D. llunt

2l

tc. ^'l-: . P<- .i. fb€ ,'- .

B-l

r

llr

; i

lrh-:...-

TIME AND ENERGY:THE ECOLOGICAL CONTEXT FOR THE EVOLUTION OF BIPEDALISM

F\:.,'.

)|j,'

. .

RobertA. Foley and SarahElton Lr E ' : tual lrr-

Human EvolutionaryBiology ResearchGroup Departmentof Biological Anthropology University of Cambridge Downing Street Cambridge,CB2 3DZ England

lr-.: h.. F.

1. INTRODUCTION This paper is concernedwith explanationsfor the evolution of bipedalism. Its general point is a very simple one-that the occurrenceof bipedalism is context specific. The pattern of hominid evolution, as much as that of any other lineage, reflects the costs and benefits of the way an animal is structuredand behaves,and this ratio is entirely dependent upon when and where it is occurring. Historically the context for bipedalism has been the general characteristicsof the environment-savanna grasslands,open environments, patchy woodland versus forest. This remains important, but here we shall add a new consideration, that of time budgets, which provides a more specific ecological context for consideringthe energeticsof bipedalism. The paper will first discuss the various contexts that do need to be taken into account when considering the origins of bipedalism. It extends a model developed earlier (Foley, 1992), based on a cost-benefit analysis, which suggestedthat the advantagesof bipedalismshouldbe placedinto the specificecologicalcontextof the Pliocenehominids. That context was specifically the effect of more arid, seasonaland open environments with more patchy and dispersedresources.The major changepredicted in the model was an increasein day rangelength,and that the energeticsofbipedalism are directly linked to increasedranging area (Foley, 1992).ln quantitativeterms, it was suggested(Foley, 1992: Figure 5.3) that bipedalismallowed a 50 kg hominid to exploit a day rangeof l6 km for the same energy as a male chimpanzee(a5 kg) usesfor the maximally observedday range length of l0 km (Rodman, 1984),a result that Leonard and Robertson(1997) have recently replicated.The earliermodel will be developedhereto incorporatea new element, Primate Locomotion, edited by Strassere/ 4/. Plenum Press,New York. 1998

419

420

R. A. Folev and S. Elton

time. Although the energeticadvantagesof bipedalism allow longer day ranges,the extension of a day rangelength hasto occur in the contextof a finite resource,that of time, and in particulai houis of daylight (Foley, 1995).Time, as Dunbar (1992) has argued,is the hidden constraint in behavioral ecology. To model correctly the energeticsof bipedalism in an ecological context it is thereforenecessaryto constructa time budget of daily activiterties. This is done here,with particularemphasison how such activitiesare distributed just the energetic restrially and arboreally. This distribution is important as it is not the costs of but also primate, a terrestrial as bipedal being from accrue that advantages part of the day climbing. The model of daily energeticsindicates that a very substantial In the final the costs' has to be spent terrestrially before the benefits ofbipedalism exceed of origins the for part of the paper, the implications for reconstructions of scenarios bipedalismare discussed.

2. ADAPTATION AS PROBLEM-SOLVING: THE COSTS AND BENEFITS OF BIPEDALISM acEcologicalcontextis highly specific,not just in terms of the habitat,but also the environment same inthe live can primate of species Two tivities of the animal concerned. vary. The but be morphologicallydifferentbecausetheir diets,activitiesand time budgets and circumstances, upon depending maladaptive, or adaptive sametrait could thus be both greatest disparperhaps the is ofcontext lack places. The and times all no adaptationis for to ity between what might be called the pre-modernsynthesisand more recent approaches advantaare hominid evolution. The essenceof a neo-Darwinianapproachis that traits any feathat implies This terms. absolute in than rather circumstances, particular geous in to response trrreof un organism,be it behavioralor morphological'will be an evolutionary only not precise settings of time and place. Or, to put it another way, that such traits are To exuduun,ug"ousin particular environments,they are also disadvantageousin others' evolution plain the evolution of featuressuch as bipedalismduring the courseof hominid ihus requiresa detailed considerationof the contextsin which it is likely to occur' Within modern evolutionary biology there are two broad conceptual frameworks into context' that provide the basis for attemptingto place adaptiveevolutionaryevents (Maynard approach The first of these is what might be called the optimal problem-solving of survival problems Smith, 1978; Foley, 1987).Featuresevolve becausethey solve the mechaevolutionary faced by organisms in particular environments,and selection is the a wide to approach This nism that leadsto what might be called problem-solvingoptima. extensively developed been has ranging seriesof adaptiveproblemsin hominid evolution of cost-benefitalgorithms.Adaptations teof.V, 1987).Compiementaryto this idea is that costs rnuy b" treated as benefits to an organism.All such adaptivetraits, however, impose they or maintenance; on an organism.These costs may be associatedwith development or opporcosts direct may be They may be energeticor structural; they may be behavioral. tunity costs-that is, costs that come into account becausethe evolution of one feature has will reduce the opportunitiesfor other evolutionarychanges.Cost-benefitanalysis particularly strategies, evolutionary proved to be very productiveas a meansofanalyzing and in terms of foraging behavior,where direct observationand measurementof energetic in are, analysis and cost-benefit problem-solving time inputs and outputsis possible.Both dethat features that premise Darwinian the basic effect, ways of making more practicable velop over the course of evolution, featuresthat are selectedfor, provide specific advantages-adapt21isns-in particular contexts.

Time an

I specia culosk that bi The bt neura anti-p seem biped biped evolv(

may u benef debat McHt of ex that t more vanta bene the cr biped activ rame

bem in th anal used ciesbe kt

3.7

3.1

PIio trat€

whir bas than line

Time and Energy: The Ecological Context for the Evolution of Bipedalism

i.

