In Pursuit of Mobile Prey: Martu Hunting Strategies and Archaeofaunal Interpretation

Home , Acacia dictyophleba, Marginal value theorem

AQ74(1) Bird 1/2/09 10:54 AM Page 3

IN PURSUIT OF MOBILE PREY: MARTU HUNTING STRATEGIES AND ARCHAEOFAUNAL INTERPRETATION

Douglas W. Bird, Rebecca Bliege Bird, and Brian F. Codding

By integrating foraging models developed in behavioral ecology with measures of variability in faunal remains, zooarchaeological studies have made important contributions toward understanding prehistoric resource use and the dynamic interactions between humans and their prey. However, where archaeological studies are unable to quantify the costs and benefits associated with prey acquisition, they often rely on proxy measures such as prey body size, assuming it to be positively correlated with return rate. To examine this hypothesis, we analyze the results of 1,347 adult foraging bouts and 649 focal follows of contemporary Martu foragers in Australia’s Western Desert. The data show that prey mobility is highly correlated with prey body size and is inversely related to pursuit success—meaning that prey body size is often an inappropri- ate proxy measure of prey rank. This has broad implications for future studies that rely on taxonomic measures of prey abundance to examine prehistoric human ecology, including but not limited to economic intensification, socioeconomic complexity, resource sustainability, and overexploitation.

Mediante la integración de modelos de forrajeo de la ecología del comportamiento con las medidas de variabilidad en restos de fauna, estudios zooarqueológicos se han realizado importantes contribuciones para entender la prehistoria del uso de los recursos y las interacciones dinámicas entre los seres humanos y sus presas. Sin embargo, cuando los estudios arqueológicos no están en condiciones de cuantificar los costes y beneficios asociados con la adquisición de presas, a menudo dependen de parámetros de sustitución como presas tamaño corporal, suponiendo que se observa una correlación positiva con la tasa de retorno. Para examinar esta hipótesis, se analizan los resultados de 1,347 episodios de forrajeo y 649 focales de la siguiente de Martu, contemporáneo cazadores-recolectores en Australia del Desierto Occidental. Los datos muestran que la movilidad de presa se encuentra altamente correlacionado con el tamaño corporal presa y está en relación inversa a alcanzar el éxito— lo que significa que el tamaño corporal presas es a menudo una medida inadecuada representación de presa rango. Esto tiene amplias implicaciones para los estudios futuros que se basan en medidas taxonómica de la abundancia de presas para estu- diar la ecología humana prehistórica, incluyendo pero no limitado a la intensificación económica, socio-económica comple- jidad, la sostenibilidad de los recursos, y la explotación excesiva. oraging models derived from behavioral tion, including its relationship to constellations of ecology have had a dramatic effect on our changes in life history, fertility, and health (e.g., Funderstanding of variability in archaeofau- Hawkes 2003; Lambert 1993; O’Connell et al. nal assemblages. Recent reviews of the field 2002), prestige and sociopolitical complexity (e.g., demonstrate that over the last three decades zooar- Bliege Bird et al. 2001; Fitzhugh 2003; Hawkes chaeologists have, with varying degrees of success 2000; Hawkes et al. 1991; Hildebrandt and Jones and sophistication, increasingly applied the theo- 1992; Hildebrandt and McGuire 2002; Kennett retical underpinnings, attendant logic, and predic- 2005; Lupo and Schmitt 2002; McGuire and Hilde- tions of these models (Bird and O’Connell 2006; brandt 2005; Sassaman 2004), tool use and tech- Lupo 2007; Shennan 2002). Many of these studies nology (e.g., Bettinger et al. 2006; Bright et al. address long-standing arguments about the role 2002; Kuhn and Stiner 2001; Ugan et al. 2003), that hunting mobile prey played in human evolu- regional colonization and economic intensification

Douglas W. Bird I Department of Anthropology, Bldg. 50, 450 Serra Mall, Stanford University, Stanford, CA 94305-2034 ([email protected]) Rebecca Bliege Bird I Department of Anthropology, Bldg. 50, 450 Serra Mall, Stanford University, Stanford, CA 94305- 2034 ([email protected]) Brian F. Codding I Department of Anthropology, Bldg. 50, 450 Serra Mall, Stanford University, Stanford, CA 94305- 2034 ([email protected])

3 AQ74(1) Bird 1/2/09 10:54 AM Page 4

4 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 (e.g., Beaton 1991; Bettinger and Baumhoff 1982; prey types based on estimates of post-encounter Grayson and Delpech 2002; Jones 1991; O’Con- return rate (energetic yield relative to handling nell 2007; Stiner and Munro 2002; Stiner et al. costs, e/h) and assumed that the inclusion of highly 2000), origins and diffusion of agriculture (e.g., ranked prey should scale with their local availabil- Alvard and Kuznar 2001; Barlow 2002; Hawkes ity. They then demonstrated that, as predicted, most and O’Connell 1992; Layton et al. 1991; Redding resources added later in time were very low ranked 1988; Russell 1988; see contributions to Kennett and that their inclusion was a function of fluctua- and Winterhalder 2006), gender and divisions of tions in the encounter rate with higher-ranked prey, labor (Kuhn and Stiner 2006; Waguespack 2005; not their own abundance. Much of this work iden- Zeanah 2004; see contributions to Stiner 2005), tified climate- or predation-related depression or and resource sustainability and overexploitation extirpation of highly ranked prey as a principal (e.g., Barnosky et al. 2004; Broughton 1994a, 1999; determinant of increasing diet breadth and intensi- Grayson 2001; Jones et al. 2004; Jones et al. 2008a; fication. Many subsequent studies of prehistoric Jones et al. 2008b; Lyman 2003; Munro 2004; diet change and, to a lesser extent, patterns in patch Nagaoka 2002; O’Connell et al. 1982; Porcasi et use have led to similar inferences (e.g., Basgall al. 2000; Simms 1987; Waguespack and Surovell 1987; Beaton 1991; Bouey 1987; Braje et al. 2007; 2003). Broughton 1994a, 1994b, 1997, 1999, 2002; Can- The zooarchaeological utility of an approach non 2000, 2003; Codding and Jones 2007; Edwards from behavioral ecology is well illustrated in the and O’Connell 1995; Erlandson 1991; Glassow development of arguments about “Mesolithic,” and Wilcoxon 1988; Grayson 1991; Gremillion “Archaic,” or “broad-spectrum revolutions.” These 2004; Hildebrandt and Jones 1992; Jones 1995; transitions include long-recognized shifts in ter- Jones and Richman 1995; Jones et al. 2004; Jones minal Pleistocene and Holocene subsistence strate- et al. 2008b; Kennett 2005; Klein et al. 2004; Man- gies around the world, whereby many nino and Thomas 2002; Nagaoka 2002, 2005; hunter-gatherers intensified resource use and began O’Connell et al. 1982; Perlman 1980; Porcasi et al. to exploit a wider array of prey than did their pre- 2000; Raab 1992; Russell 1988; Simms 1987; decessors (e.g., Clark 1952; Willey and Phillips Szuter and Bayham 1989; Wohlgemuth 1996; 1958). Well-known explanations for this change Wright 1994; Yesner 1989, 1994; Zeanah and have made reference to increasing familiarity with Simms 1999). resource variability, technological innovation or This work on changes in resource exploitation diffusion, intensifying social relations, and declines has produced important questions about how to in the abundance of certain prey, possibly as a result establish reliable measures of resource rank (see of climate change and/or human population growth discussion in Bettinger 1991, 1993; Bird and (e.g., Binford 1968; Braidwood 1960; Cohen 1977; O’Connell 2006; Broughton and Grayson 1993; Flannery 1969; Lourandos 1983). Though all these Lupo 2007). Some of these problems have been hypotheses are plausible, and combinations of them addressed satisfactorily for sessile resources (e.g., are probable, none of them has generated clear pre- plant foods, shellfish) through a combination of dictions or well-defined tests about the nature and ethnographic, ethnohistoric, and actualistic studies direction of intensification. (e.g., Barlow and Metcalfe 1996; Bettinger et al. Early applications of foraging models, espe- 1997; Bird and Bliege Bird 2000, 2002; Bird et al. cially the encounter-contingent prey choice model 2004b; Cane 1989; de Boer 2000; Jones and Rich- (PCM), changed this situation (e.g., Bayham 1979; man 1995; Kelly 1995; Madsen and Schmitt 1998; Beaton 1973; Botkin 1980; O’Connell and Hawkes O’Connell and Hawkes 1981, 1984; Petruso and 1981). One of the principal predictions of the PCM Wickens 1984; Raab 1992; Reidhead 1976; Simms is that the range of resource types exploited will 1985, 1987; Simms and Russell 1997; Smith et al. increase with declines in overall foraging efficiency 2001; Talalay et al. 1984; Thomas 2002; Thomas (energetic yield relative to search and handling 2008; Thoms 1989; Ugan 2005a). However, the sit- costs [see the subsequent section and Stephens and uation for mobile prey has proven less tractable, Krebs 1986]). In accordance with the PCM, early and thus far, few zooarchaeological analyses have archaeological applications generated rankings of incorporated actual (or actualistic) return rates for AQ74(1) Bird 1/2/09 10:54 AM Page 5

Bird et al] IN PURSUIT OF MOBILE PREY 5 mobile animals (Egeland and Byerly 2005; Lind- sentation of large-bodied prey in the archaeofau- strom 1996; Lupo 1998, 2006; Madrigal and Holt nal assemblages shows no significant diachronic 2002; Thomas 2008). This is due in part to (1) the decline, the frequency of different taxa of small- fact that in many cases the opportunity has long bodied prey changes dramatically. Slow-moving, passed to record ethnographic return rates for the easily captured prey types (tortoises in the Mediter- resources of interest and (2) the presumed difficulty ranean; the flightless duck, Chendytes lawi, in Cal- of attempting to evaluate variability in pursuing ifornia) are common in the early components but and capturing mobile prey through replicative decline in relative frequency over time, probably experiments (see Thomas 2008 and Simms as a result of human overexploitation. In fact, in the 1987:43–46 for thorough treatment). California case the relative proportion of large- To circumvent such problems analysts have bodied artiodactyls increased over time, while the often used proxy measures of post-encounter return flightless duck eventually went extinct by about rate, the most common being prey body size, 2500 B.P. (Jones et al. 2008a). In both cases,as prey assuming that rank scales closely with size and that with low pursuit costs declined in importance, fast- prey of similar size have similar rank (e.g., Bay- moving prey (e.g., lagomorphs in the Mediter- ham 1979; Broughton 1994a, 1994b, 1997, 1999, ranean; sea otters in California), presumably lower 2002; Butler 2000, 2001; Byers and Broughton ranked because they were more difficult to capture, 2004; Byers and Ugan 2005; Byers et al. 2005; increased proportionally in the assemblages. Grayson 1991; M. E. Hill 2008; Janetski 1997; The primary problem, well recognized in these Nagaoka 2001, 2002; Szuter and Bayham 1989; studies, remains: we have very little observational Ugan 2005a, 2005b). A number of researchers have data on how post-encounter return rates are deter- criticized this approach, pointing out that the cir- mined by the behavior of game relative to hunters cumstances and technology associated with col- (though see Hawkes et al. 1991), leaving analysts lecting resources en masse can make some small little more than their intuition to generate expecta- prey more profitable than would be predicted by tions about the effects of such on the ranking of body size (Grayson and Cannon 1999; Madsen and prey in archaeological contexts. This is less likely Kirkman 1988; Madsen and Schmitt 1998; Schmitt to be a problem in cases similar to those docu- et al. 2004; Sutton 1995). Ugan (2005a) has mented by Stiner and Jones given the extreme dif- addressed this issue with data from a variety of ferences in the types of prey in those assemblages: sources, arguing that when compared with indi- the behaviors of tortoises vs. lagomorphs and flight- vidual acquisition, mass collecting offers no sub- less ducks vs. sea otters are easily, almost iconi- stantial increase in the return rates of small animals, cally, dichotomized. However, as we show in the except for certain resources such as fish and insects analysis that follows, under many circumstances (also see Lupo and Schmitt 2005). more subtle differences in prey behavior and mobil- Recently some researchers have incorporated ity interact in complex ways with prey size and the critical components of prey behavior in generating pursuit and capture success of hunters, which in estimates of return rate rankings (Jones et al. 2008b; turn have dramatic effects on return rates and prey Lyman 2003; Stiner 2002; Stiner and Munro 2002; ranking. Stiner et al. 1999; Stiner et al. 2000; see also Jochim In this article we examine these effects relative 1976). Most important here are the costs of pursuit to the foraging practices of contemporary Martu and capture, which are highly variable across taxa hunter-gatherers in Australia’s Western Desert. Our and at best loosely related to prey size (e.g., O’Con- aim is twofold: (1) to illustrate, by way of quanti- nell et al. 1988; Smith 1991:230–231; Winterhalder tative ethnographic observations, some general 1981:95–96). The significance of such has been processes that influence relationships between prey well illustrated in a number of zooarchaeological behavior and hunting strategies; and (2) to provide studies, including those of Stiner and colleagues some tools for generating more reliable zooar- (for Middle-to-Upper Paleolithic transitions in the chaeological measures of prey diversity and rank Mediterranean Basin) and Jones et al. (for trans- in a wide array of contexts. We begin with a brief Holocene variation in central California), both of review of foraging models and their problems of which demonstrate that while the overall repre- application in circumstances where prey are highly AQ74(1) Bird 1/2/09 10:54 AM Page 6

