sign in sign up
<<
Home , Holocene extinction

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/304073030

Pyrodiversity and the : the role of fire in the broad spectrum revolution

Article in Evolutionary Anthropology Issues News and Reviews · May 2016 DOI: 10.1002/evan.21482

CITATIONS READS 6 178

3 authors, including:

Douglas Rebecca Bliege Bird Pennsylvania State University Pennsylvania State University

67 PUBLICATIONS 2,217 CITATIONS 53 PUBLICATIONS 2,840 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Rebecca Bliege Bird on 11 July 2016.

The user has requested enhancement of the downloaded file. Evolutionary Anthropology 25:105–116 (2016)

ARTICLE

Pyrodiversity and the Anthropocene: the Role of Fire in the Broad Spectrum Revolution

DOUGLAS W. BIRD, REBECCA BLIEGE BIRD, AND BRIAN F. CODDING

The Anthropocene colloquially refers to a global regime of -caused environ- Binford10 was among the first to mental modification of systems associated with profound changes in patterns draw on Boserup to argue that diversi- of human mobility, as well as settlement and resource use compared with prior eras. fication in hunted resources, food Some have argued that the processes generating the Anthropocene are mainly asso- processing, and storage facilities may ciated with growth and technological innovation, and thus began only in result from an intensified use of the the late under conditions of dense sedentism and industrial .1 environment that significantly pre- However, it now seems clear that the roots of the Anthropocene lie in complex proc- dates agriculture. Flannery’s11 hypo- esses of intensification that significantly predate transitions to agriculture.2,3 What thesis to account for such ‘‘broad intensification is remains less clear. For some it is increasing economic productivity spectrum revolutions’’ (BSR) then that increases , the drivers of which may be too diverse and too local paved the way for using intensification to generalize.4,5 For others using Boserup’s ideas about agrarian intensification, in an explanatory sense. For Flannery, increasing density in hunter-gatherer populationscanproducedeclinesinsubsist- the BSR was driven by an imbalance ence efficiency that increase incentives for investing labor to boost yield per unit area, between resources and , 6–8 9 which then elevates Malthusian limits on carrying capacity. As Morgan demon- resulting in an increase in labor and a strates in a comprehensive review, the legacy of such Boserupian intensification is decrease in foraging efficiency. Faced alive, well, and controversial in hunter-gatherer archeology. This is a result of its with diminishing returns, Epipaleo- potential for illuminating processes involved in transformations of forager socio- lithic populations in the Near East political and economic systems, including those dominated by harvesting more spread into less productive environ- immediate-return resources and high residential mobility as well as those character- ments, which necessitated increased ized by more delayed-return material economies with reduced residential mobility, a reliance on a wider array of more broader spectrum of resources, degrees of storage, and greater social stratification. labor-intensive, but more predictable, Here we detail hypotheses about the processes involved in such transitions and and . Flannery sug- explore the way that anthropogenic disturbance of ecosystems, especially the use of gested that this set the stage for (or landscape fire, could be fundamentally entangled with many broad-spectrum revolu- backed hunter-gatherers into) the tions associated with intensified foraging systems. social and technological processes that eventually lead to domestication and agriculture.

Douglas Bird is Associate Professor of Anthropology at Penn State University. He has spent much of the last twenty-five years working in Indigenous communities in . Recent publications of his research on the socio-ecology of human subsistence have been featured in Current Anthropology, International, and American Antiquity. Email: [email protected]

Rebecca Bliege Bird is Professor of Anthropology and the Inter-College Program in Ecology at Penn State University. She has published widely on gender and fire ecology in Indigenous societies, including recent articles in Proceedings of the National Academy of Sciences and Philosophical Transactions of the Royal Society B.

Brian Codding is Assistant Professor of Anthropology at the University of Utah. His research is both ethnographic and archeological in , with a focus on present and past human-environment interactions in Australia and . Recent articles on his research in both regions have been published in venues such as Proceedings of the National Academy of Sciences, Journal of Archaeological , and .

Key words: intensification; behavioral ecology; disturbance ecology; Aboriginal Australia

VC 2016 Wiley Periodicals, Inc. DOI: 10.1002/evan.21482 Published online in Wiley Online Library (wileyonlinelibrary.com). 106 Bird et al. ARTICLE

This longstanding explanation for ity or patch, patch-residence time economic and socio-political inten- the conditions that lead to the ori- models15 provide solutions to foraging sification and changing patterns of gins of agriculture has been formal- and traveling trade-offs. These models mobility and colonization under a ized in behavioral ecological (BE) predict the threshold at which a forager broad range of circumstances. Recent models that allow for the generation should switch from foraging (searching examples illustrate how tests of model and testing of clear predictions tied and handling) for one suite of resour- predictions give new insight into clas- to the theory of by natural ces and travel elsewhere to search for sic archeological problems.22–27 We selection.12,13 While many tests have and handle other suites of resources. suspect that processes of intensifica- been upheld, some have used theo- The more foragers are tethered to a tion illustrated in the logic of foraging retical and empirical exceptions to central place, the more salient are theory are important in many transi- challenge both the original formula- trade-offs in transporting resources. tions to domestication and the requi- 5 16 tion and its BE formalizations. Spe- Central-place foraging models pre- site social institutions of farming that cifically, these critics argue that the dict thresholds at which individuals have radically transformed earth classic BE models of resource use should move resources to their base or systems.28,29 ignore the development of social their base to the resources. Scaling up institutions necessary for the transi- to the landscape as a collection of hab- THE PROBLEM: PROPERTY AND tion to delayed return resources and itats, models of ideal distribution17,20 elements of niche construction neces- then generate expectations about how ANTHROPOGENIC HABITAT sary to account for habitat modifica- foragers should distribute themselves; CONSTRUCTION tion associated with agriculture. After that is, where they should be based Models in behavioral ecology have detailing the classic behavioral eco- and when they should move depend- proven useful in unveiling many of the logical approach and its critiques, we ing on the suitability of available habi- dynamics that likely accompany inten- show how these challenges can be tats in an environment and the density sification. However, recent work has incorporated into behavioral ecologi- of conspecifics already occupying those highlighted two issues. First, straight- cal frameworks to provide a compre- habitats. The models predict the order forward ideas about how decreasing in which habitats will be colonized and hensive model for broad-spectrum subsistence efficiency leads to intensi- relative differences in population den- revolutions that lead to agriculture. fication need to grapple with under- sity and despotic intensity. standing the development of the Taken together, foraging models THE BEHAVIORAL ECOLOGY social institutions implicit in the transi- frame testable hypotheses about the OF INTENSIFICATION tion from immediate-return or public- potential processes of intensification. goods to delayed-return or private-goods In behavioral ecology, such ideas They generate expectations about the based economies. about intensification have been cast in direction of economic transformation Bowles and Choi30 have recently various formal models of prey choice, with shifts toward a reliance on a developed a model suggesting that patch use, central-place foraging, and broader spectrum of resources charac- shifts to a delayed-return economy ideal free distribution, which empha- terized by higher handling costs, how depend on the number of individuals size the trade-offs organisms face with resource depression may link with pat- variability in foraging efficiency and terns of mobility and residence, how in any social group who act on the link habitat quality. The models can be human populations are likely to spread between labor and product in such a thought of as tiered solutions to prob- and increase with greater reliance on way that those who labor have priority lems a hypothetical forager would face more costly but abundant resources, in controlling the distribution of mate- in finding and using resources across and how resulting circumscription and rial associated with their labor. Norms different temporal and spatial scales the colonization of less suitable habitat and institutions that manage private of resource distribution. At the most may result in declining residential property and storage are thus critical local (patch scale), potential resource mobility and increased socio-political components in the maintenance of a 31 types can be ranked on a scale of util- stratification.9,12,18,19 There are com- delayed-return material economy. ity according to their post-encounter mon patterns in relationships between Along with these shifts toward delays yield and requisite handling costs (for reduced residential mobility, increas- in material return are institutions example, energy per unit of time to ing investment in processing tech- such as households as units of eco- pursue and process). If an individual nology, and shifts to lower-ranked nomic production, more centralized is concerned with the opportunity resources.13 As foragers spend propor- governance to police ownership and costs of foraging and experiences tionally more time handling lower- redistribution, specialized scaled lower rates of encounter with high- ranked resources and face associated labor, and inequalities based on the ranked resources in a given patch, the declines in the benefits of moving else- accumulation of material wealth. If prey choice model14 predicts that where when neighboring habitats are one individual stores her harvest selectivity should broaden to include more densely settled, it becomes more instead of sharing it, she inevitably resources of lower utility in rank worthwhile to invest in technologies to produces inequalities between pro- order. Scaling up to the habitat as a reduce handling costs.20,21 ducers, increasing the social costs of collection of patches as foraging effi- Foraging models have thus pro- hoarding that we argue shape nonreci- ciency declines in a given type of activ- vided a set of generic tools to explore procal sharing in many immediate ARTICLE Pyrodiversity and the Anthropocene 107

