LAND US E I N PRE-COLUMBIAN AMERICA by PEDRO ARMILLAS Museum of Anthropology, University of Michigan, Ann Arbor, US A EARL Y STAGES Human occupation of th e Western Hemisphere is comparatively recent, h u t much older than wa s aecepted a few deoades ago. Th e spread of man over North Amenea during t he last (Wisconsin) glaciation is well attested, au d th e terminal Pleistocene age for remains found in southern South Amerloa is, likewise, heyond question. Th e gradient ofthe known carhon-14 dates au d other considerations Bupport t1 generally held view o í a u -earIy drift migration from north to south, through the Central American funneI au d along th e Andes, fanning ou t into th e Argentinian grasslands an d Brazilian parklands, aIthough well substantiated findings oí corresponding antiquity have no t heen made in th e intervening area so far, no douht through want of systematic search. I t is generally assumed that th e early Americans avoided th e jungle regions, which would have heen penetrated b y ma n only with th e development of adaptative technologies. Th e presence of man in America during th e last glacial epoch, and in terms of absolute chronology since 12,000 years ago; is proven geologically an d palaeontologically on th e basis of numerous findings and dated b y sufficient c a r b o n ~ 1 4 tests. On th e basis ofunquestionable archaeological evidence, some North American hunting cultures ar e certainly earlier than that date an d an antiquity even greater is no t i m p r o ~ hable for some of the oldest known South American hunters, whose traces have been found in recent years in th e Argentinian grasslands. Furthermore, a few c a r b o n ~ 1 4 tests would make th e minimal antiquity of man in this continent considerably older: possihly 37,000 years ago [91, p. 58]' fo r fires burnt in basin-shaped hearths near Lewisville, Texas, with bones. of th e American camel an d horse, mammoth, an d bison of an extinct species, an d tools; a similar assemhlage of camp remains, with different yielded an antiquity 'oIder than 23,800' for a site, Tule Springs, near • • . 'natural resources' are in ¡act cultural appraisals. SAUER La s Vegas, Nevada. Whilst there are good reasons to view such very ol d datings with caution until corroborated hy additionaI evidence, a minimum date of 25,000 years before current times for th e first entry ofman to th e Ne w World seems quite likely on various considerations. Th e presence of man in th e extreme southern part of th e continent for a t least 9,000 years is dated b y th e carbon-14 test for th e surfaee of th e layer contain . in g th e oIdest remains of occupation in the Palli Aike cave, i n th e region of the Strait of Magellan. These remains, an d others found in th e same area, reveal a culture of hunters of th e American horse, ground sloth, an d guanaco. On th e hasis of typological correlations an d geological dating estimates, th e antiquity of th e earliest traces-lithic industries-of ma n southward in Patagonia seems to be even greater. There is no douht that th e first people t o move into th e Ne w World through th e Bering gateway were in th e predatory-foraging stage of economy-as were, b y th e way, their followers oflater times-and we may add, on th e hasis of what is known of th e prehistory of north-eastern Asia, that they possessed th e knowledge of generalized Ol d World Palaeolithic technologies for th e exploitation of land resources, rather than specialized equipment [20] (hut see also Tolstoy [79]). Th e diverse palaeo-Indian cultural traditions whose pr o files were hecoming quite distinct-eertainly so in North America, an d probably so in South America as well-even before 10,000 B.C., appear to represent indigenous developments out of that ancestral generalized hackground. I n N orth America, an antiquity of weIl over 11,000 years (Iater in th e east) is indicated for th e origins of the big-game hunting tradition, which extended mainly east of th e Rocky Mountains, to th e Atlantic seaboard an d froID Canada to central Mexico, at least. Also, in th e south-west of th e present United l. Th e figures in brllckcts refer to the bibliography at the end of the chapter. 255
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
States and in the intermontane area, a few findings
seem to indicate that big-game hunting might have
heen important there before th e dcsiccation of th e
area. In th e plains an d th e south-west, prohoscidians,
camel, and horse, were at first normal prey; Iater, they
might havc beeome searee, for th e eeonomy of the
Folsom hunters (at about 8000 B.C.) seems to have
depended mostly on the pursuit of Bison antiquus.