6 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 mobile. We then explore the determinants of vari- that will be pursued on encounter) is reached by ability in the post-encounter return rates of Martu adding prey types in rank order according to their prey types across adult foraging bouts and follows post-encounter return rate until the marginal gain recorded between 2000 and 2007, with special in overall foraging efﬁciency is negative; that is, attention to the dynamics of prey mobility and a when the post-encounter return rate of a given prey hunter’s investment in pursuit and capture. Results type is less than the overall return rate expected from this work have important implications for from searching for and handling higher-ranked prey commonly applied measures of prey rank and (Stephens and Krebs 1986:17–24). Predictions zooarchaeological methods used to analyze rela- about diet breadth are generated by comparing the tive abundance and diversity of prey types. We post-encounter return rate for each prey type (e/h, highlight the most important of these ﬁndings, espe- energy per unit handling time) with variability in

cially relative to relationships among prey behav- the overall foraging return rate (E/Tf, energy per unit ior, body size, and handling costs and illustrate how foraging time, where foraging consists of search- these can be incorporated in future studies of sub- ing and handling). An item of the highest-ranked sistence change in archaeofaunal assemblages. prey type should always be handled on encounter because no opportunity can be lost by its inclusion, Foraging Theory and Prey Mobility and lower-ranked prey are handled independent of their own abundance as a function of the encounter Foraging theory is a family of models in behavioral rate with higher-ranked prey. If encounters with ecology used to describe and explain shifts in higher-ranked prey decline, either through forag- resource exploitation relative to variability in ing pressure or environmental change, overall for- resource value and availability (Charnov 1976; aging efficiency will drop. This will broaden Charnov and Orians 1973; Emlen 1966; MacArthur selection to include prey types in rank order as their and Pianka 1966; Schoener 1971; Stephens and post-encounter return rate (e/h) exceeds the declin- Krebs 1986; see Bird and O’Connell 2006 for ing return rate expected from searching for and

archaeological applications). Two of these models, handling higher-ranked prey (E/Tf). Conversely, as “prey choice” and “marginal value,” are routinely encounter rates for higher-ranked prey increase, used in zooarchaeology. The components of the diet breadth narrows as low-ranked prey are models are well known; here we briefly review increasingly dropped from the diet. them relative to problems of application in cir- One of the most important constraints in the cumstances where foragers pursue mobile prey. PCM is that encounters with prey are a sequential The prey choice model or PCM (aka diet breadth Poisson process: finding an item of a given type model, contingency prey model) predicts the suite does not change the probability of finding another of resources likely to be handled among an array item of that type. In other words, the PCM assumes of options. The model assumes that the goal of for- that search costs are shared across all prey types aging is to maximize the efficiency of nutrient while foraging. But where the probability of intake (usually energy) and that foragers have good encountering prey types changes predictably, working knowledge of prey distribution and search should be biased by heterogeneous prey dis- exploitation costs and benefits. The PCM is con- tribution. Under such conditions prey are pre- cerned with the trade-offs involved in searching for dictably “patchy,” such that searching for some and handling prey: when a predator encounters an types of prey involves a loss of opportunity to search item of a particular prey type, handling it (pursu- for others. After entering a patch, returns may fall ing to capture and process) costs time that could be over time as a function of resource depression spent searching for other prey. Should a forager pur- (Charnov 1976), a decrease in the instantaneous sue that item, or pass it over to continue searching? gain within a patch. With resource depression in a The solution to this trade-off provides the “optimal given patch, at some point an additional unit of diet breadth,” the range of prey types that if han- gain (the marginal value) in the return rate expected dled will maximize the average return rate from for- by continuing to search for resources to handle is aging. The optimal “diet” (not technically what a less than what can be had by traveling to another forager prefers to eat but the range of prey types patch. The marginal value theorem (MVT) predicts AQ74(1) Bird 1/2/09 10:54 AM Page 7

Bird et al] IN PURSUIT OF MOBILE PREY 7 this point, the rate-maximizing patch residence time sile prey, the decision to stop search in order to han- (Charnov 1976). According to the MVT, a rate- dle mobile prey carries an inherent probability of maximizing predator will choose the residence time pursuit failure; such failures are rarely included in in which the marginal rate of gain for a given patch prey ranking schemes. The problem is well illus- is equal to the average return rate available from trated in Sih and Christensen’s (2001) meta- the habitat as a whole. If the average overall return analysis of 134 nonhuman foraging studies. They rate from the habitat declines faster than the return evaluated the degree to which foraging models pre- rate of a given patch, the opportunity costs of trav- dicted observed resource exploitation as influenced eling to another patch increase, and patch residence by a wide range of factors, including the type of times are predicted to be longer, increasing the pos- study (field or lab, experimental or naturalistic), the sibility of overexploitation or extirpation of high- degree to which the models’ assumptions are met ranked prey. Conversely, the probability of (e.g., for the PCM, encounters are a sequential Pois- sustainable resource use increases with escalating son process), the type of forager (invertebrate, opportunity costs of continuing to harvest in a given ectothermic vertebrate, or endothermic vertebrate), patch. This can occur with increasing overall for- and the type of prey (type of plant or animal, degree aging efficiency or counterintuitively, when in- of mobility from sessile to highly mobile). Of these, patch resource depression causes returns from a only prey mobility consistently explained mis- given patch to fall faster than those expected from matches between the models’ predictions and the habitat as a whole. observed foraging responses. The PCM, MVT, and In combination, the PCM and MVT predict that related models have generally done an excellent job increasing residence time in patches characterized explaining patterns of resource exploitation for for- by resource depression increases the reliance on agers that feed on sessile to slightly mobile prey, low-ranked resources—resources that are typified regardless of the type of forager: 74 percent of these by high handling costs relative to their yield. Thus, studies had a close fit between predictions and investments in improving handling efficiency will observations. Conversely, only 37 percent of the have a larger effect in situations where handling is studies that focused on predators exploiting mobile the dominant component of foraging time, that is, prey were consistent with the models’ predictions, where exploitation practices have already broad- and none of the studies that included predators of ened to include low-ranked resources (see Hawkes highly mobile prey corroborated expectations. and O’Connell 1992). As demonstrated above, Similar results are common in studies of human archaeologists routinely use evidence of such foraging strategies and are often associated with investment to identify and define economic inten- gender differences in occupation (see Bliege Bird sification and broad-spectrum transitions. 1999; Hawkes and Bliege Bird 2002; Smith 2004; Winterhalder and Smith 2000). Where they actively When Do Foraging Models Fail? hunt game, women often focus on resources and patches with more reliable return rates and less Despite obvious successes, tests of predictions mobile prey, and the range of prey types they exploit derived from the PCM and MVT have met with varies closely with foraging opportunities and their mixed results. The mixed results are instructive; we respective costs and benefits (e.g., Bliege Bird and suggest that patterns in the lack of agreement Bird 2008). Where foraging models fail, they often between predictions and observed foraging behav- do so relative to predicting men’s decisions to ior actually demonstrate the importance and util- exploit highly mobile prey. The Meriam of Aus- ity of the approach and should be of special concern tralia’s Torres Strait offer an illustrative example to zooarchaeologists. (Bird and Bliege Bird 2002; Bird et al. 2004b; Bliege Common to both ethnographic and nonhuman Bird 2007; Bliege Bird and Bird 1997, 2002; Bliege applications are the ways in which the models often Bird et al. 2001; Smith et al. 2003). Meriam women fail to predict the decisions of foragers when hunt- and children focus most of their foraging effort on ing for mobile prey. Most important is the differ- reef shellfishing and hand-line fishing for small ben- ential effect of pursuit in calculating the handling thic omnivores in the nearshore. Their decisions costs of different kinds of prey. As opposed to ses- about switching between patches in these habitats AQ74(1) Bird 1/2/09 10:54 AM Page 8

8 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 and prey choice within them are highly sensitive to Desert (Figure 1). The term Martu is now com- temporal and spatial variability in foraging returns, monly used as self-reference by about 1,000 peo- consistent with the quantitative predictions of for- ple from a number of dialect-named groups whose aging models. Conversely, while hunting for marine homelands surround Lake Disappointment and the turtles offshore, deep-sea fishing for pelagics, or Percival Lakes. The heart of their country lies in spearfishing on the reef, Meriam men consistently the Warnman estates of Karlamilyi, situated in pass over the opportunity to handle encountered Rudall River National Park, where two of their prey or switch to other foraging activities that would “outstation” communities, Parnngurr and Punmu, dramatically increase their overall foraging return are located. A third outstation, Kunawarritji, is rate. Human behavioral ecologists have developed northeast of the park at Well 33 on the Canning a number of hypotheses for why such patterns are Stock Route. Martu established these communities common (see Winterhalder and Smith 2000), in the 1980s when they returned to their homelands including how, for some foragers, the value of purs- following a mid-twentieth-century exodus into mis- ing certain kinds of prey is influenced by the way sions and pastoral stations (see Davenport et al. that the costs of acquisition and distribution can 2005; Tonkinson 1991). The outstation movement guarantee widespread social attention (see Bliege provided a foundation for Martu to gain Native Bird 2007; Bliege Bird and Bird 2008; Bliege Bird Title to much of their homelands in 2002. At any and Smith 2005; Hawkes and Bliege Bird 2002; given time each outstation has a population num- Hawkes 1991; Hawkes et al. 1991). bering between 50 and 200 people, and about 500 The results of these studies raise a number of Martu live in the towns of Newman to the west and questions directly relevant to zooarchaeological Port Hedland on the northern coast. While most applications of foraging models: Martu families keep a permanent “camp” in one of the outstations, they maintain high residential • What is the relationship between prey mobility mobility between communities and are often away and variability in foraging decisions? for extended periods of time traveling throughout • How are post-encounter return rates affected by the Western Desert and Pilbara regions to fulfill tra- the different components of handling, specifi- ditional religious and social obligations. cally variability in the costs of pursuit, capture, Ethnographic data on Martu social organization and processing? and history are available from a variety of sources. • How does patterned variability in such costs Tonkinson (1974, 1978, 1988a, 1988b, 1990, 1991, scale with commonly used zooarchaeological 2007) provides a comprehensive account of Martu proxies of prey rank, specifically prey size and identity, religion, gender, politics, and change with return per unit processing? European incursion over the last 50 years. Veth and • How might the determinants of variability in Walsh have been concerned with Martu prehistory, handling costs affect the methods zooarchaeol- subsistence ecology, and mobility (Veth 1987, ogists routinely use to monitor changes in diet 1989, 1995, 2000, 2005; Veth and Walsh 1988; breadth, specifically abundance indexes? Walsh 1990). And recently Davenport et al. (2005) In the following quantitative analysis of contem- have described the events surrounding Australia’s porary foraging among Martu Aborigines we nuclear missile testing program in Western Desert directly address these questions to demonstrate how and the direct effect it had on the uncontacted Martu the consideration of such issues can inform and bands that remained on their estates through the improve our methods of measuring key variables 1960s. Elsewhere we have written about Martu and ultimately refine our interpretations of vari- subsistence in relation to mosaic burning, season- ability in prehistoric resource use. ality, gender, and age (Bird and Bliege Bird 2005a, 2009; Bird et al. 2004a, 2005; Bliege Bird and Bird The Martu and Data Collection 2005, 2008; Bliege Bird et al. 2008). Here we will focus our description and analysis on the nature of Martu (aka Mardu, Mardujarra) are the indigenous contemporary foraging practices among the owners of estates that comprise about 150,000 km2 Manyjilyjarra, Kartujarra, and Warnman Martu of the northwest section of Australia’s Western who live in the heart of the desert. AQ74(1) Bird 1/2/09 10:54 AM Page 9

Bird et al] IN PURSUIT OF MOBILE PREY 9

Figure 1. The Martu homelands and contemporary communities.

Methods and Definitions On any given day, depending on the season and habitat, Martu can choose from a number of dif- We began working with Martu in 2000 in prepara- ferent types of hunting and gathering activities. tion of their Native Title claim (which began in 1992 Here we use Martu definitions for different forag- and was awarded in 2002). Since then we have ing activities, which correspond to mutually exclu- spent 27 months in residence (consisting of 72 sive tasks associated with acquiring particular suites researcher-months across all seasons) in the Martu of resources (see Table 1). Martu identify a given outstations, based mostly out of Parnngurr (Figure foraging activity by the term for its primary 1). Data reported here were collected in the course resource type and a directional suffix (e.g., parna- of daily foraging trips, usually within a 50-km jarlpa [sand monitor] -karti [to]). They often group radius of one of the communities or a temporary foraging activities into categories differentiated by residential camp. During foraging trips we recorded meat/hunting (kuwiyi/wartilpa) and those focused the identity of each participant in the foraging party on plant/collecting (mirrka/nganyimpa): hunting (the group that leaves the community/camp includes those activities with active pursuit of together), the time they spent traveling to foraging mobile prey; collecting focuses on immobile locales, route taken, foraging location, and forag- resources. We also use the emic definitions for ing bout details of all party participants. We defined resource types, which in most cases correspond to a foraging bout as the time each participant in the a given species (see Table 2). While some of the party spent searching for, pursuing, capturing, same resources can be found across different for- transporting, and processing wild food resources. aging activities, by choosing a particular activity a Following each bout we recorded the number and forager trades off the opportunity to focus search whole weight of each type of resource that each for- on another suite of resources. Thus, according to ager acquired and monitored subsequent food dis- the logic of foraging models, different foraging tributions. AQ74(1) Bird 1/2/09 10:54 AM Page 10

10 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 , . y ) a s c y r s / t r n r 6 3 5 9 9 e e u 6 e d t a 3 6 9 0 0 c i / 1 1 1 1 1 1 1 – 1 o 9 a l p c 4 6 8 . 9 2 c l . . . . . o B u R ﬁ u o f S T c e

(

n d i n 1 1 4 7 4 4 6 7 9 7

a e 2 r A e 0 0 9 2 4 4 0 0 4 0 n e 0 0 0 0 r i 0 0 4 . 0 0 0 0 0 1 2 s

N a ...... a

M l j d g a a . l e i a t a t

7 a c g o n e c a 0 a n T c r a c k e 0

7 6 3 1 1 3 8 9 4 1 5 8 5 u o a 2 2 A W

1 9 3 3 4 5 0 1 5 5 5 1 0 m

d % F

0 t o ; 0 . 0 0 0 0 0 0 0 0 0 5 0 0 o N o t s l ...... r i

u r r p W

g a p l y u i A n A – a n 1 4 1 4 3 8 2 5 4 4 8 6 m i – e y 0 0 0 1 0 5 1 0 0 3 6 9 0 0 0 s m r e

0 4 0 0 0 0 0 0 0 0 2 1 a ...... n M e a u u r b J n

a l a g d d j J l n a n e t n a i u , e t 3 9 8 5 2 4 2 6 1 3 1 3 5 9 ) a o g o 6 t m n 3 4 2 0 3 0 0 2 1 1 0 1 1 0 m , a i T 0 e w a

r 0 0 0 0 . 0 0 0 0 0 0 0 7 0 0 o 5

T ...... o t w 0 W i / %

W F t

0 t n o 2 i c

h r e s ( e p n 4 8 3 9 5 3 6 9 8 8 6 2 4 5 l 2 i l s a b 8 1 0 5 2 0 0 1 0 0 1 4 8 0 a 0 r u . 1 0 0 0 0 0 0 0 0 0 0 5 0 0 A r r m s ......

