st widespread of beasts of burden, and pursuit time ¢ 1960:150). The can scavenge for ‘especially in the hnirelman 1994). on diet—as both 1 increase in for~ he use of traps o se gatherers to do © other food re devices such as large numbers of and bunting nets ction and mainte ‘hunting nets re 87; this study did Foraging models, labor investment uble use of forag- ogy open many relationships of spinel expecta- eget, the use of Ba vw say fom care and epro- conomic analysis FORAGING AND MOBILITY |4 “When tm a kid we're alays moving. Never stay around one place for toa 'we got to move, olurwite we find no food. Bven then sometines Shores no food for while, 20 people in amps go hungry. Whereret food, wel, we got to move (0 that pace oscar man (Nelson 1966273) {We donot ke siting one place all dhe ie ke whise men. Ks sn (Homann 1949102 he “There is hardly a more romantic image in anthropology than that of a small band of hunter-gatherers setting off through the dunes and scrub, their few belongings on theic backs. Mobility, in fct, has long been considered one of the distinguishing characteristics of hunter-gatherers. At the “Man the Hunter” conference, Lee and DeVore assumned that all hunter-gatherers Smove around a lot” and that this was a strong determinant of the foraging sway of life (196811), In one sense they were correct; mobility docs exert strong influence over other elements of hunter-gatherer’ ives, Anthropolo- tists have recognized this for some te, Marcel Mauss, for example elated the Eskimos’ seasonal mobility to their moral and religious life (Mauss and Beuchat 1906) Sablins saw mobility as conditioning cultural arsitudes toward material goods (1972). ‘Yet hunter-gatherers move in different ways, Some do indeed “move around a lot” but others move hardly at all. (And huncer-gatherers are not the only ones who move; many horticulturaists, and virewlly all pastoralists, “As in subsistence, there is considerable variability in how, are also mobile)" 1), Many students of and how much, hunter-gatherers move (see table 4-1 lnchropology think of hunter-gatherer movement in terms of a systematic seasonal rund, wherein hunter-gatherers move between different locations 35 ur“Table 4. Hanter-Gatheror Mobility Primary ‘Average Total Total Logisicat Prisma lomo Residential distance dance aren mobility biomass Group (gym) swoves/yr Com) am) Ghee) days) Reftremcs crow Gog/=) finland mit tow @ me 35000 sth 977 Naha Ona Gene) 178 © (Gunde 19 Sta (Soot ne wa Netingait Low 6S boo Bale 1970 ‘Nunamioe low SS sHo-a0s00 AAmulea 197 Squamish 2s Binrd i978 1. Thngie ws a0 Scale 2978 ore Gul 17 Cia ave > rary Michell an Doral ‘sl ros ste Ssh ‘lla Coola sea os Sate sors sanity) (Wasa) ‘owikeno as «9 Sea 108 Seis Sas ovat (Wi Sunich) 192 Micasini ree 187 we sos Rogers 1963.16 2 Nooksack 204 8. Thing =r 3 1988 Sth on Micmse 98 Berne ier 187 so ges 1907, 196 Ojibwa Gand Victor Cree 198 0 Rogers 190% Sano a Sagi Bs x0 Ropes 1967, 19 “Tamanians, wr (Osb) NW le sar 4000 Shale 978 wise 9s “Taehian 25 os seas0 Scale 197 Michell Kidd sn Donald 1988 (6 Paite) Hes 203 os Salt crow oe (Book) 1S ka wot 2 oon 5 190 Sc 1978 ‘param SW 208 Quilese 200 ms Sealk 1978 Ne Pex 18a Blacktor ‘Cheyenne (Sia) ms wos Eves 1955 iow set Quint 205 0 Sch 2978 Slew York 196 (Ropes) m3 oeah as mr Stalk 1998 sa6 ‘ice apd Dowld Maida m8 1988 wascninceee 90 40 Roges 1957, 19698 Tm, lest ow : ye Come og; Lag ‘Opreray 208 tote Karke mr Monags a se % a0 “Taner 1944 La ‘Temanion Big 208 9s River Chinook 204 us Seba ros bab 1939 ae) iow 194 2 8 86 i o8_Whtaabe 96 2973 Kiows lama ‘ nots Mash 37 Gather 90; Spier ade Gi {930s Hane gto an (oobi) Tans 96 ‘ a an yal. wo wt Nisa Uppersiagie 19 ““Table (1. Hunter-Gatheror Mobic (times) Primary ‘Toul ‘biomass Residenial distance sree Reference Group (g/m) co) Reference Hoeach 977 ‘Naachahmalts Gunde 1984S (Soot) no ss aches 195 re ‘Machel ané Donald Bali 970 88 Aden 77 Suan 20s Michel and Donald Bineed 1978 it Sele 1078 ner Galt 7 ou Sole 108 Michell and De Sash ‘98 sss Ssh Sehak978 Te Sunich) 193 ” ‘Scie ton Michell nd Dom 1988 Schalk 1978 Seas Sait i Sanch) 192 ‘Seta ror Michell Ropes 19851 og Doma 1988 wn oats 24 a6 Schalk 1978 Sell 1978 seme 9 1000-5200 ‘Wil and Weis Ropes 1907, 195s: Dens 1908; Tete 1910: Speck Rogers 1967. Ray 1932 Ropes 190% a6 Jone 974 Stal 1978 B Sohal 1978 Schalk tod Don Kelly 1952 Schalk 197s ea fp sate Nikekor 1967 Seale 1978 wo #6 Jones 974 ‘Schalk 2978 ae Haines 1935 Ganow 1954 nett 985 Shak ws Cook Sohal 978 976 as ‘Stalk usr Cook Schl 3978 1976 ‘ischll and 8 assis Dion 1905; Bele sos 3 Roget 295 ee Cone #804 ee ca Jones 974 198 a1 f Schale1978 Tamer 944 Sgt nate Jones 974 Schae ons u m6 Letind 98 Kelly 1939 1964 arabe 198 20 Kroeber 1939 Gace 895 Ge a 00 “Tana 980 93: Bart a) o rey itcheook and bere Mitchel and 19; Hiscock 1982 rst 4 ss rocber 1935 Sth 1978 6 ‘Tle 2981972 ‘ 6 260 Gould 1968, 150 Sota sors Dae 1986“Tale x, Hhmeer-Gatherer Mobily (csoned) a meng foal Bo tage SS cu ec ee ee ee a ee ee ton i 7 a is a SSece Seti sion ss 2 “ wenn Sos panes ac Si saa sige oe fee ae (ches win lager oe : Sree CE geet ram to Henkin tise iMueteanr) 18 sie Hate Smet a eee ee Eee eure ro Hien siethde ews fimo a8 6 Hn one Alyawaea 5 3500 7 O'Connell, Late, and ane Sensor ‘Dorobo (Okeek) 6 “Huntingsford 1929 senennne 0 oe a ee oars came sper vision st yee x pte moun tigre Birhor a4 s 105 90.3 ue, 5-6 Williams 1074, floors, where seeds a at " onan gan to ripen there. se tyme aera cou wen a sama a ic com foe es oo in the early fill, mae . pion mus an eT oe camps ain pene ee gee gee aarp ete Deon ae oa Bf ab a ke fen ‘one rounds bt © = ee Sere wlnionip a OU a kt he ae Aka By 8 7 co 350 ‘Bahuchet 1988; 1979 We fy ‘Glyde River 336 5 a5 42, Peterson 1973 ships between Ss eee Som feeging mai = "s ‘Muti “ws 3 7 50-780 ‘Bicchiei 19695 ‘result in longer oF: "a ka meena eines ‘Semang ory 6 m3 pen ars Schebesia 1929 between foraging: whether huntFORAGING AND MOBILITY 115 “table 4. Hanter-Gatherer Mobily (onic) ‘Lopincal Primary ‘average Total Toul Logbtcal moby Pay Resdeotinl disance distance aren mobility “daye) Reference Group wore Towelye em) Goa) Gur) (Gam) __—_—Raference Achmann 999 Andemance Spencer an Gen (Ong inland) 08 . wo Rae Drown ‘927 92s; Cooper 1950 Peaenon sd Long Andomase 1986 cc) ss 25 ‘Cooper 1980 Gas 1972 Pout as ws Hin 1999 (©°Cosnel, Hawes, Shas 6 3385 Dison 1907 sn Bho Joes wens Valley 1988; Woodburn Pai s 1.984 steward 935 ‘96, 1972 Waibo 90, aa Downs 1965 McGee 158 Ry ee eee srw Wi Ade vw age tse ds fon ego res. A eel met inh ciel muy GN 0 itchodk and be i iw i aaron igh secre ind ton nl eS ou deed wy an ea a, he po sce ob een mob ong wt fF : Hache 9870 inchcock nd her 198 Hizchoock snd bert vote Conn, La, and Burne 1985 Hansson 1959 Siete 1973, sof Pookajoen 1985 1 Wms 1974 Cane 19875 “Toakinso 1978 Haber 1930 Stearman 980 rer Hamendor 968 ‘Chas and Sato wh rss 1983 ‘anoverbergh 1925 er 197: 1987 ech 1982 Bache 188; 1977 Perenon 97) Sigman ont Seeman 1911 Blecher 1560 “Tan 176; Hare 176: Tubal 1972 Sehebess 1939 ao a ee compen goer men oben aleve come-incl £1 SA i oan te ma he gow ay hae stopped fer ooo ce dy 22 Se ae einai tamk cr op. Taco ae im br pl, ot et EP a Cy aa mle wee ot ee ey ingen, NS Pn aes May ow ra rap ame: Mon ca debe a enti oe rotnd ae oe by ene ar ca a ped mone een ae whe] ough PoP ar tpi nd rapa ya No caval ble oe ere pane a kt aed pup For exp Ns dan coe rs cd Col fA bn i et a te gf Anaae Wee Doe, o en tf kano Nein resources come and go with the seasons. The Great Basin Shoshone, for “cample, spent the winter in vilages in the pion and juniper forests ofthe mountains (Bute 4-1). As spring came, they moved down to the valley oom, where seeds had begun to ripen, later moving upslope a8 seeds be- gun t0 pen there, In the late sunumer, they might move fo ver where cone were ninning, or to a marsh where waterfowl were passing through toe their way south or where bulrush seeds could be harvested. Later still, fn the ey fl, they might move back into the mountains to collect pion nuts and hunt deer or bighor sheep, and establish their winter ‘camps again. ‘ ar our purpose in ths chapter is not to document variability i seasonal rounds but ro describe variation in the organization of movements and the ‘elatonships between environment, individual foraging, and camp mover ‘rent. We frst discuss concepts and ethnographic data that point to relation= thipe beeween mobility and the environment. We then present 2 simple forging mode that considers how different foraging environments could result in longer or shorter individual forays, and more or les frequent group ‘ovement Following this, we discus sedentsm and, Finally, the relationship between foraging and encuturation, and how foraging studies shed light on ‘wether hunter-gatherers manage oF Conserve FeSOUICES.116 THE FORAGING SPECTRUM (Gross section along ceunty ne Figure 4. Settlement patter ofthe Tlidipihunupi, the Reese River Valley Sho- shone, From Steward 1938 (figure 8) as redrawn by Dennis O'Brien, in Thomas 1981, Courtesy ofthe American Museum of Natural History. MOBILITY AND THE ENVIRONMENT ‘Varibiley in hunter-gatherer mobility has not gone unrecognized. One carly scheme that received wide atention divided hunter-gatherer into four categories: fee wandering groups, which have no territorial boundaries and are characteristic of colonizing populations; retried.sundering groups, which live under higher population densities and are constained by tersitorial bounds, entabased wandering group, which seasonally ret 0-2 spe- ee a “Bopicl res ws Topi! ubopia deers a Tenpentedoets 154 Temperate feast eral Fro Murdock later nomadic, semisedentary, rwsed these gross cat environment, and usher mobility, especially Binford showed ddock’s settlement types and extreme Arctic tend! mobility is seasonal the winter becomes ma consrains the Binford described the tens with ewo set group from one caitip & viduals or small task regions resource loos to key locations not m wood are available) and_ (Gute 4-3) n gene mobility and invests make few residential m Bt not al foragers (aFORAGING AND MOBILITY 117 “Table 42: Biotc Zones and Murdoch's Setdement Patterns ‘Setement pattern Folly Semic Serie ET'range sora nomadic sedentary Sedentary Mean Zone “Topic forts PT ee ee ‘Topcal/ ebuopicl deers oa) Gs) 1G) “rupert desens Noa) 21s) 393) S080) 38t “Temper fress 1O) feo) G2) $00) 238 Borel es May Nae) Gs) 70538 ee surg eas) 3060 GN) 19 ‘Sane Bind tbe 2 Ste nd ate > een eng The met te mean tf rae Gn aera aen pares cng (aly nese = « Seensry = 4) ii vile; and semiermannt sede groups, ich ocapy ila Yat- Sound but move i every few years (Bearley & al. 1956), ‘Murdock later modified these categories into fully nomadic, semi- owaig,semisedentary, and fully sedentary, respectively (1967). Biaford eed shove grove categorizations to demonsrate tha mobility is telat the ae gonment, and ushered ina renewed round of interest in hunter-getheret ropa, expecially among archacologist (198). Using ET a + measure, vrtord showed a sytemaic relationship between environments and Mur- J speck’ seulement types (able 4-2). Hunter-gatheress inthe tropical fess aotlexneme Arce tend to be very mobile. In temperate forests and desers,\ avi seasonally constrained, especialy a the use of stored food during {| - auth the winter becomes more important, oF a5 the distibution of water sources comtrains the movements of desert forages. al dosribed the variability he sw in hunter-gatherer setement sy- tema with eo seement types, forages and cles. These types rested on the concepts offesidntal mobili) (movement of dhe enite band or Foal group from one camp to-anoth |, and(eistcal mobi movements of indi- asta oe wall takespecific groups out from and back to 3 residential van Fagenore consumes co food esouess and hs map ome 3 region' resource locations (Ggure 4-2), while Foleetors move residentially recognized. One to key locations not necewarly defined by food (where water OF fire- atherers into four ae re avalbie) and use long logistical forays to bring resources wo camp 3} boundaries and (Gate 3) Tn gen, infor suggested tat foragers have igh resins Fo°% ing Broups, which mobility and invest lie effor in logistical movements, while collectors bed by territorial ey weidential moves and frequent, often lengthy logistical forays. serum to.aspe- "But not all forages (as Binford defined them)* are highly mobile, nor are _ SSFigure 4-2. Charcteristics of a foraging subsistence-setlement system. From Bin= ford 1980, drawn by Dana Anderson, Reproduced by permission of the Society for American Archaeology from Ameria Antiuity 45(2), 1980.Figure 4-3. Characteristics ofa collector subistence-stiement sytem, From Bin~ eee ford 1980, drawn by Dana Anderson. Reproduced by permision ofthe Society for sion ofthe Society for ‘American Archaeology fom American Antiquity 451), 198.