In the Atlantie region, mammoth anq. the otherspeeies mcntioned were apparently rare, instead
elk, deer, an d hear were hunted. In central Mexico,
t1 e early hunters ehased Archidiskodon imperator,
roaming in Iarge numbers hy the shores of th e lakes
in about 9000 B.C. In th e rest of North America
the Columbian mammoth (Parelephas columbi) was
more common.A different tradition, which includes a large variety
of cultures, is noticeable before 8000 B.C. in th e
Plateau-Great Basin area, an d becomes more defined
from 6000 B.C. onward. It s distrihution reaches
from th e eastern foothills of the Rockies to the Pacific
coast an d from Oregon to Mexico. Th e common charac
terization of this pattern is very generalized, an d bestseen in contrast to the concentration on big-game
hunting formerly descrihed. This occidental tradition
was based on diversified food-gathering, adaptable
to the -maximal utilization of available resources,
plant as well as animal. Ecological adaptations to agreat diversity- of environments produced a variety
of economies, including adjustments to arid country,l
and to riverine, lacustrine, and littoral habitats.
Th e oriental Archaic tradition, similar in its
approach to the natural resources, tending to broad
diversification dependent upon local eonditions, was
also developing east of th e plains, towards thc Atlantic
shores, during the 8000-6000 B.C. period, succeedingthere th e early h1:lnting pattern.
During the period of Altithermal climate, the
Great Plains became a zone of extreme aridity,
indicatedhy th e formation ofloess deposits in Nebraska
hetween 5500 and 2000 B.C. Much of th e area nrighthave been deserted and other parts very sparsely
populated, for archaeological evidence of occupation
however, must be considerahly older. In fact, at that
comparatively late date indireet evidence-that
will he rcviewed in the next section-suggests that
experiments in eultivation and the production of
cultigens had already hegun in highland Mexico and
Guatemala. I f so, the Chalco culture economy must
have combined incipient farming an d food-gathering.
In South America, th e paralleI cultural developments
of early post-Pleistocene times are still littIe known.Nevertheless, a mixed hunting-food-gathering pattern
seems to he represented by the Ayampitín culture.
Its vestiges were first found in the sierras of Córdoha
and San Luis, in north central Argentina; lately, it
has been reported from th e north-west of that country
and around San Pedro de Atacama, in northern Chile.
A putative date around 6000 B.C. is currently attn
buted to that culture.
Man has ha d a long time to adapt bimself culturally
to varied and changing environments, and to modify
by bis actions the American landscapes. However,
although the activities oí pre-agricultural human
groups, living by hunting and food-gathering, may
affeet natural resources,2 the adaptation of primitivepopulations to lile on arid an d semi-arid lands, wbich
is our concern here, seems to be rather narrowly
ecologicaIly conditioned, specially in the case of
pre-Columbian America, where th e p-astoral wa y
of life never developed as sueh, fór th e distrihution
of tamed grazing anímals was limited to the Andean
arca an d was there subordinated to intensive agri
culture.
It was with th e growth of cultivation an d th e im"provementof agricultural techniques that th e American
Indians hecame capable of mastering many of th e
arid and semi-arid zones of the eontinent, an d in
two independent cases to huild, upon th e foundations
of intensive agriculture, civilizations comparable to
t10se of th e pre-Classic Old World : Mesopotamia,
Egypt, th e Indus, and north China.
DEVELOPMENT OF CULTIVATION
is lacking or rather obscure. With the subsequent 1 he appearance of cultivation an d particularly its
improvement of ecological conditions, when the logical, hut not immediate, consequence, th e achieve
modern type of climate hecame prevalent, camp ment of estahlished agriculture, signalled a profound
remains indicate a re-orientation of th e economy change in th e relationsbip of man to land. The convertowards greater exploitation of plant resources and
small game-reflecting, prohahly, a comparative
scarcity of hison-although hig-game hunting was
always an important part of the subsistence hasis,
and a dominant one in some localities or at someperiods [81, p. 85].
In - central -Mexico, th e food-gathering culture
upon wbich plant-domestication could have developed,
is helieved to he represented hy the still ill·definedChalco industry. A carbon-14 date places this complex
of lithic tools at ahout 4500 B.C.; it s heginnings,
256
1. Increasingly arid eonditions, rcsulting in tb e desiceation oflakes, aud eb:eaman d wind erosion, cte., bceame prevalent in tb e Great Basin, th e aouth-west,
and southern California about 5000 D.C. (Xerothermio or Altithermal phase,lasting in force to about 2000 D.C., when climate beeame somewhat cooler
and moister, comparablo to th e present time); hence th e generio name of'Desert Culture', which is nowadays p r e C e r r e ~ for th e tradition. However,since th e extension oí the pattcrn trespasses th e limits oC th e arid zonc,sorne less eommitting designlltion JD.ight be advisable. On th e definitionand extension oCtbe' Desert Culture' see [38, p. 69-72; 41, p. 276-280; 61].For the tatter-day survival DE the dcsert Cood-gathering economy oC th eGreat Basin, see Steward's [76] cIassic ecological study.
