competent critique of evolutionary principles applied to humans. In more recent years human behavioral ecology has gained acceptance, and has made a good deal of advances. Irons and Cronk briefly describe some of the more recent developments, many of which have paralleled work in evolutionary biology and ecology. They cite several areas that need improvement: field methods, models, methods of analysis, and a broader scope to tackle questions not yet asked. They conclude that the field has increased knowledge, and that it is still a viable and expanding discipline.

The importance of this chapter is that it represents a slightly different sub-discipline of evolutionary ecology than the majority of papers presented here. Essentially human behavioral ecology is equivalent in theoretical perspective to evolutionary ecology (looking at differential selection of phenotypes by natural selection, using predictive modeling to understand behavior) but the former tends to focus on issues like parenting, sociality, and mating. Those who label themselves evolutionary ecologists tend to be more often concerned with optimal foraging or prey choice models (and hence subsistence) rather than kin selection or parental investment strategies. Importantly, there is definitely overlap in these two slightly differing focuses, e.g. Sosis (2000) below. When we look at subsistence practices, food is certainly not the only thing we’re interested in; as archaeologists we are trying to understand more than just what people ate, we want to know how this articulated with other parts of prehistoric people’s lives. Human behavioral ecology is definitely compatible and overlapping with the studies that more often call themselves evolutionary ecology, and further integration of the two focuses (i.e. subsistence and cooperation as Sosis demonstrates) might be fruitful. Although both approaches primarily focus on ethnographic studies, they also incorporate testable models for past humans. It is likely that the interests of human behavioral ecology have been given less attention by archaeologists because remains of subsistence activities are much more visible than traces of kinship or parenting. However, that does not mean innovations from this area should not be considered by archaeologists, and in a case like the grandmothering hypothesis it has opened many new doors.

1992 The Evolutionary Ecology of Food Acquisition. In Evolutionary Ecology and Human Behavior, edited by E.A. Smith and B. Winterhalder, pp. 167-202. Walter de Gruyter Inc., New York.

The goal of this article is to discuss the various evolutionary models used to understand food acquisition, and to then present and critique research done in these areas. Kaplan and Hill hope to assess progress in two areas. The first is explaining observed variation in diet and food acquisition strategies among modern humans, and the second is developing models of behavioral decision making. Kaplan and Hill begin their discussion by reviewing the major types of optimization theories that have been used to understand foraging. These include prey choice, patch choice, or a combination of both. Some of the major assumptions the models employ include the idea that all organisms will behave in order to optimize some fitness-related currency, and that that the organism will make decisions based on currencies and restraints. Optimization of food sources should occur when one of the following is true: more food would lead to increased fecundity, more time spent in non-foraging activities would increase fecundity, or more time spent foraging exposes the individual to dangerous conditions lowering fecundity.

Kaplan and Hill discuss the model components of prey choice and patch choice, looking at the formulaic representations of the concepts, and discussing limitations of the models. They highlight the fact that predictions will probably not perfectly match the patterns of real foragers, because models are designed simplistically and have certain restrictions. For example, the model does not take into account individual foraging capabilities (i.e. between an adult and a child). The authors also discuss modifications of models, such as incorporation of central place foraging, acquisition of information, or sensitivity to risk. The further integration of these modifications could allow models more explanatory power. Kaplan and Hill also extrapolate on linking models of individual behavior with long term change, as would be useful in archaeology. Although they point out that data requirements for portions of the model are not regularly obtained by archaeologists, they do cite examples such as the move to agriculture that can utilize foraging models to understand general patterns. Another important point that Kaplan and Hill make is that “energy” may not be the most appropriate currency for analyzing human diet choice; it has been found that some groups favor meat or protein consumption, and thus a more nutrient-sensitive model needs to be employed.

Although in many contexts the foraging models Kaplan and Hill discuss are successful, some applications fail. This chapter is important for evolutionary ecology because it questions some of the models evolutionary ecology is based on, which could be especially critical in their application to archaeology. There are varying reasons why a specific group may vary from a model. Some issues may deal with what empirical units are chosen. As Kaplan and Hill discuss, some “tests” have failed to choose appropriate units for the questions being asked, or have applied inappropriate models; issues such as encounter rates or types need to be fully considered. These are portions of the model that may be impossible to determine archaeologically, because it would be difficult to have detailed enough paleoenvironmental reconstruction to determine encounter rates precisely. What then can evolutionary ecology offer for archaeological purposes, if simplistic optimal foraging models cannot be relied on for every context? Although Kaplan and Hill cite examples of archaeological uses of prey choice models, they don’t discuss problems that might be associated with this. One answer may be in the use of qualitative tests rather than quantitative assessments, which, from Kaplan and Hill’s discussion, tend to have associated problems with specificity. For example, part of the optimization model is that low-ranked resources will drop out of the diet when search costs decrease; the introduction of a new transportation technology, seen in the archaeological record, could then be compared to faunal data for previous and contemporaneous time periods. The general pattern could be seen archaeologically, and hence technology and subsistence could be tied together.

1987 Natural Selection, Ecology, and Behavior. In An Introduction to Behavioral Ecology, pp. 5-23. Blackwell Scientific Publications, Oxford.

Krebs and Davies state the purpose of their book as to explore the relationships between animal behavior, ecology and evolution. They want to look at animals within their specific ecological context and then use evolutionary theory to understand their behavior. Krebs and Davies emphasize that there are four ways to answer a question about why an animal performs a certain behavior, originally stated by Tinbergen: in terms of survival value or function, in terms of causation, in terms of development, and in terms of evolutionary history. Answers in all of these different terms may be correct but differ at each level, so it is important to specify what level is being discussed. Krebs and Davies use differences in male and female lions as an example of these different levels of explanation.

Krebs and Davies address the issue of how much genes affect behavior. They cite various studies which show that gene mutations can affect the behavior of an animal, and that even mating preference or feeding differences can have a genetic component. What they do not discuss in this section is the difference between phenotype and genotype, an important distinction as any anthropologist will tell you; my opinion is that Krebs and Davies are addressing an introductory audience and are keeping things simplistic. Another topic the authors discuss is individual versus group selection. They clearly argue that selection is at the level of the individual, and cite studies done with the clutch size of the great tit. This long-term study shows that the birds will maximize efficiency for the environment they live in. Optimization will occur when the most offspring can be raised for the least reproductive effort. Krebs and Davies end with three arguments that will be further extrapolated in their books. The first is that natural selection will favor individuals that maximize their gene contribution to future generations. The second is that, for a particular environment, there will be optimal trade-off for survival and reproduction. Finally, an individual’s survival is highly dependent on its behavior, and therefore natural selection will favor those animals that maximize efficiency in all aspects.

I chose this chapter specifically because it is an introduction to concepts directly from the source field, behavioral ecology. This is useful because it does not come with any anthropological intellectual baggage. Although sources such as Smith (1987) represent an even-handed and self-critical approach to the adoption of evolutionary biological theory, Krebs and Davies deal with issues specific to the field that they specialize in. Reading this allows us to determine what issues are important for archaeology, and moreover which might be applicable and testable in archaeology. One point Krebs and Davies make that is useful to anthropology is the distinction of the four “why” questions delineated above. The difference in these questions can be seen in heated debates in anthropology over proximate versus ultimate causation. Again, understanding how the source field deals with these problems can help us resolve our own difficulties. Krebs and Davies main point is that for differing ecological settings, differing life history strategies will be favored. For archaeology, this type of hypothesis can be tested by comparing environmental and subsistence data. Of course human societies tend to have complexities of behavior far beyond that of birds. The application of behavioral ecology to archaeology requires both anthropological concerns (how do things like culture affect optimization) and empirical concerns (is it possible to get sufficient data from the archaeological record to test the hypotheses). Both these topics are the focus of much modern debate.

1983 Anthropology, Evolutionary Ecology, and the Explanatory Limitations of the Ecosystem Concept. In The Ecosystem Concept in Anthropology, edited by E.F. Moran, pp. 51-85. Westview Press, Boulder.

Smith takes a critical look at the ecosystem concept, from the point of view of evolutionary ecology. His primary stance is that the ecosystem concept within the framework of systems ecology is not able to make more than very limited contributions to anthropology. He critiques these concepts on several levels and then shows that evolutionary ecology is a more viable framework for understanding the causes of human variation and diversity. This is an extremely important discussion for archaeology because often, as Smith said twenty years ago and is still true to some degree today, people tend to lump ecological approaches; in truth, systems ecology and evolutionary ecology are fundamentally different and invoke very different reasons for cause and change. Understanding these differences will lessen confusion from conflation of approaches.

The first critique that Smith levels at the ecosystem approach is that it is conceived as a cybernetic, self-regulating and self-organizing system. Smith says that systems ecology analyses construct descriptive generalizations, that there is very little interest in developing explanatory theory, and that these approaches are focused on short-term, system-contingent, and inductive strategies (54). As such, there is limited explanatory power for this approach in understanding human social and ecological variance. Systems ecology also looks at a higher level of organization than a human population; they are interested in entire systems and single species are not considered individually. Smith describes this as a top-down view of causality, seeing the ecosystem as self-organizing. Clearly this is problematic for anthropological interests, and also has teleological overtones. What exactly is driving the system to self-organize and self-regulate? Cause becomes a part of this teleology rather than understanding cause at the level of human populations.

Evolutionary ecology differs from this approach in several ways. Firstly, evolutionary ecology uses neo-Darwinian evolutionary theory. This framework allows evolutionary ecologists to use postulates of natural selection theory to build testable models of human social and ecological behavior. Foraging, life history, and group formation strategies can all be considered, tested, and explained in evolutionary ecology, and therefore this clearly has more potential for anthropology than systems ecology. Foraging theory specifically deals with prey choice and diet breadth, and this area is where the majority of studies have focused. The ethnographic and archaeological testing of the models Smith presents can be seen in many of the case studies presented below. Understanding these models allows a better grasp of how to account for diversity of foraging techniques while retaining an interest in general theory.

In sum, Smith does not believe that systems ecology or the ecosystem concept has the ability to posit causal explanations at the level anthropologists are interested in. He says that systems ecology is interested in prediction, while evolutionary ecology is interested problem-oriented research and explanation, and thus is more suited for anthropological studies. These ideas are reflected in Winterhalder (1994). I believe this is an important article for anthropology because it shows why systems ecology is not an appropriate theoretical framework for understanding human variation through time, while evolutionary ecology clearly has a greater potential. Although it seems that most recent ecological approaches do deal with evolutionary ecology, it is still important to understand these differences, to lessen confusion and avoid conflation of divergent approaches. Moreover, this is an important distinction when evaluating the use of ecological theory in case studies.

1987 Optimization Theory in Anthropology: Applications and Critiques. In The Latest on the Best: Essays in Evolution and Optimality, edited by John Dupré, MIT Press, Cambridge.

The goal of this paper is to review at a basic level the tenets of optimization models within anthropology. Smith reviews the logic of the application of these models, where the models come from, and critiques of the use of optimization as an explanatory framework. He begins by saying that the purpose of model building is to simplify the real world to some workable level; goals of the research and the empirical data available are some determinates of that level. Furthermore, models must be couched in some larger body of theory. In this way, the purpose of optimization models (and models in general) is to break a complex world into understandable and testable chunks, while the larger theory binds together various research to provide an explanatory system. Smith lists four defining features of optimization models: an actor, a strategy, a currency, and a set of constraints. Each of these intertwining factors needs to be considered, and they cannot be understood independently.