421

Bipedalismis a classicexample.Among mammalsbipedalismis very rare' and as a specialiied adaptationin primatesoccursuniquely in hominids.The changesin the musculoskeletalsystemthat result from bipedalismare very extensive,and it has been argued that bipedalismis the basisfor many other non-locomotorychangesin hominid evolution' The benefits that arise from bipedalism range from increasedmanual dexterity, enhanced greater neural and general thermoregulation,decreasedlocomotor energeticexpenditure, times at can benefits of array an Such anti-predatoryvigilance,intra-specificdisplay,etc. of costs the when only primates. among seem at odds with the rarity of bipedalism why clearer it become disadvantages--does bipedalism are considered-that is, the not bipedalismmay both have evolved among horninidsand, which is equally important, benefits' as the dependent context aS are in turn' evolved alnong other species.The costs, that Discussionsof bipedalismhave generallyfocusedon two types of advantage energetic is, the that advantage, locomotor the direct may accrue. The first of th"t" is considerable benefits that arise from being able to walk upright habitually' There has been and Rodman 1973;Rowntree, (Taylor and exist debate as to whether these advantages stream a second result, a as and, 1996) Young, and Isbell McHenry, 1980; Steudel,1994, been argued of explanationsrelate to the non-locomotory advantages.In particular' it has and thervigilance, (Darwin, 1871), dexterity manual greater that bipedalism leads to adindirect These l). 199 1985, 1984, Wheeler (Newman, 1970; rno..guluto.y efficiency and wiil not be incorporatedhere. Instead,in modeling the direct energeticcosts vanta=ges of some least at is that first The developed. be will particular in benefits two novel themes is that The second the costs imposed by bipedalism arise from a reduced ability to climb' a daily scheduleof of the context in occurs behavior, locomotor any with bipedalism,as set of pau"iiuiti"r, and models shoutd therefore be built around such a time-budgeting rameters. is bound to Application of cost benefit analysis to problems in long term evolution in the field or be more froblematic than analysesof living populations.Direct observation part any such of in the laioratory of actual energetic costs is obviously an important will be that analysis,and with animals long extinct this is not possible.The alternative speliving from drawn used here will be to build u rnod"l using data and observations should approach an such of cies-in particular, chimpanzeesand humans.The limitations be kept in mind.

3. THE CONTEXTSOF BIPEDALISM Severalcontexts for assessingthe costs and benefits ofbipedalism can be described'

3.1. Taxonomic Context the There is now a relatively rich fossil record for hominids within Africa covering provide illusrepresentinga number of distinct taxa. The evidencethey Plio-Pleistocene, tratesthe diversity of hominid locomotor strategies' (> 4-4 Myr), for 3.1.1.Ardipithecusramidus.This is the earliest known hominid (White et al', 1994)'On the which only limited postcranialmaterialhasyet beenpublished was more bipedal basis of describedmaterial it cannot be ascertainedwhether A' ramidus the hominid within than extant apes, and it may be that this taxon either does not belong anatomy' locomotor in lineage or that it representsa form prior to any major changes

422 R. A. Foley and S. Elton

3.1.2.Austraropithecus anamensis.This hominid is also over 4.0 Myr, and has a relatively primitive.cranialmorphology(Leakey et al., 1995).The assocrated tibia, however' has elementsthat link it with other early hominids, which may indicatesome level of bipedalism. 3'l'3' Australopithecusafarensis.This is the most extensively known of the early hominids; two lines of evidenceindicatethat it is a relatively bipedalhominid. The postcranium of AL-288, especiallythe pelvis, indicates mo.e habitual bipedalism(Lov'ejov, 1979)'while the footprint trail at Laetoli also suggests a bipedalhominid. Theseobservations lead to the conclusionthat at leastsomelevel of bipedalismwas presentrn the early hominids by shortly after 4.0 Myr' but there is considerabledebateas to whether this is universalamong hominids,is fully established, precludesother locomotorbehavior,and is different from that found in modern humans(Lovejoy, l9g0; Jungers, l9g2; senut and Tardieu, 1985)'The most likely consensusis that some;".i"iHivere activebipedaltyon the ground for some of the time' but that their overall body size and shapestill shareda number of featureswith African apes(seepapersby Hunt u.rd tuttl" et al., this volume). 3'l'4' Australopithecus africanus.In overall morphology this taxon is likely to be relatively similar to or less like that of modern humans tnai z. aJizrensis.Interpretation has been strongly inflt'encedby the completeinnominate stsl4, *tri"h rho* clear similarities with later bipedalhominids.A recentdiscovery of a pa.tiul foot showing a divergent hallux has led to some questioning of the extent of bipedalism in l. africanus, althoughthis position is complicatedby uncertainty concerningthe taxonomicintegrity of specimensassignedto this taxon (Clarke,l9g5). 3'1'5' AustralopithecusrobustusandAllies. There is little doubt that these more ro. bust and later (< 2.7 Myr) taxa are at leastas bipedal as the earlieraustralopithecines, and many elementsof their dietary adaptationshave indicated a predominantly terrestrial way of life (Susmanand Brain, lggg; Grine, l9g9) 3'l'6' Homo habitis.The evidenceof the olduvai specimenshas generallyled people to supposethat earlyHomo was fully bipedal, but this position has been called into question to some extent by reinterpretations of the oH36 ulna and the postcranialanatomy of oH62, both of which are relativelyape-likein morphology(Aiello and Dean, r990). 3'l'7' Homo ergaster.The partial skeleton wTl5000 from west Turkana provides the most completeevidencefor hominid postcrania for the plio-pleistocene,and although there are a numberof small differences,nonetheless it is clear that by 1.6 Myr at leastone lineage,as representedbyH. ergaster,is fully adapted,"u,p.a"ri# in ways that are not dissimilarto modernhumans(Walkerand Leakey,i19:). In overall terms, therefore,the paleontological evidenceshows that the context for consideringthe evolution of bipedalismlies during the plio-pleistocenein sub-Saharan Africa' At the beginningof the Pliocene,the remains from Aramis might be taken to indicate essentiallynon-bipedalor partially bipedalhominids. The early australopithecines are best thought of as being ape-likein morphology, but with a considerablygreaterset of adaptations for bipedalism than the living African apes. The later australopithecinesare probably fully terrestrialand bipedal, buiwittr some adaptivedifferencesfrom later Homo, while early Homo itself shows a mixed suite of adaptaiions.Full human bipedalism was presentby l '6 Myr' It shouldbe stressedthat the recognitionof muliipte taxa changesthe

d \. t.lron d : r . r sa B. iort l c r g 1e 1

l€ i'.trlv lc- irr)StO\ ij lO\'.

b . . ' r ra le llrlt 't:::.

is

lu:.lltd bt ', ,rn lar-i!i a Ini,

t:,,he B l . : i:' ) n f.:::]d:'.l:frt'i.i'

rr:, ,'l

Time and Energy: The Ecological Context for the Evolution ofBipedalism

423

way we should look at bipedalism.It is likely that during the pliocene,when the hominids were diversifying' so too was their locomotor behaviJr. The presenceof evidencefor bipedalismin one taxon doesnot necessarilyimply that it was unlversalamons all.