e a d j e v i e l r o 3 w 5 3 7 3 1 9 3 4 5 a e

. 1 2 4 7 3 t t N 5 4 8 4 8 3 1 4 9 ) Y t D 6 0 6 3 – 7 u – 7 4 3 2 8 7 1 6 6 , n t a : t , , , , , , , , , S 7 3 4 4 b 2 s o c u s 5 4 2 4 4 4 2 1 1 i s , 1 n t o u

e . o a g h t B s t k c / a u n i 9 2 1 9 5 3 1 2 6 3 6 2 l a e r o 8 7 a 1 2 5 6 5 8 6 7 9 0 4 2 l

e a e A A w l 7 4 e s

c 1 3 3 0 3 8 4 3 0 0 9 3 n ,

, , , , , , , , , , , , w N r 6 4 d k l A

1 3 9 2 6 3 5 2 3 4 1 2 M u n a a 0 t e

a u e t 0 , r q m u 2

i t n

o e

1 9 0 2 3 5 9 3 7 8 9 9 7 7 0 r D T u o g

b i h e y 6 8 3 4 8 2 5 9 5 3 1 3 8 7 7 ) o S n t l t a g b i

n a

h B i n t r g e / i i g m t a o e m g p e a u r f i t ( l n e

a o o m c 2 8 9 3 5 9 0 4 a r p u 9 r y i 6 5 5 1 0 1 e F

l 4 3 5 1 5 9 9 9 e e c o 1 7 4 6 9 5 6 r T l e n 1 1 1 2 1 1 1 2 S F 1 p a e

M o – t r c

r h n h . t r t a c e y

r o t r t c i o a c n a r 1 0 3 1 2 3 j t e v i n e 6 4 6 4 0 9 i 3 5 7 8 1 6 n l 7 D n M t a i 8 6 3 – 3 6 1 – P t 6 3 3 0 6 5

) c c , , , , , , S ,

1 2 9 8 2 5 e u f r Y d a 4 3 1 3 1 3 , 2

n c

o o h e r h 0 d

/ u n d l r B u c 0 o i d / o t a e u o 2 v e o e c

i d s t m r g t

r k

d a R s p s (

m g

O n 1 2 1 8 4 2 3 u a s R l

n o i l 4 7 7 8 1 6 a 6 8 3 7 5 4 0 g i ,

A A i n c 5 5 9 1 4 7 e r 7 4 1 3 5 8 2 i

u d , , , , , , , w N N w 4 5 6 4 6 4 a a , n

A 1 8 1 5 1 3 1 o M d s l a s e – r l

r n e y o e e g m r l f g n o a a w i a ) a c c r s o u r s

e s o n t ﬂ r u 9 2 e

b a F t 6 u 1 8 6 5 8 9 1 2 9 1 a

c e

8 n 1 4 6 J r j s 2 5 7 1 2 2 1 4 3 2 9 o t o h 2 6 a l a , a t r t B

u 1 h ( p n M

h d 0

a g c 0 g e t A n

W 2 b n h a i

t i 3 t W t p r T y a c

s l t , u . . ( e r i n u

l d g n a b e l s

M M M g i W n e t a v e o

i u g , , , , i i n r c t n n W S S H W S S – W S S W R S S M

g e A i i a r

s y – v u a k i e m e

e t b

r a

v v n g

n n c e e p v I c o o k k k o u s u

i p o A t a a a r . h J

a N N A

e e e a o s u , s N g – – r – n O e p p p n o t t 0 r o – n

e y c – t

c i c c l l l a 0 r e a r h c y m a a a k g r p t 0 u

O O e a l

a a u u u a e t 2 O M o d r , , e s s s e u

a y s , l n e e e w M o c t u u u p y T k i

i a g l w

n n n n r F r l b , , , , , a

e k i

u n u , n u u u p a e e e e e e e e e . t a i = u s J S J J J l l l l l l l a )

p n W t n

F A e ] A J n –

– – – b b b b b b b . r h a i o . – – l l l l p d 1 – – e t a a a a a a a l 3 a i i i t s W a y a i i i i i i i y y , r r r 8 v v n n p l r 0 r r r r r r r a p l l l ,

u a a

o p p p o l l l 9 M u o t n a a a a a a a 0 e e e s

e J s 1 . o l 2

g V V S M A V N M V A V A u V A S – A V A N

o 1

b – s n n

g i a a e 2 a

i M n r , l

t k 0 s i r

t . . t b r u a 0 y a a y w c V a e v a o 2 c e k a e

: l b T d b l , , c l m r

l , , n H i a n n o i r r u t

o m

d n g o v

o r t o o o o a e n d

y c i n o n e r n t t o s

i c a c h a r i i n a r n h = a t

a t o l y n y e t

g s t ; a i n n a o e n r G y o

e a s o h c y , g o o g s l R D e a s

r p r c : e c e l )

e m p e d r n t r i , – r p d

, o s , o e r r a m m n n t , r o a s f t s e

( p

u t d e n k a n

i f o , n e k s k e e t u a

o d d o n

o t e n

n w 6 n e b n s i s o s h e i l v p a g n n n c r i o i o t 3 l o l l r r e u u d m a a e k l y

a i e m r k r 2 i b a d C l n n

e t – – p S – A T f l S h m S s – – – F – e R p – B S – a ’ o i t

e e l n

s i o p p O s a o m e a e s [

e e w d s

a d v s

u s S d : r v s t o u c t a

i r u i a a , u

h s l e i u t s p t s ) l h o / s g t i i y c

e u t l s o o l : o a a a y : . : l a o i n d i m m i l p a n r d b e r p r

b u o l l e r e p n i r b n a

u u o c : e e d o d y : o m A l e k

o u a t : u p e s o / t f t o s g e r m n n r l : k r s h

n u u t a

i v s e F : o i

n c s d s , a a a s u s

r m r d A i s e o b e o n n m g g i l l : i i = 0 e n H d

i o a e

a a g

e l d

t o t a c

s e u : s a h o a l l o o 0 s e l e i s s a r c e n s i e : e a s

o n y a ( b

y r ﬂ b a p d l y i s l l

r S S r e r 3 u i r s p r u u u o a c o

a M r e a

t m d l t e y o

x s

s a a a e 7 t e u

f i a r i a : : – i , s n n o a r s k a

r n l , e i a r r t t r r s o e e h s t

c 4 c o d s g g s 0 l i e t u i i e t a a s e c e a n w r w l n e e s s s m l d s p a a p e 3 r r n s n l r r a n c v 0

e b u i u n d o a g o a v , o i e r r u o e u y l t n c r m l r e e e r r a l r y i a a a a u i 5 e

u i p 1 ﬂ s e ( c l B B P C P T F F R P F G k V a M V s C c E d V H m E s A S E F H T

y r

a s ﬁ u a n e ﬁ

g r T e a r e e

s e k n c o g n v i d e m f h a g g t n

o n g e i g c f n n g e c t w H a d g g n b i i e

i t t e n l n ; i n n

t e i l m t x t s i i c c y l t u i m y o o l t t n n s e o e e e t n t t t

i c g p h e c c l l f o

i i u c u g g g s

a u l l i e d l r

n e e i

g n l h g i n n n f i l m l o o d h e o o n

l d a i i i

t l l

m g n i o u b a e e c o t t t c h r r e e y i o

n o o

o r n a p g i s t c c c h S a e l r t t c i l g A r

t m i c c s u a y t s

e e e i s

a : s a a a – g i i n l l l

v e c t g h t a r h t b e r l t d e s l l l i b e

a = e y e e n t s l o t s n c s v e t

a u r o b n n h r o o o e a c a o e t e r r u u u o o a F r e a c c c o S N u u F A

p B F B P K A G T O N a b c d e B M D C N G R S h H AQ74(1) Bird 1/2/09 10:54 AM Page 11

Bird et al] IN PURSUIT OF MOBILE PREY 11 t - r d a e i p t h e p t n r c

u o x y t o e e

, c r s y n p t u i e , l s

t o t i i i s / w n s b 1 5 8 3 8 5 2 9 u e v e o a o t l s o 9 0 7 6 8 9 0 8 c f

b – – 1 1 1 1 1 1 1 1 1 l r a

c 8 3 4 9 6 8 3 3 n o o ...... m u

= u R e

F h P e k S t 5 t i a

a t ; r

) w s e s , i r ) d o 1 o a t

2 3 1 0 i k 1 8 7 7 0 7 0 2 e n m 8 3 2 0 r n t l 6 2 9 2

r 3 7 2 7 2 7 5 7 i 7 r D 4 7 3 5 o b – – 8 9 4 5 i u = 1 , 2 , 3 3 1 5 3 4 , , . u 2 t , , , , , , , ,

S a 1 2 5 5 ) 8 1 6 5 s m m e ) 5 8 1 1 2 2 3 1 3 T

r e

5 1 3 1

e ; s u p R e h r

c e / y P e r e l

a s t T e t f a m

( t r i n

u c e t n u u d o k c e s u ( h r e h

t t , o m u i . e a c . c o

o g ) e 3 8 3 2 s n . s m t l n n W 2 4 9 9 6 1 2 1 2 i l

3 8 7 4 e e , a 4 2 6 2 x E a 2 5 6 5 0 1 8 9 6 ( e o T 3 1 9 6 - y e h 5 6 7 1 – – e

R t 4 , 7 , 1 4 0 0 4 4 4 , , t l t c t

, , , , , , , , , s

4 4 4 9 a

s 8 1 8 8 e f . e r 5 9 2 2 5 6 8 4 1 M e s o 6 2 5 1 s o R c e r e

v P s n e

y w ( a r v

d o t r n o l e s t e s i o g t

n d c e s n

u t y a u d r h a

o n o 6 6 0 C e 1 0 7 1 7 0 1 0 2

c ) a n w h 6 0 7 r 6 2 9 3 r l r 8 1 9 0 7 2 8 5 7 d n 7 s , D 1 2 2 a a 8 9 4 1 – – h

u e ) , , 7 9 8 , 5 3 1 0 3 4

n 2 / t , , , , , , , , , r

S t 1 2 5 5 , l a g 1 2 2 f e e a 4 9 4 1 1 2 3 3 1

y a n 2 1 1 o t s v h i R c

i t t

l o

e k w i r

c s ( r p o e b e

e l t l e y l o l t c t n w

r o o n o u m e u F c l

u o

h o l h t c w o b l s

n o e u a F d 1 3 l A

/ r r n 1 4 5 9 1 9 6 8 2 1 2

n 6 0 e 4 2 6 7 n a - g s 3 4 5 7 9 5 0 7 8 9 6 t = a o t 2 4 ( a s 5 6 7 8 – – e

t 9 , 8 4 1 , 4 4 0 3 4 4 4 a s

, , , , , , , , , , , h o 4 4 4 8 2 n 0 0 a o t r

d 4 3 3 5 1 2 5 5 8 4 1 R M

o ; 1 2 p c P n r r s e o e a m A t t

i p e s s y n s s a

e n u e f i r y

) r a o r e b c g o g

r o l ( n n s

a c a

l e

0 i

t / i n 1 4 6 a e t C y 7 r a

e n 0 1 3 s a r 9 0 3 u 3 e a 9 z h d o e 6 0 0 – – – – – – – – – – – e a r 4 , 3 5 d i t

1 v , , , e i h

4 3 7 k

1 P e S t - 5 2 2 M g v l y

y i 2 i i l a t e k i r d a b r l n r o i n o p e u I b F , l

n m

o y f e t s i m o g a m l i

n n d y y b y

i I t t o / l

w i i o n i n . e l k t o a v i ) 2 9 1 a v i l n m . . . u 5 5 4 3 – – – – – – – – – 1 – – e i r t l b d

a t b 3 3 2 a c o o o h i c M R v r a o g F e

i f M f

d n g f h e

n o e n l

i a = C b

g =

k t a 4 t

r 1 n n e ;

i o e )

s g f k s c

a l e d r r

S e a e v

e o l t . ( 9 h 5 9 6 4 6 8 2 2 1 9 8 5 8 6 1 9 t w g 4 3 a P t i y o a 1 8 2 5 3 0 9 9 8 8 0 5 2 1 1 1 0 o

a c 3 0 n l g r T f 3 . 1 0 0 . 0 1 1 1 1 1 0 0 0 0 0 0 0 o l

k k ...... o o o t o r o l

F r f % r r m o i

e e n h f d m d h c

e r e r c r a

e O o e w t

) u n n r t e i b r ) o a r (

u i e u s d t y b w i e e m p 1 8 , o ( 3 3 8 7 6 2 5 9 6 6 9 6 4 0 c . z n a w 1 2 6 4 6 l i r 5 2 1 2 1 4 3 1 1 2 2 2 6 1 i l g c r u 1 1 .

u y o k o i e d o

( F g s e s

w ( e e r z a t u i

e l a k R t e

a c n r a i

u e t s u v c i r r h e h

t a

a p i y l a t s

l d t

e m s i i . u p u a s b s 2

r a o s

a o u s a c e o h o t e e l a t s d e s r c ﬂ m i i c l y s b e i

i o e u r

r s a a h u . y w a d p b b u r r e y

t l s c e p T s t o l u r o e o r d a u i o l a p s r u s n i e l s i v t a r

a P o b t s i e s e l s y e i s o r

l s e t a g t n . r d c

n a e e o a u

u a

u )

n e i s n u l r s o t c s d e e b i p k t r a e i y h

t c m m x n u a z n t s

o e a m

g s s x i o r e u

u a r n i a o a u r a t . y c u l o o

H n n c

l a g

r s p a f r o r l n

e d

a i

r e i G a a a : i a e i g l g s l a s l l r s r p n e a l l : c t i n c e p a e u i u o o a s a M r t t s V e s E t t r t g y a ( s

n c i S S a a o i

: e s : c : c t s u E m

a a d C c e : r : v g o V e

e c i

e r l A n u : d : o o n s

n o i a : n u e p u t t a s i i a r i : q r

s t s n t i v s t n i s a

r s c V d p a r

r m d y

l A i o n

n s a i A e d i a e : a g a t a a y a o A i d : m o l u e l e T r r : c e n a e n y

o n : d

o i s w l M i r

h e e M o n : a t o t a h r

t m l l a d

l y y s

t w u

r r

i i : i r F : k s g o k a k a

i e n

e h h s

t s

e e e l k o d c r e s v e e : e h n e l l y i n s s s t r s s t w t s r t i n a e p h l n n l h r e e i t h o i i t e r r a u u o e o u o a u r y a e e k o y r i p t C K S M D O N w w R B H C C O B H C T B P G G P P S T AQ74(1) Bird 1/2/09 10:54 AM Page 12