Smith reviews the sources of optimization models, citing economics and evolutionary biology as the prime source fields. He believes that models derived from economic theory have been more widely applied to production and exchange, while models from evolutionary biology are applied to foraging and reproduction. Describing some differences in borrowings from each field, Smith concludes that evolutionary theory has a wider explanatory scope, and is therefore more useful to anthropologists. Advantages of optimization models include a rigorous basis for generating and evaluating hypotheses from general theory. Smith focuses on four areas to support the use of these models in anthropology: land tenure and spatial organization, foraging group size, reciprocal food-sharing, and optimal birth spacing.

Critiques of optimal foraging are that the model does not incorporate perceptual or cognitive mechanisms, and that the assumed correlation with fitness is not often tested. However, as Smith points out (and many of the case studies below demonstrate), optimal foraging models have been generally successful in accounting for the empirical data. Smith also discusses the sharing of food by family members. He uses an optimization model that incorporates concepts of risk aversion to understand why sharing might take place. While factors such as storage and trade need to be considered, the model as applied to the Ache generally supports the concept that sharing would decrease risk; variation in harvest is correlated with the amount of sharing taking place. Smith follows these discussions of optimal foraging strategies with a look at wider theoretical problems. He reviews the role of optimization models in the process of natural selection, adaptation, and whether optimization models are realistic. In these sections, a variety of concerns, from within anthropology and beyond, are addressed.

The importance of this article is that it specifically deals with anthropological concerns for applications of optimization models. It is a very comprehensive review that begins by detailing all assumptions of the models, and asks a variety of questions as to how the models are viable or unworkable for certain research questions. Some of the issues Smith raises, such as group fitness versus individual fitness, have been hotly debated and many are not fully resolved. Smith does an excellent job of presenting both sides of issues, and leaves many questions unanswered; these are areas that require future research. Smith’s article is useful to archaeologists in that it offers a critical review of optimization models which might lead archaeologists to question their units of analysis or their analytic techniques. Smith himself is clearly an evolutionary ecologist, but that does not mean he thinks the field is perfect. If archaeologists are going to borrow the ideas of evolutionary ecology, we need to stay updated on the problems and advances within the field.

1994 Concepts in Historical Ecology: The View from Evolutionary Ecology. In Historical Ecology: Cultural Knowledge and Changing Landscapes, edited by C.L. Crumley, School of American Research Press, pp. 17-41. Santa Fe, New Mexico.

Winterhalder wrote this chapter in order to offer his perspective (from an evolutionary ecology background) on concepts in historical ecology. He stresses the importance of a clear definition, an awareness of the difference between concept and theory, and an understanding that preexisting ideas may impede rather than progress. In addition, he stresses that we are writing at a time of high concern for environmental well-being, and that this will of course sway our interests. The primary goal of this article is to emphasize the need for historical perspectives in ecological approaches. To further clarify, he breaks down the term “historical ecology”, and discusses what “history” and “ecology” mean. While ecology is fairly straightforward, he explains that history can be a more difficult area. It is the processes that are historical in nature; we aren’t looking to find history as a thing. This is especially important when you consider that no one is going to try to do an ahistorical ecology.

From this Winterhalder begins to ask why a historical approach is important; what can we discern from historical knowledge of something that cannot be derived from its present state? I find this an interesting discussion because it seems that ecological explanations are often considered functional rather than historical, and historical ecology clearly tries to transcend this problem. Winterhalder discusses that it is necessary to understand the history of evolved beings because their entire history is not shown in their phenotype (as it would be on a rock). Similarly you cannot look at an ecosystem and really understand it without a historical perspective. Winterhalder concludes this discussion by saying “historical ecology undertakes the temporal (diachronic) analysis of living ecological systems that in principle is necessary to analyze their structural and functional properties fully” (Winterhalder 1994: 23).

Winterhalder proceeds to discuss the importance of concepts, calling them comparable to the laws of physics. Concepts are a framework for understanding facts selectively. The terms niche and ecosystem are examples; they help organize our thoughts, but are not facts or theories. Winterhalder discusses the ecosystem concept, pointing out that viewing these systems as cybernetic does not allow for the importance of history. Moreover, the ecosystem concept implies that they are attempting to achieve homeostasis at a higher level than neo-Darwinian evolution allows. These arguments are closely related to Smith (1983). The importance of this discussion is that social scientists need to be wary of such ideas, primarily because they are not able to incorporate a necessary historical perspective.

Winterhalder is interested in more than a narrative ecohistory, and he believes that to achieve scientific goals we need to have a theoretically minded approach for understanding environmental variability. He believes that concepts from evolutionary ecology such as patch and patchiness, persistence, and predictability are well suited to this. Winterhalder concludes with a section on adaptive management, which is important to his thesis of the importance of history because it implies a lack of predictability. This discussion highlights some points such as the importance of history for understanding ecosystems, unpredictability in ecosystem behavior, and that focusing on change can be more useful than looking at stability (i.e. the cybernetic self-correcting concept is not helpful). Winterhalder’s discussion is important for evolutionary ecology and historical ecology because it highlights the importance of history for explanation in archaeology, and the benefits of an evolutionary ecology approach. Although he is discussing the social sciences more generally, there are clearly heavy implications for an archaeology that wishes to be scientific and historical. How we use “concepts” as opposed to theory or method is an important point and one that needs to be more thoroughly understood.

2000 Analyzing Adaptive Strategies: Human Behavioral Ecology at Twenty-Five. Evolutionary Anthropology 9: 51-72.

Winterhalder and Smith wrote this article as a comprehensive overview of the origins, successes, failures, and future directions of human behavioral ecology. They begin with a review of the source fields for the development of the discipline, showing graphs of the exponential growth of human behavioral ecology since the 1970’s. The authors show that a complete explanation requires the use of both models of circumstance (how socioenvironmental factors affect costs and benefits of behavioral alternatives) and models of mechanism (how natural selection will affect costs and benefits). Although most models involve complex applications of these two necessities, it can be seen from the outset that archaeology has the potential to fulfill such requirements. Winterhalder and Smith proceed to detail the prey choice model, including ethnographic and archaeological applications. This review is followed by complications for the model, such as how individuals behave in groups. The most interesting example he gives for archaeological applications is the use of the grandmothering hypothesis, which questions the importance of big-game hunting in early hominids and offers alternative models using evolutionary ecology. They also consider mating and parenting strategies, as well as reviewing human behavioral ecology as a progressive research tradition.

There are several fruitful directions that Winterhalder and Smith see for future research. The original models derived from evolutionary biology are being revised with new discoveries and applications. In addition, old models are being enhanced by new methods such as game theory or dynamic programming approaches. New attempts at applying models to archaeology have produced some new data requirements, in both hunter-gatherer and other production-based groups. In addition, human behavioral ecology can help to explain more complex social strategies than has been attempted in the past, and how adaptively directed individual decisions can help explain complex social institutions and processes. Finally, they see a need for synthesis; the discipline needs to incorporate this wide body of work under a more holistic understanding of human adaptive behavior.

The importance of this article is as a synthesis of the work that has been accomplished in the last twenty five years. It is an extremely comprehensive piece, giving background information, problems, and new directions that human behavioral ecology might move in. From an archaeological perspective, this article is especially helpful because of the discussion of non-forager applications. The large amount of work done in human behavioral ecology has been increasingly reflected in archaeology (as seen in their graphs showing percentages of studies done in human behavioral ecology), and hopefully the more recent focus on horticulturalists and agriculturalists will offer archaeologists new ways to look at old questions. In addition Winterhalder and Smith discuss the problems with applying optimal foraging theory to the archaeological record, because most models require data on individual decisions, taken in behavioral time, as well as compounding many foragers over a larger time scale, and taphonomic problems. However, there are many areas where optimal foraging models can provide a helpful framework, such as looking at residential movement, transport of items, intensification of resources, and even domestication and agricultural origins. Evolutionary ecology (or human behavioral ecology) cannot provide for archaeology the fine-tuned quantitative explanations that it can for ethnographic studies; however, it does offer archaeology a testable explanatory framework. Although issues of taphonomy, scale, and time are difficult to overcome, Winterhalder and Smith’s brief treatment of potential archaeological applications is hopeful.

1999 Risk-sensitive adaptive tactics: models and evidence from subsistence studies in biology and anthropology. Journal of Archaeological Research 7(4): 301-48.

Winterhalder at al. claim that anthropological literature rarely uses formal models of risk-sensitive adaptation, even though analysis of risk-sensitive subsistence strategies is an important part of many anthropological and archaeological studies. An analysis of this kind is needed when outcomes of subsistence practices are unpredictable or when consequences for fitness or utility are not direct, and hence could be of great use for many case studies. Their goal is to develop a general conceptual model of risk, using advances in both biology and anthropology to assess the validity of their model. It has been shown in biology that risk-sensitive behavioral capabilities exist in a variety of animals, making it possible that similar adaptations are present in the hominid line. Winterhalder at al. point out that although they focus on subsistence for the purposes of this paper, these concepts could be applied to a much wider range of behaviors. They also stress that there is a critical need for an integrated view of models, and that both biology and anthropology are necessary in the development of this. They warn that anthropologists have neglected to adhere to formal models in the past and ignored similar cases in biology; this has limited our ability to understand the role of risk in subsistence and other adaptations.

The authors first give a definition for risk, which is unpredictable variation in the outcome of a behavior, with outcomes for an animal’s fitness. They argue that the majority of anthropologists are unfamiliar with literature on the concepts and models necessary to analyze risk-sensitive adaptations. Without a theoretical perspective in this area subsistence studies will continue to be seriously lacking. Optimization models carry the assumption that an animal will experience average conditions by which it makes decisions for food procurement, with minimal variation in behavioral and result. Risk-sensitive models are more complex in that they incorporate a stochastic environment, which is probably more closely approximates real life. Performing a risk-sensitive analysis requires two steps. The first is to determine a probability distribution for outcomes associated with a particular behavior. The second step is to determine the fitness value of each outcome. Winterhalder et al. describe the equations formalized in the 1950’s in for this type of problem, and go into detail about the relationships of several of these equations to one another. They conclude that a sigmoid or concave-convex value function is the most likely to be important for subsistence studies. The form of this equation can be constructed to reflect the behavior being studied and the life history of the animal, as well as the environment it resides in.

Winterhalder et al. discuss applications of these models in biology and anthropology. Very helpfully, they list tables of all the risk-sensitive subsistence studies that have been done to date. They discuss various factors that are important in considering risk, including ways shortfalls are dealt with or avoided. Variance in crop types or herd composition, as well as location is one tactic. Others involve diversification of economic activities, sharing and exchange, or varying consumption rates. Storage can also be a major factor. In sum, risk-sensitive analysis allows an empirical basis for evaluating subsistence in a stochastic environment. It requires the specification of a value function (equations discussed above) and a set of possibilities based on behavioral options. This field has a long way to go, and is probably hampered by the complex concepts, which are probably more difficult to apply than optimization theory. In biology and anthropology theory, models, and field studies need further work. The cognitive or perceptual mechanisms an individual uses to assess a risk situation are not understood and should be more closely in anthropology for a better comprehension of this subject matter.