3.2. PaleoenvironmentalContext The paleoenvironmental contextof thesehominidshas itself beensubjectto considerablediscussion.An earlierconsensusthat all the hominids existedin relativelyopen environments has been somewhat eroded in recent years (see Reed, 1997, for a recent summary)' Earliesthominid sites from the Middle Awash and Hadar have all been interpreted as showing at least some tree cover, and recently the same interpretationhas been appliedto the southernAfrican sites.The earliesthominidsare now consideredto haveoccupied habitatsthat were not fully forest, but would certainly have contarnedat Ieastsome tree cover' The amount of tree cover is likely to have varied considerablyfrom location to locationand throughtime. Thereis little doubt that somehominids after 2.0 Myr were living in fully open,semi-aridenvironments.overall, the paleoenvironmental contextfor the evolution of bipedalismis thus Iikely to have beena mixed tropical African environment: savannain the sensethat grass would have been extensive,but also with significant levels of tree and bush cover,and in all likelihoodin relativelycloseproximity of water (seeTable l)' Theseenvironmentsare likely to haveresourcesthat weie highly seasonalin distribution, patchily and unevenly distributed, and occurring both in tre"esand on the ground. In addition,the level of thermalstressis likely to havebeenconsiderable.

3.3. PhylogeneticContext ,ra :,'S- ,::lJ il .i.,.

I n., d:: , l[r''] .

)r

All evolutionary changeis a function of the interaction between the existing components of a speciesand its current environment.Initial phylogenetic conditions thus play an important role in determining the nature of evolutionary .lung., for the existing phenotype provides the framework for the level of costs and benefitslmposed by any paiticutar strategy. Phylogenetic context is almost certainly one of the reasons why baboons and hominids have such divergent locomotor strategiesdespite very similar habitats, for the costs and benefits of changing the existing phenotypa would have been very different (Foley, 1987).with regardto bipedalism,there is a great deal of uncertaintyabout what might be the phylogeneticstartingconditions.Thereis virtually a completeatsenceof ap-

1 1l r l : .

lo-:'

[ , ' '€::,

Table l. Reconstructionsof the environmentsof early hominid taxa in Africa. Reconstructionsbasedprimarily on Reed (1997) Taxon

1I' B:.' ir:.: S .'f.. J ;-l|*r: tt .. J : : .

A. ramidus A. anqmensis A. afarcnsis A. afi'icaruts A. aethiopicus A. robustus A. boisei H. habilis + H. rudolfensis H. ergaster

Habitats Woodland I

Closedto open woodland,with edaphic grassland/shrubland/deltaic fl oodplain/waterand treesDresenr Woodlandand bushlandwith riverine forest Bushlandto open woodland/edaphicgrassland Open or wooded-bushedgrassland/edaphic grassland Woodlandto scrubland/edaphic grassland Open grasslands/dryshrublands/edaphic grasslands Shrubland/riparianwoodland/aridopen landscapes

424

R. A. Foley and S. Elton

propriately situated fossil material, so what is known is largely estimatedfrom modern humans,later fossil hominidsand extantapes.It hasbeenproposedthat the last common ancestor of the hominids and African apes was variously a knuckle-walker, brachiator or generalizedclamberer/climber(see Hunt, this volume, for a discussionof this problem). While it is difficult to test for thesevariousmodels,it would seemmore probablethat the ancestorsof the first hominids either approximatedthe more generalizedlocomotor behavior of Pan or were even lessspecialized.

3.4. Behavioral Context but Bipedalismis usually treatedas an anatomicalproblem by paleoanthropologists, the that end being an end, means to is a problem Bipedalism is behavioral. the heart at of an overall and the balancing food, and of mates predators, the acquisition of avoidance energybudget.Bipedalismwill evolve when the costsof moving in this way are lessthan the benefits. The key context is, therefore,the way in which an animal spendsits day; time is the 'hidden constraint in primate behavioral ecology' (Dunbar, 1992). lt is essentialto realizethat the key context in which bipedalism must be assessedis how a hominid would have budgeted its day under different ecological, phylogenetic and energetic contexts. Dunbar (1992) has developeda model in which a primate day is partitioned up into four activities-feeding, traveling, resting and socializing.Each of theseis essentialfor the satisfactory functioning of a population. Feeding and traveling time are the highest priority, but there are also limits on how much resting and socializing time can be varied in a typical activity budget. For example, among savannababoons the range of time budgets is: = feeding = 23-56oh, traveling = 17-360/o,resting = 540%o, and socializing 22*57o/o. Teleki (1989, quoted in Williamson, 1997)gives for chimpanzeesfrom Gombe a feeding time of 42.8%, traveling time of 13.4%,resting time of 18.90, and24.9oh socializing time. Williamson (1997) gives for both species of Pan and for different environments rangesof 29.147% for feeding, 13-27.5% for traveling, and3043o/o for resting/socializing. When consideringthe energeticsof bipedalism,it is for this scaleof activity budget that energeticefficiency has to be considered. The backgroundfor modeling the costsand benefitsof the evolution of bipedalism can thus be approximated as that of a generalizedape living in a relatively well-treed environment in Africa. This ape would be capable of both arboreal and terrestrial foraging and movement, and would have to acquire food efficiently acrossa patchy and hot environment. The problem is to model the energeticsof such an animal under different eco'cornmittedbiped'. logical conditions,to determinewhen it may pay to becomea more The novel element is placing this setting into the framework of a daily time budget such as would be expectedfor a typical social primate.