=~ { 120) THE FORAGING SPECTRUM 2 all collectors nearly sedentary. Binford’s typology focuses not onthe ‘made each yeas; (2) quency of movement but onthe opanizatin of eonp mavement iv fs each year, 4) the ‘Bing svt The Anbura of Now Asta, for example, move only 2 length of logistical fen ces a yar, but make fFequent, relatively short forays to hunt and fish, T analyzed these ‘and to collect shellfish, roots, and water (government rations provide about dance and distrib 0 percent of their calories). The Semang of Malaysia make frequent residen- ‘mass for their mea al moves, but they too usually make only daily forays from camp. Both, Here we introduce however, are foragers in Binford’s sense because they move consumers 10 matter in an en recnurces The difference in the fequency of movement is rated co the pars of plants (ts fod density of their respective environments, bus the relations between the ‘of high primary bio jncvidval forager and group movement remain the same. and the capture of inford did not intend that his two sewlement-system types be used to primary production pigeonhole ethnographic or archaeological cases (though they have Bees tops of tes or at es for ouch punpose) Instead, he sa these eyes sa simple way of de- ants inves less ehbing continuum of tiem forms and posibiie| Forages and col- reproductive t _Joleton sve the eeme end of conaum dt Bird sw gener biomas et paralleling other scales of seasonal differentiation and patchiness of food re~ droughts and omces, Where resources are homogeneously dstbuted and where food is within limits, ‘ rraiable mote or less year ound, afrager patter is more likely: where the plan: food. 1s Cn eneite conditions held tra, 3 collector pattem can be expected. AS Since arma in Stee in chaper 6, this simple pater isnot unexpected, for where re feed in tee tp) sources are patchily distributed, maximum foraging efficiency is obtained by yet “with effective sercmmaghina ceneal place and sending out foraging paies. Where r= sure ofthe p ‘“ Seer aie more homogeneously distributed, maximum foraging efficiency Primary <5” comes fom group dispersion io reroute locales general resources bs- primary qe Some more aggregated in space and more constrained in their seasonal ava and humid ot Ghiey ftom the equator to the Arctic, except forthe high Arctic, where rmore than 400 there are some groups who do not have acces to migratory Ssh or lage ffom 8 ro 12504 migratory cabou herds (ee below). Therefor, che pater Binford ob those with ET" renred in Murdock categorcs and ET shows the expected parallel between within the ET the fonger-collectorcontinuura and resoarce dstibution* meters a year of “ZV Binford’sforager-collector continuum makes the case that mobility is e- luted to the environment. Ethnographic data also demonstrate this point. ETHNOGRAPHIC DATA ON MOBILITY ‘Rather than rely on a typological scheme, I have used five separate variables to measure_dimensions of mobility: (1) the number of residential moves where X= net.= = FORAGING AND MOBILITY 121 made each year; (2) the average distance moved; (3) the total distance moved. teach year; (4) the total area used over the course ofa year; and (3) the average eample, move only 2 length of a logistical foray (Kelly 1983; table 4-1). to hunt and fish, analyzed these five dimensions of mobility in relation to the gross abun provide about dance and distribution of food, using effective temperature and primary bio equent esiden- nase for their measurement. We defined effective temperature in chapter 3, from camp. Both, Here we introduce primary biomass as the rotal amount of standing plant ‘move consumers to ‘matter in an environment. For the most part, humans eat the reproductive ent is related to the pares of plants (nuts and sees) or their sored carbohydrates (bers) In areas clations between the of high primary biomas, plants invest more energy in structural maintenance and the capture of sunlight than in reproductive parts oF storage, resulting in types be used 10 primary production that is largely inedible or dificult to reach (.e~ atthe they have been top of tees or a the ends of branches). In aress of low primary biomass simple way of de~ plans inves Tess energy in srtural maintenance and growth and more in uges and col- Feproductve dss (seeds). In addition, many plants of dry; low-primary- ord saw as generally biomass environments have large subsurface eubers (an adaptation to hiness of food re- Groughts and range fires). Therefore, primary biomass is, in general and and where food is svithin limits, inversely correlated with the effective abundance of eible likely; where the plane food. It is also inversely related to faunal bundance and distibution Since animals in high-primary-biomass settings tend to be small (o they can bbe expected. As we feed in tree tops) or, if large, few in number and widely spaced. Coupled ected, for where re" rained by swith effective temperature, primary biomass provides a rough relative mea sure ofthe potential return from foraging in a given environment ‘Primary biomass is calculated from two regression equations welizing the primary production and biomass of major biomes that are grouped into arid and humid categories (Kelly 1983). Humid environments are those with hhigh Arctic, where| more than 400 millimeters of precipitation per year and either ET values gratory fish or lange from § 012.5 or from 19.5 t0 26 degrees centigrade, Arid environments are pattern Binford ob-| those with ET values of 12.5 to 19.5 degrees centigrade, as well t those cred parallel between} ‘within the ET ranges of humid eavironments which have les than 400 mill~ a qncters a year of precipitation (tundra and Arctic environments excluded) ‘that mobility is re~ onstrate this point ‘Arid Environments logy Primary Biomass (g/m?) = 2.66 + .0009X Humid Environments: : Togs Pinay Biomas (g/m) = 42 + o0013X | five separate variables Seto he = eso pinay ofan pe122 THE FORAGING SPECTRUM NUMBER OF RESIDENTIAL MOVES PER YEAR 5 — on o 8 © © & & PRIUARY BIOMASS (clequare rats) Figure 4-4, Number of residential moves per year plotted against primary biomast ava proxy measure of resource density for tropical and subtropical forages. This fgute shows both tropical forest and desere cases (Fill Pandaram, Neadagjaa, Hades, #*Kade, Kua, G/wi, Aranda, and Ju/hoans) a4 Tintin aos & as my = ih ay Nacho a © Sy long as a group does not depend heavily on aquatic resourees (able 4-3)- \Dependence on Aquatic resource)is almost always associated with low resi- "denial mabilsy(Yesner 1980); in fc, che Few avaible dat suxges thst ‘group is heavily dependent on aquatic resources, the number of residential moves is invenely correlated with primary biomass (Kelly 1983292). In chap- ter 3, we suggested that aquatic resources may be used in lew of terrestrial game. In temperate, and especialy in Gold setGing) where primary produc- tion i lower than inthe topics, ame become more important to diet and 2 ore imporant determinant of mobil But inthe same environment, large game become more dispersed and less abundant as primary biomass increases and/or temperature decreases. Dependence on aquatic resources should increase along a gradient of increasing primary biomass asthe cost of hunting increases. Note that in table 4-3, dependence on aquatic resources does increase with increasing primary biomass for those groups les than so percent dependent on aquatic resources. For groups heavily dependent on aquatic resources, those living in lower-primary-biomass settings (et. Kla~ math) are more mobile than those in high-primary-biomass settings. All126 THE FORAGING SPECTRUM groups with heavy dependence on aquatic resources and low residential mo- bility live in high-primary-biomass settings. (Most of the cases, however, come fiom a single culture area, the Northwest Coast of North America, and thus are not conclusive.) While hunter-gatherers frequently do choose campsites on the basis of foraging conditions, they must also take into account such things as fire- ‘wood, tree boughs for bedding, shelter, water, and how dirty the present camp has become. The location of other groups of hunter-gatherers (or non hunter-gatherers) can also condition movements, cither by attracting oF te- pelling foragers, depending on the nature of the relationship. 