This article is important for anthropology because of a need for an empirical basis for studying risk-sensitive adaptive strategies. Risk is not often quantified as other aspects of subsistence are in optimization theory. Stevenson et al. (2002) is an example of a case study that would have greatly benefited from applications of ideas in this article. Although the theory doesn’t seem to be refined enough yet to be easily applied to archaeological data, it is most certainly a feasible endeavor, and one that should be looked into in future examinations of risk in prehistory. I would liken this article to Charnov (1976), as a piece that will be cited heavily in future literature when the advantages of such models are realized for biology, anthropology, and archaeology.

1983 Optimal Foraging Group Size for a Human Population: The Case of Bari Fishing. American Zoologist 23: 283-290

The goal of this article is to analyze the fishing strategies of the Bari in terms of their optimal foraging group size. The Bari are a tribe of swidden agriculturalists living near the border of Colombia and Venezuela. Bari diet is dependent on cultivated foods for caloric content, but spear fishing makes up nearly the entire source of protein. Because of their reliance on hunting fish with spears, their success is largely dependent on the clarity and depth of the rivers, and hence returns are correlated with monthly rainfall. Beckerman is using data collected on the Bari in the early 1970’s, only a decade or two after the group was initially contacted by the outside world. The author specifically mentions that the data collection at the time was not concerned with optimal foraging theory, and that his model requires further data for testing.

Beckerman notes three stages of the hunt that need to be considered for his model. First is the travel time to the portion of the river that will be used. Travel time is clearly independent of the number of people in the hunting party. The second stage is damming the river. The Bari usually fish by finding an island mid-river, damming with stones from one side of the island to shore, and then, the final step, spearing the trapped fish. The more men in a group the faster the river can be dammed, but the fish per person will then be reduced. Based on these observations, Beckerman derives an equation that looks at the number of fish remaining in a dammed-off patch at a point in time, the area of space between dams, and the density of fish per unit. He notes that this equation is not a reflection of moment to moment actions, but rather an average of an approximate ten-minute period. He also assumes, as is standard for optimal foraging theory, that fishing will stop when overall return per man-hour begins to decline.

One result of Beckerman’s model is that more men on an expedition results in less time actually spent spearing, but it also results in a declining curve for the number of fish taken per man. Therefore the optimal size for a group of fishermen is found when total time per man and catch per man-hour meet. This means that when a group is smaller than the optimal size, adding an extra person will lessen total returns per person but significantly lessen the amount of time spent fishing. The extra time could be spent fishing or performing some other activity. A larger group has the advantage of being able to stop and redirect energy to other subsistence or social activities once returns are adequate, while a too small group has to spend extra time to reach the same returns. Finally, Beckerman makes testable predictions from his models. These include group size increasing with increased patch size, but decreasing with increased fish density. Beckerman also predicts that adding more people to a group should not be problematic, but failure of people to participate (and thus small group size) might be met with hostility.

The significance of this article is its focus on group size for a particular activity. It is an interesting observation that group size for fishing is about the same as the number of men in a Bari longhouse. Archaeologically, it would very difficult to use a similar model, especially when the equation averages ten-minute periods rather than tens or hundreds of years. The archaeological remains of these activities would probably be strictly confined to fish bone, since the wooden spears and temporary dams are unlikely to survive. What might be seen is the seasonal change in fish use, represented by the move from lower to higher altitude longhouses (at higher altitudes you would see fewer fish remains, indicating the seasonal changes in rainfall and related fishing productivity). Archaeologically a single site might show that higher rainfall decreased productivity, but figuring an appropriate group size for fishing would be difficult without ethnographic correlates. This case study shows group size will vary with differing availability of resources, a prediction which has the possibility of being extended archaeologically; however, broader qualitative statements would be more feasible considering the restrictions of archaeological data.

1992 The Use of Optimal Foraging Theory in the Understanding of Fishing Strategies: A Case from the Sepetiba Bay (Rio de Janeiro State, Brazil). Human Ecology 20(4): 463-475.

The goal of this paper is to compare observed behavior of Gamboa fishermen to optimal foraging theory. Begossi found that there are aberrations from the model and proposes explanations for those differences in this specific group, including stock availability, minimal number of patches for shrimp, and competitive aspects of fishing. Begossi begins by briefly reviewing the origins of optimal foraging theory and explaining that central place foraging deals with the special case when travel time is not zero. This is a necessary consideration for fishing, where the boat type can determine the amount of time needed for a trip. Begossi reviews travel time to patches as well as time traveling between patches, time spent in patches, patch productivity, and foraging periods.

Several forms of ethnographic data were collected. Firstly, interviews with fishermen and their wives were conducted. Begossi also participated in 22 fishing trips, a small sample size restricted by limited space in the small canoes. Data was collected on travel time, time spent moving between patches, time spent in patches, number of fish and shrimp caught in each patch, and net trials by patch; in addition the catch was weighed at the end of the trip. Begossi separated analyses by examining weight in kg by patch of non-marketable or very cheap species separate from the marketable shrimp (the item fishers are specifically looking for). Linear regression analyses were performed to test predictions of the optimal patch choice model; energetic returns should be a function of the time spent in a patch, and the optimal patch residence time will increase with travel time into a patch.

Begossi finds that for a significant portion of the trials, foragers were not behaving optimally. Rather, the fishermen were remaining at a patch longer than might have maximized returns; the cost/benefit ratio would have been better elsewhere. There are several possibilities that Begossi offers as explanations for behavior differing from expectations. Firstly, the unpredictable environment of maritime fishing is cited, and Begossi states that fishermen might need more time to assess the patch productivity, and thus stay longer than necessary. However, it seems to me that this would be a consideration for all fishers, not just the Gamboa, and yet other groups follow the expected model. Another possibility presented by Begossi is that because limited patches of valued shrimp exist, it may be more profitable to remain at a patch than to move to one possibly depleted by other fishers. The final and most convincing possibility is that of competition with industrial fishers. Commercial competition is high because the Gamboa rely on shrimp sales for their subsistence; staying longer at one patch may reflect a perceived need to thoroughly fish a patch so that someone else cannot. Depletion of a resource can cause intensification, especially when patches become scarce.

Begossi's sample size is small, and results may be an indication of this. However, the importance of this paper can be seen in the emphasis on competition between industrial and artisanal fishermen. These present day ethnographic studies could be used in historic archaeology to explain behavioral choices of foragers adjusting to outside forces such as colonization and westernization occurred. Additionally, the behaviors of Gamboa fishermen might provide clues for prehistoric changes in resource use. Although time spent foraging at a patch will never be available archaeologically, we might see an increase in a resource over time that does not fit within an optimal foraging model. Social reasons or outside competition may need to be considered as causal factors in these cases; in this way evolutionary ecology has the potential for getting at social and economic change. For the Gamboa, I would not expect them to follow an optimal foraging model because of the influence of needing to sell rather than consume their catch. Although this would essentially put shrimp at the top of a rank order chart, changes in the model based on the differences between actually consuming versus selling a catch seem necessary.

1994 Late Holocene Resource Intensification in the Sacramento Valley, California. Journal of Archaeological Science 21:501-514.

Broughton argues that in the central California late Holocene there were substantial decreases in foraging efficiency over time, supporting a model of resource intensification. This type of modeling can allow for the prediction and evaluation of temporal changes in subsistence activities, but at the time Broughton was writing no fine scale tests had been conducted. Therefore his purpose in writing this article is to test such models using archaeological vertebrate data from the Sacramento Valley. Broughton uses the fine-grained prey choice model in order to understand cost-benefit ratios for different resources, and he describes the model and its tenets. He details the relationship between body size of prey and energy gain per prey item, energy cost per prey item, and prey rank. The basic relationship is that the greater the body size, the better the energy gain, and thus the greater the prey rank up to a certain size. Handling costs are greatest for excessively small or large prey, being smallest for intermediate body size. Broughton argues that prey body size is thus the best measure of prey rank. Although he mentions problems such as group capture with netting, the use of body size has been validated by considerable theoretical and empirical support.

Broughton uses data from nine archaeological sites in the Sacramento Valley. The prehistoric inhabitants would have had access to at least four vegetational zones: freshwater marsh, grassland, oak woodland, and riparian forest. The author discusses various analytical and recovery issues, such as use of different screen sizes and only MNI rather than NISP being available for two sites. More problematically, only a coarse grained temporal view can be obtained because of uneven distributions of radiocarbon dates and a lack of reporting of stratigraphic layers for excavations. Because of this Broughton is forced to compile “mean dates” of occupation, with change only visible between occupations. While eliminating finer scale variability, this technique does ensure that any patterns seen will be reflections of adaptively meaningful change.

Broughton uses simple taxonomic ratios to test his hypothesis of increased resource intensification. The first index he creates is the mammal/fish index. To examine changes within mammals, he uses artiodactyl index, which compares artiodactyls (large mammals, hoofed) with lagomorphs (medium mammals, bunnies). He also creates the fish index, comparing large and high ranked anadromous fishes smaller freshwater taxa. The mammal/fish index is negative, linear, and highly significant, well supporting Broughton’s hypothesis. Taking into account problems that might be a result of differential screen size, he gets similar results. The artiodactyl index does not show any trend through time, and there is no relationship between sample size and the artiodactyl index. The same result can be seen for the fish index, with no apparent pattern. However, when Broughton controls for spatial and seasonal variation in the availability of anadromous fishes, the temporal pattern correlates with the mammal/fish index, again supporting his hypothesis.

Broughton concludes by discussing a variety of other studies that also indicate an increasing reliance on smaller prey items during the late prehistoric. Although some of the species he looks at vary from context to context, the general trend fits the predicted pattern of resource intensification. He also mentions that demographic data on mean and maximum ages of individuals should be studied as additional indicators of impacts on prehistoric faunas by humans. This is an important article because it looks at changes in a wide range of animal taxa over time, and the majority indicate similar patterns. Aberrations from the model as seen with the artiodactyls and lagomorphs are areas where further work into taphonomic problems might resolve the differences; there could also be other reasons for the differences that need to be considered within the model. Other archaeological case studies could benefit from examining some of the problems Broughton had with being forced to group many dates into temporally coarse occupations. This stresses the importance of good excavation techniques that consider taphonomic issues as well as time depth. Evolutionary ecology tends to have high data requirements, making precise excavations imperative.

1997 Widening Diet Breadth, Declining Foraging Efficiency, and Prehistoric Harvest Pressure: Ichthyofaunal Evidence from the Emeryville Shellmound, California. Antiquity 71: 845-862.

The goal of this article is to show that prehistoric people living in the San Francisco Bay had a significant impact on their environment, specifically the sturgeon populations at the Emeryville Shellmound. Broughton discusses notions held about Native American populations; people were thought to have lived "in harmony" with nature, and that only the incoming European populations caused substantial changes in the environment. This view has been more recently rejected on both theoretical and empirical grounds, and Broughton posits to add evidence for this rebuttal of traditional thought. In California the declining abundance of large bodied mammals over time supports models of resource intensification in the late Holocene. Another avenue for exploring evidence of resource intensification is through the mean and maximum age and size of exploited vertebrate populations. Broughton proposes to use both taxonomic abundances and age/size structure to test resource intensification models with ichthyofaunal evidence form the Emeryville Shellmound.