2. Namely, man's possihle contribution to th e extinction oí arge, roving, slowbrccding anllnals_ Also, 1 think oC tb e oCten debated question oíthe efi'eets oíme on the extension oC grass-lauds; ío r sorne tecent statements pro an d conand supporting bibliogrsphies on this subject, see [71, p. 12-18; 72, p . 54-56;77, pauim, specially p. 129; 82].
an d some sort of runner hean, perhaps, wild, were
found in the lowest strata of refuse. Two c a r b o n ~ 1 4 dates place these remains in th e seventh millennium
(M-49B an d M-500, [24]).
In th e following Ocampo Phase th e importance of
hunting as a supplement to vegetal food diminished.
The diet depended mostly on wild pIants, hut to the
cultivated species listed aboye common heans were
added. Direct carhon·14 dates for Ocampo matt{rialrange from 3700. to 2600 n.c., an d c r o s s ~ d a t i n g with
th e Sierra de Tamaulipas sequenee o.f artüacts (see
heIow) support the pIacing of this phase in th e third
and fourth millennia.
Panicum, Amaranthus, an d a primitive type of
maize related to that ofBat Cave (see helow) increased
th e numher of cultivated plants before 2000 B.C.(Flaceo Phase, one carhon-14 date 3947 ± 334 B.P.),but these people were still primarily gatherers of
wild plants, with very little hunting an d snaring as
supplementary economic activities. However, during
the succeeding Guerra Phase2 cotton and perhaps
C. moschata were introduced,3 and very soon, at
about th e middIe of the second millennium, a g r i ~ culture with a numher of varieties of maize, teocentli
(= teosinte, Euchlaena mexicana, prohahly as aweed, cf. DressIer [25, p. 150]), common, lima an d
runner heans, C. pepo an d C. moschata, gourds, an d
cotton, hecame established as th e hasis of th e local
economy with th e ensuing Mesa del Guaje Phase,
at th e same time that pottery appears for th e firsttime in that archaeoIogical sequence.
1. In hill clanio Whaf Happ6ned in Hútory (cbaptcr 11[: Neolithicbarbarism').2. Oddly dated by a single radiooarbon test over 2750 B.C., to be adjusted
perbaps to about 1900·1600 B.C. (cf. Crane and Griffin [24, p. 1104J dating oCsBmples M·504 witb M·567, and comments on M-50S; a1so MacNeish 152,p. 199].).
3. While, perhaplI signiñcantly, neitber Panieum nor Amaranlhu& are mentioncd in tbe published 8ummary [52, p. 168]; improved maize may havercplaced tbem [see 71, p. 72].
In th e Sierra- de Tamaulipas, a detached monntainrange rising ahoye th e Gnlf oC Mexico coastal plainin th e -south central section oC the State, th e Nogales
culture, still based upon a Coraging economy, succeeded
-wi th an apparent time gap which would correspondto th e Infiernillo Canyon earliest phase-the Lermaprimarily hunting culture mentioned aboye; however,
i t
was wild plant life which constituted mostoC
theNogales people's food-supply, hunting heing only asupplementary economic activity. It seems that thereis no!)concrete" evidence Cor cultivation during thisphase, albeit-because oC the continuity oC culturaldevelopment with th e Collowing one and by parallelismwith early Ocampo-MacNeish thinks that cultivatedplants ma y have contrihuted a very small proportionoC the people's diet in the last part oC th e period.
The ensuing La "Perra Phase, seemingly cross-datingwith late Ocampo oC the Sierra Madre sequence,
certainIy represents a stage oC incipient Carming, withmaize an d C. pepo constituting maybe from 5 to 10pe r cent oC th e Cood consumption, th e remainderheing supplied hy wild plants (ahout 80 pe r cent) and
th e products of hunting and insect catching. Th e cornremains found belong to two now extinct varieties,classed by Dr. Mangelsdorf of the Botanical Museum oC
Harvard University as Primitive Nal-tel A and Btypes; these are pod-pop corn, with two-inch longeight-row cobs having each kernel of every row surrounded by pod-like leaves, and seem to be ancestralto lowland adapted Middle American maize of latertimes. A radiocarbon test dates La Perra Phase atahout 2500 B.C.