4. THE MODEL There are two componentsto the model, one related to time budgeting and one to energetics.The first component is a partitioning of a l2 hour day into percentagetime spent feeding and traveling. For baboonsand chimpanzeesthesetwo activities may acc o u n t f o r u p t o h a l f o f a d a y ( D u n b a r ,1 9 9 2 ; I s b e l l a n d Y o u n g1,9 9 6 ;W i l l i a m s o n1, 9 9 7 ) . As feeding and traveling are the most energeticallyexpensiveactivities, and the most affected by locomotor abilities, the time spent on these activities is varied in the model. Both feedins time and travelins time are varied between a minimum of 20Yoand a maxi-

lltu

ing ma ple mo cor cru fac

zee on tua me sur

mo chi

(le

hor

S. I lton

lr:: iru-

Table 2. Parametersused in the model, showing rangeof valuesand coefficients applied in the simulations

loir .inll\'l- ()f

Parameter

brl::r t.

Length of day (tropical environments) Body weight Feedingtime ranger Traveling time ranger Resting/socializingtime rangel Energeticcostsof feeding on ground (kJ/min)l

hr: lhc l(r:1r'-

Dl\..111

Energeticcosts offeeding in trees(kJ/min)l

in3:irc Or r'r.ill Energeticcosts oftraveling on ground (kJ/min)

!: :it.in ). i:rllr' l\1

Dl:.i

Energeticcostsoftraveling in the trees(kJ/min)

I rr ,.'..,.1 rlta\I>. It :,'ur hc ..rt-

Energeticcostsofresting and socializingactivities (all habitatsand species)(kJ/min)'

a : ' .n l -

' :\ B.-: l-:fecc::rS ial:,'::r:: nnlani\ Dc..i.r1b-c:ct td.r.:.::: RCC'::. Drr: :l! Dti:.'':-

n: ,, ,b: le .:' SU-^:.

| .'- -

I

t

Dt.' ,

r' I

-

Organism

modemhuman early biped quadrupedalape modem human early biped quadrupedalape modem humana early bipeds quadrupedalape6 m o d e mh u m a n T e a r l yb i p e d T q u a d r u p e d aal p e "

Value/range 720 mtn. 30 kg 2V0% 2M0% 4M0% 5.15 5.67 6.44 9.41 9.41 7.06 10.29 11.32 12.86 18.8t 18.81 l4.ll 2.91

'Ti'r. .ungestaken from Dunbar ( I 992). 2Energetic cost of leeding on ground estimatedat half the tenestrialtraveling rate. 'Energetic cost offeeding in treesestimatedat halfthe arborealtraveling rate. aEnergetic cost oftraveling on the ground calculatedusing data obtained Elton et al. (in press;seeTable 3). 'Energetic cost of traveling on the ground calculatedas I 0olomore costly than for the modem human. 6Energetic cost of traveling on the ground calculatedas 25olomore costly than for the modern human. 'Energetic cost oftraveling in the treescalculatedusing data ftom Elton et al. (in press;seeTable 3). nEnergetic cost of traveling in the treesestimatedas 75Yoof the modem human cost. 'Resting and socializing costs estimatedusing the modern human standing cost, using data from Elton et al. (in press;seeTable 3).

Jlrrl.i\.

! c ' D.r. .,

425

Time and Energy: The Ecological Context for the Evolution ofBipedalism

, . dl

s,

mum of 400lo(Dunbar, 1992;Williamson, 1997;Table2). Acontrol valueof 25o/ofor feeding time and20ohfor traveling'timeis usedwhen the other is being varied.A casecan be made that social behaviormay also be sensitiveto patternof locomotion (see,for example, Jablonski and Chaplin, 1993),but this is not pursuedhere. A further constrainton the model is that it only exploresstrategiesthat reducethe costsof locomotionin particular contexts; ways in which energeticbenefits through accessto additional resourcesmay accrue have not been considered,although a case may be made that this was a significant factor in the evolutionof bipedalism(Hunt, 1996). The secondcomponentof the model is the energeticsof the hominids and chimpanzees as a comparativeanimal for a quadrupedalape. There are considerabledata available on human energy expenditure(see Ulijaszek, 1995, for a recent summary). There are virtually none available on the energy costs of climbing, however, and therefore an experimental study was carried out on the energeticsof standing (which is used here as a surrogate for resting/socializing),walking and climbing in a mixed sex sample (see Elton et al., in press, for details; Table 3). Data on chimpanzeeenergy expenditure are scarce, most of which returns to an early study by Taylor and Rowntree (1973) on a juvenile althoughCaldwell et al. chimpanzee.There are no dataon climbing costsin chimpanzees, (1972) found that moving vertically was around double the energeticcosts of traveling horizontally.

R. A. Foleyand S. Elton

426

Ti

Table 3. Energeticsof the activitiesusedin the model. Theseestimates are derived from a mixed sex study involving controlled exercise(Elton et al., in press) Model Energy per kg Small biped (30 kg)