plete solutions to the problems with existing intensification models. While Bowles and Choi argue a necessary co-evolution between subsistence and social institutions governing property and ownership rights, Zeder and Smith’s approach presupposes such property rights in the design or man- agement of ecosystems for sustain- able benefit. Likewise, while Zeder and Smith’s model includes a focus on the ecological dynamics of human- environment interactions, the ab- sence of ecological dynamics in the Bowles and Choi model does not allow for the incorporation of proc- esses that shift property norms. Nei- ther model addresses issues of fundamental conflicts of interest Figure 1. Intensification with anthropogenic habitat modification. A simple visual repre- inherent in individual decision- sention of the way foraging strategies that depend on intermediate disturbance to making and the collective action maintain immediate efficiency can feed back to increase patchiness, habitat heteroge- required to sustain property regimes neity, insentives to store, and the opportunity costs of mobility. that govern storage and delayed mate- rial return.41 32–34 5 return contexts. However, if every- Along these lines, Zeder and In what follows, we suggest that 37–39 one is storing and not sharing, the Smith have recently pointed out both of these modifications to the clas- social costs are minimal. What Bowles that in some archeological contexts it sic intensification explanation of the and Choi show is that the maintenance is not clear that a BSR in foraging BSR can be integrated by understand- of a delayed-return economy character- economies is preceded by population ing how the behavioral ecology of indi- istic of farming requires cultural insti- pressure or evidence of declines in for- vidual decision-making interacts with tutions to manage material property aging productivity. In response, they the ecological dynamics of anthropo- for storage and redistribution. The have developed an approach that calls genic disturbance that structure how question they pose but do not address on niche construction and complex resources are distributed in space and 40 is how societies get to the threshold at systems theory to propose that anth- time. These dynamics affect the patch- which it pays for individuals to invest ropogenic construction and man- iness and predictability of some types in maintaining institutions that man- agement of ecosystems to improve of resources, in some cases reducing age material storage as opposed to subsistence lie at the heart of the the cost of acquisition so that assert- sharing? BSR. While offering a more nuanced ing ownership over them becomes The second issue centers on pre- perspective on human-environment worthwhile (Fig. 1.). Whether or not sumptions about the direction of interaction, this approach introduces resources are worth acquiring and environmental change and human additional problems. Perhaps most then defending depends to a great behavior. As opposed to farming, in importantly, it provides no framework extent on the history and practice which environments are engineered for developing hypotheses about of forager-environment interactions. for food production, foraging is decision-making with regard to re- Ecology shifts the relative costs of often portrayed as a merely extrac- source use beyond notions of inten- sharing versus storing in ways that tive activity in which food is har- tionality in innovation. Actors who affect most individuals within a social vested from a nonanthropogenic intentionally modify habitats gain via group, thereby setting the benefits for environment. Variability in foraging future, delayed payoffs. This delay sets resource defense and reduced mobil- strategies is thus often modeled as a up a collective action problem: How ity. Both the marginal valuation of static, linear relationship: Either can individuals in immediate-return defense at the scale of the individual environmental change results in societies, which typically operate harvest42,43 and the marginal valua- behavioral change or behavioral under forms of common property ten- tion of defense at the scale of the patch change results in environmental ure regimes, ensure that their labor to or landscape44 are predicted to affect change. We know that, even among improve future resource availability the benefits of staying put and defend- foragers, the process is more will not induce free-riders to take ing claims to exclusive use rights over dynamic. Patterns of embodied for- advantage of their forethought, leav- a resource. Individuals are predicted aging practices and values emerge in ing them with little future benefit to to share more and store less when complex, nonlinear interactions of show for their trouble? there is high intertemporal variation anthropogenic construction, cogni- The co-evolution and niche con- in foraging success (stochasticity), tion, and material engagement.1,36 struction approaches are both incom- package size and perishability is high, 108 Bird et al. ARTICLE and the degree of interforager correla- cal (“nonequilibrium”) perspectives interactions58–60 and provide tion in harvest success is low.45 consider interacting networks of spe- rescaling and habitat protection effects Both empirical observations of cies and include the study of indirect for species that require a variety of hab- sharing among hunter-gatherers and and positive effects and how they are itats for both food and shelter.61–63 experimental work on the links involved in the persistence and stabil- This, in turn, may increase between sharing norms and unpre- ity of entire ecological communities. stability, allowing more species to per- dictable rewards support these pre- Considered at the scale of the ecologi- sist with more stable populations than dictions. More synchronously and cal community, negative effects like in the absence of such disturbance,64 predictably acquired resources, such are a form of disturbance and, particularly, affect the composi- as small , tend to be character- that can result in positive outcomes tion of communities by increas- ized by a greater percentage kept for for the community as a whole. These ing patch density and habitat edges. the consumption of acquirers and positive effects come about indirectly, This would favor populations of dietary their families.46,47 Experiments in in- without agent intent or management, or habitat generalists, which would see dustrialized societies (Japan and the when one organism has an effect on their foraging returns increase through U.S.), corroborate this pattern, sug- another via a third organism (for reductions in the cost of travel between 65 gesting that the links between expect- example, wolves shape riparian dy- patches or encounters with prey. This ations about sharing versus keeping namics by reducing elk populations). would make it easier for such species can be manipulated by providing These effects also can include direct to switch to alternative prey when resources that are associated with an interactions, like mutualism and com- high-ranked prey became scarce, facili- unpredictable link between labor and mensalism. Positive effects, whether tating the persistence of prey popula- 66 reward. Windfall (stochastic) re- direct or indirect, are increasingly tions. Reducing the cost of accessing sources are commonly shared more viewed as a significant force in the a wider range of resources and increas- widely than are resources with strong assembly of ecological communities ing the number of encounters with any links between labor and production,48 and may be critical in sustaining pop- one species would also reduce dietary ulation stability in a complex food suggesting similar processes behind variability and the possibility of harvest web.49,50 These positive effects are such norm development. This sug- failure across individuals. often associated with “keystone spe- gests that when everyone has access The co-evolutionary ecological model cies,” those that have a high number to economically defendable resources we explore here suggests that the emer- of trophic connections,51 construct gent properties of human-environment (when resources are spatially and niches for other organisms, or provide interaction, with intermediate levels of temporally predictable, with patchy disturbance at intermediate temporal ecosystem disturbance, shape land- distribution decreasing the costs of and spatial scales.52 Such species can scapes and ecological communities in defense), and when the amount of have positive effects on a wide range ways that cause shifts in species distri- time invested predicts harvest size, of other species, so that their removal butions,whichinturnshapehuman ownership may be likely to emerge can cause extensive population extinc- social organization in ways that favor and spread in a population. We