Th e Almagre late sub-division of the La PerraPhase, may have lasted well into th e second millen
nium. A blank follows in ou r knowledge of thissequence after which, the Laguna Phase, dated by
correlation with th e Huastec area to the south atahout the middle of the fust millennium B.C. [52,
p. 172 and p. 198], represents a tradition of fnllyestablished agriculture, evidently introduced to thisregion from th e south. Not only in the Sierra properbut also on the level land south and west of thesemountains, villages of 200-400 circular" huts and ~ t h ceremonial mounds arranged around plazas, revealthe new social pattern-which is known to he mucholder in central and southern Mexico an d in Guatemala.Three new races of maize, C. ficifolia, an d Manihotdulcis, appear for the first time in that area. Th eraces of maize are 'modern' types showing evidence ofEuchlaena introgression; they must have heen deve
loped outside th e area and introduced to the Sierrawith other innovations such as pottery.l M. dulcis,
a cultigen oC generally undoubted South Americanorigin, appears here for th e first recorded time reachingthe northern limit of it s pre-Columbian distrihution(cf. [70, p. 508; 71, p. 62]).
In spite of th e fact that there are still some blanks,uncertainties, an d incongruities to he smoothed out
258
in these sequences, together they constitute the mostdetailed case history to provide a firm foundationupon which to huild the study of th e origins of nativeAmerican agriculture. .
Evidence from caves in west central New Mexico
indícates that th e period from perhaps the Courthmillennium to a few centuries before th e heginning
of th e Christian era was one of slow shift from thefood-gathering economy of th e Cochise 'desert'culture to a sedentary way of life based on cultivation.Primitive maize an d C. pepo were added to th e subsistence hasis by at least 2000 B.C.,2 as an innovationderived seemingly from Mexico, and around 1000 B.C.
'modern' races oC maize-hybridized with Tripsacumor Euchlaena-and th e common hean appear. Established agriculture, however, díd not replace th eancient food-gathering pattern until about 300 B.C.,
coinciding with th e appearance of pottery [38,p.74-78].
It is generally accepted that th e introduction ofcultivation in Tamaulipas an d New Mexico indicatesan even greater antiquity of it s practice in th e high
lands of central and southern Mexico an d Guatemala.Th e hotanists agree that the last mentioned regions
are th e hearth where ~ a n y of the most importantAmerican cultigens were developed. Of th e plantsmentioned in the previous paragraphs, th e origins
of Phaseolus vulgaris, P. lunatus, an d P. coccineus
(or multiflorus), an d C. moschata, seem to he certainIytraced to that area. Two species oC Amaranthus,
A. leucocarpus an d A. cruentus, are tentativelyassigned to central Mexico and th e Guatemalanhighlands respectively [25, p. 153]. Th e case forC. pepo is dubious; it is domÍnant in the cool highlandsof Mexico, where it is a staple Cood, hut it is morediversified in the eastern United States [71, p. 67].
Considering its very early occurrence in Tamaulipas,i t seems that the possihility oí an origin to th e north hasto be discarded, since, on the basis of our presentknowledge, cultivation hegan much later there; thearchaeological" evidence, then, adds weight to Dressler'sstatement [25, p. 131] that th e endemic forms found inMexico anel Central America should he consideredwith regard to the origin of this cultigen.
The case for Zea mays is no w seen in a new light.Th e discovery oí polI en, apparently oC maize, at greatdepths in the subsoil oí Mexico City, would indicatethat a wild ancestor of the cultivated species grewin the highlands of Mexico (and prohahly Guatemala
l. Th e preceding summary is based on MacNeieh [49, 50, 51, 52].