1 Cesers) presenta spoil probe in hit regard. Given inmand! alest “aily need for water, the location of water is often more critical than foraging 7 * considerations as a determinant of residential movement in deserts. Meehan Gescribes hovr tension and anxiety permeated Anbarra camps in northern ‘Australia as water became more scarce and had to be carried from longer and longer distances; movement of camp was often predicated upon the se- ‘verity of these arguments (1982). Taylor labels groups whose movements are restricted by water sources at tethered foragers (1064, 1973). In desert settings, the relationship between residential movement and primary biomass i con ditioned by the availabilty of water. Since a number of local fictors can determine the availabilty of water (springs, subsurface geology, topography, etc), we could expect a fiir amount of varibiley in subtropical mobility with no correlation to primary biomass (figure 4-4). ‘The Dobe Ju/‘hoansi and G/wi demonstrate the effect of water on desert mobility. One of the key diflerences between these to groups is that the Ju/‘hoansi have access to extensive pans that hold water during the dry sea- ‘on. The geomorphology of the G/vis habitat prevents the formation of ‘water pans. The Ju/*hoansi are less mobile during the dry season, preferring to make long individual foraging tips from campsites near water pans. Data in table 4-4 show the nature of Ju/“hoansi movements during the late wet and early dry seasons. During the wet season, however, families become more mobile, making foraging treks from Dobe since surfice water is more available (Yellen 1976; villages near Dobe also encourage the Ju/"hoansi to remain at the water hole). Conversely, the G/wi are mobile during the dry season and obtain water from the rumens of game animals and tame melons (Silberbaer 1972, 1981a,b; Tanaka 1980). To insure a constant supply of these items, the G/wi must move camp more frequently than the Ju/hoansi during the dry season, ‘Where water sources are localized in deserts, we could expect foraging ba “Table 44. Tethered they alway cone they are than expected. resources Wjow residential mo- he cates, however, of North America, tes on the basis of soch things as fie wr dirty the present c-gathercrs (or non by attracting oF re- ship gen humans’ almost ccal than foraging rin deserts. Mechan , camps in northem carried from longer dicated upon the se- dhose movements are 3). In desert settings, mary biomass is com cof local factors can of water on desert groups is that the during the dry sea the formation of season, preferring ‘water pans. Data during the late wet r, families become water is more the Ju/‘hoansi to during the dry and tsama melons 2 constant supply of than the Ja/hoanst expect foraging, Xe FORAGING AND MOBILITY 327 rane 4-4 Tethered Forging Trips of the Dobe Ja/oent Mean duration ofeamp ays at Dobe Tap Dove camps occupation (ay) ‘pon return 5 ” oe ‘See Yale oa 5. Seen pay wy eh Toe mes etn (eee sepa aan eae 7 HATER ficiency wo be scced in ivr of retaining casero a wale sur, CO soc rer saligs Wester Deset, where several thousand evare RIOTS ve onan only 2 anal of water Sources (6 Cane 1990137) Here, teople such as the Nendaina tend to stay at 2 wath Role un ie dries up Pear gon another. They als tend to use water source in OS vag order of labilcy (Gould 1991; pensnal commaniion 1989), s0 that they aways Know that the ext water source is ove dependable than the sae hoy ave curren singin ase the dry season should Bs POO and drier aoe nspecte, Water-esheed forages may mks fall we of exploitable tha ein frapng Stance ofthe water sue, Jeng OO YD Faery reach zero, Is my impression that water-tthere Sree fore ve ter from di camps for resoures (eluding Pn food) than do 2 ed ese forge The Jo/ Hox for example fame C° vor more Kilometers from camp, while the G/w7 ‘tend to forage no more than 2 Fleer fom a amps. Diet may ao e acted Vy ‘water-tethered seine, Arong the Boro ofthe central Baja California des “ag water ed op, forcing the concentration of population 2 2 POTS vraag on mow of the game ranges was reduced > the VANS point am ple dss would be eset uted ou before ess of ‘Re degeal resources were evident” (Pschmane 103990) Ts 0" 0 tame became scare, men rurned tir effort fo collecting plant food slong- aa pe yomen, Bororo die 2s well a mobiry was affected y the need t9 reduce the energetic cost of acquiring Water ree gmmporary huner-gaierets flow 2 patter of tere forag- sng Along with water, however living hunter gatherers ae frequently ceth- iN eral pls (heir wn o those of sedentary neghbon) sourw8 THE FORAGING SPECTRUM | g 8 AVERAGE DISTANCE/RESIDENTAL MOVE (im) Figure 4-7. The average dance moved per residential move plotted against effec sive temperature ata proxy measure of the distance between resources, of wage labor, cigarettes and alcohol, government agencies and welfare ra- sions, or mision posts. In ths respect, many contemporary hunter-gatherers are very similar to tropical horticuleuralsts, such as those in the Amazon, ‘who petioically make treks into the forest that lst from several days to several weeks in order to collect forest products and food, especially meat {Gee papers in Hames and Vickers 1983). Average distance per residential move ‘We have discussed how residential mobility enables hunter-gatherers to posi~ tion themselves relative to food and other resources. We could deduce that the average distance moved berween residential locations should be closely related to the distribution of food resources. In general, resources tend 19 become more spatially segregated along a gradient of decreasing tempera ‘ure cherefre there should be an increase in the average distance per resi- dental move with decreasing ET. The relationship between the two variable is depicted in figure 4-7. “This figure, however, shows a second pattem in addition to the expected ‘one. This consists of groups that da not move very far residentially om aver- age although they live in environments with low to medium ET. These exceptions fill into three groups: the extreme Arctic (Netsiingmiut and Baffinland Inuit), Perce, Blackfoot), kab, Souther Unlike many land Inuit have animals, fish ofthe ocean and they exploit are sources in the gatherers in the resources around: new Foraging example, bunt group moves to72), thats, ai very far in 25plotted agains effec- cies and welfare ra- hhunter-gatheress se in the Amazon, om several days to especially meat er-gatherers to posi~ We could deduce that ons should be closely al, resources tend to cistance per resi~ berween the cwo tion to the expected residentially on aver medium ET. These (Netslingmiut and FORAGING AND MOBILITY 129 Baffinland Inuit), equestrian bison hunters (Crow, Cheyenne, Kiowa, Nez Perce, Blackfoot), and coastal fishers (Nuuchahnulth, Klamath, Aime, Ma~ Jah, Southern Kwakiut) ‘Unlike many other Arctic groups, neither the Netsilingmiut nor Baffin~ land Inuit have access to large herds of caribou.’ Instead, they hunt solitary animals, ish small streams, of, during the winter, live on the frozen surfice ‘ofthe ocean and hunt seals through breathing holes in the ice. The resources they exploit are dispersed, and their istrbution resembles mote that of re~ Sources in the tropical forest than that of other Arctic regions. Like hunter gatherers in the tropics, those of the extreme Arctic exploit dispersed Fesources around their camp, then move so that they are at the center of a new foraging area. While hunting seals in the winter, the Copper Inuit, for ‘cxample, hunt an area & Kilometers in radius. Afer the aea is depleted, the group moves about 16 kilometers to a new hunting area (Danas 19652. 1572), that i, a move of pwice the foraging radius. Its not surprising, then, that the Netiingmiut and Bafinland Inuit have shore residential moves like tropical forages. Plains bison hunters used horses to move camp frequently, but not always _very farina single move. Larocque's 1805 journal (Ewers 1955:147), for ex: imple, shows that the Crow moved 47 times in only 76 days, moving 2 miedian distance of 15 kilometers, but with a range of 5 to 38 kilometers per ‘move. Dunbar traveled about 644 kilometers with the Pawnee during a win- ter hunt of 1834-35 that ated 156 days. During this time the Pawnee made 433 camps for an average of 20 kilometers per move (Roper 1991). By taking esources that were mobile and dispersed, Plans hunters were in one respect similar to Arctic and tropical forest foragers: they must have experienced rapid drops in foraging eturns soon after occupying 2 camp. Consequently Phins groups moved short distances frequently Factors other chan bison dis- tribution, such as the availabilty of firewood, plants, forage for the horses, or the presence of enemies undoubtedly affected camp movement, but the hhorse probably lowered the cost of moving suficientiy so that Plains groups could afford to move more fiequently and for shorter distances than we might otherwise expect. ‘Many northem fishing societies move short distances because they are territorially constrained and cannot move very far without trespassing Living in small territories with high population densities, they move from winter villages, often located near the coast at the mouths of rivers, to nearby Spring/ssmmer fishing, shelfshing, or plant-gathering camps, then back t0 the winter village. Figure 4-7 suggests that if these coastal societies were not330 THE FORAGING SPECTRUM territorally constrained or relying upon aquatic resources, they would move very long distances, perhaps especially in winter. “These thre sets of exceptions show that whereas ET's not a perfect mea sure of resource distsbution, it nonetheless points to foraging considerations 4s significant factors in structuring hunter-gatherer residential mobility. Logistical mobility and territorial coverage “To this point, we have only discussed movements of the residential camp ‘Yer, if we were to rack an individual's movements, we would find that the :najority ofthe ime he oF she spent moving was not spent in moving camp, bbut in logistical forays to hunt or to gather plant food. Though foragers also make trips © procure raw materials or firewood, to visit, or to gather information, these tasks are often embecied in food-getting forays (Binford 1982; Lee 19792214; O'Connell and Hawkes 1981). We willassume that food collection i the primary purpose of logistical forays from camp. Given the general trophie pyramid of any eavironment, carnivores must normally use larger territories than herbivores. Holding other factors con- stant for the time being, we would expect to see an increase in the area of land exploited by hunter-gatheres as dependence on hunting increase. Fig- ture 4-8 shows that there is a strong relationship between the relative depen- dence on hunting and the total area exploited (IN = 36, df= 34. 