argue 53 tions. Keystone species tend to greater benefits of ownership and pro- that such conditions, with profoundly 54 be top predators, super-generalist ducer control over resource distribu- dynamic networks of nonlinear inter- omnivores with many weak links to tion. The model (Fig. 1) links ecological action, are more likely to emerge 55 other species. They also tend to be disturbance creating patchiness and where foragers regularly and system- 56 large-bodied or to modify the physi- predictability in certain resources with atically disturb landscapes in ways 57 cal environment in significant ways. social processes of sharing and cooper- that increase patchiness and habitat When such species modify environ- ation. We propose that shifts in heterogeneity. We can think of no ments with disturbance at intermedi- ecological conditions that increase pro- anthropogenic disturbance in human ate temporal and spatial scales, it duction of small prey, reduce variabili- foraging more salient than landscape tends to produce more positive ecolog- ty in acquisition, increase group size, burning. ical effects, often increasing species reduce mobility, and thus favor storing diversity or the stability of ecological over sharing can be an emergent pro- DISTURBANCE AND THE communities.58 pertyofthewayhunter-gatherers ECOLOGICAL DYNAMICS OF Disturbing vegetation at an interme- interact with and animal com- 67 MOBILITY AND OWNERSHIP diate scale would theoretically have munities. This would be so especially two potential effects: it might increase in savannahs and , where The notion that the BSR can be in-patch species diversity (alpha-diver- applications of fire at a patch scale can explained by either resource depres- sity) by interrupting processes of plant dramatically improve immediate sub- sion or intentional strategies to man- succession that result in domination sistence returns. age it emerges from an ecological by a few competitive species and mi- focus on simple (dyadic) negative in- ght increase diversity at larger scales FIRE AS A SOURCE teractions between species, mainly by increasing the heterogeneity of OF DISTURBANCE involving competition for resources, patch types across the landscape predation, and the direct effect of (beta-diversity). Intermediate levels of The idea that landscape fire as a one species on another. New ecologi- landscape heterogeneity may stabilize source of environmental disturbance ARTICLE Pyrodiversity and the Anthropocene 109 might be fundamentally involved in evolutionary processes, so that where associated with anthropogenic fire, subsistence intensification is not new. Aboriginal burning was interrupted by which fragments landscapes, shapes Lewis68 explored the idea in an often colonial forces the result was an species distributions, and facilitates overlooked but important paper. abrupt .79,80 ecological interactions that support What he was not able to articulate distinctive habitats and landscapes. was the process by which fire affected BROAD SPECTRUM INSTITUTIONS Many Australian Aboriginal groups the immediate returns from foraging IN AUSTRALIA’S WESTERN DESERT have maintained or reinstituted cus- and how this affects the likelihood of tomary burning practices in their intensified production and all of its There is now considerable archeo- homelands. Despite dramatic changes, concomitant social structures. logical support for the view that while critical components of their ecosys- In seasonally dry forests, wood- people have occupied greater Australia tems remain intact and contiguous, lands, savannahs, and grasslands, the for at least 45k years, something and fundamental cultural institutions use of broadcast fire at a patch or important happened in many Aborigi- continue to be supported by their stand scale is widespread among for- nal societies during the Holocene and treatment of landscapes with fire. Fig- agers. Indeed, it probably is the princi- that the mid- to Late-Holocene in par- ure 2 is a map of all anthropogenic ple form of anthropogenic ecosystem ticular was a turning point in Austra- fires from April through August 2014, disturbance with a deep evolutionary lian prehistory, with diverse along with the location and usual pop- history in our lineage.69,70 In 1963, signatures of social and economic ulation of discrete Aboriginal com- Stewart71 argued that application of intensification in many locales.78,81,82 munities across the Western and fire to landscapes in traditional sub- In the Central and Western deserts, Central Deserts of Australia. We pro- sistence systems was a fundamental Codding,83,84 Edwards and O’Con- duced the map by outlining the foot- component of processes shaping over nell,85 Smith,78 and Veth86 have print of every new fire visible in a quarter of the world’s ecosystems. shown that the latter half of the Holo- Landsat imagery across the entire He reprimanded both anthropologists cene, especially the last 2,000 years, region during the season (cool-dry) and ecologist for not considering was marked by a dramatic increase in when there are no sources of ignition indigenous practice and knowledge reliance on plants, especially grass other than people. The vast majority with regard to landscape burning as seeds, that require significant process- of the fires that dominate the remote fundamental in understanding wide- ing, and a proportional increase in region are lit in the course of daily spread social and ecological inter- small game, especially monitor lizards trips out from the remote actions.71 Stewart’s argument was and small . These shifts communities, on travel between com- largely ignored and his full treatise on occurred in association with the munities through the region, and in indigenous fire remained unpublished archeological appearance of distinc- prescribed burning in communities until long after his death.72 tive elements of ethnographic desert with ranger programs. The remote Stewart, however, was right, and his tool kits, specifically woodworking region — the vast interior with no proposal has been especially relevant adzes, seed-grinding implements, and rural centers and only small Aborigi- to understanding the nature of indige- composite weaponry. Late Holocene nal communities — is dominated by nous social and ecological integration and ethnographic technology are desert hummock grasslands and com- in California73,74 and Australia. Unlike more “curated” tool kits, with imple- prised almost entirely of contiguous elsewhere, in Australia it has been dif- ments purpose-built for specific func- Native Title, Indigenous Protected ficult to ignore the central role of fire tions, more suited to logistical than Areas, and Aboriginal Reserve Lands.88 in Aboriginal economic, social, and residential mobility. Smith78 argues Hunting-firesbeyondthisremote ecological systems. The immediacy of that these indicate clear signatures of region are restricted by administration contact among ecologically bent mobility reorganization in the Late of non-Aboriginal land and by wood- anthropologists, historians, archeolo- Holocene, with key sites occupied land/shrubland dominated habitats gists, and traditional owners who more often by groups with reduced where the use of fire for hunting is less maintain their burning practices residential mobility and increased common. forces our attention away from the investments in logistical mobility Here we explore the contingent destructive side of fire and toward the (more frequent and longer subsistence social and ecological processes of fire way in which intentionally applied fire trips out from a residential base, sensu among Martu foragers in the Western constructs anthropogenic landscapes. Binford87). That these “broad- Desert (see Fig. 2 for the Martu Native Such landscapes comprise the “land- spectrum revolutions” in the desert Title). Previously, we have shown that esque capital” of classic Aboriginal occurred independent of the reorgan- Martu mosaic burning strategies are estates.75–78 The distinctive institu- ization often associated with the adop- maintained in support of immediate tions that manage property, regulate tion and spread of classically defined foraging efficiency.89,90 From simple resource distribution, shape mobility, farming points to the dramatic diver- rules to increase immediate hunting and facilitate social exchange are fun- sity of human social and economic returns with fire, anthropogenic sys- damentally linked to the nature of an systems during the Holocene.100 Work tems emerge with radically rescaled anthropogenic niche that emerges with contemporary Aboriginal forag- fire regimes. The emergent land- from regular practices of firing the ers suggests that broad-spectrum scapes are buffered from climate- landscape. These are bound in co- desert economies are fundamentally driven cycles of drought and wildfire, 110 Bird et al. ARTICLE