2. Incipient cultivation of these plants is represented by remains found at th eBa t au d Tularosa caves.
At Ba t Cave, th e aucce6sion of corn aamptes staita with tho tiny ORre oí avery primitivo pod-pop type oímaize. M a . a g o l ~ d o r f [ S 3 , p. 409] says' althoughthis ellrliostBatCave m me may have been cultivated, t was oo t ía r removediu ita botanical characteristics írom wild maizo'. The hyhrid coro mentiouediu the text appeared in upper levela of tbe excavation: Tbere is a series ofcarboo-14 date. 'presumllbly conelatod witb tb e dovelcpment oí mllizeculture' [47. p. l11J. beginning c.4000 D.C. However, tho validity oí theoldest dates is not yet generally acceptod. Tb o specialists in tb e area dooo t commit tbomselve!l more tban to montion tb e minimum ago of 2000 n.c.,or ¡eave it vaguely botween 4000 sn d 2000 D.C.
too) at th e time when man might have begun experiments in cultivation [25, p. 150-152; 4]. This, andthe late advent ,?f th e cultigen to the Peruvian coast(see heIow), which leads us to discard th e hypothesisof a South American origin, make highly probableit s domestication within the Middle American nucleus.
The mixed food-gatheringjincipient-farming eco
nomies have not yet heen identificd with certaintyin the hearth oí secd culture (but see my remark aboye,on the Chalco food-gathering pattern). Dated remainsfroID central Mexico to Guatemala indicate a minimumantiquity for sedentary farming communities between1500 and 2000 B.C.; circumstantial evidence points toth e third millennium for the heginnings of estahlishedagrlculture, as distinct froID incipient cultivation.
Along th e arid Peruvian coast a good number ofmiddens, sorne oí them attesting hy the accumulationof refuse to a prolonged penod of abode, reveal anoccupation by a people with a mixed economy basedon fishing, gathering (shell-fish an d wild plants)an d an incipient cultivation (probably in m'oist areasat the valley mouths, [see 22, p. 19]; manne mammals
formed an ':lppreciable complement of th e subsistencebasis, but there is no evidence of hunting land animalsin th e northern sites [11, p. 24; 78, p. 41] and ouly alittle elsewhere [26, p. 134]. In area, their presentlyknown range extends from th e neighhourhood ofPacasmayo (Iatitude S. 7°20') to past th e OcoñaRiver (Iatitude S. 16°30') [8, p. 118; 26, p. 73-85];in time, from about 2500 B.C. (based on c a r b o n ~ 1 4 dating) to the latter centunes of th e second milIenniumwhen, after th e introduction of maize,l archaeologicalevidence from th e only w e l l ~ s t u d i e d oasis, th e VirúValley, indicates a revolutionary change in foodproduction, a shift of population away from th e oldfishing stations into sections of th e lower valley
more favourable for estahlished agriculture (probablybased on river Hoodplain farming) an d an increasein nurnhers-a trend gaining momentum in the earlycenturiea of th e first milIennium B.C. [86, passim,especially p. 390].
The record of cultivated an d possihly cultivatedor semi-cultivated planta during this period at Huaca.Prieta, on th e Chicama Valley shore line, includes :
cotton, gourd, squash,3 chile pepper, jackbean (Canavalia ensifoTmis) , tw o varieties of pacae or guaba(rnga sp.), achira (Canna edulis). Truly wild eatentubers, roots, an d fruits were : 'papas de junco'(Scirpus ripanus or americanus) c a t ~ t a i l , a sedge(Cyperus sp.), lúcuma (Lucuma obovata), 'ciruela
de fraile' (Bunchosia sp.), and guayaha (Psidiumguajaba). A few of the plants found are no t identifiedfo r certain but maize remains were conspicuouslyahsent from this assemhlage [11, p. 24].
While th e techniques of cultivation-that aIlowed
th e pre-ceramic pattern of mixed subsistence economyto develop-must have come from the north [cf. 71,
p. 44], from north-western South America an d
Land use in pre-Columbian America
beyond, Engel makes an interesting suggestion toplace th e ancestral horne of these coastal people onth e ' lomas' ecological zone, between Chala an d Arica[26, p. 142].
A striking conclusion stands ou t of th e availablefactual information on th e remote prehistory ofnative American agriculture, that th e penod of
experimentation might have been a long one, withmixed food-gathering/incipient-farming subsistenceeconomies persisting long after plant domesticationbegan. A truly sedentary way of life based on farming,was only achieved when th e development an d diffusion
of high:-yield varieties of maize induced ma n to devotehimseIf f u l l ~ t i m e to th e pursuit of cultivation [12,
p.21].
THE GREAT PLAINS
In th e n o r t h ~ w e s t e r n Great Plains of North America,th e hunting-gathering pattern mentioned in the introduction persisted, without important modifications
of th e ecological balance, until it was altered by th eintroduction of th e Old World horse an d th e development of th e highly mobile methods for equestrianhunting and warring (eighteenth century).