12 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 activities are analogous to different patches (see For all types of game except kangaroo, in order Smith 1991). to convert whole weight acquired to edible weight, During each foraging trip, at least one researcher we measured the proportion of refuse to whole present asked permission to conduct a focal indi- weight of specimens (n = 46) representing all small vidual foraging follow, where one of us accompa- game taxa, as well as four bustards (Ardeotis aus- nied a forager for the duration of a bout, recording tralis). For hill kangaroo (euro, Macropus robus- all time allocated to search and the different com- tus), proportional meat and marrow weights for ponents of handling: pursuit, capture, and pro- each body part, male and female, were obtained cessing. On many follows we also recorded the from O’Connell and Marshall 1989: Tables 3 and foraging path using GPS, and as with all bouts, at 4. Edible weights of all resources according to the end of the follow we weighed and counted each species and body part were then converted to kilo- type of resource acquired. Martu most often for- calories using Brand-Miller et al. 1993. These age separately, but occasionally people cooperate. records constitute two primary data sets used in the During these episodes we divided the returns by description and analysis below: 1,347 adult forag- the number of individuals that foraged together, ing bouts and 649 adult focal individual foraging although if people were working closely together follows. an experienced observer often could conduct more than one focal follow at a time. Contemporary Martu Hunting In keeping with the PCM, we considered handling to start when a forager began pursuing a spe- Some of the most vibrant aspects of Martu iden- cific item, or collecting multiple items, of a tity are expressed in their foraging ethos and prac- particular resource type, thereby trading off the tice. Martu established the outstations in the 1980s opportunity to continue searching for other in response to the threat of dispossession wrought resources. Accordingly, a resource encounter could by institutionalized racism and processes of be either direct or indirect, the latter being when assimilation. Living in the outstations allowed foragers found indications (usually tracks) that Martu to assert ownership over their homelands caused them to begin following a particular prey and maintain their ritual and socioeconomic oblig- item. We thus defined pursuit as time spent in track- ations to the desert. These obligations are mani- ing, stalking, or chasing an item (see Winterhalder fest in the regular maintenance of their estates 1981).1 These are distinct from the costs of cap- through daily hunting, gathering, and mosaic land- ture, defined as the time foragers spent retrieving, scape burning. While often unacknowledged or collecting, extracting, and dispatching resources devalued by “mainstream” society and policy, for- during a follow (e.g., probing for and excavating aging is the dominant occupation of Martu men, burrowed game, digging geophytes, picking fruit, women, and children, and many hunt or gather extricating cossid larvae). Processing is a substan- every day. Most families receive government wel- tial proportion of handling for only a few Martu fare, which is often used for gambling and travel resources, mostly tree and grass seeds. Processing to sustain social and ritual imperatives or to pur- costs, including those in and out of camp (e.g., field chase household goods from town or the small preparation of game, in-camp butchery, winnow- shops in the outstations. ing of bulbs or seeds), were charged when the focal Depending on seasonal and individual obliga- forager participated in preparing the harvest. This tions, foraging provides outstation residents with is the case for most plant foods and small game, 25–50 percent of their daily calories per capita and but larger game (e.g., kangaroo) are often butchered over 80 percent of their meat calories; participants and distributed by someone other than the hunter in foraging parties (including all women, men, and (see Bird and Bliege Bird 2009). In these cases we children) acquire an average of 1,700 kcal/capita/ did not charge the acquirer with in-camp process- day from bush foods (Bliege Bird and Bird 2008). ing time. Also in keeping with the logic of forag- The remainder, mostly in the form of commercial ing models, because foragers could usually perform flour and canned goods, is purchased from the shop other activities while food cooked, we did not count or during visits to town. Typically a foraging party cooking time as a component of handling cost. leaves the community in the late morning, usually AQ74(1) Bird 1/2/09 10:54 AM Page 13

Bird et al] IN PURSUIT OF MOBILE PREY 13 following household or community business. Vehi- aging time and produce 51 percent of all calories cles are regularly (although not always) used to from foraged foods. As for other Aboriginal groups access foraging locales, but on arrival much of the (e.g., O’Connell and Hawkes 1981), grass seeds hunting and gathering is conducted on foot. Par- were a traditional staple but have dramatically ties average 8.0 ± 3.3 participants (2.3 ± 1.1 men, declined in importance. Today, for both men and 3.6 ± .8 women, 2.1 ± 1.4 children),2 although women, hunting is far more important than col- groups of over 20 (in a single vehicle!) are not lecting plant resources (Bliege Bird and Bird 2008): uncommon. 81 percent of all foraged calories and 87 percent of A given foraging location is decided by con- all foraging time are devoted to the four most impor- sensus among the adults in the party, usually involv- tant hunting activities. ing discussion about recent returns and intricate ecological and social dynamics. Game animals, Results I: The Effects of Prey especially bustards, are occasionally spotted en Mobility on Foraging Decisions route to foraging destinations, and active search often begins as soon as the party leaves camp. Adult While a description of all Martu foraging activities men are almost always armed with small-caliber is beyond the scope of this article (see Table 1 for rifles and axes. Women rarely go anywhere outside a summary; for further details, see Bird and Bliege of camp without their highly prized digging sticks,3 Bird 2005b, 2009; Bliege Bird and Bird 2005, 2008; although they occasionally carry rifles if men are Bliege Bird et al. 2008), an analysis of the more not present, and many of the women are keen shots. important hunt types will serve to illustrate the way On arrival at a foraging locale, the party designates that prey mobility structures hunting decisions. a dinnertime camp: this is the site where all participants will gather after foraging to process, cook, Sand Monitor Hunting and share resources prior to traveling back to the The most frequent foraging activity is sand moni- community. In our sample, dinnertime camps were tor (goanna, Varanus gouldii; parnajarlpa in used on over 90 percent of the foraging trips; this Martu-wangka) hunting, taking up 54 percent of is partly due to firewood constraints around the out- all foraging time across all seasons, peaking in the stations and, as Martu say, to facilitate sharing (see cool/dry season (Wantajarra, May–August; Figure Bird and Bliege Bird 2009). 2). Resources from this hunt type account for 35 Martu country is structured by four foraging percent of all foraged calories. Varanids make up habitats: spinifex grass (Triodia and Plectrachnae the bulk of this, but skink (Tiliqua scincoides; spp.) dune fields, low-lying rocky ranges, eucalypt- lungkuta) and python (Aspidites ramsayi; kanati) dominated (Eucalyptus spp.) watercourse margins, are always pursued on encounter. Although they are and mulga (Acacia anura) woodland. Depending usually pursued in other foraging activities, sand on habitat, season, and expected returns, when for- monitor hunters sometimes take feral cats (Felis sil- agers arrive at the dinnertime camp, the party splits vestris [see below]), and if encountered they often up, and participants typically engage in one of a collect cossid larvae (Endoxyla spp.; lunki), number of mutually exclusive foraging activities solanum fruit (Solanum diversiflorum; kampulpaja; (see Table 1). Men more often forage alone, while and S. centrale; jinyjuwirri), and nectar (wama) women often forage within earshot of each other, from Hakea suberea and Grevillea eriostachya sometimes cooperating in aspects of search, pur- flowers. Sometimes fresh signs of bustard (Ardeo- suit, and capture (see Bliege Bird and Bird 2008). tis australis; kipara) or hill kangaroo (euro, Macro- Typically (though not always) people engage in pus robustus; kirti-kirti) are encountered (especially only one foraging bout per day; adult foraging bouts near the interface of dune fields and mulga wood- (not including travel time) across all activities aver- land or rocky range), but these are rarely pursued age 159 ± 88 minutes (n = 1,358). Men’s and during this type of hunting (see below). women’s bouts are similar in length (respectively, Sand monitor hunting is limited to spinifex dune x = 158 ± 94 minutes, n = 521, vs. x = 159 ± 83 fields and requires an anthropogenic fire regime: minutes, n = 835), but women forage more fre- lengthy search and pursuit are conducted on foot, quently: they make up 62 percent of all adult for- and without an established system of burning to AQ74(1) Bird 1/2/09 10:54 AM Page 14

14 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

Figure 2. Sand monitor (Varanus gouldii) hunting. A view of two Martu women near Punmu outstation walking down a dune to continue hunting in an area prepared by burning off the overburden of overgrown spinifex grass (R. Bliege Bird, photographer).

clear the overburden, encounters with tracks and burrow is fresh and there is no evidence of an exit dens are too infrequent to make it worthwhile (see hole, hunters use their digging sticks to probe in Bird et al. 2005; Bliege Bird et al. 2008). Hunters wide concentric circles to locate a terminal cham- usually begin their search in a new burn (nyurnma), ber, usually found 10–50 cm beneath the surface, often following along just behind the advancing although during the hot seasons dens can be sig- flames in the clear surface of new ash. This activ- nificantly deeper. Martu probe and excavate dens ity usually involves more tracking than the larger with tremendous skill: distinguishing a subsurface game hunting described below; once fresh tracks chamber den from an entrance shaft or an escape are found, prey are typically pursued to exhaustion tunnel is very difficult, and care must be taken to or to a monitor burrow. If the entrance mound of a not misjudge the occupant, as deadly king brown AQ74(1) Bird 1/2/09 10:54 AM Page 15

Bird et al] IN PURSUIT OF MOBILE PREY 15 and taipan snakes often shelter in varanid burrows. course margins, Martu sometimes pursue the game All of the varanids are fast, and even when bur- to other habitats. The hunt type is dominated by rowed they sometimes escape through a “pop hole” tracking/chasing over long distances, with hunters and can outrun a hunter over short distances. Skink, often pursuing the lizards from shelter to shelter however, are quite slow, and while they may attempt until the prey are exhausted. If successful, pursuit to escape to a monitor burrow, hunters typically usually ends in a tree, with a hunter dispatching the capture them in spinifex hummocks after only a monitor with rocks, a rifle shot, or a quick hit with short amount of tracking. Success rates from sand a digging stick. However, pursuits often fail (over monitor hunting are high (Table 1). With an estab- 50 percent of the time/follow; Table 2), and hunters lished anthropogenic fire regime, hunts rarely fail return with nothing on 33 percent of bouts. outright, and hunters can predictably adjust harvest size with foraging time (Bliege Bird and Bird 2008; Feral Cat Hunting Bliege Bird et al. 2008). Return rates from sand Feral cats are sometimes pursued during sand mon- monitor hunting across all seasons average 641 ± itor and perentie hunting, but they more often make 519 kcal/hr searching and handling (n = 612 bouts) up their own hunt type. Typically cats are hunted but drop significantly (an average decrease per for- on a day following another foraging activity, when aging bout of 97 kcal/hr; df = 566, t = –2.25, p = a hunter returns to pursue fresh tracks previously .025) during Yalijarra, the hot/wet season that lasts encountered. As such, cat hunts have relatively lit- from January to April. This is due to changes in tle search time but very long pursuits, effectively prey mobility: the pursuit and capture costs limiting the amount of time hunters can search for involved in sand monitor hunting increase in the anything else. As with perentie, cats are hot months when all of the herpetofauna are faster tracked/chased from site to site until they capitu- and require more tracking/chasing and when prob- late. Most cat hunts occur in the dune fields, and if ing and excavating the deeper dens become more a pursuit fails (over 60 percent/follow do; Table 2), difficult. foragers routinely begin searching for sand monitors/skink/python as described above, increasing Perentie Hunting the chance that the bout will succeed (Table 1). As a consequence of seasonal changes in varanid mobility during the hot/wet season, monitor hunt- Bustard Hunting ing shifts from the dune fields to the eucalypt- As shown in Table 1, women spend as much time, dominated watercourse margins and floodplains, or more, as men in sand monitor, perentie, and cat with a focus on very large perentie (Varanus gigan- hunting. Men, however, spend far more time hunt- teus; yalapara) and Argus monitors (Varanus ing for bustard and hill kangaroo. Bustard (Ardeo- panoptes; marantu [see Bliege Bird and Bird tis australis; kipara) hunting is the second most 2005]). Average return rates for this hunt type are common foraging activity after sand monitor hunt- roughly equivalent to that of sand monitor hunting ing, accounting for nearly 20 percent of all forag- (Table 1), however, in contrast to sand monitor ing time and 30 percent of all foraged calories. hunting, efficiency increases in the hotter months Martu report that time allocated to this activity has (from 507 ± 497 kcal/hr in May–November [n = increased with vehicle and rifle use since they estab- 22 bouts] to 867 ± 1,068 in December–April [n = lished the outstations. This is because a reasonable 56 bouts]). This is because in the winter perentie encounter rate requires being able to cover a good nest in impenetrable termite mounds and caves, deal of country: the birds are highly nomadic and making capture from a den very difficult. Thus, thinly dispersed across all arid-zone bioregions, during the hot/wet season when the large varanids occurring singly, in pairs, or even in small flocks. are more accessible (less likely denned) and smaller Whenever in a vehicle, Martu are on the lookout sand monitors are more difficult to pursue and cap- for bustards, especially at the interface of dune ture, the average return rate from perentie hunting fields and mulga scrub or in regions surrounding exceeds that expected from sand monitor hunting recent burns. When the birds or fresh tracks are by over 200 kcal/hr (df = 286, t = 2.16, p = .031). sighted, the hunters make every attempt to pursue While perentie hunting is limited mostly to water- them in the vehicle, using it as a mobile blind: if AQ74(1) Bird 1/2/09 10:54 AM Page 16

16 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 aging time (Table 1, Figure 3). Unlike many other regions across the arid zone, expansive sclerophyll woodlands are quite limited in the Martu homelands, and encounters with plains kangaroo (Macropus rufus; marlu) and emu (Dromiceius novaehollandiae; karlaya) are rare. As such, except while traveling in far western parts of Martu country or in the Pilbara, large-game hunting is focused on hill kangaroos in the low-lying rocky ranges. Hill kangaroo hunting is conducted on foot with rifles and typically begins with a hunter walking along the base of a range in search of signs of kangaroo resting in acacia shade on the flats and alluvial fans. Robustus, especially males, carry a distinctive scent, and stalking often begins when a hunter smells a particular animal. If downwind, the hunter can sometimes stalk the kangaroo to fairly close range (within 30 m). Robustus have poor eyesight and if undisturbed by smells or sounds, will often remain prone until after the first shot. If missed or wounded, the kangaroo will bolt for the hillslopes, and if close enough, the hunter will typically continue the pursuit, either tracking or in direct chase. If the hunter is thereafter able to maintain visual contact with the kangaroo, or Figure 3. Hill kangaroo (Macropus robustus) hunting. A blood track indicates a good hit, pursuit will con- Martu hunter near Parnngurr outstation returns to a din- tinue in an attempt to exhaust the animal. As with nertime camp with a small hill kangaroo carcass slung over his shoulders (B. Codding, photographer). bustard hunting, average return rates are higher than that of smaller game hunting (Table 1), but due to the high variance of hill kangaroo hunting, bustards do not see a human form, they are far less the differences are not significant (e.g., sand mon- likely to fly and if disturbed by a vehicle, tend to itor vs. hill kangaroo hunting; df = 90, t = –1.51, move away slowly in their characteristic stately p = .137). The high variance is a product of the walk. The mean return rate from bustard hunting low success rate: 79 percent of all hill kangaroo is nearly double that of mutually exclusive alter- hunts fail (n = 91 bouts). This, in turn, is a func- natives such as sand monitor or perentie hunting tion of how mobile kangaroo are relative to the (Table 1), but choosing it carries a high risk of fail- hunter, which proves to be the critical factor in ure. The coefficient of variation in bustard hunting determining their post-encounter return rate. efficiency is over three times that of sand monitor hunting and double that expected from perentie Results II: Prey Ranking and hunting. This is due to the fact that while they occa- the Costs of Handling sionally acquire bonanzas (e.g., six birds in one hunt, equivalent to over 50,000 edible kcal), most Here we present an analysis of the trade-offs that of the time bustard hunters return with nothing at hunters face in acquiring commonly encountered all (a failure rate of 56 percent per bout [n = 289]). prey types, directing our attention to the following questions: (1) How are post-encounter return rates Hill Kangaroo Hunting affected by the different components of handling, Hill kangaroo (euro, Macropus robustus; kirti- specifically variability in the costs of pursuit and kirti) hunting is a common foraging activity for capture relative to prey mobility and size? and (2) men, accounting for about 20 percent of their for- How does patterned variability in such costs scale AQ74(1) Bird 1/2/09 10:54 AM Page 17