7 = 66, ‘p< .01). The equation describing this relationship, log y = .o24x + 2.06, is actually curvilinear as dependence on hunting increases, the size of the territory increases very rapidly: The same factors which act on the average distance per residential move affect the total area exploited since hunting ‘becomes more important toward the poles and since fanna need large terito= ries to support themselves in cold litudes. Additionally although our data are not adequate for analysis, group size could also factor into the equation, «with larger groups needing larger ranges. However, as we shall se in chapter 6, both empirical data and theoretical argument suggest that there is an upper limit wo foraging group size of not much more than twenty-five people. And, where local group size is large, 38 itis, for example, on the Northwest Coast, territories are often small and subsistence focuses om aquatic resources. As a group grows in size it may fision or if ths isnot possible, it may aler its settlement and subsistence strategy: Thus, group size may be ales significant factor than the food source in directly determining range size "Hunter-gatherers heavily dependent on fauna may use a large range ann ally, but they do not necessarily cover their range as thoroughly through SIZE OF TERRITORY ((og 59, kn) Figure 4-8. The: {ence on untilthey would move Tis not a perfect mea- considerations the residential camp. would find that the tin moving camp, ‘Though foragers +o visit, or to gather forays Binford menting increases. Fig- the relative depen- 36, f= 34,7 = 66, hs act on the average cploited since hunting need large tersito- y alchough our data ctor into the equation, -we shall seein chapter ese that there is an upper ive people. And, sche Northwest Coast, aquatic resources. Asa ible, ie may alter its -may be a less significant range size Juse a large range anni 4 thoroughly through FORAGING AND MOBILITY 13% 1000: 100, SIZE OF TERRITORY (10g 69. km) 10; re ee er CONTRIBUTION OF HUNTING TO DIET (5) Figure 4-8. The sizeof forages’ annual ranges ploted agaist he percent depen “dence on hunting (om chapter 3, table 3-1). As the dependence on hunting increases, so must the size ofthe exploited territory. residential mobility as groups heavily dependent on plant food. As we will fee below, the distance at which a resource can be gainfully procured is elated to the resource’s return rate (and its transportbility). High-return- tate resources can be procured at a longer distance from camp than low- Teeurn-rate resources. In general, large game provide high return rates once they are sighted. In economies dominated by gathering, foragers cannot father resources at long foraging distances, and therefore should move Shorter distances than in societies dominated by hunting, Hunters should se Jong logstial forays and cover less of their large temtores through resi dential mobility. Gatherers can be expected to cover their territory more thoroughly chrough residential mobility. A rough coverage index can be eal- culated ffom table 4-1 by dividing the total distance moved residentially by the total area exploited each year. These indices are grouped according to the dominant subsistence category (hunting, gathering, fishing; fom table $11) with the following results: for groups dependent on gathering, mean = su (N= 10; mean = .6s if Hadza and Semang are excluded), for groups “dependent on hunting, mean = .05 (N = 6). These figures support the Srgument that gathering groups cover a greater percentage of their range132 THE FORAGING SPECTRUM through residential mobility than do hunting groups.* Conversely, hunters probably cover more of their range and spend more time moving individu- ally about the landscape. Hunter-gatherers who rely heavily on aquatic re- sources also do not cover their ranges through residential mobility (mean 18, N= 6) but make long logistical trips, although, since many are made in bouts, these are not directly comparable to terrestrial forays. Residential and logistical mobility are not mutually exclusive. Investing ‘ensrgy in residential mobility does not exclude the use of extensive logistical ‘mobility as well. Both boreal forest hunter-gatherers and horse-equipped bison hunters, for example, invest a great deal of energy in both residential and logistical mobility. In general, hunter-gatherers heavily dependent on large game (especially in high-primary-biomass settings), have the potential for both high residential and logistical mobility. This brief overview of ethnographic data only demonstrates patteming be-ween environmental and mobility variables. It assumes that che nature of foraging in a particular environment affects the movements of groups. We now need to examine this assumption in more detail INDIVIDUAL FORAGING AND GROUP MOBILITY In Stone Age Economics, Sablins pointed out that the day-to-day economy of hunter-gatherer is seriously acre by the inminence of diminishing returns. Beginning in subsistence and spreading ffom there to every sector, an intial sucess seems only to develop the probability that farther efforts will yield smaller benefits. This describes the typi- cal cure of food-geting within a particular locale. A modest number of people sully sooner than later reduces the food resources within convenient range of camp. Thereafter, they may stay on only by absorbing an increase in ral costs ora decline in real recurs: ie in cont ifthe people choose to scarch farther and far- ther afield, decline in return if they ae satisfied to live on the shorter supplies or ineioc foods in easier reach, The solution, of course, isto go somewhere else, (Gablins 1972:53; emphasis in origina} Exinographic literature demonstrates the link between individual foraging and camp movement quite clearly. In the central Kalahari, for example, #Kade women begin to gather food near the campsite [and] they can complete cheir work in a tip (of co 2 ka during the first few days of their stay Then, gradually, as chey con- (Vincent 1984). primarily because f ‘even where rome o to improve their place long before For the G/wi, resources of the competition might (ilberbauer 198ta:3 1990:59). (All of marginale th “appears to be the: ina day ina varity gatherers in daly posible walking: fa camp is moved, aging? ‘The distance fro Figure 4-9 shows 1090, 1991). In th day, We will ay, that the a cost of 300 klocd 30 percent when© Conversely, hunters ‘moving individu Iheavily on aquatic re- tal mobility (mean = ‘many are made in exclusive. Investing ‘of extensive logistical and horse-equipped in both residential vily dependent on have the potential ynstrates patterning that che nature of ts of groups. We MOBILITY economy of ig in subsistence only to develop This describes the eypi- ‘number of people cent range of in real coms or a ‘archer and fir- shorter supplis or somewhere ee. individual foraging tari, for example, their work im trip cally os they con FORAGING AND MOBILITY 133 sume the plants near camp, they mast go father If he round tip for gathering fod plants exceeds 10 km or 0, convenience dictates that they move themselves With all eheir belongings to virgin tertory (Tanaka 1980166) ‘The Mbuti also move when foraging becomes difficult within 5 kilometers ‘of camp (Harako 1981:533). Williams found that 91 percent of Bithor move- ments were for foraging reasons; specifically, a camp moved when hunting within a s- to 6-kilometer radius fll below acceptable levels (1974274). The Australian Pigjiandjara move when women complain of walking too far to forage (Tindale 1972:244~45). Although the Hadza can forage for roots up ‘© 8 kilometers from camp, they generally do not go beyond 5 kilometers (Vincent 1984). They move primarily beeause food and water are les really available than they would like; and even where some other motive is present, ehey will of course atthe same time try to improve their access to food and water. However, movement normally takes place long before shortages have become in any way serions. (Woodburn 1968:106) For the G/wi, “migration co the next campsite is timed to occur before the resources of the last become depleted to the stage at which intethousehold competition might aise to threaten cooperation and dislocate coordination” (Gilberbauer 19812:250~52). When Malaysian Batek women find themselves walking an hour to find yams, they consider it time to move camp (Endicott and Endicote 1986:149), Agta move before all local resources ae used (Rai 199059). (All of these cases could, of course, have been predicted by the marginal-value theorem; see chapter 3.) A 20- to 3o-kilometer round trip appears to be the maximum distance hunter-gatherers will walk comfortably ina day in a variety of habitats." Thus, the distances walked by many hunter- tatherers in daly food-colecting trips are frequently les than the maximum posible walking distance. Since these distances clearly affect when and how far camp is moved, what conditions how fir person will walk in daly for- aging? ‘The distance fiom a residential camp at which a forager can procure re sources at an energetic gain is limited by the rerum rates of those resources Figure 4-9 shows the results ofa simple central-place foraging model (Kelly 1990, 1991). In this model, two foragers collect food for their family each ay. We will assume thatthe family requires about 14,000 kilocalories per day, thatthe forager walks at a leisurely pace of three kilometers per hour at 4 cost of 300 kilocalories per hour, and thatthe cost of walking inereases by 430 percent when returning home with food (Jones and Madsen 108). We134 THE FORAGING SPECTRUM DAILY NET RETURN FROM FORAGING els) concwav stance TO romaine aA ten) Figure 49. The relationship berween the diy et recur from foraging and di tance to the foraging area a function ofthe mean foaging-etum rate. The horizontal ins indicate the calories that a forager must provide: So percent ifs m= iy has ew foragers, 100 pereentif only one, A foragers expend more time and flor in eavling oa foraging area, the mean daly return declines, Asa frager provides for mor ofthe fails food needs, nr sche return rat fom the en- vironment decines, the ditance at which be or she con forage fom camp becomes shore. ‘will also assume that foraging activities, including the time to travel to forag- ing areas as well as the time to harvest and process the food resource, are ‘confined to eight hours a day.’ The daily return to foraging is simply: Net Return = [(8 -27)R] ~ (007 + 3907) where: travel time to foraging patch (dstance/3 km/hr) [R= mean environmental return rate (varied here from 1,000 to 4,000 keal/ hr for four different environments)" For any of the four environments, the net retum fiom foraging decreases farther from camp as the forager spends more time and energy traveling (© and from the foraging area relative to the time spent collecting and pro~ cessing food resources. As could be expected, the net return also decreasesan foraging and dis Seam rate. The de: 50 percent ifa fm- eed more time and ccines. As 2 forager rate from the en from camp becomes Bie to travel to forag- food resource, ate Bee is simply: 3007) ; rote somo! foraging decreases ‘energy traveling to collecting and pro- also decreases FORAGING AND MOBILITY 135 with a concomitant decrease in an environment’ retuen rate. The two hori~ zontal lines indicate the amount of energy needed on a daily basis by the family depending on whether the forager is gathering 50 oF 100 percent of the total caloric intake. The distance at which the forager brings home at least a day's worth of food, the intersection between the sloping net-rerurn lines and one of the horizontal caloric-nceds lines, becomes shorter 28 the return rate decreases or as the individual foragers workload increases. Let's ‘ay that our forager lives in local environment C, providing a mean return tate of 2,000 kilocalores per hour. Another resource located more than ¥0 Kilometers fom camp is collected by another family member (this is not an atypical division of labor in which the wife collects resources near camp ‘while the husband hunts at longer distances). We will aso stipulate that the foragercellecting the 2,000-kilocalories-per-hour resource i responsible for so percent ofthe dit. This forage, therefore, can collect food up to about 5.75 kilometers from camp. If for some reason the need for the resource increases, so that it makes up 100 percent of the dit (¢.g., ifthe other forager becomes ill), then it can only be collected at a nct gain up to about 1-5 Kilometers from camp. The efetve foraging radius, therefore is lngely «prod- tact of the return rates of the available resources and the degree of depen- ence on them (which, in turn, is a function of how many people are foraging for each family, family