Broughton first discusses the tenets of the fine-grained prey choice model and its relationship to resource intensification; depression of a high ranked resource will cause a forager to increasingly incorporate low ranked prey. Although body size is a good indicator for ranking resources, Broughton mentions that factors such as technological innovations can affect handling costs, and that fat content or similar variables also need to be considered. In addition to taxonomic abundances, exploited populations should undergo demographic changes as indications of harvest pressure; there is a reduction in mean and maximum age and size as animals are taken before their full potential is achieved.

Broughton's unit of analysis is ichthyofaunal remains from the Emeryville Shellmound, a deeply stratified site excavated by several archaeologists. He compiles data from these sites, focusing on ten strata with established radiocarbon dates. Broughton chooses to specifically look at sturgeon remains because these are arguably the highest ranked fish taxon in the estuarine habitat for the area. In addition sturgeon populations would reflect harvest pressure but are unlikely to vary because of any technological changes. The sturgeon index (a quantitative index of sturgeon in relation to other fishes) was calculated for the ten strata, and a decline in sturgeon through time can be seen. Spearman's rank order correlation coefficient is both negative and significant.

The decrease in the sturgeon index could be explained by human harvest pressure, a change in environment, or both. Paleoenvironmental data is obtained by looking at the isotopic composition of estuarine sediment cores. These show no significant fluctuations or directional change in salinity that might have caused a decline in sturgeon populations. To further support the hypothesis that human harvest pressure caused the decline, Broughton examines the mean and maximum age/size of sturgeon through time; harvest pressure should lower these attributes. He chooses to examine the dentary because it was well preserved and abundant. Looking at dentary widths by stratum, there is a significant decrease in size over time. The hypothesis that human harvest pressure caused declines in high ranked resources is upheld.

Broughton concludes with a section detailing other studies which have reached similar conclusions for mammals and shellfish. The overall pattern also indicates that at European contact the reverse situation occurs, most likely because of the massive decline in population with the spread of disease. Broughton indicates that this has implications for modern environmental resource management; although aboriginal groups undoubtedly have great knowledge of their ecosystems and have total rights to their lands, they are not necessarily better candidates as resource managers. This article is important for evolutionary ecology because it is an excellent example of the prey model being tested archaeologically, and offers additional evidence in the form of age/size changes; this shows that there is more than one way to test evolutionary ecology models with archaeological data. Additionally, the article shows that faunal data can help us understand population history as resources become more or less exploited. Broughton’s discussion of contemporary ideas about Native Americans shows that archaeology is important for the present.

2000 Resource Depression on the Northwest Coast of North America. Antiquity 74: 649-661.

Butler is interested in subsistence change on the northwest coast of North America before and after European contact. In an argument similar to Broughton (1997), she discusses the accumulating evidence for substantial impact on environments by prehistoric people. Prior research has tended to consider the area a “garden of Eden” and has often failed to account for change seen in the archaeological record; studies that analyze reduced prey abundance and dynamic predator-prey relationships are needed. Butler uses the prey choice model to examine the mammal and fish faunal record from several sites along the Lower Columbia River, dating from approximately the last 2200 years. Changes in human demography, specifically a dramatic decline in population due to disease at European contact, draw explicit expectations from foraging theory. An increased use of low ranked resources with population growth should be followed by greater use of high ranked resources after European contact, and this hypothesis is verified by archaeological evidence.

Butler’s data is taken from eight sites along the Lower Columbia River, which she claims is an appropriate area to investigate resource depression because of the high estimated population prior to the introduction of European disease; if population density caused resource depletion anywhere it certainly did here. She focuses on both mammal and fish remains, listing NISP by site. After considering possible problems with screen size variably used at different sites, she assigns the faunal assemblages to four phases to establish temporal order. Butler then creates indices to calculate a ratio of high to low ranked resources; rank is based on body size. The first index is the mammal/small fish index, which compares large bodied mammals to mammal and small bodied prey. The second index is the fish index, comparing large fish to small fish. All calculations are based on NISP. When plotted against the temporal phases, both indices show a drop in high ranked resources over time followed by a sharp increase in high ranked resources at European contact, just as the model predicted. Butler also considers individual sites by index and phase, finding that a similar pattern is present. Variation in these graphs point to site-specific functional variation that this coarse grained model. However, Butler points out that results fit expectations remarkably well and open the possibility for further, more fine grained, research.

Although these results are consistent with human resource intensification, it is necessary to evaluate if another factor may have similar results, such as environmental change or introduction of a new technology. Butler addresses both of these possibilities. Unfortunately there are limited sources for detailed environmental change, and this possibility needs more research. As far as technological innovations, Butler discusses how a mass capture technique, like a change in net use, could increase the frequency of smaller fish in the diet without associated high resource depression. This is because the entire group of fish caught by a net is considered the prey type, rather than an individual, and hence provides higher energetic returns. This is however an unlikely explanation for several reasons. Firstly, it is not consistent with the subsequent rise in high ranked resources after European contact. Nor is there any indication in the artifact record of a new technology. Butler’s findings fit well with other studies (e.g. Broughton 1994, 1997) that document human induced resource depression in the late Holocene. She mentions that further studies can be done by examining prey population dynamics, as resource intensification should affect these variables. A potential problem with this analysis is the use of non-animal resources such as acorns. How would an increased reliance on a floral portion of the diet influence the faunal data? Would it be possible to incorporate acorns or other plant products in a rank order abundance analysis? These are questions that could yield a more comprehensive picture. In general, Butler’s analysis uses optimization theory in an excellent analysis of faunal remains. She admits the need for a larger sample size to statistically quantify and further support her hypothesis, but overall she builds a very strong argument. This article is important for subsistence studies because it shows that such analyses can reflect demographic changes, as historically seen in the dramatic population drop at European contact. In this way, subsistence activities could reflect broader settlement patterns and population history as well as social or political change.
Gurven, M., K. Hill, H. Kaplan, A. Hurtado, R. Lyles

2000 Food transfers among Hiwi foragers of Venezuela: tests of reciprocity. Human Ecology 28(2): 171-218.

The goal of this paper is examine data collected on food sharing with the Hiwi of Venezuela. Gurven et al. are interested in the various ecological factors that produce food sharing, and why certain resources are shared with certain individuals or groups. They believe that these are issues that have not been fully resolved, and are important for further study. Some specific questions that they ask include how characteristics of food sources and acquirers determine the amount transferred to others, and how characteristics of nuclear families determine how much food gets transferred between each. They posit that reciprocal altruism is an important factor, but not in the “tit-for-tat” form most commonly used in theoretical statements by anthropologists.

The authors cite three good reasons food sharing among hunter-gatherers should be studied. Firstly, small groups of foragers often have high levels of sharing, making them good subjects for understanding these processes. In addition, sharing in these groups happens in the open and is readily observable. Finally, by understanding food sharing in modern hunter-gatherer groups we have the potential to better understand the evolution of cooperation and sociality in the past. Gurven et al. note that although food sharing has been one of the most widely discussed attributes of hunter-gatherer groups, it is rarely quantified. They cite a lack of “good theory” as a reason for the stunted development in this area, and wish to apply models from evolutionary biology and economics (e.g. evolutionary ecology) to make testable predictions about sharing. Gurven et al. hope to resolve the issue between tolerated theft and reciprocal altruism, two models that have been seen as competing in the past. They believe in a multivariate approach, incorporating biological kinship, geographical proximity, family size, and amounts shared between nuclear families. They preface their discussion of the Hiwi by discussing theory; specifically they treat kin selection, tolerated theft, and reciprocal altruism (tit-for-tat, bargaining, and variance reduction). Although these have been separated in the past, Gurven et al. note the possibility that all three might be working simultaneously; therefore, they are interested in explaining why a combination of these models might appear.

Data was collected by A.M. Hurtado and K. Hill from the Hiwi foragers in the 1987-88 field season. Every fifth resource brought back to camp was analyzed for resource type, original package size, acquirer, weights, and recipients of the food. Primary and secondary sharing events differentiate between the first distribution of a resource as opposed to later transfers. Relatedness was coded biologically between individuals, but the unit analyzed was the nuclear family (due to constraints on how closely every piece of food could be followed to consumption). Linear regression was used for univariate analysis and path analysis was used for multivariate analysis.

Gurven at al. found that several factors were important in how food was transferred. They claim that this is related to several motivations and actors interacting simultaneously. Unfortunately they were unable to quantify models of food transfers to the degree they were interested in, and instead developed more qualitative predictions. They found that the high protein packages, i.e. meat and fish, are more influenced by prior transfers and are transferred more often than other items. They hypothesize that this is due to the larger package size of meat and its greater variability in capture; sharing allows a more even distribution of these important nutrients. One specific problem that the authors discuss is their lack of knowledge of nonfood goods and services that may be traded or shared for food items.

This article brings up some interesting issues for archaeology. To what degree might we see sharing in the archaeological record? Certainly distinguishing between tolerated theft and reciprocal would be impossible to see, but a general pattern of sharing might leave a record. Ethnoarchaeological studies might reveal distinct patterns in refuse disposal in societies that share a lot versus those that do not. Also, understanding the conditions under which sharing is most likely to operate may allow archaeologists to extrapolate on possibilities for the past. Because these are models from evolutionary ecology, they have the potential to be modified for understanding long term change, much as optimization models have been reworked for archaeology.

1982 Why Hunters Gather: Optimal Foraging and the Aché of Eastern Paraguay. American Ethnologist 9: 379-398.

Hawkes et al. are interested in examining the foraging behavior of the Aché in eastern Paraguay. They discuss two competing models for the determination of hunter-gatherer subsistence patterns. The first is from Lee (1968, 1979) who argues that the dependability of plant foods makes it more efficiently exploited. However, this does not explain the archaeological record in that plants were not always heavily exploited. Harris (1977, 1979) implies that animal resources are more efficiently exploited, and that hunter-gatherers only use plant foods because meat has become scarce due to climatic changes and over-hunting. Sahlins (1976) rejects ecological factors as principal determinants, which again is not consistent with the archaeological record. The authors agree with Harris, but for reasons that can be explained by optimal foraging theory and a cost/benefit analysis. Hawkes et al. support their hypothesis with data from the Aché, who have a high proportion of meat in their diet. They also believe that this model has wide explanatory capability for hunter-gatherer groups in differing environments.

Data collection from the Aché was conducted on long range foraging trips. Hawkes and Hill made observations on women and men respectively; they collected information on time spent in travel, search, collecting and processing resources, weighing when possible. Sixty-one gathering days and 58 hunting days were recorded. Hawkes et al. give an in-depth review of the animals taken, tools used for hunting/capture, and the various roles men and women play in a day’s foraging activities. Presenting animal foods versus plant foods, it is clear that animal foods are the primary portion of Aché diet; 80% of foraging calories came from game animals. Although the use of shotguns at times may increase foraging efficiency, Hawkes et al. show that the difference is only about 4% from shotgun to non-shotgun hunts.