Bird et al] IN PURSUIT OF MOBILE PREY 17

Figure 4. Mean post-encounter return rates (e/h, kcal/hr handling per foraging follow) compared with mean overall return rates (E/Tf , kcal/hr foraging per foraging bout) for primary adult hunting activities discussed in the text. Handling is the summed time spent pursuing, capturing, and processing all items of a given resource type during a given adult foraging follow (n = 649). Foraging is the summed time spent searching for and handling all resources during a given adult foraging bout (see text for the distinction between follows and bouts). Note that while some resources are commonly passed over in the course of certain types of hunting, they are regularly pursued in separate, mutually exclusive foraging activities (see text and Table 1).

with commonly used zooarchaeological proxies of post-encounter return rates (e/h = kcal/hr handling) prey rank, speciﬁcally prey size and yield per unit higher than the return rate for the activity overall

processing? Figure 4 presents a test of the prey (E/Tf = kcal/hr searching + handling) should be choice model prediction that within a given forag- pursued on encounter. As expected, the only ing activity (i.e., patch), only those prey types with resources that Martu always pass over on encounter AQ74(1) Bird 1/2/09 10:54 AM Page 18

18 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

Figure 5. Distribution of post-encounter return rates (e/h, kcal/hr handling) for animal prey types, ranked by size. Each point represents, for all items of speciﬁc prey type, the sum of all calories acquired during a given foraging follow over the summed time spent pursuing, capturing, and processing on that follow (n = 440). Overlying box-and-whisker dia- grams show the median (solid bar in box), mean (dash), interquartile range (length of box), and extreme values (whiskers extend to the tenth and ninetieth percentiles) of the post-encounter return rates. Five bustard points lie beyond 60,000 kcal/hr, extending to 113,075 kcal/hr.

while hunting are those whose mean e/h is lower trading off the opportunity to reliably acquire more

than the expected E/Tf: these include seeds, pencil meat: even under the best conditions, pursuits of yams, and bush onions.4 However, in violation of bustard and hill kangaroo succeed only 30 percent PCM predictions, when foragers occasionally of the time (Table 2). If foraging for daily provi- encounter fresh signs of bustards or hill kangaroo sions, Martu ﬁnd that despite the potential bonanza, while sand monitor or perentie monitor hunting, the risks from pursuing these larger game are too they very rarely stop search to pursue these prey (2 costly. Elsewhere we have shown that time allo- out of 46 bustard pursuits/encounters and 3 out of cated to bustard and hill kangaroo hunting to the 16 hill kangaroo pursuits/encounters [n = 103 focal exclusion of more reliable types of hunting is bet- follows]). They do not stop even if they are carry- ter understood as a political rather than provisioning riﬂes and despite the fact that doing so should ing strategy (Bliege Bird and Bird 2008). increase the mean overall return rate. This is Figure 5 summarizes the distributions of the because while monitor hunting, the decision to pur- post-encounter return rates (e/h = kcal/hr in pursue a bustard or kangaroo will almost always entail suit, capture, and processing) for each animal prey AQ74(1) Bird 1/2/09 10:54 AM Page 19

Bird et al] IN PURSUIT OF MOBILE PREY 19 type pursued during a given focal follow, ordered teracted by the effect of relative prey mobility, from smallest (grubs) to largest (hill kangaroo). which is positively correlated with prey size (Fig- The interactive effects of prey size and pursuit failure 6b; r2 = .92, F = 72.83, p < .001). To control ures are well illustrated here; for example, while for this covariance and isolate possible conse- skink are the second smallest prey type (301 g/spec- quences of prey size, we regressed e/h on mobility imen), they have the highest average e/h (>20,000; and plotted the residual variance in e/h against prey see Table 2) with the most normal distribution size: even if we include only successful pursuits, (median e/h = 18,428, skewness = .16). This is a when we control for the mobility effect, size has result of the high rate of pursuit success (.963; Table no predictive value on e/h (r2 = .003, n = 272 fol- 2). Mean e/h values for larger, more mobile prey lows, F-ratio = .77, p = .38). The mobility effect (kangaroo and bustard) are much lower than we on e/h is a product of the way that predictable prey would expect from their size. As Figure 5 shows, behavior changes the probability that a given prey this is due to how highly skewed their e/h measures type will be captured if pursued (pursuit success are (skewness = 4.03 and 2.73, respectively): kan- rate). While the probability of pursuit success is garoo and bustard have a median e/h of 0 but occa- rarely considered in archaeological prey rankings, sional maxima that well exceed 50,000 kcal/hr (in it is a critical component of e/h and, as shown in the case of bustard, over 100,000 kcal/hr). Much Figure 6c, negatively correlated with prey mobil- of this variance in e/h is a result of prey mobility. ity (r2 = .87, F-ratio = 38.89, p < .001). Thus, when To measure the effects of prey mobility on e/h we add the real costs of pursuit in handling time we use an ordinal index of animal prey types that for prey of varying mobility, including pursuits that describes the upper limit of how fast prey can be lead to both successful and unsuccessful captures, relative to the hunter: (1) effectively sessile on there is no relationship between prey size and e/h encounter (cossid larvae); (2) low mobility— (Figure 6d; r2 = .003 for mean e/h of eight animal always slower than a hunter (skink); (3) moderate prey types; r2 = .002 for observed e/h by prey size mobility—torpid prey that can still potentially of animal prey types handled on each follow; n = escape a hunter over short distances (cool/dry sea- 439, F-ratio < .01, p = .98). These data show that son sand monitor, python; hot season skink); (4) if a hunter stops search while foraging in an attempt high mobility—prey that are faster than a hunter to handle an encountered prey item, the probabil- over short distances (hot season sand monitor, cat, ity of failing to capture that item (which is a func- perentie/Argus monitors); (5) fast—prey that will tion of prey mobility) dilutes any predictable effect always escape a hunter without technology that of prey size on e/h. allows a hunter to capture prey from a distance (bustard, hill kangaroo). Figure 6 presents a sum- Discussion mary of these effects, along with prey size, on post- encounter return rates (e/h). As discussed earlier, Prey mobility has profound effects on Martu hunt- most archaeological applications of foraging the- ing decisions. It determines whether or not a pur- ory have used prey body size or yield and pro- suit is likely to lead to a capture, which structures cessing costs, without considering pursuit costs, to post-encounter return rates and prey choice within generate expectations about e/h rank and changes a given patch (i.e., mutually exclusive foraging in resource use. Figure 6a shows that in the absence activities), patch choice, seasonal shifts in patch res- of pursuit costs (that is, when we measure e/h as idence time, and gender differences in time devoted kcal/hr spent capturing and processing, by sub- to different types of hunting. We propose that such tracting time spent tracking, stalking, and chasing results have implications for questions about broad- prey from handling time on follows with success- spectrum transitions and zooarchaeological mea- ful captures), body size is a reasonable predictor of sures of prey rank and abundance. increasing e/h values (r2 = .51 for mean e/h of eight animal prey types; r2 = .32 for e/h by prey size of Declining Abundance Indexes and animal prey types captured on each follow; n = Alternative Goals for Pursuing Larger Game 271, F = 127.91, p < .001). However, this rela- The method that is perhaps most commonly used tionship between size and e/h is consistently coun- in zooarchaeological investigations of prehistoric AQ74(1) Bird 1/2/09 10:54 AM Page 20

20 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

Figure 6. The counteractive effects of prey size and prey mobility on mean post-encounter return rates (e/h, kcal/hr handling per foraging follow) for animal prey types (linear ﬁt with 95 percent conﬁdence interval curves). (a) Mean e/h regressed by log prey size (see Table 2), where e/h is calculated as kcal/hr spent capturing and processing per follow, sub- tracting time spent in pursuit from handling during follows with successful captures of a given prey type. (b) The covariance of log prey size and ranked mean prey mobility (see Table 2). (c) The post-encounter pursuit success rate per follow (the probability that attempted pursuits of a given prey type lead to at least one capture on a follow) regressed by prey mobility. (d) Mean e/h regressed by log prey size, where e/h is calculated as kcal/hr spent pursuing, capturing, and processing a given prey type per follow. Results of regressions on actual (nonmean) data points are presented in the text.

resource depression, especially in North America, pogenic) in the encounter rate with large-bodied posits within-patch prey rankings based on prey prey (e.g., Broughton 1994a, 1994b, 1997, 1999, size and then calculates an abundance index (AI) 2002; Butler 2000, 2001; Byers and Broughton of the large-bodied prey relative to the sum of large- 2004; Byers and Ugan 2005; Byers et al. 2005; and small-bodied prey (Bayham 1979; see Lupo Grayson 1991; M. G. Hill 2008; Janetski 1997; 2007 for discussion). In accordance with foraging Nagaoka 2001, 2002; Szuter and Bayham 1989; models, declines in AI are thus said to imply shifts Ugan 2005a, 2005b). However, the Martu data toward a broader-spectrum economy with intensi- show that as a result of predictable differences in ﬁed investment in handling lower-ranked prey, the mobility of prey types, there is no relationship which results from depression (usually anthro- between prey size and the ranking of resources by AQ74(1) Bird 1/2/09 10:54 AM Page 21

Bird et al] IN PURSUIT OF MOBILE PREY 21 post-encounter return rate (e/h, kcal/hr handling). Increasing Abundance Indexes Quantitative evidence from a number of disparate and Resource Depression ethnographic and environmental contexts also con- Conversely, in circumstances when AI increases tradict the postulated prey size and e/h correlation over time, some researchers have treated the pat- (Bliege Bird 2007; Bliege Bird and Bird 1997; tern as a violation of the prey choice model requir- Hawkes et al. 1991; O’Connell et al. 1988; Smith ing either appeals to climatic conditions (e.g., Byers 1991:230–231; Winterhalder 1981:95–96). We sus- and Broughton 2004; Byers et al. 2005) or revisions pect that under many circumstances, this is caused to the model (e.g., Hildebrandt and McGuire 2002; by differences in prey mobility and its effect on the McGuire and Hildebrandt 2005). Based on the variance around return rates (that is, the risks of pur- results detailed here, we suggest that the absence suit failure). Larger prey, simply because they are of evidence for resource depression of larger prey large, do not always have higher return rates. If is not always a pattern that requires a special expla- researchers do not use a more accurate measure of nation. That is, in some cases, it may be that a prey rank, declines in AI may not generate predic- diachronic increase in the proportion of larger- tions about the intensified use of less profitable bodied prey in an assemblage is predicted under prey. While a decrease in the frequency of the resource depression. If the e/h for larger prey is still remains of large game may indicate overexploita- higher than the overall return rate for a specific for- tion or habitat modification, this does not neces- aging activity in which that prey would be encoun- sarily reflect the goals assumed in most foraging tered (as with bustards and hill kangaroo in the models (i.e., maximizing foraging efficiency). Martu case; see Figure 4), an increase in the AI of Where large, highly mobile game are pursued for that prey may conform to the model’s predictions, purposes other than provisioning, such as when the indicating a decrease in less mobile but higher- very costs of their pursuit and distribution guaran- ranked prey. This does not imply that abundance tee honest and widespread advertisement of a indices focused on larger prey are meaningless but, hunter’s qualities that are difficult to assess other- rather, that for issues of conservation, overex- wise (e.g., Bliege Bird and Bird 2008; Bliege Bird ploitation, and anthropogenic resource depression, et al. 2001), standard foraging models are unlikely there may be smaller, less mobile but higher-ranked to provide accurate predictions about variability in prey that are better indicators of intensification the frequency of large game (see Hildebrandt and (e.g., Jones et al. 2008a; Jones et al. 2008b; Stiner McGuire 2002; McGuire and Hildebrandt 2005). et al. 1999; Stiner et al. 2000). In a concrete exam- In other words, sometimes we may be right to infer ple, if we were to assume that the prey ranking (e/h) increasing hunting pressure from decreasing AI, but of fauna from the California coastal site of Diablo for the wrong reasons (reasons other than those Canyon (CA-SLO-2; see Jones et al. 2008b) scaled hypothesized in classic foraging models). We now with prey body size, the trans-Holocene increase need archaeologically relevant predictions that can in artiodactyl remains and concomitant declines in differentiate the two broad hypotheses (i.e., forag- some smaller prey (including the flightless duck, ing for food vs. foraging for prestige, or both) and Chendytes lawi) would be consistent with the PCM more sophisticated links between those hypothe- prediction that diet breadth contracts with increas- ses and group dynamics, cooperation, and compe- ing abundance of large game. In this case, our inter- tition (e.g., Hildebrandt and McGuire 2002; pretation of the ecological dynamics of human McGuire and Hildebrandt 2005). A first step may predation would be misleading at best: we now involve the incorporation of secondary measures know that some of the smaller prey at such sites, of prey abundance (e.g., Cannon 2003; Jones et al. the flightless duck in particular, were hunted into 2008b). Further progress can be made by estimat- extinction (Jones et al. 2008a) and are probably use- ing potential effects of prey mobility on return rates ful markers of long-term subsistence intensifica- (e.g., O’Connell 2000), which, even when very tion, the exact opposite of the conclusion we would rough, will provide analysts with tools to more reach if we assumed that larger game are always clearly identify violations of foraging models and highly ranked. If, however, relative prey mobility the determinants of diachronic decreases in abun- were considered in the prey rankings, the PCM dance indexes. should predict that a relatively immobile flightless AQ74(1) Bird 1/2/09 10:54 AM Page 22