size, and per capita caloric needs) As average resource-retuen rates decline (as would happen ifower-ranked resources are ‘added to the diet) and/or as the amount of food a forager must bring back increases, the effective foraging radius becomes shorter and he family will probably move moze frequently and for shorter distances.” “The relationship between return rate and the effective foraging radius has an effect on diet. Central-plce foraging models suggest that the farther 2 forager tavels from camp, the more restricted his or her choice of resources ust become. Only high-return-rate resources can be taken atlong distances from camp; a wider diversity of food can be taken close to camp (Kaplan and Hill 1992; see Vickers 1939 for an ethnographic example and Speth and Scott 1986 for an archaeological one). However, a resource’ transportability also figures in here. Ifa resource is bulky compared to its caloric value, it cannot be transported easly and this will offet ts high return rate. We noted in chapter 3 that under the right conditions grasshoppers can be collected at very high return rates. However, Kevin Jones and David Madsen demon- ‘trate that grashoppers cannot be carried long distances at an effective rear rate, since a relatively small weight takes up a rather large volume (1989)” So far we have discussed foraging 28 ithe decision to move or not is based136 THE FORAGING SPECTRUM ae TPT Boe NET FORAGING RETURN RATE (kal 1800] % Te 6 ‘ONE WAY DIGTANCE TO FORAGING ARER (om) Figure 4-10. The relationship between che return rate experienced within a forag- ing area relative to that which could be expected if the foragers moved 0 anew aren, The model predicts camp movement a the point at which foragers are travel. ing about 3 kilometers from camp in order to find food, nly on the nature of foraging around the immediate camp. However, Sahlins pointed out, hunter-gatherers must weigh the cost of remaining ‘where they are and foraging farther out (or using progressively lower-return- rate resources nearby) against the potential benefit of moving to a new area. Imagine a foraging family living in an environment where 2 4,000- kelocalories-per-hour resource is homogeneously distributed across the land- scape. For the sake of simplicity, lets say each family has one active forager who must, therefore, collect 14,000 kilocalories per day to feed the family. ‘Assuming an eight-hour workday, this means that the forager must gather the resource at a minimum overall daily return rate of 14,000/8 = 1,730 kkilocalories per hour. Making the same assumptions as above, the net return rate (RR) decreases with increasing one-way foray distance (figure 4-10): 2T) — (ool + 390 T) ‘This gives am effective foraging radius of 6 kilometers the area encompassed “with ths 6-kilometer radius of camp isthe effective foraging patch, ‘We can ako compute the return rate ifthe family were to move toa new foraging area after exploiting the resources within a given radivs ofthe ste. Since food is homogencously distributed, we initially wil assume that they sects, and so tion of the ‘constant at $ each after-camp. However, as the cost of remaining Jvely lower-return= Fmoving to a new area. it where a 4,000- thas one active forager -day to feed the family. orager must gather ‘of 14,000/8 = 1,750 above, the net return ice (figure 4-10): 7) the area encompassed foraging patch. -were to move to 2 new given radius of the site. ‘will assume that they FORAGING AND MOBILITY 137 move the minimum distance to position themselves in a pristine foraging area, that is, ewice the current foraging radius, The after-move return rate of the individual forager, allowing an hour for camp breakdown and setup (more on this variable below) is figured as: pp = PUR 2D = 30007) “The after-move line in figure 4-10 shows the daly return rate ifthe forages were to move camp after foraging within 1, 2, 3... kilometers of camp. Note hat a a return rate ofjust under 3,000 kilocalories per hour (achieved at a foraging distance of about 3 kilometer), the net afer-move return rate is equal to the within-pateh return rte. After foraging within about 3 Kilo- mneters of camp, the family would do better to move to the center of a new foraging atea (6 kilometers away). Even with the move, the forager would achieve a higher return rate for that day (and would return to a 4,000-Kcal/ hnrrate the following diy) than if they had not moved, This simple model and the ethnographic eases cited above suggest that if maximization of foraging ‘efficiency isthe general goal of foraging behavior, then cental-plae foragers should ty to minimize travel time (Orans and Pearson 1979). "At the heare of the relationship berween daily foraging and group move ment are the perceived cost of camp movement and foraging. In the above model, we assumed that the location ofthe next camp was a function of the foraging radius of the current camp. But campsites can be determined by many diferent fictors, such as water sources, firewood, shade, shelter, in~ fects, and #0 on. Tn figure 4-11, the distance to the next patch is nota fanc~ tion of the current foraging radius, as it was in figure 4-10, but is held constant at $ and, for comparative purposes, 7 kilometers; thus the slope of teach affer-move line is zer0, since the cost of moving is now constant rather than being function of the site of the foraged area. As we would expect, if the next camp is 5 kilometers away, a forager should forage within 3.9 kilo~ ‘meters of camp before moving; ifthe next camp is 7 kilometers away, he or the should forage within 4.9 kilometers of camp before moving (creating some overlap in foraging areas and decreasing the amount of food available at the next camp). The predicted differences in foraging distance may seem to be minor. But note that increasing the effective foray distance from 3 te it Kilometers (a 33 percent increase) increases the foraging area, and the length of time a camp can be occupied, by 77 percent (assuming homoge- neous resource distribution). Distance to the next patch, however, i only one variable affecting the138 THE FORAGING SPECTRUM g 5 3 8 [NET FORAGING RETURN RATE fel) Y ¥ T 23 e_£ ‘ONE-WAY DISTANCE TO FORAGING AREA (km) 3 Figure 4-11. The relationship berween the return rte experienced wihin a forag- ing area relative to hat which could be expected ifthe foragers moved to a new area 5 0° 10 kilometers away. As the distance to the next patch increases, we could expect the curren foraging patch to be occupied for longer and longer periods of ‘ime (asthe marpinalvalue cheorem alo predict) cost of moving. The difficulty of wavering the terrain also figures into the calculations (Fgure 4-12). For example, 10 kilometers of mukeg in the spring is harder co cros than 10 kilometers of pati. Increasing the moving {2st to 1.300 kilocalories per hour in the model while leaving the foraging ost at 300 klocalories per hour predicts camp movement afer foraging a 4. rather than @ j-kilomeser radius, If the walking cost of both foraging and ‘moving is 1,200 kilocalories per hou, then the group should move afer foraging within only about 2.25 kilometers of camp. Thus, group nobility responds to the cost of group movement and foraging effor. This will be affected not only by the terain to be crosied, but also by whether draft animals and transportation technology are avilable (c.g. dogsleds or horses; see Binford r990). A moving cost of 1,200 kilocaloies per hour is excenive but we have chosen it to make a point: as the cost of moving increases rela. tive to the cost of foraging, residential mobility is expected to decrease. The com could inchude other factors in addition to those affecting the physical ‘movement of belongings. For example, ifthe anticipated next campsite is already occupied, the cost of moving, would have to include the cost of dis, placing current residents, perhaps through waite.y in) Senced within a forag- moved to a new increases, we could and longer periods of also figures into of muskeg in che Increasing the moving Jeaving the foraging after foraging a 4- ‘of both foraging and should move after Thus, group mobility effort. This will be aso by whether draft , dogsleds or horses; per hour is excessive ‘moving increases rela~ sd to decrease. The affecting the physical sd next campsite is lude the cost of dis FORAGING AND MOBILITY 139 Figure 4-12, A band of Tree River (Agianmius) Copper Inuit move across the tun= dea; women and dogs pull the sleds, The cost of moving i affected by the terrain to be crossed, the weather the amount of material to be carried, and the type of hous ing. Photograph by J.J. O'Neill, October 1915, courtesy ofthe Canadian Museum cf Civilization, no. 38571 ‘Housing also affects the cost of moving. The time required to break down and set up camp is seldom discussed in the literature. ‘The models as first developed used one-hour camp breakdown/setup time based on what litle information was available (Robert Hitchcock, personal communication, 1986; Peter Brosius, personal communication, 1989). Camp breakdowns ‘may be quicker for many tropical groups, but could be slower for Arctic peoples due to differences in the amount of goods carried (ee Burch 1988:107 for an example of a one-hour camp-setup time for an Arctic ‘group)."" We can see, however, that increasing or decreasing this time greatly alters the mobility solution by changing the length of the working day (fiz~ lure 4-13): a camp-breakdown time of two hours means that it is not worth- while to move before exhausting neatly all food within a 6-kilometer radius, while a breakdown/setup time of one-halfhhour predicts movement at a 1.5 Ikilometer foraging radius. Ie appears, then, that as the movernent of camp itzelf becomes more difficult and time consuming, hunter-gatherers may re- ‘main in their current foraging area for longer periods of time. Conversely if140 THE FORAGING SPECTRUM F sols e hswe E y Bo a 3 2 2000: posi ees te v y y (ONE-WAY DISTANCE TO FORAGING AREA (en) Figure 4-13, The relationship between the resurn rate experienced within a orag~ ing ara relative to that which could be expected ifthe foragers moved to a new area at diferent camp-move times, from one-half hour to ewo hours. As the cost of camp breakdown and setup increases, s0 does the predicted foraging radius and length oftime the current camp is occupied. 