Standing kJ/min/kg

Walking kJ/min/kg

Climbing kJ/min/kg

0.097 2.91

0.343 t0.29

0.627 18.81

In the absenceof unequivocal data the following estimateswere used. Standing, walking and climbing costswere calculatedin kJ per minute per kg using the results of the study by Elton et al. (in press;Table 3). Energy expenditureduring feeding time was calculated at half the walking rate for terrestrial feeding, and half the climbing rate for arboreal feeding. This rate was selectedto reflect the probability that feeding would involve some movement on whatever substratewas being used. Energy expenditureduring traveling time was estimatedusing the walking rate for terrestrial travel, and the climbing rate for arboreal travel. Energy expenditureduring resting and socializing time was estimated for modern humans using the rates for energy expenditurewhilst standing. Overall daily (12 hour) energy expenditure is the sum of time spent in each activity. In the model, the percentageof time spent on the ground was varied from20%oto 100%. This affected overall energy expenditureas traveling and feeding time spent in the trees were more energetically expensivethan that on the ground. Three types of 'creature' were modeled: one that basically has the same rate of expenditure as a modern human, using the rates describedabove; a less efficient, early biped; and, a chimpanzee-like quadruped.The climbing costs of the less-efficient biped "modern human" rate, and the cost of bipedalism in this creaturewas 10%o were set at the more than that of the modern human (Table 2). The chimpanzee-likequadruped'sterrestrial efficiencywas lessthan that of a biped, following Rodmanand McHenry's (1980) results showing that chimpanzeesare around a third less efficient on the ground than are humans. A more conservativeestimateof a 25o/oincreasein energy expenditurewas used here. Conversely,it was assumedthat this creaturewould have been more efficient in the trees than a modern human, so the climbing rate for the chimpanzeewas estimatedat75o/o of the modem human rate. In the model describedhere a body size of 30 kg was used. This is a relatively small body size, and is appropriatefor some of the very earliest hominids and for chimpanzeefemales. Increasedbody size would, in these models, not affect the outcome other than by increasing overall energy expenditure across all model creatures (but see Steudel, 1994,for a discussionof the allometric factors relating to locomotion in primates).The main parametersof the model are shown in Table 2.

4.1.Results Applying the model to the creaturesdescribedabove shows how differences in activity patterns affect energy expenditure,and these can be used to explore the effects of time budgetson the costsand benefitsof bipedalism. 4.1.1.Feeding Time.As discussedabove,in practicalterms,using baboonand chimpanzeeanalogues,a large social primate can expectto feed for about 20o/oof its time when conditions are relatively good and to increasefeeding time to as much as 40ohwhen food is either of poor quality, takes time to process,or is hard to find. Figure I shows the effect

Fig exp cen but the wht

I t-lton

Time and Energy: The Ecological Context for the Evolution of Bipedalism

427

Earlybiped(30 kg) Modernhuman(30 kg) Quadrupedalape (30 kg)

n \ :r ng . o:-rhe l..'rlf art tr! \ ( \ ] \e

40

Fi':\ !'l-

60

80

30%feedingtime

lg l-.ite imlrcd

I J.rrlr bl. t:rc I t r , ,c r lr!fl l -

J

(, = =

5000

''-)..>_

tt g o CL x o

c', L

o tr

IIJ

3500 3000

40%feedingtime '"'")-

20 ) r::

hr

lc^ l

.i'

F ' I c:'

40

60

80

100

% Time on ground Figurel. Modeleddailyenergyexpenditureofbipedsandquadrupedalapes.Theverticalaxisshowstotalenergy expenditureacrossall activities (feeding,traveling,resting,socializing).The horizontal axis shows increasingpercentage of time spent on the ground. As time spent terrestrially increases,overall energy expendituredecreases, but that ofbipeds at a faster rate. All graphsuse 20o/o|jlaveltime. The top graph shows the situation when 20% of the time is spentfeeding (relatively good conditions);the middle graph when 30% is spentfeeding, and the bottom when 40% is spentfeeding (very harshconditions).

R. A. Folev and S. Elton

constant of increasing feeding time for the three model creatures,with traveling time held chimpanzee-like and bipeds, early (modern humans, at20Vo.As can be seen,for all three This quadrupeds),the costsof foragingdecreasewith increasingtime spenton the ground. (and feeding the of most When model. the in feeding reflects the greater costs of arLoreal enera considerable has chimpanzee the then trees, the in proportionately,travel) is done spent on the ground increasesthe gap between ietic advantage.ns the percentageof time the bipeds have an energeticadvanand eventually ihe bipeds and the quadiupedscloses, of bipedalism' the question is' the evolution for context tage. From the poini of view of the more than 60oh of activities is when answer The occur. at-what stage does the cross-over for a less efficient biped' 70"/o than more and biped, modern for a are spent on the ground and increasing the time)' (20% feeding biped inactive a relatively This is the case for but has little effect amount of feeding time raisesthe overall levels of energy expenditure, quadruped'Even a active equally an to relative occurs on the point at which the crosspver it spends on the less efficient biped will havlan energetic advantagewhen the time ground exceeds75%o.

4.2.Traveling Time (Figure 2)' The transiThe effects in relation to increasingtraveling time are similar occurs when over again to bipeds advantage to an quadrupeds tion from an advantageto and at'75o/ofor a efficiency, of levels human ior modern ground the 60,,/otime is spent on feeding time constant less efficient biped. Similarly, as travel time increases(now holding at20yo),overall levels ofenergy expenditurerise for all model creatures'

Time an

J

o

.: c o c ,( c

E c

E

u

4.3. Integrated Patterns critical zone, in what these results indicate is that, other things being equal, the occurs when a biped to terms of time budgets,for the advantageshifting from a quadruped may bring factors other terrestrial activity lies between 60 and 80% (Figure 3). Bringing in range of the for stable this value down, but the transition point doesappearto be relatively daily activities a typical social primate is likely to employ'

5. DISCUSSION data and a It should be rememberedthat theseare models basedon limited energetic impliinteresting firm application of the principle of uniformitarianism;nonetheless,some cations arise from consideringbipedalism in the context of time budgets. to bipedalism The first ofthese is thai the results hinge on there being an advantage activity is inarboreal as a mode of locomotionon the ground,and an increasedcost when Rowntree, (Taylor and volved. The first of these has been relatively well documented Robertson, and 1973; Rodman and McHenry, 1980; Foley, 1992; Steudel,1994;Leonard we do not know 1997).Climbing for a biped is energeticallyexpensive,but at this stage prinBiomechanical animal' whetherthis is very much greaterthan for a chimpanzee-like efficiently, more to climb able be ciples would seem to indicate that chimpanzeeswould costs,and risk of und tr,"y certainly have major advantagesin terms of speed,opportunity with bipedalassociated costs accidents.Given as an assumptionthe additionalclimbing activity proterrestrial of levels higher ism, it is interesting,if not unexpected,to note that mote bioedalism.