22 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009 bird would be highly ranked and especially sus- have a predictable impact on post-encounter ceptible to overexploitation. returns. Both large and small prey with slow life histories whose mobility is predictably constrained Seasonal and Technological by season or technology should be especially sus- Effects on Prey Mobility and Rank ceptible to overexploitation and serve as key indi- The Martu data also draw our attention to the kinds cators of economic intensification. Conversely, prey of local conditions that make prey mobility an with high mobility and relatively fast life-history important determinant of variability in resource strategies (such as deer acquired by direct pursuit rank and patch choice (see O’Connell 2000 for an rather than snares) are likely to be lower ranked as illustrative example of similar pursuit costs else- subsistence resources, and their populations should where in Australia). Here we highlight two of these: be less susceptible to anthropogenic depression or seasonality and technology. The behavior of some extirpation (Whitaker 2008). Such effects are rarely game is highly conditioned by season, making a considered in archaeological resource rankings but single taxon effectively two different prey types could be estimated in actualistic studies designed (Smith 1991; Winterhalder 1981). For example, as to evaluate the proportional differences that ethno- we show, time allocated to Martu hunting activi- graphically known tactics and technologies have on ties shifts dramatically from sand monitor hunting pursuit costs of different prey. This is not a simple to perentie monitor hunting as the mobility of sand solution, but even rough devaluation of the rank- monitors and the costs of pursuing them increase ing for highly mobile prey relative to variables that in the hot months. This, in turn, structures gen- are likely to affect pursuit success will significantly dered differences in foraging, as many Martu men improve the predictive power of commonly used are more inclined to pursue the larger, more mobile foraging models. perentie. Elsewhere we have argued that this is often more related to concerns about reputations of Concluding Remark skill and generosity than foraging efficiency per se (Bliege Bird and Bird 2005, 2008). Similar seasonal The single most important dimension of zooar- differences in prey mobility are common elsewhere, chaeological applications of foraging models especially for animals that rely on nests or rook- involves establishing a reliable means for ranking eries, and are likely to generate large differences in resources based on the post-encounter return rate post-encounter return rate and resource rank (e.g., of different resources. Doing so is essential to many Bliege Bird et al. 2001; Lyman 2003). Pursuit and powerful arguments about variability in prehistoric capture technology can have similar effects on prey resource use and its implications for understand- ranking: snares, traps, blinds, and projectile ing phenomena as disparate as the evolution of weaponry can dramatically change the effective human life histories, sociopolitical complexity, gen- mobility of prey. For example, hill kangaroo and der, and sustainable harvesting. The Martu data bustard are too mobile to even be considered as clearly demonstrate that prey body size and yield potential prey (let alone be high ranked) without relative to processing costs are poor predictors of weapons that allow capture from a distance (at the the post-encounter return rates of highly mobile very least boomerangs or spears). As demonstrated prey, a pattern that we suggest is common else- here, even with rifles and vehicles, these prey are where. How then should zooarchaeologists mea- too mobile to be highly ranked. At contact Martu sure prey rank? We argue that any proxy measure readily adopted these technologies, which effec- must incorporate critical components of prey tively reduced the relative mobility of bustard and behavior, mobility being one of the most important hill kangaroo, enhancing the probability of their for predicting handling costs. Future work could capture without significantly dampening either the focus on measures that consider yield (as body size skills/knowledge that can be displayed in ensuring does) scaled by estimates of pursuit costs that are encounters or one’s ability to be generous in their largely dependent on mobility. This idea is not new distribution. Likewise, a deer captured by snare is (Jochim 1976) but has yet to be applied broadly in a very different type of prey than one captured by analysis of faunal assemblages (Stiner and Munro atlatl, and the adoption of either technology would 2002). AQ74(1) Bird 1/2/09 10:54 AM Page 23

Bird et al] IN PURSUIT OF MOBILE PREY 23

Acknowledgments. First and foremost, we would like to ing. Journal of Archaeological Science 24:887–899. thank all of the Martu at Parnngurr, Punmu, Kunawarritji, Bettinger, Robert L., Bruce Winterhalder, and Richard McEl- and Parnpajina for facilitating this project and making us reath part of their family for so many years, particularly Nyalanka 2006 A Simple Model of Technological Intensification. Journal of Archaeological Science Japartujukurr, Nyerri Morgan, Muuki Taylor, Waka and 33:538–545. Binford, Lewis R. Kanu Taylor, Desmond Taylor, and all of their children. This 1968 Post-Pleistocene Adaptations. In New Perspectives in is their work as much as ours, although the mistakes and mis- Archaeology, edited by Sally R. Binford and Lewis R. Bin- interpretations are ours alone. Brooke Scelza, Eric A. Smith, ford, pp. 313–341. Aldine, Chicago. and Chris Parker also deserve a big thanks for their help in Bird, Douglas W., and Rebecca Bliege Bird diverse aspects of this project, from grant writing to field- 2000 The Ethnoarchaeology of Juvenile Foraging: Shell- work. Our work with Martu would also not be possible with- fishing Strategies Among Meriam Children. Journal of out Robert and Myrna Tonkinson and Peter Veth, who Anthropological Archaeology 19:461–476. introduced us to the community and have generously sup- 2002 Children on the Reef: Slow Learning or Strategic For- Human Nature ported us over the years. The ideas, description, and analysis aging? 13:269–298. 2005a Evolutionary and Ecological Understandings of the presented here have benefited greatly from comments and Economics of Desert Societies. In Desert Peoples: Archae- discussion with James O’Connell, Jamie Jones, Sarah ological Perspectives, edited by Peter M. Veth,Mike Smith, Robinson, Adie Whitaker, Bill Hildebrandt, Kelly McGuire, and Peter Hiscock, pp. 81–99. Blackwell Scientific, Lon- David Zeanah, and four anonymous reviewers. The Martu don. project is supported by funding from the National Science 2005b Mardu Children’s Hunting Strategies in the Western Foundation (BCS-0127681 and 0314406), the Leakey Desert, Australia: Foraging and the Evolution of Human Foundation, the Christensen Fund, Stanford University, and Life Histories. In Hunter Gatherer Childhoods, edited by the Stanford Archaeology Center. Barry S. Hewlett and Michael E. Lamb, pp. 129–146. Aldine Transaction, New Brunswick, New Jersey. 2009 Competing to Be Leaderless: Food Sharing and Mag- References Cited nanimity Among Martu Aborigines. In The Emergence of Leadership: Transitions in Decision Making from Small- Alvard, Michael S., and Lawrence Kuznar Scale to Middle-Range Societies, edited by John Kantner, 2001 Deferred Harvests: The Transition from Hunting to Kevin J. Vaughn, and Jelmer W. Eerkins. School of Animal Husbandry. American Anthropologist Advanced Research Press, Santa Fe, in press. 103:295–311. Bird, Douglas W., Rebecca Bliege Bird, and Christopher H. Barlow, K. Renee Parker 2002 Predicting Maize Agriculture Among the Fremont:An 2004a Women Who Hunt with Fire: Aboriginal Fire Use Economic Comparison of Farming and Foraging in the and Fire Regimes in Australia’s Western Desert. Australian American Southwest. American Antiquity 67:65–88. Aboriginal Studies 2004:90–96. Barlow, K. Renee, and Duncan Metcalfe 2005 Aboriginal Burning Regimes and Hunting Strategies 1996 Plant Utility Indices: Two Great Basin Examples. in Australia’s Western Desert. Human Ecology Journal of Archaeological Science 23:351–371. 33:443–464. Barnosky, Andrew D., Paul L. Koch, Robert S. Feranec, Scott Bird, Douglas W.,Rebecca Bliege Bird, and Jennifer L. Richard- L. Wing, and Alan B. Shabel son 2004 Assessing the Causes of Late Pleistocene Extinctions 2004b Meriam Ethnoarchaeology: Shellfishing and on the Continents. Science 306:70–75. Shellmiddens. Memoirs of the Queensland Museum, Cul- Basgall, Mark tural Heritage Series 3:183–197. 1987 Resource Intensification Among Hunter-Gatherers: Bird, Douglas W., and James F. O’Connell Acorn Economies in Prehistoric California. Research in 2006 Human Behavioral Ecology and Archaeology. Jour- Economic Anthropology 9:21–52. nal of Archaeological Research 14:143–188. Bayham, Frank E. Bliege Bird, Rebecca 1979 Factors Influencing the Archaic Pattern of Animal 1999 Cooperation and Conflict: The Behavioral Ecology of Exploitation. Kiva 44:219–235. the Sexual Division of Labor. Evolutionary Anthropology Beaton, John M. 8:65–75. 1973 The Nature of Aboriginal Exploitation of Mollusk 2007 Fishing and the Sexual Division of Labor Among the Populations in Southern California. Unpublished Master’s Meriam. American Anthropologist 109:442–451. thesis, Department of Anthropology, University of Cali- Bliege Bird, Rebecca, and Douglas W. Bird fornia, Los Angeles. 1997 Delayed Reciprocity and Tolerated Theft: The Behav- 1991 Extensification and Intensification in Central Cali- ioral Ecology of Food Sharing Strategies. Current Anthro- fornia Prehistory. Antiquity 65:947–951. pology 38:49–78. Bettinger, Robert L. 2002 Constraints of Knowing or Constraints on Growing? 1991 Hunter-Gatherers: Archaeological and Evolutionary Fishing and Collecting Among the Children of Mer. Theory. Plenum Press, New York. Human Nature 13:239–268. 1993 Doing Great Basin Research Recently: Coping with 2005 Human Hunting Seasonality. In Primate Seasonality, Variability. Journal of Archaeological Research 1:43–66. edited by Diane K. Brockman and Carel P. van Schaik, pp. Bettinger, Robert L., and Martin A. Baumhoff 243–266. Cambridge University Press, Cambridge. 1982 The Numic Spread: Great Basin Cultures in Compe- 2008 Why Women Hunt: Risk and Contemporary Forag- tition. American Antiquity 47:485–503. ing in a Western Desert Aboriginal Community. Current Bettinger, Robert L., Ripan Mahli, and Helen McCarthy Anthropology 49:655–693. 1997 Central Place Models of Acorn and Mussel Process- Bliege Bird, Rebecca, Douglas W. Bird, Brian F. Codding, AQ74(1) Bird 1/2/09 10:54 AM Page 24

24 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

Christopher H. Parker, and James H. Jones 2001 Changing Fish Use on Mangeia, Southern Cook 2008 The “Fire Stick Farming” Hypothesis:Australian Abo- Islands. International Journal of Osteoarchaeology riginal Foraging Strategies, Biodiversity, and Anthro- 11:88–100. pogenic Fire Mosaics. Proceedings of the National Byers, David A., and Jack M. Broughton Academy of Sciences 105:14796–14801. 2004 Holocene Environmental Change, Artiodactyl Abun- Bliege Bird, Rebecca, and Eric A. Smith dances, and Human Hunting Strategies in the Great Basin. 2005 Signaling Theory, Strategic Interaction, and Symbolic American Antiquity 69:235–256. Capital. Current Anthropology 46:221–248. Byers, David A., Craig S. Smith, and Jack M. Broughton Bliege Bird, Rebecca, Eric A. Smith, and Douglas W. Bird 2005 Holocene Artiodactyl Population Histories and Large 2001 The Hunting Handicap: Costly Signaling in Male For- Game Hunting in the Wyoming Basin, USA. Journal of aging Strategies. Behavioral Ecology and Sociobiology Archaeological Science 32:125–142. 50:9–19. Byers, David A., and Andrew Ugan Botkin, Steven 2005 Should We Expect Large Game Specialization in the 1980 Effects of Human Exploitation on Shellfish Popula- Late Pleistocene? An Optimal Foraging Perspective on tions at Malibu Creek, California. In Modeling Change in Early Paleoindian Diet. Journal of Archaeological Science Prehistoric Subsistence Economies, edited by Timothy 32:1624–1640. Earle and Andrew L. Christenson, pp. 121–139. Acade- Cane, Scott mic Press, New York. 1989 Australian Aboriginal Seed Grinding and Its Archae- Bouey, Paul D. ological Record: A Case Study from the Western Desert. 1987 The Intensification of Hunter-Gatherer Economies in In Foraging and Farming:The Evolution of Plant Exploita- the Southern North Coast Ranges of California. Research tion, edited by David R. Harris and Gordon C. Hillman, in Economic Anthropology 9:53–101. pp. 99–119. Unwin Hyman, London. Braidwood, Robert J. Cannon, Michael D. 1960 The Agricultural Revolution. Scientific American 2000 Large Mammal Relative Abundance in Pithouse and 203:130–148. Pueblo Period Archaeofaunas from Southwestern New Braje,Todd J., Douglas J. Kennett, Jon M. Erlandson, and Bren- Mexico: Resource Depression in the Mimbres-Mogollon? den J. Culleton Journal of Anthropological Archaeology 19:317–347. 2007 Human Impacts on Nearshore Shellfish Taxa:A 7,000- 2003 A Model of Central Place Forager Prey Choice and Year Record from Santa Rosa Island, California. Ameri- an Application to Faunal Remains from the Mimbres Val- can Antiquity 72:735–756. ley, New Mexico. Journal of Anthropological Archaeol- Brand-Miller, Janette, Keith W. James, and Patricia M. A. Mag- ogy 22:1–25. giore Charnov, Eric L. 1993 Tables of Composition of Australian Aboriginal Foods. 1976 Optimal Foraging, the Marginal Value Theorem. The- Aboriginal Studies Press, Canberra. oretical Population Biology 9:367–390. Bright, Jason, Andrew Ugan, and Lori Hunsaker Charnov, Eric L., and Gordon H. Orians 2002 The Effect of Handling Time on Subsistence Tech- 1973 Optimal Foraging: Some Theoretical Explorations. nology. World Archaeology 34:164–181. Mimeo, Department of Biology, University of Utah, Salt Broughton, Jack M. Lake City. 1994a Declines in Mammalian Foraging Efficiency During Clark, J. Desmond the Late Holocene, San Francisco Bay. Journal of Anthro- 1952 Prehistoric Europe: The Economic Basis. Cambridge pological Archaeology 13:371–401. University Press, Cambridge. 1994b Late Holocene Resource Intensification in the Sacra- Codding, Brian F., Terry L. Jones, Judith F. Porcasi, and Jereme mento Valley, California: The Vertebrate Evidence. Jour- Gaeta nal of Archaeological Science 21:501–514. 2008 Integrating History and Behavioral Ecology in Archae- 1997 Widening Diet Breadth, Declining Foraging Effi- ological Interpretation: Case Studies from the Central Cal- ciency, and Prehistoric Harvest Pressure: Icthyofaunal Evi- ifornia Coast. In Ecology, Evolution, and Human Behavior: dence from the Emeryville Shellmound. Antiquity Perspectives from California and Great Basin Archaeol- 71:845–862. ogy, edited by Adrian R. Whitaker and Deanna Grimstead. 1999 Resource Depression and Intensification During the University of Utah Press, Salt Lake City, in press. Late Holocene, San Francisco Bay: Evidence from the Cohen, Mark N. Emeryville Shellmound Vertebrate Fauna. University of 1977 The Food Crisis in Prehistory: Overpopulation and California Press, Berkeley. the Origins of Agriculture. Yale University Press, New 2002 Prey Spatial Structure and Behavior Affect Archaeo- Haven. logical Tests of Optimal Foraging Models: Examples from Davenport, Susan, Peter Johnson, and Yuwali the Emeryville Shellmound Vertebrate Fauna. World 2005 Cleared Out: First Contact in the Western Desert. Archaeology 34:60–83. Aboriginal Studies Press, Canberra. Broughton, Jack M., and Frank E. Bayham de Boer, William F. 2003 Showing Off, Foraging Models, and the Ascendance 2000 Between the Tides: The Impact of Human Exploita- of Large Game Hunting in the California Middle Archaic. tion on an Intertidal Ecosystem, Mozambique. Universal American Antiquity 68:783–789. Press, Veenendaal, the Netherlands. Broughton, Jack M., and Donald Grayson Edwards, Douglas A., and James F. O’Connell 1993 Diet Breadth,Adaptive Change, and the White Moun- 1995 Broad Spectrum Diets in Arid Australia. Antiquity tains Faunas. Journal of Archaeological Science 69:769–783. 20:331–336. Egeland, Charles P., and Ryan M. Byerly Butler, Virginia L. 2005 Applications of Return Rates to Large Mammal 2000 Resource Depression on the Northwest Coast of North Butchery and Transport Among Hunter-Gatherers and Its America. Antiquity 74:649–661. Implications for Plio-Pleistocene Hominid Carcass For- AQ74(1) Bird 1/2/09 10:54 AM Page 25