Figure 2. The spatial distribution of early dry-season fires between April and August, 2014, around Aboriginal communities across the Central and Western Deserts of Australia. Fires were visually detected and hand-digitized using a ratio of Landsat 8 OLI infrared bands 7 and 5. and are characterized by increased foragers, including but not exclusively to late-1960s. Today, families based density of habitat heterogeneity, com- people who refer to themselves as out of Parnngurr maintain a mixed pressed cycles of vegetative succes- Martu, continue to maintain funda- economy based on hunting and gath- sion, and modified distribution and mental aspects of their customary life ering, social security, and craft density of animal populations, all of ways and livelihoods. These practices income from painting and traditional which feed back to improve hunting include economic pursuits focused on arts. and gathering return rates and the hunting, gathering, and patch mosaic On and off for the last 15 years, our predictability of key subsistence burning, as well as pursuits that are research team has been living with resources.79,90–92 Here, we examine fundamentally intertwined with the Martu based out of Parnngurr. Over Martu use of fire as an important maintenance of social relations and that time, we have amassed a large source of disturbance to investigate ritual obligations.93 In the mid-1980s, quantitative observational dataset on whether habitats shaped by human many Martu returned to their home- contemporary foraging, including fires produce higher hunting return lands after a 20-year exile in missions more than 4,500 observation hours of rates for small animals, reduce inter- and settlements on the desert fringe, foraging during more than 1,900 indi- 32 forager harvest variability, reduce establishing three remote commun- vidual foraging bouts. Martu spend mobility, and increase an acquirer’s ities, Parnngurr, Punmu, and Kuna- 25%-30% of their days hunting or ability to control distribution in ways warritji (Fig. 2). Among the more gathering; they rely on foraged food that could facilitate ownership. We recently contacted were Martu who for 35%-50% of their daily diet, de- 94 argue that these are conditions com- took up permanent residence at the pending on the season. While they mon across broad-spectrum foraging site of a uranium lease at Parnngurr continue to collect a wide array of economies and shape emergent prop- rockhole, owned then by an Austra- plant and insect resources (especially erties of interactions requisite for lian subsidiary of Rio Tinto, an inter- Solanum spp. fruit, Hakea spp. and farming and conditions that led to the national mining company. This group Gevillea spp. nectar, and Endoxyla Anthropocene. of about 70 people included several spp. grubs), men and women are pri- families from the Kartujarra, Manyjil- marily hunters. The main prey, espe- jarra, and Warnman estate owners, cially of women, are several species of CONTEMPORARY MARTU 93,95 who felt that their claim to the area monitor lizards. Bustards (Ardeo- In the remote desert regions of was quite strong. Most in this group tis australis) and hill kangaroo (Mac- Western Australia, many Aboriginal were full-time foragers until the mid- ropus robustus), mostly hunted by ARTICLE Pyrodiversity and the Anthropocene 111