If returns from hunting are so high, why do the Aché still forage? One idea is that they are supplementary; on a bad hunting day you can always collect some more seeds. However, Hawkes et al. look at the correlation coefficient for plant calories and meat calories and find no correlation. Instead, they use optimal foraging and patch choice theory to understand the Aché foraging behavior. After a brief discussion of the tenets of these theories, they graph the ratio of calories returned to handling time for each resource, ordered by rank. Plant and insect rank order is separated from animal rank order. The result is a graph showing the optimal diet breadth, where the meat line meets the plant line. This shows that the plant and insect resources in Aché diet actually increase the caloric intake for time invested. This model allows testable predictions. For example, if encounter rates increase, search time is reduced and lower ranked items should move out of the diet. Hawkes et al. conclude with a section detailing the affects of patches on optimality models.

This paper was written at a time when evolutionary ecology models were gaining a hold in explaining subsistence activities. Hawkes et al. provide models and testable hypotheses for the Aché, although their analysis is perhaps not as sophisticated as later case studies. Some possibilities they do not consider are resource depression over time and how that might affect foraging patterns. The Aché they were working with lived at a mission, thereby creating a centralized area that could become over-exploited. Hawkes et al. deal with the patch choice model in depth, and further studies could more thoroughly explore how patchiness in the environment affected Aché foraging. This article is also important because it looks at both plant foods and animal foods in the diet. Many of the archaeological case studies presented here deal with zooarchaeological remains, but do not incorporate the floral portion of the diet. Considering that plant resources are a major portion of many hunter-gatherer diets, this could be a problem in optimal foraging models. Pollen and macrobotanical remains are recovered with increasing frequency from archaeological remains, and it would be interesting to see a study that used rank order for both like Hawkes et al. have done in this ethnographic case.

2002 The Ascendance of Hunting during the California Middle Archaic: An Evolutionary Perspective. American Antiquity 67(2): 231-256.

Hildebrandt and McGuire argue that an increase in large game hunting during the California Middle Archaic can be explained through its importance for male social status and prestige. At this time (between 4000 to 1000 BP) there is an increase in population in California, and a corresponding intensification in subsistence activities. With this intensification an optimal foraging model predicts an increase in hunting smaller prey, and a decrease in the hunting of higher risk large animals. However, the faunal record throughout California is exactly the opposite, with an apparent increase in large game hunting. Hildebrandt and McGuire’s goal is to use an evolutionary ecological approach to explain not only this anomalous increase in large game hunting, but also other phenomena such as an increase in lithic production and rock art, and possibly other cultural aspects.

As stated above, an optimal foraging model does not predict the archaeological data presented by Hildebrandt and McGuire. Alternatively, Hildebrandt and McGuire argue that without the traditional formulation of optimal-foraging concepts, large-game hunting can be viewed as more of an organizing principle, which might have more far-reaching affects for various cultural systems and processes. They posit that rather than more efficiently provisioning families, increases in hunting large game improved hunter social status and prestige; once basal subsistence needs are met, men have the option to behave in ways that will enhance their social status and therefore their individual fitness. Hildebrandt and McGuire also place importance on the ethnographic correlates to this model as they detail their evolutionary ecology approach.

Their primary unit analysis is faunal remains. They use data from a variety of case studies to show that over time there was population increase and a significant increase in big-game hunting. Hildebrandt and McGuire also describe the intensification of obsidian biface production as a hunting-related technology; in other words, it can be explained by the increase in large game hunting by men. They suggest that the making of a toolkit by an individual (rather than by trade) was important and possibly signaled the symbolic importance of hunting implements. Additionally, the authors dedicate a section of their paper to rock art associated with the Middle Archaic, arguing that there is a strong connection between the depicted rock art events and the increase in big game hunting.

An alternate explanation, the Central Place Foraging Patch Choice approach, also offers a rationalization for the increase in large game hunting associated with higher population. This model posits that an increased, more centralized population would deplete nearby resources, so that it is necessary to go farther to procure game; the farther the source, the more a hunter will require higher-ranked species to make the trip worthwhile. The authors clearly state that by looking at the archaeofaunal record alone it is probably not possible to distinguish the Central Place Foraging Patch Choice approach from their model highlighting the social aspects of large game hunting.

Overall Hildebrandt and McGuire build a strong argument, but it seems that evolutionary ecology should offer a way to test the two models. For example, other articles (e.g. Butler 2000, Aswani 1997) discuss the trade-off between travel times versus returns, and conclude that searching for higher ranked resources far away is unlikely. This seems like a perfect test for an evolutionary ecology model; if long distance travel for large game did not fit an optimal foraging model then Hildebrandt and McGuire’s hypothesis would be better supported. This article is important because it attempts to incorporate various lines of evidence (i.e. rock art, technology) to build a more comprehensive theory that is concerned with subsistence and social practices. In addition, their hypothesis could potentially be further supported by applying evolutionary ecology models to further test competing theories, as well as their own.

1997 Fremont hunting and resource intensification in the eastern Great Basin. Journal of Archaeological Science 24(12): 1075-88.

The primary goal of this article is to review trends of resource intensification in the eastern Great Basin and Northern Colorado Plateau. This data is being added to an already growing body of literature for California which documents diminishing foraging efficiency over time. Janetski’s discussion is interesting because it looks at the Fremont, who were farmers, but still reliant on hunting and gathering as a significant contribution to their diet. As farmers, the Fremont were committed to a central place foraging strategy, and given increasing populations they would have experienced resource depression through time. Janetski tests the expectation of a reduction in higher ranked resources during the Fremont period using faunal evidence.

Janetski reviews the prey choice models and assumptions it holds, as well as the concept of resource intensification. Predation efficiency should be empirically measurable by a large animal to small animal ratio; a negative correlation over time shows declining efficiency. To test his hypothesis he uses the artiodactyl index, which compares artiodactyls to lagomorphs (see Broughton 1994). Both of these animals are regularly recovered in Fremont sites. Janetski lists the twenty previously excavated Fremont residential sites that he will use for analysis, breaking down sites first by age and then by NISP of taxa, followed by the calculated artiodactyl index. Possible problems with comparing these twenty sites are in differential use of screen size, variation in ecological setting by location, functional differences, and dating. These problems are largely dealt with through the specific use of artiodactyls and lagomorphs; all sites reported this data. Other factors aside, when Janetski plots the artiodactyl index against time, a clear correlation for a decrease in artiodactyls as compared to lagomorphs is upheld. This supports Janetski’s hypothesis of increased population pressure causing resource depression through time.

Environmental or technological factors could have similar results as resource depression by humans. Janetski looks at climate change during the Fremont period and finds that if anything, conditions were increasingly favorable for artiodactyls. He considers the bow and arrow and mass capture techniques as technological innovations with the potential to affect hunting ability and prey type. However, the artiodactyl decline begins well before the introduction of the bow. Net fragments may decrease rather than increase, suggesting that mass capturing was not a factor; however, this issue remains problematic because of the unclear relationship between technologies preserved in the archaeological record and mass capture. Janetski also looks at butchering and transport as major factors; his detailed treatment of this reveals a need for greater taphonomic understanding of processes affecting transport and deposition. He also discusses the few sites that do not conform to a model of resource depression, suggesting that some of the differences may be explicable by a large screen size failing to retrieve lagomorphs.

The importance of this article is as growing evidence for population increase and resource depression in the eastern Great Basin, in addition to trends seen in California and the Northwest coast. This suggests a widespread pattern across varying environmental situations and heavily supports optimality models as useful tools in understanding prehistory. This discussion is also important because Janetski considers various factors other than human population pressure that may have caused similar changes in subsistence. His article is especially detailed in the treatment of environmental and technological factors, and also considers butchery and transport problems that other articles fail to look at. Although he is clearly in agreement with other authors that more work needs to be done, especially in taphonomic areas, his conclusions are well supported by the faunal data and optimality models. His interest in groups that are farmers with hunting and gathering still making up a portion of their subsistence base raises interesting questions about the how different groups (specifically farmers versus foragers) would deal with increased resource depletion.

1998 Mass Collecting and the Diet Breadth Model: A Great Basin Example. Journal of Archaeological Science 25: 445-455.

The goal of this article is to examine energetic return rates as affected by density of the resource. A low ranked animal (i.e. an animal with small body size) might have higher return rates than a high ranked animal if it is easily collected in mass (e.g. with nets). Madsen and Schmitt look at the consumption of grasshoppers at Lakeside Cave in northwestern Utah. When grasshopper abundance is high, the hunting of higher ranked resources like bighorn sheep may have declined as people were able to more efficiently collect a low ranked prey. They argue that this is an important point because the increased use of smaller prey is usually seen as a decrease in foraging efficiency, and mass collecting needs to be seen as an alternative in archaeological cases.

Madsen and Schmitt argue that the use optimization models in the past have conflated and misused concepts within the theory. They discuss two general predictions that result from the diet breadth and optimal diet models: firstly, that body size is correlated with return rate, and secondly that the abundance of lower ranked resources is irrelevant in determining diet breadth. Mass collecting creates problems for both of these assumptions. Other articles presented here (e.g. Butler 2001) take this problem into account by ensuring that there is no technological change (i.e. nets) associated with changes in subsistence. However, this could potentially be problematic if the artifact does not preserve well. Madsen and Schmitt believe that limitations can be overcome by treating the entire swarm or school as an individual prey item. They also discuss the problem of conflating prey type and food type; the former is a category including all prey items with the same return rate, while the latter refers to a food item such as oranges or sheep. Thus, a change in technology to mass collecting could change the prey type, but food type would remain the same. Furthermore, as Madsen and Schmitt point out, higher ranked resources could be displaced from the diet with no change in their actual abundance if mass collecting of lower ranked resources became prevalent. Prior studies that have not taken this into account assume that larger-sized animals represent higher foraging efficiency, an assumption Madsen and Schmitt wish to disprove for their case in prehistoric Utah.

The cave they are examining has cultural deposits spanning 5000 years, broken into 9 “sets” (layers) for comparative purposes. Sample columns were taken for the recovery of small animal remains, and grasshoppers appear to have been a large part of the diet (as also seen in recovered human feces). Large mammal and grasshopper NISP were calculated. Grasshopper NISP is looked at in 1 m² samples and 12 m² estimates. When rank order abundances for these high ranked versus low ranked animals are calculated, it can be seen that large shifts in subsistence occurred during the occupation of Lakeside Cave. Madsen and Schmitt explain this by the periodic appearance of tens of thousands of grasshoppers shifting focus from large game to the temporary highly efficient small bodied prey. This makes sense within the foraging theory because the huge amounts of grasshoppers that periodically are deposited on the beaches of Salt Lake make a much more efficient prey type as a group than a single large mammal. Thus these smaller animals appear in the diet of the occupants of Lakeside Cave in abundance.

This study is extremely important for archaeological studies using evolutionary ecology because it looks at some of the assumptions of optimization models. Madsen and Schmitt are not questioning the validity of evolutionary ecology theory, but rather are pointing out misuse in its application by archaeologists. Periodic changes in the abundance of a low ranked resource can affect subsistence choices, but this does not necessarily point to increased diet breadth or resource intensification. It is critical that archaeologists and anthropologists fully understand the theory they are deriving models from because variations within the model can be critical to interpretations of data. This article recognizes that the simple application of a standard optimization model will not work without a more in depth understanding of the concepts behind the model; explanatory ability is based on this understanding.