Bird et al] IN PURSUIT OF MOBILE PREY 25

aging and Site Use. Journal of Taphonomy 3:135–158. the California and Oregon Coasts. Journal of Anthropo- Emlen, Stephen logical Archaeology 11:360–401. 1966 The Role of Time and Energy in Food Preference. Hildebrandt, William R., and Kelly R. McGuire American Naturalist 100:611–617. 2002 The Ascendance of Hunting During the California Erlandson, Jon M. Middle Archaic: An Evolutionary Perspective. American 1991 Shellfish and Seeds as Optimal Resources: Early Antiquity 67:231–256. Holocene Subsistence on the Santa Barbara Coast. In 2003 Large Game Hunting, Gender-Differentiated Work Hunter-Gatherers of Early Holocene Coastal California, Organization, and the Role of Evolutionary Ecology in Cal- edited by John M. Erlandson and Robert H. Colten, pp. ifornia and Great Basin Prehistory. American Antiquity 101–111. Institute of Archaeology, University of Califor- 68:790–792. nia, Los Angeles. Hill, Matthew E., Jr. Fitzhugh, Benjamin 2008 Variation in Paleoindian Fauna Use on the Great Plains 2003 The Evolution of Complex Hunter-Gatherers: Archae- and Rocky Mountains of North America. Quaternary ological Evidence from the North Pacific. Plenum Press, International (DOI:10.1016/j.quaint.2007.10.004), in New York. press. Flannery, Kent V. Hill, Mathew G. 1969 Origins and Ecological Effects of Early Domestica- 2008 Paleoindian Subsistence Dynamics on the North- tion in Iran and the Near East. In The Domestication and western Great Plains Zooarchaeology of the Agate Basin Exploitation of Plants and Animals, edited by Peter J. Ucko and Clary Ranch Sites. British Archaeological Reports and G. W. Dimpleby, pp. 73–100. Duckworth, London. International Series, Oxford. Glassow, Michael A., and Larry R. Wilcoxon Janetski, Joel 1988 Coastal Adaptations near Point Conception, Califor- 1997 Fremont Hunting and Resource Intensification in the nia, with Particular Regard to Shellfish Exploitation. Amer- Eastern Great Basin. Journal of Archaeological Science ican Antiquity 53:36–51. 24:1075–1089. Grayson, Donald K. Jochim, Michael A. 1991 Alpine Faunas from the White Mountains, California: 1976 Hunter-Gatherer Subsistence and Settlement: A Pre- Adaptive Change in the Great Basin? Journal of Archae- dictive Model. Academic Press, New York. ological Science 18:483–506. Jones, Terry L. 2001 The Archaeological Record of Human Impacts on 1991 Marine-Resource Value and the Priority of Coastal Animal Populations. Journal of World Prehistory 15:1–67. Settlement: A California Perspective. American Antiquity Grayson, Donald K., and Michael Cannon 56:419–443. 1999 Human Paleoecology and Foraging Theory in the 1995 Transitions in Prehistoric Diet, Mobility, Exchange Great Basin. In Models for the Millennium: Great Basin and Social Organization Along California’s Big Sur Coast. Anthropology Today, edited by Charlotte Beck, pp. Unpublished Ph.D. dissertation, Department of Anthro- 141–151. University of Utah Press, Salt Lake City. pology, University of California, Davis. Grayson, Donald K., and Francoise Delpech Jones, Terry L., William R. Hildebrandt, Douglas J. Kennett, 2002 Specialized Early Upper Paleolithic Hunters in South- and Judith F. Porcasi western France? Journal of Archaeological Science 2004 Prehistoric Marine Mammal Overkill in the North- 29:1439–1449. eastern Pacific: A Review of New Evidence. Journal of Gremillion, Kristen J. California and Great Basin Anthropology 24:69–80. 2004 Seed Processing and the Origins of Food Production Jones, Terry L., Judith F. Porcasi, Jon M. Erlandson, H. Dallas in Eastern North America. American Antiquity Jr., T. A. Wake, and R. Schwaderer 69:215–234. 2008a The Protracted Holocene Extinction of California’s Hawkes, Kristen Flightless Sea Duck (Chendytes lawi) and Its Implications 1991 Showing Off: Tests of an Hypothesis About Men’s for the Pleistocene Overkill Hypothesis. Proceedings of Foraging Goals. Ethology and Sociobiology 12:29–54. the National Academy of Science 105:4105–4108. 2000 Big Game Hunting and the Evolution of Egalitarian Jones, Terry L., Judith F. Porcasi, Jereme Gaeta, and Brian F. Societies: Lessons from the Hadza. In Hierarchies in Codding Action: Cui Bono? edited by Michael Diehl, pp. 59–83. 2008b The Diablo Canyon Fauna: A Coarse-Grained Center for Archaeological Investigations, Occasional Paper Record of Trans-Holocene Foraging from the Central Cal- 27. Southern Illinois University Press, Carbondale. ifornia Mainland Coast. American Antiquity 73, in press. 2003 Grandmothers and the Evolution of Human Longevity. Jones, Terry L., and Jennifer R. Richman American Journal of Human Biology 15:380–400. 1995 On Mussels: Mytilus californianus as a Prehistoric Hawkes, Kristen, and Rebecca Bliege Bird Resource. North American Archaeologist 16:33–58. 2002 Showing Off, Handicap Signaling, and the Evolution Kelly, Robert L. of Men’s Work. Evolutionary Anthropology 11:58–67. 1995 The Foraging Spectrum: Diversity in Hunter-Gatherer Hawkes, Kristen, and James F. O’Connell Lifeways. Smithsonian Institution Press,Washington, D.C. 1992 On Optimal Foraging Models and Subsistence Tran- Kennett, Douglas J. sitions. Current Anthropology 33:63–65. 2005 The Island Chumash: Behavioral Ecology of a Mar- Hawkes, Kristen, James F. O’Connell, and Nicholas G. Blur- itime Society. University of California Press, Berkeley. ton Jones Kennett, Douglas J., and Bruce Winterhalder (editors) 1991 Hunting Income Patterns Among the Hadza: Big 2006 Human Behavioral Ecology and the Origins of Food Game, Common Goods, Foraging Goals, and the Evolu- Production. University of California Press, Berkeley. tion of the Human Diet. Philosophical Transactions of the Klein, Richard G., Graham Avery, Kathryn Cruz-Uribe, David Royal Society, London, Series B 334:243–251. Halkett, John E. Parkington, Teresa Steele, Thomas P. Vol- Hildebrandt, William R., and Terry L. Jones man, and Royden Yates 1992 Evolution of Marine Mammal Hunting: A View from 2004 The Ysterfontein 1 Middle Stone Age Site, South AQ74(1) Bird 1/2/09 10:54 AM Page 26

26 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

Africa, and Early Human Exploitation of Coastal 1988 Hunting Hoppers. American Antiquity 53:593–604. Resources. Proceedings of the National Academy of Sci- Madsen, David B., and David N. Schmitt ence 101:5708–5715. 1998 Mass Collecting and the Diet Breadth Model:A Great Kuhn, Steven L., and Mary C. Stiner Basin Example. Journal of Archaeological Science 2001 The Antiquity of Hunter-Gatherers. In Hunter- 25:445–457. Gatherers: An Interdisciplinary Perspective, edited by Mannino, Marcello A., and Ken D. Thomas Catherine Panter-Brick, Robert H. Layto, and Peter 2002 Depletion of a Resource? The Impact of Prehistoric Rowley-Conwy, pp. 99–142. Cambridge University Press, Human Foraging on Intertidal Mollusk Communities and Cambridge. Its Significance for Human Settlement, Mobility, and Dis- 2006 What’s a Mother to Do? The Division of Labor Among persal. World Archeology 33:452–474. Neanderthals and Modern Humans in Eurasia. Current McGuire, Kelly R., and William R. Hildebrandt Anthropology 47:953–978. 2005 Rethinking Great Basin Foragers: Prestige Hunting Lambert, Patricia M. and Costly Signaling During the Middle Archaic Period. 1993 Health in Prehistoric Populations of the Channel American Antiquity 70:695–712. Islands. American Antiquity 58:509–522. Munro, Natalie D. Layton, Robert H., Robert A. Foley, and Elizabeth Williams 2004 Zooarchaeological Measures of Hunting Pressure and 1991 The Transition Between Hunting and Gathering and Occupational Intensity in the Natufian: Implications for the Specialized Husbandry of Resources. Current Anthro- Agricultural Origins. Current Anthropology 45:S5–S34. pology 32:255–274. Nagaoka, Lisa Lindstrom, Susan 2001 Using Diversity Indices to Measure Changes in Prey 1996 Great Basin Fisher-Folk: Optimal Diet Breadth Mod- Choice at the Shag River Mouth Site, New Zealand. Inter- eling the Truckee River Aboriginal Subsistence Fishery. national Journal of Osteoarchaeology 11:101–111. In Prehistoric Hunter-Gatherer Fishing Strategies, edited 2002 The Effects of Resource Depression on Foraging Effi- by Mark G. Plew, pp. 114–179. Boise State University ciency, Diet Breadth, and Patch Use in Southern New Press, Boise. Zealand. Journal of Anthropological Archaeology Lourandos, Harry 21:419–442. 1983 Intensification:A Late Pleistocene–Holocene Archae- 2005 Declining Foraging Efficiency and Moa Carcass ological Sequence from Southwestern Victoria. Archae- Exploitation in Southern New Zealand. Journal of Archae- ology in Oceania 18:81–94. ological Science 32:1328–1338. Lupo, Karen D. O’Connell, James F. 1998 Experimentally Derived Extraction Rates for Mar- 2000 An Emu Hunt. In Australian Archaeologist: Collected row: Implications for Body Part Exploitation Strategies of Papers in Honour of Jim Allen, edited by Atholl Anderson Plio-Pleistocene Hominid Scavengers. Journal of Archae- and Tim Murray, pp. 172–181. Division of Archaeology ological Science 25:657–675. and Natural History, Research School of Pacific and Asian 2006 What Explains the Carcass Field Processing and Studies, Australian National University, Canberra. Transport Decisions of Contemporary Hunter-Gatherers? 2007 How Did Modern Humans Displace Neanderthals? Measures of Economic Anatomy and Zooarchaeological Insights from Hunter-Gatherer Ethnography and Archae- Skeletal Part Representation. Journal of Archaeological ology. In Neanderthals and Modern Humans Meet? edited Method and Theory 13:19–66. by Nicholas Conard, pp. 43–64. Kerns Verlag, Tübingen. 2007 Evolutionary Foraging Models in Zooarchaeological O’Connell, James F., and Kristen Hawkes Analysis: Recent Applications and Future Challenges. 1981 Alyawara Plant Use and Optimal Foraging Theory. In Journal of Archaeological Research 15:143–189. Hunter-Gatherer Foraging Strategies: Ethnographic and Lupo, Karen D., and David N. Schmitt Archaeological Analyses, edited by Bruce Winterhalder 2002 Upper Paleolithic Net-Hunting, Small Prey Exploita- and Eric A. Smith, pp. 99–125. University of Chicago tion, and Women’s Work Effort: A View from the Ethno- Press, Chicago. graphic and Ethnoarchaeological Record of the Congo 1984 Food Choice and Foraging Sites Among the Alyawara. Basin. Journal of Archaeological Method and Theory Journal of Anthropological Research 40:435–504. 9:147–179. O’Connell, James F., Kristen Hawkes, and Nicholas G. Blur- 2005 Small Prey Hunting Technology and Zooarchaeolog- ton Jones ical Measures of Taxonomic Diversity and Abundance: 1988 Hadza Hunting, Butchering and Bone Transport and Ethnoarchaeological Evidence from Central African For- Their Archaeological Implications. Journal of Anthropo- est Foragers. Journal of Anthropological Archaeology logical Research 44:113–161. 24:335–353. O’Connell, James F., Kristen Hawkes, Karen D. Lupo, and Lyman, R. Lee Nicholas G. Blurton Jones 2003 Pinneped Behavior, Foraging Theory and the Depres- 2002 Male Strategies and Plio-Pleistocene Archaeology. sion of Metapopulations and Nondepression of a Local Journal of Human Evolution 43:831–872. Population on the Southern Northwest Coast of North O’Connell, James F., Kevin T. Jones, and Steven R. Simms America. Journal of Anthropological Archaeology 1982 Some Thoughts on Prehistoric Archaeology in the 22:376–388. Great Basin. In Man and Environment in the Great Basin, MacArthur, Robert, and Eric Pianka edited by David B. Madsen and James F. O’Connell, pp. 1966 On Optimal Use of a Patchy Environment. American 227–240. Society for American Archaeology,Washington, Naturalist 100:603–609. D.C. Madrigal, T. Cregg, and Julie Zimmerman Holt O’Connell, James F., and Fiona B. Marshall 2002 White-Tailed Deer Meat and Marrow Return Rates 1989 Analysis of Kangaroo Body Part Transport Among and Their Applications to Eastern Woodlands Archaeol- the Alyawara of Central Australia. Journal of Archaeo- ogy. American Antiquity 67:745–759. logical Science 16:393–405. Madsen, David B., and James Kirkman Perlman, Stephen M. AQ74(1) Bird 1/2/09 10:54 AM Page 27