Figure i expendi centage but that of the tir bottom \

I some I getic a that wr can ac

429

Time and Energy: The Ecological Context for the Evolution ofBipedalism I l h,rn

Earlybiped(30 kg)

2oo/o travelingtime

! s :. : ' :

Modernhuman(30 kg)

R-..ri'

ape (30 kg) Quadrupedal

l.::l\1

!

e : 1 . ' :f\\

.J .J ll

;..;.;-_

f,".r't-

io:- .. ir:: -'. brl.'.i lg 1:1.' €::t.: .:

:\ ii'

J

tn :''

o = .= t tr o

CL

x o c'|

E.::'. ll ,'.I

at,

tr ul

54,)fl-'-,'

D n .. 3'r\

-i

y." ]tllr

20 I

I : : :.

100

total energy Figure 2. Modeled daily energy expenditureofbipeds and quadrupedalapes.The vertical axis shows perex"penditureacrossall activities (feeding,traveling,resting,socializing).The horizontal axis shows increasing decreases, expenditure energy overall increases, terrestrially time spent ground. As the on centage of time spent when20%o but th;t of bipeds at a faster rate. All graphs use 25%feedingtime. The top graph shows the situation and the traveling, is spent graph when 30% middle the (relatively good conditions); ofthe time is spent traveling bottom when 4070 is spenttraveling (very harshconditions).

':

1 ) . l " .

).-'. t . ' l. :Ic,." I r -. )tt'. lit .

80

% Time on ground

:" -l

B|j;

40

flt

The significanceof the model lies in the fact that when time budgetsare considered, some level of quantification is possible.A time budget setsconstraintson the level of energetic advantagethat is required.Although other things may have varied in the past in ways ihut *" cannot comprehendtoday, ultimately all diurnal primates are limited by what they can achievein a twelve hour period of daylight. A social primate must spendtime maintain-

R. A. Folev and S. Elton

430

8000

3 5

7000

E

6000

o

o ct x

o

5ooo

Et

o LrJ 4000

80

100

% Time on ground Figure 3. Summary graph. The upper limit for the quadrupedalape is set by a large male chimpanzeeunder conditions ofhigh energy expenditure,the lower limit by a small female chimpanzeeunder low energy expenditure conditions. The two lines for the biped set the outer limits to energy expenditure.The critical zone, derived from the model discussedin text, is where the switch to bipedalismis most likely to be beneficial.

ing social relationships,and an overly active animal is likely both to run into energy deficit and to increasethe probability of accidentsand predation,and hencethere are limits to the amount of tirne that can be spent feeding and traveling. Those limits set the real ecological constraintswithin which the costsand benefitsof bipedalismshouldbe considered. This model specifiesa link betweenthe habitat and the behaviorof early hominids that may go some way towards resolving the discussionsabout bipedalism and habitat. The clearest result shown here is that bipedalism does have a significant advantageover quadrupedalism,but only when a very considerableamount of time is spenton the ground. As long as the ancestralpopulations were feeding and traveling in the trees for as much as 40Yoof the time, then bipedalism will be more of a cost than a benefit. In other words, it would take a predominantly terrestrial life to lead to bipedalism,unlessother factors came into play. Furthermore, traveling time has a greateraffect than feeding time on the switch to bipedalism with increasedterrestriality,and thus if traveling time is high and feeding time low, then bipedalism might develop if most of the traveling is on the ground, as might be the casein scatteredwoodland. What implicationsdoesthis model have for our thinking on the evolutionof bipedalisrn among Plio-Pleistocenehominids? The first is that the major change is a behavioral one. The strategyof feedingon the ground is likely to be independentofbipedalism for a considerableperiod of time. As long as the ancestralpopulationswere still feeding at least half the time in the trees,then bipedalismwill be more of a cost than a benefit. By the time bipedalismdoes evolve then it is likely that the hominidswould have been well established as very competentterrestrialprimates. Although evolutionary change can occur very rapidly, it is likely that a long history of terrestrialactivity may precedethe first evidenceofbipedalism in the fossil record. The secondimplication is that it is travelingtime that is critical here.The model did not explore in detail minor variations in time spentdifferentially on the ground when feed-

Time

ing , criti crea and boor rest mor( thes dwe play and time espe time rang bud6

obse cons cost ben time cour migl tory and I

to di in th hom loco hom mod mate were gued may othe tent ' pithe relat thes anat ferre this. Hont aust later whic

|\. t lton

ts. F lc

rC.:.r: i:, :::a ) 1 . ' - ' . 'l )Ir. :..!:.

ba^:.,: 9c . i'r

p'

-:

lL.i

.r.

Of-.

:i

l5 ..::--J

s"\ :.:' [e.- : lll,l.

.:'

ipr .:., l\

'.:

, F ' , . lr , '

B. te I t - - . -

15: !c l : - - :