4 group must be mobile for energetic reasons (low return rates that result in a short effective foray distance), their housing should be tailored to their ‘mobility needs, Obviously, housing and mobility are expected to be system atically related to each other (Binford 1990)" ‘This simple model assumes only that hunter-gatherers move as families, snot necessarily as hands. From ethnographic data we know that some hunter- gatherer social units can have extremely fuid composition, with individuals and families moving a different schedules. Anthropologists frequently atti- bute this fuidity to the need to relieve social tension. Though this may be the proximate cause, subsistence can often be the ultimate source of this tension, Agta bind members intensely debate for hours or even days whether to move or not, with foraging effort playing a prominent role in these de~ bates (Rai 1990:39). Large families could have higher camp move costs (See below), and thus may wish to remain in a current camp longer than smal families. On the other hand, foragers with large families need higher mini- imum daily returns than ehose with small families, and chus have shorter effective foraging radi. Ifa lage family has remained in camp out ofa social need of obligation, the foragers ofthis family may forage past the point atced Within 3 forage moved to a new hours, As the cost of foraging radius and rates that result in be tailored to their 10 be system- move as families, thatsome hunter: sition, with individuals fogists frequently attsi- = Though this may be akimate source of this s or even days whether gnent role in these de- ecamp move cost (see amp longer than small dics need higher mini- and thus have shorter in camp out of a social orage past the point at FORAGINGAND MOBILITY 141 which they could forage more efficiently by moving, and even reach the ‘edge of their effective foraging radius before the foragers of smaller families, resulting in tension and posibly group fision. Therefore, large and small families should move on different schedules, Family size will not be im- portant in determining camp movement, however, where everyone's subsis~ tence is tied to the same resource (e.g, fish runs, communal hunting) “We should also point out that since many plant foods provide returns lower than those of large game—but ate much more reliable resources— the effective foraging distance for plants is shorter, in general, than it is for large game. Since large game is usually procured by men (with some excep- tions; see chapter 7), women's foraging should by and large determine when ‘camp is moved. Among the Agta, “[since] hunting depends on mobile ani- mals, itis not an important consideration [in determining moves). Men and ‘women freely voice their opinion on residence change, but women, who ‘must carry out the most gathering, have the final say” (Rai 1990:39). This is important when considering the effect of reduced residential mobility on ‘womer's and men’s nonforaging activities (see below), Other variables: return rates “The effects of alterations in the return rates used in figure 4-10 can be easily compated. Ifthe retum rae of the resource to be used in the next camp is lowered relative to that of the resources used at the current camp, the forager is predicted to remain atthe current camp for a correspondingly longer pe~ riod of time. However, ifthe return rate ofthe resources of both the current and potential camp is changed, but held constant, the within-patch and after~ move return rates always equal one another ata foraging radius of about 3 kilometers. They do not alter the predicted outcome even if, as is true for a return rate of 1,000 Kilocalories per hous, i is not possible to forage at an energetic gain at a disance of 3 kilometers. Under circumstances of low return rates, camp-breakdown time and the amount of material posessions ‘would have to be extremely low (and/or family size would have to be very sama) In these cases, some people may set out before others, using the camp move 2s a foraging trip as well In some tropical forest cases, men head out immediately for the new camp, hunting along the way, while women pack, move camp and gather plant food as they travel (eg, Morris 1982:176; Hill and Hawkes 1983). “Tis leads to an intriguing contradiction berween the marginal-value the orem and the foraging-radivs model used here. In chapter 3 we introduced142 THE FORAGING SPECTRUM CUMULATIVE er RESOURCE HARVEST Teeleaine oy oe TIME IN PATCH Figure 4-14. The curve in this illustration depicts a general depletion curve: the nce return decreases the longer a forager stays in 3 patch, Recalling the MVT, where ‘mean environmental recur rates are low (the slope of ine A) the forager should re- ‘main in the current patch longer (7. than when return rates are high (Line B, Tj. Drawn by Dan Delaney. the marginal-value theorem (MVT) to predict when a forager should leave one foraging patch and move on to another. To review the most general case, a forager enters a patch and collects resources at a net gain intially, but slowly depletes the resources, lowering the net gain over time (Figure 3-5) ‘The MVT states thatthe forager should leave the patch when the rate of recur in the patch equals the mean environmental rate of return (taking ttavel time into account) Figure 4-r4 shows how different mean environ ‘mental return rates produce different patch-resdeney times. Line A repre~ sents an environment with a low return rate while Line B, with a steeper slope, is an environment with a high return rate. Holding other variables constant, and though it may seem counterintuitive, inthe environment with the higher mean return rate the forager would leave the patch sooner than 3 overall return rates were low (E. Smith 1991:255). In either case, they ‘would leave before patch depletion occurs. The foraging-adius model sug- gests that under low return rates (and holding other factors constant) foragers will reach their effective foraging radius before reaching the moving-on threshold and could completely deplete a patch before moving on, ‘There are several reasons for this. The MVT does not hold, strictly speak ng, in the cae at hand since in the present foraging-radius model resourcesP : sion curve: the net the MVT, where the forager should re- ee high (Line B, T,) = Be he mos pene et gain intially, but time (igure 5-3) when the rate of rte of return (taking mean environ times. Line A repre= ine B, with a steeper ing other variables fhe environment with patch sooner chan In either cae, they radius model sug- constant) foragers the moving-on moving on. hold, strictly speak- ‘model resources FORAGING AND MOBILITY 143 are homogeneously rather than patcily distributed, Moreover, the foraging radius model assumes that foragers are aware of the resource distribution and fare more o less vacuuming up resources close to camp, then those further ava, then those still further away, and so on until they reach their effective foraging radius or the point at which they would do berter to move camp. ‘The MVT assumes random search and encounter. Also, the model used here docs not use resource density. Resource density is central to the MVT since the diminishing-retums curve is based on gross returns over time that de- ‘crease 28 a fonction of increased search time that is in tum a function of decreasing abundance. Finally, the foraging-radius model is a central-place ‘model, in which foragers go out from camp, procure resources, and then return, rather than consuming resources as they encounter them while for~ aging. In the foraging-radius model, 2 decline in return rates result fom investing time in traveling to and from the foraging area. In the diminishing returns ease to which the marginal-value theorem is applicable a decline in _eturn rates results from increasing search time. Foraging situations could fill under the conditions of the central place or MVT conditions. This raises the ‘question of whether the pattern of increasing residential mobility relative to primary biomass depicted in figure 4~4 a product of decreasing retara rates (as measured by the proxy variable of primary biomass) and decreasing net per capita guins, or a product of increasing return rates (since primary Pro- duction increases with primary biomass) and higher consequent opportunity ‘costs. The ecological perspective taken above argues for the former situation, ‘but hard data on foraging returns for a diversity of groups along a similar environmental gradient are needed for atest ‘We might also note that the rate of decrease in a resource's net return contelates with the slope of the net-return line. That is, in figure 4-9, net returns from collecting in a 4,000-kilocalories-per-hour environment decrease more rapidly relative to foraging distance than in the 1,000- Iclocalories-per-hour environment, Simms also found that the returns from hhigh-ranked resources decrease more rapidly with decreases in abundance than do returns from low-ranked resources (1987:50-55). This means that Ihunter-gatherers are likely to perceive and respond co changes in the avail- ability of high-ranked resources before they respond to changes in Jower- ranked resources. This makes sense in light of the central prediction of ‘optimal-foraging theory's diet-breadth model discussed in chapter 3: that jnclusion of low-ranked items in a die isa funetion of changes in the avail- ability of high-ranked items,144 THE FORAGING SPECTRUM ‘Time frames In the model, we assumed that some foraging is done on the day of the move. But this may not be true in many cases, Where no foraging is done fon the day of the move, the after-move retum rate of the day of a move ‘would be zero. According to the model used here, this would mean that the forager should not move until eating everything within the 6-kilometer ra- dius of camp. In other words, the effective foraging patch should be com- pletely depleted before moving. This is clearly not supported by the cethnogeaphiec data cited at che beginning of this section. ‘The answer may lie in the time frame used. In the model used here we assumed an hourly retumn-rate maximization, a common assumption, as we noted in chapter 3, of simple foraging models. However, san hour the time period over which hunter-gatherers evaluate resouree returns? Or, recognia- ing that there is always some daily variance in returns and that food may not be gathered every day, do they respond to the environmental return rate averaged over several days, or even weeks? Moving before exploiting every- thing (in our model) within a 3-kilometer radius of camp may result in return-rate depression fora single day, but hunter-gatherers may accept this temporary loss knowing that the day after the move the return rate will be 4,000 kilocalories per hour. Let's assume that the forager achieves a 4,000-kilocalories-per-hour retarn rate for several days after moving, but does not forage on the day of the rove. As figure 4-15 shows, depending on the time period over which rates are averaged, the forager stays for different lengths of time than those pre- dicted by a model assuming foraging on the day of the move. For example, if after-move returns are averaged over only two days (0 for the day of the move + 4,000 keal/hr for the next/2 = 2,000 keal/ht), the forager should remain at the current camp until consuming everything within s.s kilome- ters of camp. Ifa four-day average is used, then the mean return is increased. {0 3,000 kilocalories per hour, and ehe forager would remain in the current camp for a much shorter period of time. Ethnographic data, unfortunately, are not sufficiently detailed to provide much guidance on what time frames ‘may be used under different conditions but its likely thatthe relevant period is several days in length. Risk ‘The model used here also assumes that hunter-gatherers have a perfect knowledge of the environment and can know the state of anticipated re-one on the day of the sc no foraging is done = of the day of 3 move gs would mean that