1992 An Optimal Foraging Analysis of Prehistoric Shellfish Collecting on San Clemente Island, California. Journal of Ethnobiology 12(1): 63-80.

The goal of this article is to explain an archaeological change in subsistence through an optimal foraging model of predation. Raab’s work was conducted on San Clemente Island, California, where shellfish collecting was a major portion of aboriginal diet. His focus is the time period from approximately 200 B.P. to 2500 B.P., and he is interested in explaining an increase of turban snails associated with a decrease in abalone shells through time. Raab begins by presenting the major types of shellfish available to people in this area, isolating black abalone (Haliotis cracherodii) as the most accessible and widely used. The black turban snail (Tegula funebralis) is common in similarly accessible areas, but has a higher handling time with less meat and therefore, according to optimal foraging theory, should be procured less. Raab further prefaces his data and methods by emphasizing the importance of shellfish in the prehistoric diet, since he believes shellfish studies in the past have been more commonly ignored or under-analyzed.

Raab’s data is derived from excavations at three coastal midden sites. For the purpose of this study he presents five classes of faunal remains identified from the excavations: Haliotis, Tegula, sea urchins, mammal bone, and fish bone. The rest of the analysis focuses on the data for Haliotis and Tegula. Looking at 10cm intervals within the excavation, Raab uses a ratio of Haliotis to Tegula to see if there is a change over time in the frequency of these shells. From each of the three separate excavations, it is clear from these ratios that the use of Tegula in relation to Haliotis is significantly increasing; the ratio goes from 1.8:1 at the lowest (oldest) level to 20.9:1 at the highest (most recent) level in one excavation unit. By carefully calculating meat yield Raab also shows that Tegula becomes an increasingly important source of food over time.

Raab wants to explain this shift to a higher use of Tegula. Why would people increase use of a food item that is more difficult to process with less meat? His answer is that this trend conforms to an optimal foraging model. Firstly, by looking at handling time and meat intake, Haliotis and Tegula are the only species that would be worth collecting. Secondly, any decrease in the more productive species (Haliotis) would result in an increase in the use of other species (Tegula in this case). Raab shows that the size of the abalone shells especially later in the sequence are around the size that the animal emerges from rock crevices and becomes an easier target for humans; in other words, the animals are being taken as soon as they are available. If humans were overexploiting this resource and there was a resulting decrease in Haliotis, then optimal foraging theory predicts an increase in Tegula, which is exactly what the archaeological record shows. Because abalone was easily depleted, the diet breadth had to be dynamic; in addition it was necessary to maintain mobility to move away from declining resources. Raab cites thousands of small midden sites as evidence for this mobility.

Raab very clearly presents the taphonomic factors that many case studies ignore or discount. He begins by discussing MNI (minimum number of individuals) and associated difficulties with estimating food yields from archaeological sites. He also describes many other problems such as leaching of the shells that can cause problems; all in all he does a thorough taphonomic review for his data. This would be an excellent paper for someone interested in shell analysis to review. Another important aspect of this study is that Raab shows the dynamic nature of food collecting. Over-exploitation by humans resulted in fewer abalones and an increased reliance on turban snails. Other cases of a change in diet may be explicable as a result of over-exploitation, and this case study shows the importance of considering subsistence and the environment as a dynamic process. Although this is a very simplistic study of two food sources that are directly related, it has implications for more complex assemblages where there might be several interrelated items. Raab has demonstrated the possibility that even much more complicated assemblages can be explained with an optimal foraging model. In addition, his discussion of high mobility settlement dynamics for dealing with an easily depleted resource shows that this type of study has wide explanatory power.

Case Studies: Oceania

1997 Roviana Foraging: Patch Choice and Patch Time Allocation. In Customary Sea Tenure and Artisanal Fishing in the Roviana and Vonavona Lagoons, Solomon Islands: The Evolutionary Ecology or Marine Resource Utilization, pp. 296-361. PhD Dissertation, University of Hawaii.

The purpose of this chapter is to analyze the foraging decisions of Roviana and Vonavona fishers using models from optimal foraging theory. Specifically Aswani is interested in answering this question: Do fishers make decisions to maximize short-term efficiency or will they refrain from over harvesting resources to maintain a sustainable harvest? Aswani uses a narrow definition for such a conservation strategy, saying that it requires a reduction of time allocation to areas with declining productivity. This definition is important because, according to Aswani, it allows him to address his question while avoiding the problematic concept of intentionality. One possible problem Aswani considers is being able to differentiate a conservation strategy from an optimization strategy. He explains that to tell these potentially equifinal strategies apart, it is important to know how much time is spent during foraging.

In order to understand how foragers choose what patches to exploit and how long to stay in a particular area, Aswani uses the patch choice and the marginal value theorem patch time allocation models. He is using ethnographic data (observed intake and disposal of fishers) to answer his questions, and potentially could better understand the application of these models for archaeology. These models look at patch types in rank order, and understand use of certain patches as a reflection of return rate balanced by foraging range. In other words the farther you have to go for an item the more highly ranked it should be to be worth it. These models are of course dependent on more complicated factors such as prey mobility and changing patch productivity; both models have problems in their details that Aswani discusses. Aswani proceeds to the in-depth analysis of fishing patterns between hamlets in the Roviana and Vonavona Lagoon, considering patch time and patch time allocation for each area. This intra-regional study is important because it looks at variability in productivity within village habitats, and what influences on foraging strategies that might have. In the next several sections he details the analysis of each village, and finds that although general results from the villages are comparable, factors such as local hydrology and size of neighboring reefs affect local foraging responses. He also outlines the foraging history of a single individual, in order to look at an even smaller scale, and results are comparable.

From his evidence, Aswani concludes that Roviana fishing strategies are not designed to conserve resources; rather, they act to maximize foraging efficiency, and their behavior is consistent with foraging model predictions. When productivity in a patch decreases, Roviana fishers increased time foraging at that patch, even under conditions of diminishing returns, suggesting an optimal foraging strategy. These results show that although conservation (or alternatively, depletion) of resources may occur, this is independent of the variable strategies that Roviana fishers employ. Decisions about whether to stay at a particular patch are unconnected to the long-term sustainability of the environment.

Aswani’s paper is important for several reasons. Firstly, these results show that although local small-scale variation was present, the general analysis was comparable for all the villages studied, as well as the single individual. This suggests that although local variation might be difficult to see archaeologically over time, general trends should be available. Secondly, Aswani does not use any form of intentionality in his study; results are not dependent on whether the Roviana believed they were conserving or not. Because their activities can be explained with evolutionary ecology models it offers a strong device for explanation in the archaeological record. One important factor that Aswani is not concerned with is how taphonomic factors might affect the application of such models to an archaeological assemblage.

1997 Contemporary Shellfish Gathering Strategies among the Meriam of the Torres Strait Islands, Australia: Testing Predictions of a Central Place Foraging Model.Journal of Archaeological Science 24: 39-63.

Bird and Bird are interested in better understanding how the central place foraging model can be accurately used in archaeology. They do an ethnographic study on Meriam intertidal shellfish gathering strategies, field processing practices, and patterns of resource transport to see how these might affect the composition of shell assemblages archaeologically. The Meriam live in the eastern Torres Strait of Australia and are of Melanesian descent. Bird and Bird’s tenet is that the previously mentioned factors may increase or decrease the appearance of different kinds of shells at a site, and need to be considered. As an explanation for why there might be differential gathering and disposal strategies, they discuss the central place foraging model and cost-benefit ratios as used in behavioral ecology. Specifically they discuss that they are testing a model on ethnographic data derived from a well-established theoretical framework rather than from ethnographic observations alone. Their goal is to look at the role of faunal assemblages in understanding site function, settlement patterns, etc., and how processes like transport behavior might affect these interpretations. They present a prediction and hypothesis to be tested, which may have a wider application for understanding past subsistence. Firstly they are interested in the dissimilarities of deposition for certain shellfish resources, especially the contemporary deposition of shells at residential areas. Their hypothesis is that locations of processing activities can be predicted by attributes of the resources.

As mentioned, Bird and Bird are using ethnographic data to investigate their hypothesis. Data collection involved focal individual follows. They thoroughly discuss the different collection and field processing strategies, detailing their collected data. By sampling shell accumulations at randomly selected households, they were able to see which shells were over or under-represented. Because this is an ethnographic study, they know that this results from differential field processing and transport of different prey types. Bird and Bird then use central place foraging models to explain why this differential processing occurs. They do this primarily by looking at the costs and benefits that travel time and field processing time might confer for a particular prey item. For example, if foragers transport the optimal load size, a formula can be used to predict the travel time beyond which it is better to field process. The authors proceed to detail their observations on different species and families of prey types under this model.

Bird and Bird conclude with three implications for archaeological data. Firstly, their study highlights the importance of understanding transport decisions and transport costs. Their findings show that the Meriam data closely follow the predictions of a central place foraging model. This implies that the model could be applied to archaeological cases. Secondly, Bird and Bird discuss that deriving settlement patterns from relative types and amounts of shellfish is risky because foraging range, etc. may have more effect on the assemblage. They emphasize the need for regional approaches when constructing settlement patterns. Finally, they discuss the importance of littoral resources in prehistoric diets. Their results show a pattern of relatively high-ranked resources being under-represented in residential deposits while lower-ranked shellfish are over-represented.

One of the strong points of this article is its use of ethnographic data. Bird and Bird discuss that many critiques of ethnoarchaeology, such as its contingency on historical and cultural context, are valid. However, they feel that their study is legitimate because they are not using the Meriam as an analogy but rather as a test to understand factors that influence variation in subsistence behavior. Their test of the central place foraging model allows an understanding of factors that likely heavily influence faunal assemblages, but which are more easily understood ethnographically than archaeologically. Understanding the cost/benefit ratios for processing and travel time for certain species of shellfish might allow an archaeologist to better explain subsistence patterns. Again, this study has major implications for taphonomic issues, as well as possibilities for the application of evolutionary ecology models to archaeological sites.

2001 Changing Fish Use on Mangaia, Southern Cook Islands: Resource Depression and the Prey Choice Model. International Journal of Osteoarchaeology 11: 88-100.

The goal of this article is to examine changes in fish procurement over time on Mangaia using a prey choice model. The central thesis that Butler promotes is that a shift from higher ranked resources to lower ranked resources occurred during the period from about 1000 to 1700 AD; she posits that this shift is a result of human predation rather than technological or environmental factors. Because fish are a major source of protein on Pacific Islands, Butler considers this an appropriate focus. Butler uses foraging models as a theoretical framework for testing her hypothesis. One prediction of foraging models is that when higher ranked resource abundance is depressed (by humans or other factors), there will be an increase in the abundance of lower ranked resources being taken. In order to see if this pattern is upheld at Mangaia, Butler ranks fish taxa by their profitability. As Broughton (1994, 1995) has demonstrated, this can be done by using body size as a measure for prey rank.