Figure 3. The effect of fire on the mosaic of vegetative succession in the Martu homelands. The stand ignition map shows the cumula- tive effect of both Martu and lightning fires between 2000 and 2010. Fires were visually detected and hand-digitized using a ratio of Landsat 7 infrared bands 7 and 5. Light colors indicate more recent fires; dark shades indicate older fires. Regions across the study area are stratified into four categories of foraging intensity from 4,106 adult foraging hours observed between 2000 and 2010: Category 1 includes regions with 0-0.05 forager days/km2,25 0.06-0.25 forager days/km2,35 0.26-2 forager days/km2,45 >2 forager hrs. A) The number of patches at different stages of vegetative succession/km2 regressed by foraging intensity. B) shows the number of old-growth patches (patches that remained unburned)/km2 regressed by foraging intensity. men, make up the bulk of the ual spent searching for, pursuing, cap- acquired in this foraging activity remainder.32,95,96 turing, and processing resources from make up more than half of all hunt- In the following analysis, we com- a given foraging activity, and associ- ing time and provide nearly half of bine previously reported metrics on dif- ated yield by resource type. all the bush food people consume.32 ferences between Martu modified Sand monitors burrow during the landscapes and those that emerge from MARTU FIRE, FORAGING, cold winter months and live off stored 79 a nonanthropogenic fire regime with AND MOBILITY fat. The lizards are hunted with fire in spatial variability in foraging intensity the winter and tracked in recently and efficiency.90 Details on methods The immediate function of Martu burned ground in the summer. Hunts are provided in those publications but, burning operates at the scale of the in winter target old-growth spinifex briefly, we use a composite of 10 years patch and is designed to clear off grasslands last burned more than 5 to of vegetative patch heterogeneity large stands of hummock grass (spi- 10 years ago, which provide shelter to (interpatch variability in succession fol- nifex, Triodia spp.) in order to monitor lizards. Also, only past this lowing fire) measured from satellite increase the efficiency of searching stage is vegetation thick enough carry imagery between 2000-2010. We then for and tracking small animals, pri- a fire. Sand-monitor hunting returns map spatial differences in foraging marily burrowed ones. Today, the are high in both winter and summer, return rates and variability in foraging major prey hunted with fire is the averaging between 1,500-2,000 kcal/ time onto the composite heterogeneity sand monitor lizard (Varanus goul- forager hr until about one year after map. Time allocation and return rates dii). However, while searching for fire (given rain). Hunting returns then from different foraging activities were these lizards, hunters often pursue a decline rapidly as the burned patch is recorded in systematic observations of range of resources on encounter revegetated.90 Hunting without burn- 1,831 adult foraging bouts between (mostly other species of monitor, as ing in patches older than about 5–7 2000-2010.32 A foraging bout consists well as Tiliqua spp. skink, snakes, years is inefficient, averaging only of measures of time per day an individ- and Solanum spp. fruit). Resources between 25–100 kcal/forager hr. 112 Bird et al. ARTICLE