the bin the 6-kilometer =~ ch should be com- sot supported by the ‘model used here we assumption, as we isan hour the time returns? Or, recogmiz~ nd that food may not sal return rate exploiting every- camp may result in xs may accept this 1 return fate will be jes perchour return on the day of the 1d over which rates time than those pre- move. For example, {@ for the day of the the forager should within s.s kilome return is increased semain in the current ie data, unfortunately, ‘on what time ffames the relevant period sets have 2 perfect ‘sate of anticipated re~ FORAGING AND MOBILITY 145 Jae Gl & Eee & seoo| tO0rmens a S| saws Re 4 | vue ‘ eee ‘ONE-WAY DISTANCE TO FORAGINIG AREA fm) Figure 4-15. The relationship beeween the resum rate experienced within a forag- ing area relative to that which could be expected ifthe foragers moved to a new acca ith after move return ratet averaged over ow t0 six-day periods. As the period over which afer-move return rates are averaged increases the predicted for- aging radius and camp-occupation time decreases sources. This is rarely true, and so the cost of moving must include a risk factor. If the anticipated resource is not a certain one, then the cost of ‘moving will, in effect, be higher, and we could expect hunter-gatherers to stay longer in their current camp. Many desert hunter-gatherers elect to remain at a water source at the expense of decreasing foraging-return rates because they are uncertain of the condition of other water holes; they may remain at the current water hole either until it rans out or tuntil che status of other water sources is ascertained. Some Australian ‘Aborigines, in fact, will accept extremely low caloric intakes and forage up to 15 kilometers ffom camp rather than move from a secure water source (Cane 1987; Gould 19603). ‘One way to look at risk is in terms of return-rate variance. As noted in chapter 3, holding resource type, density, and even forager capabilities con- stant, there can be a great deal of variance in day-to-day return rates. How might this affect foragers’ decisions to move? First, this isa question about perception. Given that hunter-gatherers do not carry notebook computers, how low must an average return rate be before the foragers decide to move, knowing that variance in the after-move146 THE FORAGING SPECTRUM return rate can also be expected? At what point would foragers perceive that they can do bewer (maximize foraging efficiency) by moving? Figure 4-16 shows the within-patch and ater-move return rates depicted in figure 4-10, but with a 200-Kilocalories-per-hour variance limit around the within- patch line and a *400-kilocalores-per-hour variance limit around the after move line, In dis case, chere i greater variance in the after-move return rate than in the within-patch return rat, aldhough the average returns are equal [Note that at a foraging radius of just over 4 kilometers, the lower variance limit ofthe within-patch retara sate is about equa to that ofthe after-move return rate, Is at this point, after using everything within 4 kilometers of the current camp, that the foragers may move, knowing that they ae guaran- teed to do better than their current average effort. The higher the variance in the aftcr-move revur rate, the longer foragers are predicted to remain in the current camp. However, variances and means are normally inversely related to one an- ‘other; low mean return rate will normally havea high variance. As we have already seen, if the retum rate of a resource to be exploited at the next potential camp is lower chan that of the current camp, the foragers should remain in the current camp for longer periods of time than if the resource return rates were the same. It may not be possible, or necessary, therefore, to separate the effects of mean return rates from variance, Second, there is a question about goals. We know very litle about whether hunter-gatherer judge resources in terms of mean return rates or variance (see Gragson 1993 fora recent effort). But if forages wish (0 go for 4 resource that can occasionally provide very high remus, they may elect to seek out resources with high variances intentionally. For reasons we will discuss in chapter 5, some hunter-gatherer (pechaps especially male hunter) do indeed seck out riskier resources. If forager wished to try and maximize his foraging return, he might move when the highest posible retarn after roving is higher than the highest posible return of foraging from the eur- rent camp. Referring to figure 4-16, this would be at a foraging radius of less than 2 kilometers, implying 2 shorter occupation time than our original model predicted. Unfortunately, foragers’ perceptions of resources in terms ‘of variables that we, as anthropologists, consider important is not well known and merits farther feld study. Linked tothe issue of resource variability is the isue of storage. Storage obviously results in the accumulation of food at one or more locations, in~ creasing the patchiness of an environment, and could encourage decreased residential mobility. Food storage among hunter-gatherers may principallyHoragers perceive that “moving? Figure 4-16 epicted in figure 4-10, round the within- around the after returns are equal. the lower variance that of the after-move thin 4 kilometers of related to one an- ‘variance. As we have ‘exploited at the next the foragers should shan if the resource necessary, therefore, ‘very little about ‘mean return rates OF ‘wish t0 go for they may elect £0 For reasons we will male hunter) 10 try and maximize at a foraging radius of ‘ime than our original of resources in terms st s noe well known 1 of storage. Storage ‘or more locations, n= ‘encourage decreased rors may prineipally FORAGING AND MOBILITY 47 g a 4 q [NET FORAGING RETURN RATE hee) g i Cm eee re eet ‘ONE. WAY DISTANCE TO FORAGING AREA (ers) Figure 4-16, The effects of retuen-rate variance on decisions to remain inthe cur ‘ent camp or move, To maintain 2 high a ream rate as posible, the forager should ‘remain in the curent patch until the lower variance limit ofthe after-move rate equals the lower variance limit of the within-potch rate. Alternatively, to seck esky resources, the forager may move much sooner, when the upper variance limit of the affer-move rate exceeds that ofthe within-patch rate bbe a way to cope with resource seasonality: data show that the volume of stored resources increases with decreasing ET (Binford 1980; see also Keeley 1988), The decision to reside atthe location where food is stored oto trans port the resources to another location, however, depends on the return from moving one set of resources versus the return rate expected fiom each loca tions resources (Jones and Madsen 1989; Rhode 1990). Thus, food storage does not necessarily cause a reduction in residential mobility. The decision to move also depends on the variance of the return rate of the next anticipated resource. Figure 4-16 suggests that, given a choice of several resources with similar return rates, hunter-gatherers may move from their current foraging area to that of the resource with the least variance, If resource variability Jnereases over the long term, hunter-gatherers could be expected to invest more in storage, and decrease mobility, since there is no variance in the acquisition of food once it is stored (see Rowley-Conwy and Zvelebil 1989). [As noted above, not all sesidential movements are directly controlled by sub sistence, People may move to gain access to firewood, raw materials, or be- cause insects have become intolerable at the current camp. Movements ean448 THE FORAGING SPECTRUM be socially or politically motivated 2s well, as people seck spouses, lies, ot shamans, or to distance themselves from sorcery, death, or politial forces, People may move to cieve socal tension, to vit friends and relatives, trade, gamble, participate in rituals, or just catch up on news." However, some ‘movements made for social reatons can ulkimately be related to foraging, For cxample, during period of drought-induced food stress, /ai/xai Bushmen stated that they were going elsewhere to trade, but this decision followed two weeks of bickering over food (Wiessner 19826). While the model pre~ sented here is limited to the extent that itis based only on considerations of foraging efficiency as we pointed out in chapter 2 this permits analysis to determine when fictors other than foraging eficency are at work in a par- ticular ease SEDENTISM, ‘To this point we have discussed the factors chat keep hunter-gatherers mov ing. Yet we know from archaeology that at certain times and in certain places, hunter-gatherers settled down, and ceased to move residentially. The origins of sedentary communities is an important question in anthropology, for itis with the appearance of sedentary communities that we find evidence of nonegalitarian sociopolitical organization—social hierarchies and heredi- tary leadership, politcal dominance, gender inequality, and unequal acces to resources, a well as changes in cultural notions of material wealth, privacy individuality, and cooperation (Wilson 1988). We examine the possible ela. tionship between sedentism and sociopolitical organization in chapter 8. In this chapter we are concerned with understanding the eanse(s) of edentism, ‘The term sedentom means different things to diffrent people. For the ‘most part, sedentiam refers to the process “whereby human groups reduce their mobility to the point wire they remain residentially stationary year- round” (Hitchcock 1987a:374) or as sctlements where “at least par of the population remains at the same location throughout the entire year” (Rice, in Raffery 10852115). Sedentism is usualy thought of 5 relative diference rather than a static condition, thus sedentary settlement systems are “less ‘mobile than previously” or become “increasingly sedentary over time” (Kelly 1992). Frequently, definitions conflate several dimensions of mobility including seasonal movement of the residential base camp, movernent of in dividuals around and between residences, movement of a group's yearly range or aggregation site (eg., winter villages on the Northwest Coast or ‘wetseason villages in seasonal tropical forests), and the permanence of facilicck spouses allies, or B. oF political forces. ds and relatives, ade, ws. However, some e to foraging. For PXsi/sai Bushmen ds decision followed le the model pre~ fon considerations of permits analysis to at work in a par residentially. The in anthropology, ‘we find evidence cies and heredi- unequal access to wealth, privacy, the possible rela~ in chapter 8. In (5) of sedentismn, people."