Butler uses NISP of fish taxa for her study, splitting the data into zones (stratigraphic cultural layers). Because the prey choice model assumes that there is fine-grained search (prey are taken when randomly encountered in a relatively homogeneous environment), Butler divides her study into freshwater and marine studies. She then creates a Marine Fish Index and a Freshwater Index, using NISP to look at the ratio of large to small fish. Plotting this figure against zone, it is clear that there is a decrease in the ratio over time, indicating the increased use of smaller (lower ranked) resources. Butler uses Spearman’s rank order correlation coefficients to show that there is both a negative and significant correlation of her indices by zone, empirically verifying a change from low to high ranked resources. This pattern would be expected as the result of human exploitation lowering the encounter rate for high ranked resources.

Butler addresses several issues that could also produce the pattern described above, other than human exploitation. If multiple individuals are pursued at the same time, such as with netting, then many small fish could be taken very quickly; therefore, the biomass of the entire group should be considered in rank ordering. New technologies, then, could affect the rank of taxa, potentially making low ranked prey more profitable than high ranked prey. To test this, Butler examines frequency of fish hooks by zone and compares this to the frequency of other artifacts; a very similar pattern appears. In other words, fish hooks do not vary in any significant way over time. Additionally, there is no directional change in the carnivorous/herbivorous ratio of fish, indicating that there was no major change to a technology that would favor one of these fish types (i.e. fish hooks for carnivorous fish and nets for herbivorous fish). The other major issue that Butler discusses is that environmental change independent of humans could cause similar depression of resources. She addresses this question by very briefly reviewing the paleoenvironmental record, arguing that the marine fishes would not be affected by any increased erosion. This is probably the only portion of Butler’s paper that is slightly lacking; she addresses erosion but does not have information for other types of environmental change and thus discounts them.

This is a critical article for several reasons. First and foremost is the application of the prey choice model to an archaeological assemblage. This article is exemplary in its careful use of the tenets of foraging theory and attention to analytical problems. Butler is also very good about presenting data in a way that it could later be reanalyzed. If I had to recommend an article for someone hesitant about the usefulness of foraging theory in archaeology, this would be it, firstly because of its adherence to theory and secondly because of its careful consideration and testing of alternate hypotheses. Although Butler is clearly not reaching the fine quantitative scale at which many ethnographic studies use foraging theory, her model is appropriate and useful for explaining long term change on Mangaia. Butler’s approach is especially admirable in that she tests multiple hypotheses, something that non-human evolutionary ecologists seem to do often, while it seems more commonly not addressed by archaeologists. In this way she highlights one of the primary benefits of an evolutionary ecology approach – it provides a theoretical framework to test hypotheses. Because of this, Butler’s argument for human related resource depression is very convincing; importantly, it is also testable, and could be refuted or supported with further evidence.

2002 GIS-Based Analyses of Agricultural Production and Habitation in the Sigatoka Valley, Fiji. In Pacific Landscapes: Archaeological Approaches, edited by T.N. Ladefoged and M.W. Graves, Bearsville Press, Los Osos, pp. 97-124.

Field is interested in the role of environmental heterogeneity as the impetus for competitive strategies and territoriality in Fiji. She cites past research that uses differential access to resources as well as hostile social atmospheres as explanatory for settlement patterns. Her goal is to assess both of these conclusions using data from the Sigatoka Valley on Viti Levu, Fiji. From a dynamic perspective, she first addresses environmental issues, looking at temporal and spatial variation in production to assess the affects of such environmental variability. In addition she examines fortifications as an indication of competitive behavior in the archaeological record; specifically she wants to explain why competition would have been favorable and why certain regions would attract more competitors.

In her analysis, Field does not incorporate any type of hill-fort/ring-ditch typology as past studies have done, but rather examines individual features of natural and constructed defenses. An example would be the presence or absence of ditches or walls and their quantity. Eleven classes of archaeological remains constructed by Field are thus the unit of analysis for this research. These sites were identified through aerial photography of 1300 km² of the Sigatoka Valley drainage system. The subsequent creation of a GIS database allowed Field to analyze the topographic location of each habitation site to assess the degree of natural defense it offered. Soils, vegetation, and drainage networks were digitized from maps and photos to simulate variables important for crop cultivation in the area, including topography, soil quality, and hydrology. All were linked as data in the GIS. In the case of wet land cultivation, after using the aerial photography to identify irrigated terraces, these were analyzed within the GIS system to develop a point score indicating the level of potential taro productivity. Dry land cultivation was ranked in productivity by the percentage of land surrounding a habitation site that was below a 45º slope. Again, a point system was developed within the GIS database.

Field finds that 79% of the habitation sites were reliant on dry land cultivation. In addition, moderate and low yield sites make up 70% of the sample. This is an important point because of seasonality in dry land agriculture; a bad year could cause major complications for subsistence needs for a large portion of people. Field believes that this would cause competition and warfare between groups trying to recover from such shortfalls. She thus hypothesizes that groups in high productivity areas with access to wet land cultivation should display an increased amount of defensive sites as they would need to protect their crops from raiding groups, and these were probably the first areas occupied by settlers. Field creates a classification based on presence/absence of defenses and low to high productivity: a defensive strategy, a production strategy, or a defended production strategy. Defensive sites were probably only used in times of conflict as temporary safe-havens, while defended production lands probably represent an attempt at access to productive land while maintaining an aggressive stance. Production sites were generally small and may have had defenses no longer visible archaeologically.

Field effectively shows that settlement patterns in the Sigatoka Valley reflect productivity of the land as well as defensive capabilities. Her research could be greatly extrapolated on with temporal data to look at the chronological history of cultivation and defense. She suggests that evolutionary ecology theory would predict an initial population of the most productive areas followed by ecological “crunches” resulting in competitive behavior. Field’s research is important because it is interested in an evolutionary ecology approach to an agricultural economy. By looking at population pressure and resource intensification, she effectively explains the resulting aggressive behavior and abundance of defensive sites. This is an area that clearly needs more development, and in this case the critical axis of time is yet to be incorporated. However, as Field predicts, evolutionary ecology theory has explanatory potential for this type of study, and future research can further utilize the approach.

2000 Evolutionary theory and the historical development of dry-land agriculture in north Kohala, Hawai’i. American Antiquity 5(3): 423-48.

Ladefoged and Graves are interested in studying the dry-land agricultural field system in Kohala on the big island of Hawai’i. Their goal is to take an evolutionary perspective for looking at and explaining change over time in agricultural practices for the region. They review some efforts of evolutionary ecologists for understanding what particular advantages agricultural practices might confer; their specific goal is to extend these studies and attempt to look at proximate factors over time to see how the role of selection might explain differential persistence of agricultural traits. They believe that evolutionary ecology can help us see how different aspects of the environment affect the success or failure of individual endeavors. Briefly they discuss their view on the role of intention, saying that individuals, especially those with more power, have the ability generate and shape variation. As such, intention has a role in understanding the switch to selection at the group level, but Ladefoged and Graves maintain that selection still acts upon the success or failure of such groups. In sum, they believe that evolutionary ecology offers a way to provide a historical account of evolutionary change, and they do this by looking at the agricultural walls built over time by people at Kohala.

The agricultural walls, then, are the basis of their study. Ladefoged and Graves use GIS to look at the walls and the trails between these enclosures. The walls are essentially field borders, serving as windbreaks, inhibiting erosion, and slowing evapo-transpiration; in this way the walls would sustain and improve harvest output, having a function that could be selected for. Because one of the main goals is to look at differential rates of development in local areas (as opposed to earlier studies that focused more on synchronic studies) they need to control for time. Kirch had previously done this by noting the drenditic patterns of trails, which allows you to figure out which trails were built earlier than others. Unfortunately Ladefoged and Graves were not able to use trails as their unit of analysis because they were not well delineated on the aerial photograph for the GIS. Instead, they used community boundaries recorded in the mid-nineteenth century to form spatial boundaries. The length of each field border wall was used to assign it to a temporal unit (TU 1, 2, and 3). Ladefoged and Graves then proceed to describe and consider production communities, the size of fields, the level of distribution of production, the amount of variability in field size, and the spatial distribution of the fields in order to understand the temporal variation in the area.

Ladefoged and Graves come to several conclusions. They note that there was a move of human populations from optimal (good for exploiting both marine and dry-land farming areas) to less optimal areas (higher cost and energy investment in transport, etc.), correlated with increasing population density. Why would this happen? Their answer considers the shift in the scale of selection that Dunnell discusses. Political factors may have decreased individual subsistence risk by allowing chiefs to secure surplus, but there is also an additional cost of maintaining this larger scale of organization. Although it was advantageous for the original populations to have higher levels of risk but low energy input, later populations were better off with lower risk and higher input, in order to deal with a variable environment and higher population density. This is correlated with increased political organization and the move to complex society. The placement, size, variability, etc. of the agricultural walls reflect this shift.

Ladefoged and Graves make a clear argument that incorporates their archaeological GIS information to changes with subsistence and related social change. They very explicitly state their theoretical orientation and the unit of analysis, discussing how and why the agricultural walls would be selected for or against and how changes in field size, etc., relate to the social and political aspects they are interested in. In addition they are very inclusive, in that they consider both human intention and selection, in their relation to the larger picture of agricultural intensification and the emergence of social complexity. This is an important paper because it uses aspects of evolutionary ecology to explain societal change, namely a move to complex society. Evolutionary ecology has traditionally been restricted to hunter-gatherer societies, but Ladefoged and Graves show that this theoretical framework is capable of much wider, and sometimes more complicated, endeavors.

2002 Explaining subsistence change in southern New Zealand using foraging theory models. World Archaeology 34(1): 84-102.

Nagaoka’s goal is to demonstrate how methodological advances have changed our understanding of subsistence change in New Zealand using foraging theory. She first briefly discusses recent uses of foraging theory in archaeology and the few past studies in New Zealand that have utilized such models. The concept of resource depression could easily be applied to New Zealand as there is an initial use of large bodied prey such as moas and seals, with later exploitation of smaller taxa. Nagaoka claims that the past twenty years has yielded significant research on the effects of resource depression, as well as increased understanding of butchery and transport decisions. Archaeology has benefited with improved quantitative measures for documenting change, as well as more rigorous evaluations of alternative explanations. Nagaoka chose the Shag River Mouth site because it has a large faunal dataset and is from a reliably dated, stratified, and documented site, with a significant time depth to look at long term change in subsistence.

Nagaoka first separates the faunal data into three analytical patches based on location: coastal, inland, and offshore. This is necessary because the prey choice model requires homogeneous distributions; if there are patches the data need to be separated, which Nagaoka does, grouping the taxa by the three areas above. She then uses two indices to compare frequency distributions. The first is called an evenness index, and measures the proportional abundance for each taxon. For example, the inland patch shows high unevenness in the lower layers, probably indicating the high specialization of moa in the earlier time period. Increasing evenness over time is an indication of the depression of moas. The second index Nagaoka uses is of the proportion of high to low ranked resources. She looks at the moa remains (high ranked) in comparison to much smaller (low ranked) quails, and her graph indicates an early high reliance on moa followed by the late use of quail. This supports the findings of the evenness index.

Nagaoka considers factors such as technological or environmental change that might have also led to changes in foraging. No mass capture technologies or environmental variables could explain the shift as convincingly as resource depression by humans. Nagaoka also briefly reviews her similar analysis of the coastal patch data, showing that all research points to a reduction in prey size over time associated with an increase in diet breadth. However, she also mentions that there was differential use of the patches over time; for example, only after the inland and coastal patches experience decline does the offshore patch become important. This indicates that all the patches need to be considered for a holistic argument.