nificantly higher kangaroo popula- tions, likely because the benefits of an anthropogenic fire regime include increased access to patches of high- quality forage and a net increase in kangaroo at intermediate distances from Martu communities.91 The patchy landscapes created through the accumulation of several years of winter hunting fires have long- term effects on the distribution of plants and animals. Landscapes where hunting is most intense also have more patches of mid-successional , which has a significantly high density of high-ranked seed grasses such as Eragrostis spp., which were a precon- tact staple, and fruits such as Sola- num.92 Again, just as with hunting Figure 4. Martu foraging mobility is significantly lower in anthropogenically modified patches, this higher density minimizes regions. The graph shows a subsample of pedestrian sand monitor hunting focal individ- ual follows in which GPS tracks were recorded (n 5 35) comparing the proportion of for- the likelihood of having no seed- aging time individuals spent in travel between patches and search within patches in bearing patches available within a rea- remote regions (use category 1, see Fig. 3) versus anthropogenic regions (use categories sonable travel distance from any camp. 2-4, see Figure 4). Errors are 95% confidence intervals. The anthropogenic landscape thus re- scales resource patch density to reduce the cost of accessing a wider range of When iterated across the landscape, If resource patches are closer to- resources. the aggregate effect of burning to hunt gether in anthropogenic landscapes, Figure 5 shows the overall foraging has a large effect on the scale of land- this implies that mobility should return rates (kcal/forager hr) for both scape patchiness. Figure 3 shows a decline in landscapes with increased collected and hunted resources, in- stand ignition map of the region: the use. Figure 4 illustrates this with a cluding both small and large game, according to regional differences in cumulative effect of both Martu and subsample of winter foraging bouts intensity of foraging use. Returns are lightning fires between 2000 and 2010. for sand-monitor lizards where GPS highest in areas under intermediate Patches of different time since fire, or tracks were recorded. When people use, where the negative effects of successional stage, are associated with forage on foot in anthropogenic land- anthropogenic resource depression different plant communities. It takes scapes, they spend about 10% less are outweighed by the positive effects about ten years to cycle through to the time in mobile search and travel. of human fire. Even in the most used final successional stage in which spini- This scaling of patch density not regions, Martu achieve significantly fex grass will completely dominate the only reduces the costs of foraging for better returns than they do in more patch. In this analysis, we classify the Martu, it also increases habitat qual- remote regions. This is not a result of composite stand ignition map into four ity for certain species. As we have 90 foragers simply targeting areas that regions according to the intensity of shown elsewhere, Martu use of a have higher inherent successional foraging use across all adult foraging region increases the density of their diversity because, as we have previ- hours (4,106) observed between 2000 major prey, the sand monitor lizard. ously demonstrated, without anthro- and 2010, where intensity is measured Density is higher in anthropogenic pogenic burning heterogeneity as forager days per square km. As regions despite the fact that hunting dramatically declines.79 shown in Figure 3a, more foraging pro- pressure is also higher. This is likely Because resource patch heterogene- duces significantly greater density of because sand monitors are edge- ity increases with foraging use inten- different patch types (patches of differ- loving generalists, who do better in sity, at the level of the camp, each ent successional communities) across anthropogenic environments. Other forager has better access to high- the landscape. Fires are dramatically species show different responses to quality resource patches in anthropo- smaller where Martu hunt more often, anthropogenic patchiness. Larger ani- genic landscapes. Also, there is less leading to a higher density of diverse mals such as kangaroo also benefit variability in returns across individu- patches.79,92 Particularly, as shown in from the landscape diversity that als per day (Fig. 6). Everyone in a Figure 3b, the density of unburned Martu fires create: areas that are camp is roughly equally successful in patches increases significantly with most heavily hunted have lower kan- an anthropogenic environment, lead- more use, meaning the distance one garoo densities, but so do areas with ing to a lower daily coefficient of vari- has to travel to reach a suitable hunting no hunting or very limited hunting. ation in daily food production in patch for burning is shorter. Regions of intermediate use have sig- anthropogenic landscapes compared ARTICLE Pyrodiversity and the Anthropocene 113

hierarchy based on material wealth? This question gets at the heart of ubiquitous shifts toward increasingly broad-spectrum economies, reduced residential mobility and, ultimately, the origins of agriculture. The standard model of the broad- spectrum revolution is one of intensi- fication. An increase in human popu- lation density causes reductions in mobility and a decline in high ranked resources, particularly large-bodied, slowly renewing prey. This then shifts foraging toward a wider array of Figure 5. Mean foraging efficiency in anthropogenic landscapes compared to unmodi- low-ranked resources (smaller prey fied landscapes. Efficiency is measured as kcal/forager hr across all foraging bouts and handling-intensive plant foods), recorded between 2000 and 2010 (bout 5 time an individual spent in search, pursuit, which spurs investment in technolo- capture, and processing in a given foraging activity per day, n 5 1,721). Anthropogenic modification is measured by category of use intensity (4 5 regions of highest use; see Fig. gies to reduce associated handling 3.). Errors are 95% confidence intervals. costs, which are the roots of processes associated with plant or animal dom- estication and agriculture. to unmodified landscapes. Further- The Martu data indicate that people Issues with the standard model are more, the increase in small game pro- operating in anthropogenic fire reg- highlighted in arguments about whe- ductivity and the reduction in the cost imes experience increased returns ther population increases or declines of accessing patches for hunting from small-game hunting at the imme- in efficiency precede shifts to broad- spectrum economies, where it app- small game and plants increase the diate scale, which supports long-term ears, in some cases, that there are no proportion of overall camp produc- increases in subsistence returns over detectable declines in high-ranked tion comprised of small game in more longer periods. Long-term small-game resources.23 Whether or not the re- anthropogenic habitats. Small ani- returns are higher because human fire sources of concern are “high-ranked” mals and collected resources, relative increases patchiness and fine-scale (high yield for time spent pursuing to large animals, increase from 37% habitat heterogeneity, which increases and processing) needs to be demon- of camp harvest in remote regions to the predictability and density of gener- strated. The assumption that large more than 80% in the most inten- alist and staple animal prey near cen- bodied, slowly renewing prey are sively used regions (Fig. 7). Previ- tral places. This tends to reduce always ranked higher than smaller, ously, we have shown that, relative to mobility, increase reliance on small more quickly renewing prey, depends large game (hill kangaroo), small prey, and reduce variability across on the probability that pursuits will game (sand monitor) is associated individuals, which decreases sharing be successful.32,95 Equally important, with increased producer control over and increases the benefits to hunters 93,95,97 however, the standard model needs to distribution. of controlling distribution. Critically, account for the necessary develop- The effects of decreased mobility fire is not being used deliberately as a ment of norms of ownership, which also have a strong influence on the habitat management tool; rather, the make possible shifts to a delayed costs of claiming the benefits of own- use of fire is maintained by its imme- return economy more dependent on ership. In the small bush camps that diate benefits to hunting burrowed food storage.30 Martu form on a daily basis, compris- animals. Complex systems then em- Our analysis suggests that the ing from 5-20 adults from various res- erge from these interactions. emergent properties of human- idential camps within the permanent environment interaction can shape community, members do not segre- SYNTHESIS: ON THE BROAD landscapes and ecological commun- gate into separate households but, SPECTRUM REVOLUTION ities in ways that cause shifts in instead, act as a single food-pooling species distributions, which, in turn, unit. Acquirers of small game keep a These results are directly relevant shape human social organization in smaller average portion for consump- to the central question regarding the ways that favor the benefits of redu- tion than that given to others. How- social and ecological processes sur- ced mobility and producer control ever, when sharing occurs in larger, rounding mobility and subsistence over resource distribution. The model long-term camps where people have shifts in hunter gatherer economies: links ecological conditions creating segregated into multiple residential How do such societies shift from the patchiness and predictability in cer- groupings, the average portion that more immediate-return economy of tain resources with the social proc- acquirers keep for distribution to mobile hunter gatherers to a more esses of sharing and cooperation. The their own dependents is greater than delayed return economy focused on shifts in ecological conditions that that given to others. sedentism, storage, and greater social increase production of small prey, 114 Bird et al. ARTICLE