* For the ‘groups reduce stationary year- “atleast part of the centize year” (Rice, ssa relative difference systems ace “Tess sary over time” ons of mobility, ‘movement of in- fof a group's yearly wihuvest Coast or permanence of fcili- FORAGING AND MOBILITY 149 ties such as houses and fish weirs (Eder 1984; Ingold 1987; Ralerty 1985; Stak 1982). In conflating these different dimensions there is a tendency to think of mobility in terms ofa single scale of residential mobility and some times to tink of societies a cither mobile or sedentary. But it should be lear by now that mobility is not an all-or-nothing affir. Nor is there a single sale of mobility for as we have seen, mobility varies along several potentially independent behavioral dimensions. Ethnographic data show the inteplay between group and individual ‘movements as residential mobility decreases. Formerly nomadic, the Batak now maintain a central settlement at which someone is almost always present throughout the year (Eder 1984). They move this central setlement, how- ever, every seven to ten years. An individual family spends only about 25 percent ofthe year in the central ettement. The rest ofthe time is spent in field houses and forest camps. Moving among a limited number oflections, a Batak family changes location about ninety times a year, moving about 3 Jelometers each time for a total yearly residential mobility of about 270 kilo- meters. Individuals make foraging trips into the forest from these camps, adding to overall individual mobility. In other words, the Batak “shifted the burden of that [residential] mobility off of the local group as a whole and conto lower levels of social organization” (Eder 1984:851). It appears to be generally true that as residential mobility i reduced, logistical mobility in- creases (Binford 108). This process occurs among many hunter-gatherers who, for one reason or another, have become sedentary in recent decades (Gee Hitchcock 1982, 19874; Hitchcock and Ebert 1984, 1989). As Bush ‘men become sedentary (due to government coercion or the atractions of ‘wage labor) men make longer logistical forays (Hitchcock 1982). The Kala- har’ residentially mobile Kua, for example, might not travel more than 6 Jéilometes from camp, while members of sedentary Kua villages make tips ‘of up to 50 Kilometers." In brief, there i a constant trade-off between the costs and benefits of group movements on the one hand, and individual logistical movements on the other. Sedentism does not save energy, but it does reorganize it. Why should ebis happen? For many yeats, the reigning view in anthropology was that a nomadic lifestyle was not something any rght-thinking individual would want. “We have taken for granted,” wrote the pareipants in a 1955 seminar in com munity patterns, “that in general sedentary life has more survival value than wandering life to the human race, and that, other things being equal, “whenever there is an opportunity to make the transition, it will be made” (Beardsley et al. 1956:134). The opportunity was thought to be either plant1g0. THE FORAGING SPECTRUM domestication or resource abundance, The former was assumed to be the more likely cause, and sedentary hunter-gatheress, such as those of the ‘Nonhwrest Coz, were considered anomalies. However, we tow have many archaeologieal cases where agriculture preceded sedentism of where seden~ tism preceded agriculture (see Kelly 19925 Price and Brown 19836). Ethno- traphiealy, tere are many horticulrraists who are seazonally mobile (8 fhe Raramuri of Mexico; see Hard and Merrill 1992), who make long treks {eg many Amazonian societies), or who shift residence every few yeas in response to sol depletion ora decline in hunting returns Gee Vickers 1989) "Thus, the relationship between agriculture and mobility is by no means straightforward, TBinford challenged the cesource-abundance argument—what he called the Garden of Eden perspective—by raising the issue of seasonal and yearly ‘esource variability (1983). Hunter-gatheress remain mobile, Binford ar- fed, not because they dont have the opportunity to settle down, but in eater to maintain information about resources in order to have backup oP- ont in case an expected resource is not available, Maintaining knowledge ‘Sf eurrent and potential future states of resources —the status ofa water hole, signs of game and plant food—is critical for group planning, In the early “pring Nunamiut men, for example, “travel widely attempting c0 find mov- dng caibos ... [and] to gather information as to the number of animals and the probable timing of movement so they may plan thei intercepestaegy” {Binford 1978°165). The Arctc’s Netlingmive and che Australian Arands fdo much the same thing (Balikei 1970; Horne and Aiston 1924). Hunter- taherers use this information to ass various aematives, much asa hess player does: in asesing the cost of exploiting resource, the band considers not only its next wpa but the whole series of migrations in the foreseeable future... dhe sim is not to plot the coming seo tnerary in detail butt work outa series of moves dat wel permic the ban the widest choice of subsequent sites. (Siberbauer 19814249) Consequently, many hunter-gatherer feel compelled to maintain knowledge “ofenormous ateas. The Nunamiut maintained knowledge of nearly 250.000 Square kilometers Binford 1983:206); the Australian Pintupi have know tage of over $2,000 square kilometers (Long 1971)” No one in either of these groups uses ll of his or her territory ina single year_—a Numativt $2 fact, may penonally use oly one-tenth ofit in his or her lifetime. Nonethe- less, mobility does help individuals maintain knowledge of enormous areassumed to be the . as those of the ¢ now have many or where seden- n 198sb). Ethno- nally mobile (c.g, > make long treks every few years in see Vickers 1989). is by no means —what he called sasonal and yearly bile, Binford ar- ele down, but in have backup op- sining knowledge us of water hole, ning. In the early sing to find mov- per of animals and ntercept strategy” Australian Aranda mn 1924). Hunter- s, much 2s 2 chess not only its next c.. the sim is not series of moves that ecbaver 19812:249) sintain knowledge “of neatly 250,000 upi have know!- 2 one in either of —a Nunamiut, in ifetime. Nonethe- of enormous areas, FORAGING AND MOBILITY 181 thers point out that mobility also helps hunter-gatherers to maintain social ties that form insurance networks of affinal kin, trading and religious part- ‘nets, and instructs children in the resource geography of a region." However, the extent to which hunter-gatherets must maintain informa~ tion about and ties to other areas is related to the degree of temporal and spit variation in resources. Whereas no environment s perfectly constant, cavironments fluctuate on different scales. We could expect some variability in the need and ability to maintain information o social tes relative to the degree of resource fluctuation (ee chapter 5). Where resources are constant and reliable, mobility will nt be needed to gather information, Also, main- taining knowledge of other areas does not require residential movement— sedentary hosticulturaists maintain information networks without moving their villages. Leaving aside the issue of information, then, we return to the relationship between foraging and residential mobility and ask: is sedentism a product of resource abundance? Recall the foraging model discussed above (figure 4-10). Exactly how long. 4 forager could remain in the 6-Kilometer-radivs patch depends on the den- sity of food within the foraging area. Assuming a homogeneous caloric yield of .25 kilocalories per square meter, the 6-kilometer foraging area could potentially be occupied for upwards of 673 days by a band of 25 people. But if the foragers leave afer eating everything within about 3 kilometers of camp, the camp is occupied for only 167 days. Even though they could remain where they are and forage at an energetic gan for ncarly (wo years, this band of hunter-gatherers should leave after a few months if they wish to ‘maintain as high a daily return rate as possible." It would seem that even. in a Garden of Eden, and leaving aside other factors that could encourage movement, foragers should still move. In fact, this model suggests that in an environment of homogeneously distributed resources, the only apparent reason hunter-gatherers would net move i if there is no place to move to— that is, if population density rises to the saturation point, packing foraging groups into a region. As population density rises, residential mobility would involve the additional cost of displacing a group already using.a region. Since this could involve physical violence, it may tip the scales in favor of re- ‘maining in place and thus encourage sedentism. Where resources are dense and retuen rates high, hunter-gatherers may be very mobile initially; but hhigh per capita return rates (or high rates of recovery; see Winterhalder and Goland 1993) could result in rapid population growth that could quickly constrain residential movements (se chapter 6) ‘What about environments where resources are not hoanogeneously dis-152 THE FORAGING SPECTRUM ‘trbuced? In our discussion above (refer back to figures 4-10 through 4-16), ‘ve saw that as more energy or time was needed to move camp, relative to the cost of foraging at the current camp, the camp would be occupied for longer periods of time. The frequency of residential movements decreases as resource patches become more spread out, while the length of logistical for- 2 increases. We have already diseussed several factors that are important haere: the distance to the next camp, the terrain to be crossed, the amount of saterial that must be carried, the time required to construct housing, and the anticipated retum rate (and variance) of resources at the next location. ‘Nonetheless, these variables all converge on the single question raised by Shlins and reiterated by behavioral ecology: what is the cost (and bencft) ‘of staying in one place versus the cost (and benefit) of moving somewhere che? From our models we can deduce that hunter-gatherers should stop ‘moving residentially ifthe anticipated return rate of the next patch minus the cost of moving is greater than the anticipated retum rate of the patch currently occupied. Stated more generally, sedentism can be product of local abundance in a context of regional scarcity. Ths is different than say- ing that resource abundance results in sedentism. While itis clearly neces- sary that resources be continuously availble year round, or be available in sufficient bulk to be stored for a seasonal lean period, in order for hunter- ‘stherers to be sedentary (and continue to hunt and gather, rather chan farm), it is also likely that the decision to become sedentary is based on regional, not just local, resource conditions. Sedentism may also have a domino effect. If there is some overlap in ‘groups’ ranges (as frequently appears to be the case ethnographically) when fone group elects to become sedentary, for example at the mouth of a pro- ductive salmon stream, they in effect remove a resource patch from others. ‘Tais makes the environment more patchy, and increases the cost of moving to an occupied resource area (ince it would require displacing the now- sedentary group). Once established, then, a single sedentary village could encourage neighboring groups to become sedentary. Therefore, sedentism may be 2 regional phenomenon, with sedentary communities occurring in batches, rather than singly” FORAGING, MOBILITY, AND SOCIETY ‘The mobility ethos Among many hunter-gatherers, mobility is highly valued; there are even a few cases of hunter-gatherers, such as the Australian Anbarra, who move