Nagaoka also discusses changes in butchery and transport patterns. By looking at skeletal element transport, she could determine how selective people were being about what they brought back to the site (as based on central place foraging). Over time people became more selective about what portions of the moa they returned with, indicating more field processing of the resource as depression occurred and travel distance increased. This is seen in increases of higher utility elements over time. Seals, however, show the opposite trend, indicating an intensification of the resource over time.

This article is important because it shows that some of the methodological problems with applying evolutionary ecology to archaeology can be overcome. For instance, documenting resource depression has become more quantitative as methods for statistically testing measures of foraging efficiency and diet breadth have been developed. Nagaoka notes that these models are useful in a variety of contexts, and thus have a potentially wide use. Perhaps more importantly, these models can lead to identification of alternate explanations that are then testable. As Nagaoka points out, these models have been widely used in the Americas, and there is no reason they can’t have explanatory power in the Pacific. Like Butler (2001), Nagaoka uses clear and explicit analysis to address and test a hypothesis, although she does not present evidence in a way that it could be easily reanalyzed independently. Further studies such as Nagaoka’s do have the potential to rigorously compare sites across variable environments, and thus be very useful in explaining subsistence and related processes over time.

2000 The Emergence and Stability of Cooperative Fishing on Ifaluk Atoll. In Adaptation and Human Behavior, edited by L. Cronk, N. Chagnon, and W. Irons., pp. 437-472. Walter de Gruyter, Inc., New York.

The goal of this article is to test theoretical models of cooperative foraging using empirical data from Ifaluk Atoll (Yap, Micronesia). Sosis uses this case study to discuss the emergence and stability of cooperative foraging, which he cites as a subject of interest to both biologists and anthropologists, in the study of human evolution and today. Sosis uses equations for mean per capita net rate of energy capture to observe that cooperative foraging will emerge when the return rate of cooperative foragers is greater than the return rate of solitary foragers. Sosis is also interested in optimal group size. Finally, he wants to see what conditions will lead to the stability of cooperative foraging; theoretically, stability is met when the net benefits for free riders are less than the net benefits for cooperators.

Sosis is using approximately 5 months of ethnographic observations on solitary and cooperative fishing on Ifaluk. He watched and meticulously recorded all fishing activity, including triple-checking weights and species of fish caught and the subsequent division of fish to various people/groups on the atoll. Fish is the primary source of protein and fat for the people on Ifaluk, so it a good focus for studying primary subsistence; if horticulture played a larger role it might be more difficult to isolate factors for communal foraging because of conflicting subsistence interests. Sosis used formulas to look at distributions of caloric intake, to compare intake between cooperative and single fishers, and to look at differences between free riders and cooperative fishers.

Sosis’ data shows that cooperative fishing accounts for 87.7% of the fish caught, and that only a very small amount were brought in by solitary fishing. While solitary fishers keep all of their catch, it turns out that cooperating is more profitable. When Sosis employs the equation mentioned above, he finds that the per capita net production and consumption rates of cooperative fishing are much greater than that of solitary fishing. Distribution of fish after a cooperative hunt is variable but Sosis did find some consistencies. Importantly, in every combination of distributions the pay-off for cooperators was always higher than the pay-off for free riders (i.e. those who did not participate in the fishing). This suggests that cooperators have a way to control distributions and that cooperative fishing is stable. The payoffs for cooperative fishing are also determined by the number of men who go out at once, since, as expected by the model, after a certain point too many men decreases average pay-off.

The importance of this discussion is that it relates subsistence practices (fishing) to larger social processes (cooperation, distribution of foods, etc.). Sosis discusses how Western influences have greatly affected cooperative fishing on other atolls in the region because motor boats and refrigeration change the optimal foraging strategies. This shows that a new technology could have greatly affected cooperative foraging prehistorically, and would be an interesting model to test archaeologically. Testing these models in an ethnographic setting can help us better understand what we see in the archaeological record. Potentially an archaeologist could determine the optimal foraging group size and then try to see evidence of that in the archaeological record. How well this would work with an archaeological sample is debatable, since this case study shows that distribution after capture is a major factor, and it would probably be impossible to look at who is actually doing the fishing versus who is profiting from the fishing. In other words, it would be difficult to tell archaeologically how many people were freeloading, to discover if there was stability in cooperative foraging or if it was a fluctuating occurrence. Although the units and methods used by Sosis are excellent for understanding an ethnographic case, application of the models to archaeology would have to incorporate both a much greater time depth, taphonomic problems, and some sort of variable for factors like freeloading that can’t be seen archaeologically.

2002 Productive Strategies in an uncertain environment: prehistoric agriculture on Easter Island. Rapa Nui Journal 16(1): 17-22.

The purpose of this paper is to relate environmentally differing agricultural zones on Rapa Nui to the development and maintenance of a hierarchical social organization. Stevenson et al. begin their paper by briefly reviewing Dunnell’s waste hypothesis, the idea that highly variable environments might favor people who put time into “wasteful behaviors” (i.e. monuments) instead of into reproduction; this behavior keeps the population low and thus avoids hardship in bad years. In this way, the authors associate agricultural practices with a hierarchy necessary to oversee such monumental “waste” behavior. Why a hierarchy is necessary for this activity to take place is not treated. They continue by briefly reviewing the environmental constraints in Rapa Nui, the two most important for agricultural being wind and rainfall; these vary by altitude. The broad environments for Rapa Nui are also shown and briefly described in terms of agricultural risk. Three questions are proposed to be discussed: risk factors for rainfall agriculture and its variation, variable landscape use across habitats, and the chronology for agricultural technologies.

Stevenson et al.’s unit of analysis is the morphologically distinct agricultural features found throughout the island. The authors believe that a major problem in landscape interpretation has been the recognition of these features as part of field systems. The six features he identifies are mulched soils, veneer surfaces, stacked boulder concentrations, pu (steep sided depressions), manavai (stacked stone gardens), and planting circles. Each type of agricultural feature has advantages, such as reducing wind damage or increasing water retention. Stevenson et al. also detail which plants most likely would have been grown in the different feature types. They then look at two distinct regions, the dry lowland Heki’i District and the Akahanga/Vaihu upland region. Agricultural features as above defined are identified in the differing zones of production (coastal, lowland inland, lowland interior, and upland interior) for each region. Radiocarbon and obsidian hydration dates from previous studies on ahu and agricultural features were used to assign dates; figures are not listed. Stevenson et al. find that the coastal zones were occupied since initial settlement, while the inland zones had limited use until around AD 1400. It is suggested that differential access to resource zones played a role in social developments. The authors believe that more labor-intensive agricultural activities correlate with more intense ceremonial construction. In addition, their final position is that because of low returns on high labor investment, there needed to be a strict hierarchy in place early in the settlement sequence.

This article is a somewhat different take on an evolutionary ecology perspective, because it focuses on agricultural aspects of subsistence rather foraging techniques; I used this article specifically because of this differing concentration. Although Stevenson et al. do not explicitly use evolutionary ecology models, they are interested in the effects of ecological variation and are using risk as a potential determinate for behavioral strategies (they would benefit from Winterhalder et al. 1999). The authors take an important step in examining morphologically distinct agricultural features that apparently have been previously ignored. However, the article could have been greatly improved by the explicit use of evolutionary ecology modeling. For instance, they look at varying features of Rapa Nui environments to see which are more agriculturally amenable. However, these ideas are never quantified in any manner, and although they discuss distance from coastal areas as a factor they fail to extrapolate on its relative significance. The application of a revised patch choice model could be used to incorporate traveling distance and production at different elevations. There are undoubtedly many associated problems with finding empirical archaeological data to test such models, but if possible it would yield much more accurate and verifiable answers than Stevenson et al. have produced. Their association of ecological information with an “ideologically strong and strictly ranked hierarchy” (21) seems vague at best, and untestable at worst. The article tends to skip from point to point, making a logical argument difficult to follow. I would argue that a more specific and detailed treatment of differing agricultural areas, perhaps utilizing evolutionary ecology, could produce results less comprehensive but more verifiable results. From that foundation more extensive questions about social systems such as Stevenson et al. attempt to answer could become more feasible.

2002 An evaluation of central-place foraging among mollusk gatherers in Western Kiribati, Micronesia: linking behavioral ecology with ethnoarchaeology. World Archaeology 34(1): 182-208.

Thomas looks at foraging models (prey choice and central-place) in an ethnoarchaeological setting in order to help explain variability in the archaeological record. He is interested in how behavioral ecology can help to strengthen behavioral inferences, by looking at the variable costs and benefits of different subsistence activities. Specifically there is a focus on mollusk culling and transport, which Thomas says has lent itself to explanation by “typical” patterns in the past. Thomas’ goal is to better understand how to tie complex ethnographic information to archaeological data, and he uses foraging models to do this.

In order to conduct his study, Thomas looked at several contemporary groups exploiting shellfish on atolls in Western Kiribati, and interviewed foragers. In addition several excavations, dating to AD 1700-1920, were conducted. He used the gathered information on shellfish to review foraging and central place modeling. Looking at cost/benefit variables, he also discusses trade-offs between field processing and transport, reviewing the literature on the subject and applying relevant concepts and formulas. He then compares archaeological data (shells) with the ethnographically derived models, specifically pointing out variations from the models. Thomas makes an interesting comment regarding the debate about how much indigenous people used conservation strategies in their decisions about procuring food from the environment. While only commenting briefly on this subject, Thomas goes on to make a valid point: the low density of specific species (e.g. tridacnids) in an archaeological site could be explained by field processing (as opposed to bringing the shell and all back to a central processing area) rather than being explained as a conservation measure.

Thomas has attempted to link decision making in an ethnographic setting to possible indicators of such behavior in the archaeological record. He tries to test empirical archaeological data against models derived from ethnographic research. However, as Thomas himself discusses, “the very nature of the archaeological record… often cannot capture the variability proposed in even the simplest foraging model” (Thomas 2002: 203). Along with other problems that Thomas mentions, such as the difficulty of determining how much environmental change is due to predation, there need to be ways to explain derivation from a model. Thomas cites Yesner (1981) as an example of a way to do this. Yesner looked at discrepancies in his data and considered several different possibilities to explain the derivations, including prey type aggregation, non-food yield, and social value. Thomas concludes that Yesner’s study is important because it considers factors unrelated to return maximization by using ethnographic data (specifically analogy to modern day Aleut).

Thomas’ modeling techniques could be very useful in attempting to explain archaeological material; his study is reminiscent of Bird and Bird (1997). By deriving models from ecological contexts in the present it is possible to then test those models with archaeological data. Outside of testing models against ethnographic data, Thomas is also interested in taking the study one step further, to explain variation from the models. Why might there be unpredicted differences in the archaeological record? Yesner’s approach is interesting, but, as mentioned, the problem of the archaeological record being unable to capture variability in even simple ethnographic models was not further explored. How valid is the use of ethnographically derived models for archaeological data? This is a difficult question, which has been partially but not fully addressed by Thomas. A much more detailed discussion of how taphonomic factors might contribute to derivations from the model is necessary, prior to considering the factors Yesner (1981) mentions.