prediction that the degree to which acquirers bias sharing in ways that produce greater material benefit for their own dependents (producer pri- ority) increases positively with the daily production of others in camp. As others produce more, one should keep more of one’s own production, a necessary step toward being able to accumulate and defend access to food stores. This is the case for small ani- mals but not larger ones. Harvests of small animals show substantial con- tingency of harvest size on foraging time, leading to a stronger link Figure 6. Mean coefficient of variation (CV) in daily foraging production in anthropo- between labor and the products of genic landscapes compared to unmodified landscapes. Variability in production is that labor, while harvests of larger measured as CV in kcals produced across all individuals on all camp days recorded animals do not.95 between 2000 and 2010 (n 5 285). Anthropogenic modification is measured by category In this model, short-term foraging of use intensity (4 5 regions of highest use; see Fig. 3.). Errors are 95% confidence returns under conditions of reduced intervals. mobility are sustained through the application of broadcast fire. This reduce variability in acquisition, and broad-spectrum economy. The finer- potentially illuminates a wide range increase group size and reduce mobil- grained habitats shape the scale at of diachronic subsistence shifts, ity, and thus favor storing over shar- which foragers experience resource particularly the common trend in the ing, can be an emergent property of patchiness. Travel time between pat- Late- and Holocene to- the way hunter-gatherers interact ches is reduced, lowering the overall ward broad-spectrum economies, with plant and animal communities cost of foraging and reducing the total especially in seasonal habitats where in woodlands and grasslands: They distance traveled. lightning fires is a common occur- use fire at a landscape scale to re- The final piece of the model brings rence. Indeed, the origins of broad- shape their environments. together the ecological shifts in small spectrum economies across much of Combining the results discussed, animal density that increase average Asia co-occurred with increased fire which show how certain subsistence returns and reduce variability in those frequency as shrublands of the late resources increase in density with returns, with the social relationships Pleistocene gave way to the warmer and patchier grasslands and wood- fire, we show that these anthropo- and norms that are created through 35,98 genic landscapes shape subsistence sharing. The cost of making claims to lands of the Holocene. The com- and mobility patterns in ways that ownership decreases when variability mon increase in the proportional presage the increasing sedentism of a among individuals is lower, with the representation of small animals in the archeological record suggests that in many environments occupied by modern , co-evolutionary dynamics may have caused shifts in ecological communities, with an increase in species that are resilient to human environmental impacts or even dependent on human disturb- ance.99 We suggest that broad- spectrum economies may be shaped through co-evolutionary processes involving habitat modification, which can result in increased populations of some species (often smaller or dis- turbance-dependent) and reduce vari- ability in acquisition, shifting sharing Figure 7. Proportion of overall camp production comprised of small game versus large norms toward increasing private own- game in anthropogenic landscapes compared to unmodified landscapes. Camp pro- ership and reducing mobility. duction is measured as total kcal produced from small game (<2kg/animal) versus large We propose that Aboriginal land- game (>2kg/animal) across all foragers on all camp days on which game was acquired between 2000 and 2010 (n 5 270). Anthropogenic modification is measured by category scapes in Australia, which have char- of use intensity (4 5 regions of highest use; see Fig. 3.). Errors are 95% confidence acterized much of the desert since the intervals. mid- to late-Holocene, are emergent ARTICLE Pyrodiversity and the Anthropocene 115 ecological systems. Codding83 sug- 4 Morrison KD. 1994. The intensification of pro- the Prearchaic Great Basin. Quaternary Int gests that anthropogenic fire mosaics duction: archaeological approaches. J Archaeol 352:200–211. Method Theory 1:111–159. 26 Codding BF, Jones TL. 2013. Environmental are an epiphenomenon of foraging 5 Zeder MA. 2012. The broad spectrum revolu- productivity predicts migration, demographic, economies based on small tion at 40: resource diversity, intensification, and linguistic patterns in prehistoric California. and , requirements for forag- and an alternative to optimal foraging explana- Proc Natl Acad Sci USA 110:14569–14573. tions. J Anthropol Archaeol 31:241–264. 27 Jones TL, Porcasi JF, Erlandson JM, et al. ers with reduced incentives to main- 6 Boserup E. 1965. The conditions of agricul- 2008. The protracted Holocene of tain especially high residential tural growth: the economics of agrarian change California’s flightless sea duck (Chendytes lawi) 92 under population pressure. Chicago: Aldine de and its implications for the Pleistocene overkill mobility. Bliege Bird and coworkers Gruyter. 78 hypothesis. Proc Natl Acad Sci USA 105:4105– and Smith argue that mid- to late- 7 Boserup E. 1981. Population and technologi- 4108. Holocene reorganization in tech- cal change: a study of long-term trends. Chi- 28 Winterhalder B, Kennett DJ. 2006. Behav- nology and mobility, as well as the cago: University of Chicago Press. ioral ecology and the transition from hunting 8 Cohen MN. 1977. The food crisis in prehis- and gathering agriculture. In: Kennett DJ, Win- emergence of the ethnographically tory: and the origins of agricul- terhalder B, editors. Behavioral Ecology and documented estate-based system of ture. New Haven: Press. the Transition to Agriculture. Berkeley: Univer- land ownership, closely corresponds 9 Morgan C. 2014. Is it intensification yet? Cur- sity of California Press. p 1–21. rent archaeological perspectives on the evolu- 29 Gremillion KJ, Barton L, Piperno DR. 2014. with establishment of anthropogenic tion of hunter-gatherer economies. J Archaeol Particularism and the retreat from theory in fire mosaics. Efficient exploitation Res 23:1–51. the of agricultural origins. Proc of small game and seed resources, 10 Binford LR. 1968. Post-Pleistocene adapta- Natl Acad Sci USA 111:6171–6177. tions. In: Binford SR, Binford LR, editors. New 30 Bowles S, Choi J-K. 2013. Coevolution of core characteristics of the social- perspectives In archeology. Chicago: Aldine de farming and private property during the early economics of desert broad-spectrum Gruyter. p 313–341. Holocene. Proc Natl Acad Sci USA 110:8830– institutions, may thus be bound to 11 Flannery KV. 1969. Origins and ecological 8835. creation of the small-scale succes- effects of early domestication in Iran and the 31 Eerkens J. 2010. Privatization of resources Near East. In: Ucko PJ, Dimbleby G, editors. and the evolution of prehistoric leadership sional mosaics created by anthropo- The domestication and exploitation of plants strategies. In Vaughn KJ, Eerkens J, Kantner J, genic burning. Similar processes and animals. Chicago: Aldine de Gruyter. p 73– editors. The evolution of leadership: transitions 100. in decision making from small-scale to middle- elsewhere, where intensification will 12 Bird DW, O’Connell JF. 2006. Behavioral range societies. Santa Fe: SAR. p 73–94. continue to amplify in feedbacks that ecology and archaeology. J Archaeol Res 14: 32 Bird DW, Codding BF, Bliege Bird R, et al. increase Malthusian limits of carrying 143–188. 2013. in a continent of small game: capacity, probably are the roots of 13 Codding BF, Bird DW. 2015. Behavioral archaeological implications of Martu camel ecology and the future of archaeological sci- hunting in Australia’s Western Desert. Quater- domestication, dense sendentism ence. J Archaeol Sci 56:9–20. nary Int 297:155–166. and, ultimately, the Anthropocene. 14 Charnov EL. 1976. Optimal foraging: attack 33 Bliege Bird R, Bird DW. 1997. Delayed reci- strategy of a Mantid. Am Nat 110:141–151. procity and tolerated theft: the behavioral ecol- 15 Charnov EL. 1974. Optimal foraging, the ogy of foodsharing strategies. Curr Anthropol ACKNOWLEDGMENTS marginal value theorem. Theor Popul Biol 9: 38(1): 49–78. 129–136. 34 Hawkes K, Bliege Bird R. 2002. Showing Foremost, we thank the Martu com- 16 Orians GH, Pearson NE. 1979. On the off, handicap signaling, and the evolution of munities that have hosted us and sup- theory of central place foraging. In: Horn DJ, men’s work. Evol Anthropol 11:58–67. ported our research for many years. Mitchell RT, editors. Analysis of ecological sys- 35 Janz L. n.d. Fragmented landscapes and tems. Columbus: Ohio State University Press. economies of abundance: the broad spectrum Our research in collaboration with p 155–177. revolution in arid East Asia. Curr Anthropol. In them concerning landscape burning 17 Fretwell SD, Lucas HL. 1969. On territorial press and their contemporary foraging behavior and other factors influencing habitat 36 Renfrew C. 2012. Towards a cognitive distribution in I. Theoretical develop- archaeology: material engagement and the practices is funded by NSF (BCS- 1459 ment. Acta Biotheor 19:16–36. early development of society. In: Hodder I, edi- 880, BCS-0850664, BCS-0314406), the 18 Lupo KD. 2007. Evolutionary foraging mod- tor. Archaeological theory today, 2nd ed. Cam- Woods Institute for the Environment, els in zooarchaeological analysis: recent appli- bridge: Polity Press. p 124–145. cations and future challenges. J Archaeol Res 37 Smith BD. 2007. Niche construction and the the Leakey Foundation, and the Chris- 15:143–189. behavioral context of plant and animal domes- tiansen Fund. The work discussed in 19 Kennett DJ, Winterhalder B, editors. 2006. tication. Evol Anthropol 16:188–199. this article has been greatly improved Behavioral ecology and the transition to agricul- 38 Smith BD. 2011. A cultural niche construc- through discussion with Jim O’Con- ture. Berkeley: University of California Press. tion theory of initial domestication. Biol Theory 20 Bettinger RL, Winterhalder B, McElreath R. 6:260–271. nell, Peter Veth, and Doug Kennett, 2006. A simple model of technological intensifi- 39 Smith BD. 2011. General patterns of niche along with insightful critique from three cation. J Archaeol Sci 33:538–545. construction and the management of “wild” anonymous reviewers. 21 Hawkes K, O’Connell J. 1992. On optimal plant and animal resources by small-scale pre- foraging models and subsistence transitions. industrial societies. Philos Trans R Soc Lond B Curr Anthropol 33:63–66. Biol Sci 366:836–848. REFERENCES 22 O’Connell JF, Allen J. 2012. The restaurant 40 Odling-Smee FJ, Laland KN, Feldman MW. at the end of the universe: modeling the coloni- 2003. Niche construction: the neglected process 1 Kohler TA. 2012. Complex systems and sation of Sahul. Aust Archaeol 2012:5–17. in evolution. New Haven: Princeton University archaeology. In: Hodder I, editor. Archaeologi- 23 Kennett DJ. 2005. The Island Chumash: Press. cal theory today. Cambridge: Polity Press. p 93– behavioral ecology of a maritime society. 41 Hawkes K. 1992. Sharing and collective 123. Berkeley: University of California Press. action. In: Smith EA, Winterhalder B, editors. 2 Klein RG, Steele TE. 2013. Archaeological 24 Thomas DH. 2014. The shellfishers of St. Evolutionary ecology and human behavior. shellfish size and later human evolution in Catherines Island: hardscrabble foragers or New Brunswick: Transaction. p 269–300. . Proc Natl Acad Sci USA 110:10910– farming beachcombers? J Isl Coast Archaeol 9: 42 Blurton Jones NG. Tolerated theft: sugges- 10915. 169–182. tions about the ecology and evolution of sharing, 3 Stiner MC. 2001. Thirty years on the “broad 25 Elston RG, Zeanah DW, Codding BF. 2014. hoarding, and scrounging. Soc Sci Inf 26:31–54. spectrum revolution” and paleolithic demogra- Living outside the box: an updated perspective 43 Winterhalder B. 1996. A marginal model of phy. Proc Natl Acad Sci USA 98:6993–6996. on diet breadth and sexual division of labor in tolerated theft. Ethol Sociobiol 17:37–53. 116 Bird et al. ARTICLE