Time and Energy: The Ecological Context for the Evolution of Bipedalism

431

ing or traveling,but nonethelessit is the traveling time that is likely to be energeticatly critical. What might be significantis not the relative amountsof time, but the overall increasein the percentageoftime spentboth feedingand travelingat the expenseofresting and socializingtime. Dunbar (1992) has shown that in hotter and drier environmentsbaboons will spend more time traveling, and that this is time taken from that available for resting or socializing.Thus, a hominid living in a more arid environmentwould spend more time traveling in search of food and traveling between dispersedpatches.Overall, thesepopulationsare likely to be more active and to have larger day rangesthan forest dwelling apes.The hypothesisthat the energeticadvantages ofbipedalism are broughtinto play primarily by increasedday range lengthswas proposedby one of us (Foley, 199r), and has been supportedby some recentreanalyses(Leonardand Robertson,1997).These time budget models supportthe hypothesisthat the energeticadvantagesof bipedalism are especiallyimportant in the context of increasedday range lenglhs (Foley 1992),when time stresscan become a critical problem. It is interestingto note that a maximum day range length of around 16 km arisesfrom the day length model (Foley,l9g2), the time budgetmodel presentedhere,and Leonardand Robertson's(1997) analyses. A third implicationof someimportanceis that while bipedalismmay be the primary observablechange at the divergenceof hominids from the other African apes,it may be a consequenceof anotherand equally significantchange.Terrestrialityis less energetically costly than arboreality,but it still takestime to forageand, furthermore,food is likely to be more widely dispersed.The key change might therefore be the ability to spend more time in feeding and traveling behavior, in order to be generally more active across the course ofa day (Foley, 1995). This higher level ofactivity is the sort ofpressure that might also integratewell with the idea that early hominids have very high thermoregulatory costs and that thermoregulationhas played a major role in shifting hominid anatomy , 9 8 4 ,1 9 8 5 ,I 9 9 l a , b ) . a n d p h y s i o l o g y( W h e e l e r 1 Finally, it may be worthwhile retuming to the largertime scaleof hominid evolution to discussthe locomotor and paleoenvironmental evidencefor Plio-Pleistocene hominids in the light of these time budget models. As discussedearlier, the very early Pliocene hominid, Ardipithecus ramidus, has not been published in sufficient detail to determine its locomotor repertoire, but there is convincing evidence that the other three taxa of early horninids-l . anamensis,A. ffirensis, and A. africanus-were neither fully bipedal like modern humans, nor exhibited chimpanzee-likequadrupedalism.They perhaps approximated the inefficient small-bodiedbiped modeledhere.If so, it can be inferredthat they were likely to have been spendingat least 65%oof their time on the ground. It has been argued that they may have been living in substantiallywoodedenvironments,but while this may have been the case,it is unlikely that this reflects specializedarboreality.Put the other way, however,the level of bipedalismfound in the early australopithecines is consistent with as much as 35o/oof daily activity involving time spentin the trees.Later australopithecines, with greater robusticity and megadontic specializations,in addition to relatively good evidenceassociatingthem with more open habitats,are, on the basis of these models, expectedto be specialistbipeds,a conclusionconsistentwith the known anatomicalevidence.Homo ergasteris anatomicallyfully bipedal, and it would be inferred to have been a completeterrestrialspecialistwith a time budgetthat would reflect this. Perhapsthe greatestremainingareaof uncertaintyis whetherearly representatives of Homo, living around 2.5 Myr, would have still retainedthe pattern found in the earlier australopithecines, or whetherthey alreadypossessed the derivedcondition found in both later Homo or the robust australopithecines. Certainly models of land use at Olduvai, which have been linked to Homo habilis, imply a terrestrialway of life (Peters and

432

R. A. Foley and S. Elton

Blumenschine,1995),but, as this paperhas shown,terrestrialitymay be more widespread than bipedalism.

6. CONCLUSIONS Simulationmodelsshouldalwaysbe treatedwith greatcaution,for they simulateour the model developedand discussedhere does perhypotheses,not the past.Nonetheless, provide into the evolution of bipedalism.Stresswas laid on the some new insights haps to very specificcontextsthat the costsand for it is only in relation importanceof context, It was arguedthat the context for bipedalism be assessed. benefitsof any adaptationcan hominoid populationsdeployed particulaq in which extinct in the way was ecologicaland, By specifying the time their time budget. other words, their activities acrossa day-in possible into accountnot just the to take it was budget for a twelve hour set of activities ability. losing climbing but also the costs of benefitsof bipedalism, The time and energy model examined how the percent of time spent on the ground influencedthe adaptivevalue of bipedalismin relationto the relativeand absoluteamount of time spentfeedingand traveling.The principalconclusiondrawn was that at least60%o of daily activitieswould have to be spentterrestriallybefore the energeticadvantagesof bipedatismoutweighedthe loss of climbing ability. This result implies that, other things being equal,hominidsmay well havehad a long ancestryof terrestrialactivity prior to the evolution of bipedalism.Extensiveforaging on the ground,and traveling larger distances with greaterday ranges,even in relatively closedhabitats,may have been the ecological heritageof the first hominids,and the essentialpre-requisitefor successfuland ultimately bipedal adaptationto drier and more open environments.

ACKNOWLEDGMENTS We thank P.C. Lee for comments, Charles Fitzgerald for help with the computer models, and RAF is grateful to Henry McHenry and Elizabeth Strasserfor the invitation to contribute to the conference in Davis. Elizabeth Strasser,Kevin Hunt and a number of anonymousreviewers provided helpful commentson an earlier draft.

REFERENCES Aiello LC, and MC Dean ( 1990)An lntroductionto Human EvolutionaryAnatomy. London: Academic Press. Clarke RJ (1985) Australopithecusandearly Homo in southernAfrica. In E Delson (ed.): Ancestors:the Hard Evid e n c e .N e w Y o r k : A l a n L i s s ,p p . l 7 l - 1 7 7 . Darwin C (1871) Descentof Man and Selectionin Relationto Sex. London: Murray. D u n b a rR I M ( I 9 9 2 ) T i m e : A h i d d e nc o n s t r a i not n t h e b e h a v i o r a el c o l o g yo f b a b o o n s .B e h a v .E c o l . a n d S o c i o b i o l . 3l:3549. Elton S, Foley RA. and Ulijaszek SJ (in press)How much does it cost a human biped to climb and clamber?Ann. H u m .B i o l . : a t t e m so f H u m a nE v o l u t i o n a r yE c o l o g y .H a r l o w : L o n g m a n . F o l e y R A ( 1 9 8 7 )A n o t h e rU n i q u eS p e c i e sP F o l e y R A ( 1 9 9 2 ) E v o l u t i o n a r ye c o l o g yo f f o s s i l h o m i n i d s .l n E A S m i t h a n d B W i n t e r h a l d e (r e d s . ) :E v o l u t i o n a r y . h i c a g o :A l d i n e d e G r u y t e r ,p p . l 3 l 1 6 4 . E c o l o g ya n d H u m a n B e h a v i o r C F o l e y R A ( 1 9 9 5 )H u m a n sB e f o r eH u m a n i t y :A n E v o l u t i o n a r yP e r s p e c t i v eO. x f o r d : B l a c k w e l l sP u b l i s h e r s . " R o b u s t " A u s t r a l o p i t h e c i n e sC. h i c a g o : A l d i n e d e G r i n e F E , E d i t o r ( 1 9 8 9 ) T h e E v o l u t i o n a r yH i s t o r y o f t h e Gruvter.