Bird et al] IN PURSUIT OF MOBILE PREY 27

1980 An Optimum Diet Model, Coastal Variability, and Smith, Eric A. Hunter-Gatherer Behavior. Advances in Archaeological 1991 Inujjuamiut Foraging Strategies: Evolutionary Ecol- Method and Theory 3:257–310. ogy of an Arctic Hunting Economy. Aldine de Grutyer, Petruso, Karl M., and James F. Wickens Hawthorne, New York. 1984 The Acorn in Aboriginal Subsistence in Eastern North 2004 Why Do Good Hunters Have Higher Reproductive America: A Report on Miscellaneous Experiments. In Success? Human Nature 15:343–364. Experiments and Observations on Aboriginal Wild Plant Smith, Eric A., Rebecca Bliege Bird, and Douglas W. Bird Food Utilization in Eastern North America, edited by 2003 The Benefits of Costly Signaling: Meriam Turtle Patrick J. Munson, pp. 360–378. Prehistory Research Hunters and Spearfishers. Behavioral Ecology Series, Vol. VII. Indiana Historical Society, Indianapolis. 14:116–126. Porcasi, Judith F., Terry L. Jones, and L. Mark Raab Stephens, David W., and John R. Krebs 2000 Trans-Holocene Marine Mammal Exploitation on San 1986 Foraging Theory. Princeton University Press, Prince- Clemente Island, California: A Tragedy of the Commons ton. Revisited. Journal of Anthropological Archaeology Stiner, Mary C. 19:200–220. 2002 Carnivory, Coevolution, and the Geographic Spread Raab, L. Mark of the Genus Homo. Journal of Archaeological Research 1992 An Optimal Foraging Analysis of Prehistoric Shell- 10:1–63. fish Collecting on San Clemente Island, California. Jour- 2005 The Faunas of Hayonim Cave (Israel): A 200,000-Year nal of Ethnobiology 12:63–80. Record of Paleolithic Diet, Demography, and Society. Redding, Richard Peabody Museum Press, Harvard University, Cambridge. 1988 A General Explanation of Subsistence Change: From Stiner, Mary C., and Natalie D. Munro Hunting and Gathering to Food Production. Journal of 2002 Approaches to Prehistoric Diet Breadth, Demography Anthropological Archaeology 7:56–97. and Prey Ranking Systems in Time and Space. Journal of Reidhead, Van A. Archaeological Method and Theory 9:181–214. 1976 Optimization and Food Procurement at the Prehistoric Stiner, Mary C., Natalie D. Munro, and Todd A. Surovell Leonard Haag Site, Southeastern Indiana: A Linear Pro- 2000 The Tortoise and the Hare: Small Game Use, the gramming Model. Unpublished Ph.D. dissertation, Depart- Broad-Spectrum Revolution, and Paleolithic Demogra- ment of Anthropology, Indiana University, Bloomington. phy. Current Anthropology 41:39–73. Russell, Kenneth W. Stiner, Mary C., Natalie D. Munro, Todd A. Surovell, Eitan Tch- 1988 After Eden: The Behavioral Ecology of Food Pro- ernov, and Ofer Bar-Yosef duction in the Near East and North Africa. International 1999 Paleolithic Population Growth Pulses Evidenced by Series 391. British Archaeological Reports, Oxford. Small Animal Exploitation. Science 283:190–194. Sassaman, Kenneth E. Sutton, Mark Q. 2004 Complex Hunter-Gatherers in Evolution and History: 1995 Archaeological Aspects of Insect Use. Journal of A North American Perspective. Journal of Archaeologi- Archaeological Method and Theory 2:253–298. cal Research 12:227–280. Szuter, Carl, and Frank Bayham Schmitt, David N., David B. Madsen, and Karen D. Lupo 1989 Sedentism and Prehistoric Animal Procurement 2004 The Worst of Times, the Best of Times: Jackrabbit Among Desert Horticulturalists of the North American Hunting by Middle Holocene Human Foragers in the Bon- Southwest. In Farmers as Hunters: The Implications of neville Basin of Western North America. In Colonisation, Sedentism, edited by Susan Kent, pp. 67–78. Cambridge Migration, and Marginal Areas: A Zooarchaeological University Press, Cambridge. Approach, edited by Mariana Mondini, Sebastián Munõz, Talalay, Lauren, Donald R. Keller, and Patrick J. Munson and Stephen Wickler, pp. 86–95. Oxbow Books, Oxford. 1984 Hickory Nuts, Walnuts, Butternuts and Hazelnuts: Schoener, Thomas W. Observations and Experiments Relevant to Their Aborig- 1971 Theory of Feeding Strategies. Annual Review of Ecol- inal Exploitation in Eastern North America. In Experi- ogy and Systematics 2:369–404. ments and Observations on Aboriginal Food Utilization Shennan, Stephen in Eastern North America, edited by Patrick J. Munson, 2002 Genes, Memes and Human History. Thames and Hud- pp. 338–359. Prehistory Research Series Vol. VI, No. 2. son, London. Indiana Historical Society, Indianapolis. Sih, Andrew, and Bent Christensen Thomas, David H. 2001 Optimal Diet Theory: When Does It Work, and When 2008 Native American Landscapes of St. Catherines Island, and Why Does It Fail? Animal Behaviour 61:379–390. Georgia: I. The Theoretical Framework. Anthropological Simms, Steven R. Papers, No. 88. American Museum of Natural History, 1985 Acquisition Cost and Nutritional Data on Great Basin New York. Resources. Journal of California and Great Basin Anthro- Thomas, Frank R. pology 7:117–126. 2002 An Evaluation of Central-Place Foraging Among Mol- 1987 Behavioral Ecology and Hunter-Gatherer Foraging: lusk Gatherers in Western Kiribati, Micronesia: Linking An Example from the Great Basin. International Series Behavioral Ecology with Ethnoarchaeology. World 381. British Archaeological Reports, Oxford. Archaeology 34:182–208. Simms, Steven R., and Kenneth W. Russell Thoms, Alston V. 1997 Bedouin Hand Harvesting of Wheat and Barley: Impli- 1989 The Northern Roots of Hunter-Gatherer Intensifica- cations for Early Cultivation in Southwestern Asia. Cur- tion: Camas and the Pacific Northwest. Unpublished Ph.D. rent Anthropology 38:696–702. dissertation, Department of Anthropology, Washington Smith, Craig S., William Martin, and Kristine A. Johansen State University, Pullman. 2001 Sego Lilies and Prehistoric Foragers: Return Rates, Tonkinson, Robert Pit Ovens, and Carbohydrates. Journal of Archaeological 1974 The Jigalong Mob: Aboriginal Victors of the Desert Science 28:169–184. Crusade. Cummings, Menlo Park. AQ74(1) Bird 1/2/09 10:54 AM Page 28

28 AMERICAN ANTIQUITY [Vol. 74, No. 1, 2009

1978 Semen vs. Spirit Child in a Western Desert Culture. Prehistoric Prey Populations in Northwestern California. In Australian Aboriginal Concepts, edited by L. R. Hiatt, Unpublished Ph.D. dissertation, Department of Anthro- pp. 81–92. Australian Institute of Aboriginal Studies, Can- pology, University of California, Davis. berra. Willey, Gordon R., and Phillip Phillips 1988a Egalitarianism and Inequality in a Western Desert 1958 Method and Theory in American Archaeology. Uni- Culture. Anthropological Forum 5:545–558. versity of Chicago Press, Chicago. 1988b Ideology and Dominations in Aboriginal Australia: Winterhalder, Bruce A Western Desert Test Case. In Hunters and Gatherers 2: 1981 Foraging Strategies in the Boreal Forest: An Analy- Property, Power and Ideology, edited by Tim Ingold, David sis of Cree Hunting and Gathering. In Hunter-Gatherer Dressing Riches, and James Woodburn, pp. 150–164. Berg, Foraging Strategies: Ethnographic and Archaeological Oxford. Analyses, edited by Bruce Winterhalder and Eric A. Smith, 1990 The Changing Status of Aboriginal Women: “Free pp. 66–98. University of Chicago Press, Chicago. Agent” at Jigalong. In Going It Alone? Prospects for Abo- Winterhalder, Bruce, and Eric A. Smith riginal Autonomy, edited by Robert Tonkinson and Michael 2000 Analyzing Adaptive Strategies: Human Behavioral C. Howard, pp. 125–148. Aboriginal Studies Press, Can- Ecology at Twenty-Five Years. Evolutionary Anthropology berra. 9:51–72. 1991 The Mardu Aborigines: Living the Dream in Aus- Wohlgemuth, Eric tralia’s Desert. Holt, Rinehart and Winston, New York. 1996 Resource Intensification in Prehistoric Central Cali- 2007 Aboriginal “Difference” and “Autonomy” Then and fornia: Evidence from Archaeobotanical Data. Journal of Now: Four Decades of Change in a Western Desert Soci- California and Great Basin Anthropology 18:81–103. ety. Anthropological Forum 17:41–60. Wright, Katherine I. Ugan, Andrew 1994 Ground-Stone Tools and Hunter-Gatherer Subsistence 2005a Does Size Matter? Body Size, Mass Collecting, and in Southwest Asia: Implications for the Transition to Farm- Their Implications for Understanding Prehistoric Forag- ing. American Antiquity 59:238–263. ing Behavior. American Antiquity 70:75–90. Yesner, David R. 2005b Climate, Bone Density, and Resource Depression: 1989 Effects of Prehistoric Aleut Exploitation on Sea- What Is Driving Variation in Large and Small Game in Fre- Mammal Populations. Arctic Anthropology 25:28–43. mont Archaeofaunas? Journal of Anthropological Archae- 1994 Seasonality and Resource “Stress” Among Hunter- ology 24:227–251. Gatherers: Archaeological Signatures. In Key Issues in Ugan, Andrew, Jason Bright, and Alan Rogers Hunter-Gatherer Research, edited by Ernest S. Burch and 2003 When Is Technology Worth the Trouble? Journal of Linda J. Ellanna, pp. 151–167. Berg, Oxford. Archaeological Science 30:1315–1329. Zeanah, David Veth, Peter M. 2000 Transport Costs, Central Place Foraging and Hunter- 1987 Martujarra Prehistory: Variation in Arid Zone Adap- Gatherer Alpine Land Use Strategies. In Intermountain tations. Australian Archaeology 25:102–111. Archaeologies, edited by David B. Madsen and Michael 1989 Islands in the Interior: A Model for the Colonisation D. Metcalf, pp. 1–14. University of Utah Anthropological of Australia’s Arid Zone. Archaeology in Oceania Papers, No. 122. University of Utah Press, Salt Lake City. 24:81–92. 2004 Sexual Division of Labor and Central Place Forag- 1995 Aridity and Settlement in North West Australia. Antiq- ing: A Model for the Carson Desert of Western Nevada. uity 69:733–746. Journal of Anthropological Anthropology 23:1–32. 2000 Origins of the Western Desert Language: Conver- Zeanah, David, and Steven R. Simms gence in Linguistic and Archaeological Space and Time 1999 Modeling the Gastric: Great Basin Subsistence Stud- Models. Archaeology in Oceania 35:11–19. ies Since 1982 and the Evolution of General Theory. In 2005 Cycles of Aridity and Human Mobility: Risk- Models for the Millennium: Great Basin Anthropology Minimization Amongst Late Pleistocene Foragers of the Today, edited by Charlotte Beck, pp. 118–140. University Western Desert, Australia. In Desert Peoples: Archaeo- of Utah Press, Salt Lake City. logical Perspectives, edited by Peter M. Veth, Mike A. Smith, and Peter Hiscock, pp. 100–114. Blackwell Pub- lishing, Oxford. Notes Veth, Peter M., and Fiona Walsh 1988 The Concept of “Staple” Plant Foods in the Western 1. Some might question whether tracking should be Desert Region of Western Australia. Australian Aborigi- charged to the handling budget or is more appropriately nal Studies 1988:19–25. assigned as a search variable. The prey choice model is con- Waguespack, Nicole M. cerned with the trade-offs entailed in decisions to stop search- 2005 The Organization of Male and Female Labor in For- ing for an array of possibilities in order to handle an aging Societies: Implications for Early Paleoindian Archae- ology. American Anthropologist 107:666–676. encountered resource. For mobile prey, handling involves the Waguespack, Nicole M., and Todd A. Surovell pursuit of a particular item (or unit of items) of a specific 2003 Clovis Hunting Strategies, or How to Make Out on resource type to make capture and processing possible. As Plentiful Resources. American Antiquity 68:333–352. such, Martu tracking must be considered as a pursuit cost Walsh, Fiona because this is the point at which a forager decides whether 1990 An Ecological Study of Traditional Aboriginal Use to pursue a particular animal or continue searching for other of “Country”: Martu in the Great and Little Sandy Deserts, resources. In this case it is simply wrong to classify tracking Proceedings of the Ecological Society Western Australia. as search, because the hunter tracks a specific animal to the of Australia 16:23–37. exclusion of others. Even for sessile or less mobile prey, Whitaker, Adrian R. 2008 The Role of Human Predation in the Structuring of Martu pursuits often begin before the resource is directly AQ74(1) Bird 1/2/09 10:54 AM Page 29

Bird et al] IN PURSUIT OF MOBILE PREY 29

sighted (e.g., for cossid larvae, foragers switch from search to and thin (2–3 cm in diameter), manufactured from either pursuit on encounter with fresh droppings and subtle bulges wooden shafts of Acacia dictyophleba or metal rods acquired beneath the bark of a eucalyptus or at the base of witchetty from spare windmill pump shafts (see Figure 2). They are acacia and then follow and probe the intraphloem tunnel ﬁre-hardened and sharpened in a spatula shape at one end, tracks to extricate the grubs). Elsewhere, similar criteria, specialized to facilitate both probing and digging. including tracking, are identiﬁed as crucial for calculating 4. Note that these resources are exploited, but only as sep- handling costs and post-encounter return rates of mobile prey arate foraging activities, not when encountered during hunt- (e.g., Winterhalder 1981:82, Table 4.4 notes). ing. 2. All errors are given as standard deviation. 3. A woman’s digging stick (wana) is individually owned and highly curated, made for her by her husband, father, or a particularly close male relative. Wana are long (150–220 cm) Received April 10, 2008; Accepted June 27, 2008. AQ74(1) Bird 1/2/09 10:54 AM Page 30