44 Dyson-Hudson R, Smith EA. 1978. Human 66 Van Baalen M, Krivan V, van Rijn PC, et al. 86 Veth P. 2006. Social dynamism in the territoriality: an ecological reassessment. Am 2001. Alternative food, switching predators, archaeology of the Western Desert. In: David B, Anthropol 80:21–41. and the persistence of predator-prey systems. McNiven IJ, Barker B, editors. The social 45 Winterhalder B. 1986. Diet choice, risk, and Am Nat 157:512–524. archaeology of indigenous societies. Canberra: food sharing in a stochastic environment. 67 Wiessner P. 1982. Beyond willow smoke and Australian Aboriginal Studies Press. p 242–253. J Anthropol Archaeol 5:369–392. ’ tails: a comment on Binford’s analysis of 87 Binford LR. 1980. Willow smoke and dogs’ 46 Bliege Bird R, Bird DW, Smith EA, et al. hunter-gatherer settlement systems. Am Antiq tails: hunter-gatherer settlement systems and 2002. Risk and reciprocity in Meriam food 47:171–178. archaeological site formation. Am Antiq 45(1): sharing. Evol Hum Behav 23:297–321. 68 Lewis HT. 1972. The role of fire in the 4–20. 47 Gurven M. 2004. Reciprocal altruism and domestication of plants and animals in south- 88 Altman J. 2012. People on country as alter- food sharing decisions among Hiwi and Ache west Asia: a hypothesis. Man 1972:195–222. nate development. In: Altman J, Kerins S, edi- hunter–gatherers. Behav Ecol Sociobiol 56:366– 69 Scherjon F, Bakels CC, MacDonald K, et al. tors. People on country: vital landscapes and 380. 2015. Burning the land: an ethnographic study Indigenous futures. Sydney: Federal Press. p 1– 48 Kameda T, Takezawa M, Tindale RS, et al. of off-site fire use by current and historically 25. 2002. Social sharing and risk reduction: explor- documented foragers and implications for the 89 Bird DW, Bliege Bird R, Parker CH. 2005. ing a computational algorithm for the psychol- interpretation of past fire practices in the land- Aboriginal burning regimes and hunting strat- ogy of windfall gains. Evol Hum Behav 23: scape. Curr Anthropol 56:299–326. egies in Australia’s Western Desert. Hum Ecol 11–33. 70 Parker CH, Keefe ER, Herzog NM, et al. 33:443–464. 49 Bruno JF, Stachowicz JJ, Bertness MD. 2016. The pyrophilic hypothesis. Evol 90 Bliege Bird R, Tayor N, Codding BF, et al. 2003. Inclusion of facilitation into ecological Anthropol. 25(2):54–63. 2013. Niche construction and dreaming logic: theory. Trends Ecol Evol 18:119–125. 71 Stewart OC. Barriers to understanding the aboriginal patch mosaic burning and varanid 50 Kikvidze Z, Callaway RM. 2009. Ecological influence of use of fire by aborigines on vegeta- lizards (Varanus gouldii) in Australia. Proc R facilitation may drive major evolutionary tran- tion. Fire Ecol 10:4–9. Soc Lond B Biol Sci 280:20132297. sitions. BioScience 59:399–404. 72 Stewart OC. 2002. Forgotten fires: Native 91 Codding BF, Bird RB, Kauhanen PG, et al. 51 Paine RT. 1969. A note on trophic complex- Americans and the transient wilderness. Nor- 2014. Conservation or co-evolution? Intermedi- ity and community stability. Am Nat 1969:91– man: University of Oklahoma Press. ate levels of aboriginal burning and hunting 93. have positive effects on kangaroo populations 73 Anderson K. 2005. Tending the wild: Native in Western Australia. Hum Ecol 42:659–669. 52 Connell JH. 1978. Diversity in tropical rain American knowledge and the management of forests and coral reefs. Science 199:1302. California’s natural resources. Berkeley: Univer- 92 Bliege Bird R, Bird DW, Codding BF, et al. 53 Cottee-Jones HEW, Whittaker RJ. 2012. Per- sity of California Press. 2008. The “fire stick farming” hypothesis: Aus- tralian Aboriginal foraging strategies, biodiver- spective: the keystone species concept: a critical 74 Lightfoot KG, Parrish O. 2009. California sity, and anthropogenic fire mosaics. Proc Natl appraisal. Front Biogeogr 4: 117–127. Indians and their environment: an introduc- Acad Sci USA 105:14796–14801. 54 Smith DW, Peterson RO, Houston DB. 2003. tion. Berkeley: University of California Press. Yellowstone after wolves. BioScience 53:330– 93 Bliege Bird R, Scelza B, Bird DW, et al. 75 Gammage B. 2011. The biggest estate on 340. 2012. The hierarchy of virtue: mutualism, altru- earth: how Aborigines made Australia. Sydney: ism and signaling in Martu women’s coopera- 55 Springer AM. 1992. A review. Walleye pol- Allen and Unwin. lock in the North Pacific: how much difference tive hunting. Evol Hum Behav 33:64–78. 76 Gould RA. 1971. Uses and effects of fire do they really make? Oceanogr 1:80–96. 94 Bliege Bird R, Bird DW. 2008. Why women among the Western Desert Aborigines of Aus- hunt: risk and contemporary foraging in a 56 Laws RM. 1970. Elephants as agents of hab- tralia. Mankind 8:14–24. itat and landscape change in East Africa. Oikos Western Desert Aboriginal community. Curr 1970:1–15. 77 Jones R. 1969. Fire-stick farming. Aust Nat Anthropol 49:655–693. Hist 16:224–228. 57 Naiman RJ, Johnston CA, Kelley JC. 1988. 95 Bird DW, Bliege Bird R, Codding BF. 2009. Alteration of North American by bea- 78 Smith M. 2013. The archaeology of Austral- In pursuit of mobile prey: Martu hunting strat- ver. BioScience 38:753–762. ia’s deserts. Cambridge: Cambridge University egies and archaeofaunal interpretation. Am Press. 58 Roxburgh SH, Shea K, Wilson JB. 2004. The Antiq 74:3–29. intermediate disturbance hypothesis: patch 79 Bliege Bird R, Codding BF, Kauhanen PG, 96 Codding BF, Bliege Bird R, Bird DW. 2011. dynamics and mechanisms of species coexis- et al. 2012. Aboriginal hunting buffers climate- Provisioning offspring and others: risk–energy tence. Ecology 85:359–371. driven fire-size variability in Australia’s spinifex trade-offs and gender differences in hunter– grasslands. Proc Natl Acad Sci USA 109:10287– 59 Roff DA. 1974. Spatial heterogeneity and gatherer foraging strategies. Proc R Soc Lond 10292. the persistence of populations. Oecologia 15: B Biol Sci 278:2502–2509. 245–258. 80 Burrows ND, Burbidge AA, Fuller PJ, et al. 97 Bird DW, Bliege Bird R. 2010. Competing to 2006. Evidence of altered fire regimes in the 60 Holt RD. 1984. Spatial heterogeneity, indi- be leaderless: food sharing and magnanimity rect interactions, and the coexistence of prey Western Desert region of Australia. Conserv Sci among Martu Aborigines. In: Vaughn KJ, Eerk- species. Am Nat 1984:377–406. West Aust 5:14–26. ens J, Kantner J, editors. The evolution of lead- 61 Futuyma D. 1988. The evolution of ecologi- 81 Lourandos H, Ross A. 1994. The great ership: transitions in decision making from cal specialization. Annu Rev Ecol Syst 19:207– “intensification debate”: its history and place in small-scale to middle-range societies. Santa Fe: 233. Australian archaeology. Aust Archaeol 1994:54– SAR. p 21–49. 63. 62 Marvier M, Kareiva P, Neubert MG. 2004. 98 Bettinger R, Richerson P, Boyd R. 2009. , fragmentation, and dis- 82 Williams AN. 2013. A new population curve Constraints on the development of agriculture. turbance promote invasion by habitat general- for prehistoric Australia. Proc R Soc Lond B Curr Anthropol 50:627–631. ists in a multispecies metapopulation. Risk Biol Sci 280:20130486. 99 Schollmeyer KG, Driver JC. 2012. The past, Anal 24:869–878. 83 Codding BF. 2012. “Any kangaroo?” On the present and future of small terrestrial mam- 63 Wiggins N, McArthur C, Davies N. 2006. ecology, ethnography and archaeology of forag- mals in human diets. In: Wolverton S, Lyman Diet switching in a generalist mammalian foli- ing in Australia’s arid west. PhD dissertation, RL, editors. Conservation and applied vore: fundamental to maximising intake. Oeco- Stanford University. ooarchaeology. Tucson: University of Arizona logia 147:650–657. 84 Codding BF, Douglas W, Bliege Bird R. Press. p 179–207. 64 Gross T, Rudolf L, Levin SA, et al. 2009. Interpreting abundance indices: some zooarch- 100 Ulm S. 2013. ‘Complexity’ and the Austra- Generalized models reveal stabilizing factors in aeological implications of Martu foraging. lian continental narrative: Themes in the arche- food webs. Science 325:747–750. J Archaeol Sci 37:3200–3210. ology of Holocene Australia. Quat Int 285: 65 MacArthur RH, Pianka ER. 1966. On opti- 85 Edwards DA, O’Connell JF. 1995. Broad 182–192. mal use of a patchy environment. Am Nat 100: spectrum diets in arid Australia. Antiquity 69: 603–609. 769–783. VC 2016 Wiley Periodicals, Inc.

View publication stats