' [.lton

prr-ad

le .rur t lcrm the F .ind blr>rn It'i .'d | :lnte Sl lhe Itr Lrit d ll('-;llt I f,l r",,

;e. ,rf ht::st k :irc' l !'. i !'s

tg.i,rl ]ai;lr

F ' . :i r ic:: :,, tc: ' ,i

E fc

tr.

)-

Time and Energy: The Ecological Context for the Evolution ofBipedalism

433

Hunt KD (1996) The postural feeding hypothesis- an ecological model for the evolution ofbipedalism. S. Afr' J. Sci.92:77-90. Hunt KD (this volume) Ecological morphology of Australopithecusafarensis:Traveling terrestrially, eating arboreally. ln E Strasser,JG Fleagte,AL Rosenberger,and HM McHenry (eds.):PrimateLocomotion: Recent Advances.New York: Plenum Press,pp. 397-418. Isbell LA, and young TP ( I 996) The evolution of bipedalismin hominids and reducedgroup-sizein chimpanzees altemative responsesto decreasingresourceavailability.J. Hum. Evol. 30:389197. J. Jablonski NG, and Chaplin G (1993) Origin of habitual tenestrial bipedalismin the ancestorof the Hominidae. Hum. Evol. 24:259180. Nature Jungers WL (1982) Lucy's limbs: Skeletal allometry and locomotion in Australopithecus afarensis. 297:676478. from Leakey MG, Feibel CS, McDougall I, and Walker A (1995) New four million year old hominid species Kanapoi and Allia Bay, Kenya. Nature 376:565-571. Leonard WR, and RobertsonML ( 1997) Rethinkingthe energeticsof bipedalism.Cun. Anthropol. 38:30ut-309. 50"460. Lovejoy CO (1979) A reconstructionofthe pelvis ofAL-288 (Hadar Formation).Am. J. Phys. Anthropol. . 5 2 : 2 5 0 . A n t h r o p o l P h y s . A m . J . b i p e d a l i s m . T h e r o l e o f ( 1 9 8 0 ) o r i g i n s : Hominid L o v e j o yc o Maynard Smith J (1978) Optimization theory in evolution. Ann. Rev. Ecol. Syst. 9:31-56. Hum. Biol. N e w m a n R W ( 1 9 7 0 ) W h y m a n i s s u c h a s w e a t y a n d t h i r s t y n a k e d a n i m a l : A s p e c u l a t i v er e v i e w . 42:12--27. peters CR, and BlumenschineR ( | 995) Landscapeperspectiveson possiblelandusepattemsfor early hominids in the Olduvai Basin. J. Hum. Evol. 29:321-362. R e a d e rJ ( 1 9 8 8 )M i s s i n gL i n k s . L o n d o n :P e l i c a n . P l i o - P l e i s t o c e n eJ. H u m . R e e d K E ( 1 9 9 7 ) E a r l y h o m i n i d e v o l u t i o na n d e c o l o g i c a lc h a n g et h r o u g h t h e A f r i c a n Evol. 32'.289-322. and JGH Cant Rodman PS (1984) Foraging and social systemsoforangutans and chimpanzees.ln PS Rodman pp' (eds.): Adaptations for Foraging in Non-Human Primates. New York: Columbia University Press, I 34-l 60. Phys. Anthropol Rodman PS, and McHenry HM (1930) Bioenergetics and origins of bipedalism. Am. J. 5 2 :I 0 3 - l 0 6 . lmplications for phyloSenut B, and Tardieu C ( I 985) Functional aspectsof Plio-Pleistocenehominid limb bones: Liss. pp. 193-201' York: Alan New Evidence. Hard (ed.): The Ancestors: geny and taxonomy.In E Delson steudef K ( 1994)Locomotor energeticsand hominid evolution. Evol. Anthropol. 3:4248. the gait of Paranthropus Susman RL, and Brain TM (l98St New first metatarsal(SKX50l7) from Swartkransand robusttts.Am. J. Phys.Anthropol 77:7 16. more energy Sclence Taylor CR, and Rowntree VJ (1973) Running on two legs or four: Which consumes I 79:t86-187 . evolution of homiTuttle RH, Hallgrimsson B, and Stein T (this volume) Heel, squat,stand,stride: Function and Renoid feet. ln E Strasser,JG Fleagle, AL Rosenberger,and HM McHenry (eds.): Primate Locomotion: pp.435448' Press, York: Plenum New cent Advances. University Press' Ulijaszek SJ (1995) Human Energeticsin Biological Anthropology.Cambridge:Cambridge University Press' Harvard cambridge: Skeleton. (1993) Nariokotome The RE, Editors and Leakey walker AC, Hum. Evol' Wheeler p (1984) The evolution of bipedality and loss of functional body hair in hominids. J' 1-i:91-98. body form and Wheeler p (1985) The loss offunctional body hair in man: The influence ofthermal environment' bipedality.J. Hum. Evol. l4:2328. hominids. J. Hum' Wheeler pE (l99la) The influence of bipedalism on the energy and water budgets of early E v o l .2 1 : l l 7 1 3 6 . environments: Wheeler pE (l99lb) The thermoregulatoryadvantagesof hominid bipedalism in open equatorial The contribution ot increasedconvective heat loss and cutaneousevaporative cooling. J. Hum. Evol' 21:lO'7-116. Aramis' White TD, Suwa G, and Asfaw B (1994) Australopithecusramidus,a new speciesof early hominid from // I 2. -l :306-3 Ethiopia. Nature Williamson D (1997) Hominid Socioecology.Ph.D. Dissertation,University College London'