In th e central plains and on th e middle Miasouri
River basin, horticulture (the maize-beans-squashcomplex) was added to th e h u n t i n g ~ g a t h e r i n g basissince about A.D. 500, as an intromission of th e eastern'woodland' agricultural pattern into th e westernfringes of th e prairie zone. A shift of th e agriculturalfrontier eastwards to th e ninety-ninth meridian,evidenced by archaeology for late prehiatoric times,might as well be due to climatic Huctuations as to otherundetermined factors [81, p. 89] (see also [80])
However,i t
wa s th e spreading of th e mounted nomadilison-hunting pattern, militarily onented, that mostprofoundly affected land use in this region. Th e hortic u l t u r ~ village communities dwindled in nurnher,
l. Tb e bcginnings of maize cultivntion on the north const dnte Íl:ODl no Interthnn 1200 D.C., on th e bnsis oí ita nllBoclntion with Chnvin horizon styIecernmica [87, p. 355]; plnin potter¡ appears first in th e axchaeologicnlsequence in the for:mer half oí th e second millennium, bu t coro ia a1ll0 to heíound in deposita oí this age. On th e central coast mnize might slightlyantecede th e appearancc oí cernmics (n t Aspero, near Supe ¡see 88, p. 151;26, p. 78]) 01' nppear together with undecornted pottery at Chirn·Villa,near Lima [see 26, p. 80]); although neither its inception no r that oí pluinearthenware hnve heen precisely dnted there, sequential evidence makesth e earliest maize anterior to the spread oí th e ChaVÍD ¡¡tyIe. What is sign{_ñeant, however, is not just the introduction oí the new plant bu t tb e eonver·sion oí the economy to established agriculture, allowed by ita hlgh yield;this momentous change definitely corresponds with th e horillon oí th eChavin ceramies (1200 to 700 B.C. nccording to carbon·I4 dates).
To judge from Engel's list oí eultivated planta in ·prc.eCl'nmie· times[26, p. 145J it seems that th e common hean (Phanolu$ vulgarb) ma y hnveheen introduced at th e same time as muize, aIthough Collier [22, p. 23]places it Jater in th e sequenee.
2. Although ou r present knowledge is Vf:r:y sketchy, lo r wnnt ol exhnustivemethodieal studies sueh as those made in Viró Valley, it is m a n i f e ~ t that~ i m i l a r changes took place at th e ¡¡ame time elsewhere along th e Peruvianconsto
3. For the pre.cera.mic period Engel [26, p. 145J liste: Cucurbila jirifo/ ia andC. moschala; C. maxima appearitlg with th e oIdest pottery nnd mme,C. pepo tnrdily (see also Snuer [70, p. 504]; Dressler [25, p. 130]).
Remnins oí C. jicifolia have been uncovered in ~ o o d qunntity at HuncaPrieta, to judge from a remark by Sauer [71, p. 66j, quotiog Whltaker andBird io Amer. Muuum Nod'ales, No. 1426, a paper no t avnilable to meat th e time oí writing.
eighteenth century, trihe after tribe' were joining
in the' mounted carrousel.
Mter th e horse, firearms (and indiscriminate
huffalo hunting, ending with th e almost complete
extinction of t1 e species), open range cattle-raising,
barbed wire and th e plough, th e prairie schooner and
th e iron horse-with th e white ma n hehind all t h a t -
in rapid succession changed th e face of this landscape
in th e course of the nineteenth century, as described
by Professor Logan in a later chapter.
THE COLORADO PLATEAD
area [15]. Whatever the causes may be, th e a b a n d o n ~ ment of th e hearth of t1 e Pueblo territory at the end of
th e thirteenth century is an archaeologically proven
fact. Henceforth, th e Pueblo Indians concentrated
south an d s o u t h ~ e a s t w a r d s , in zones where different
physiograpmc f e a t u r e ~ !llad.e the farmIand less destruc
tible by guIlying. Q:v'\ II}-'CJ..t;;I.
The erosio.ty hyPothesis has been contested by
O'Br an [60], who prefers to explain th e desertion of
th e San uan area as resulting from the harassment
exerted by marauding Athapaskan invaders; however,
there is littIe archaeological evidence of widespread
violence by nomadic bands against the sedentary
people.The eomponnded elfeet oí prolonged drought and
Th e Anasazi (Basketmakers and premstoric Pueblo) loss of farmland through erosion, as th e primary
Indians lived in the Colorado Plateau physiographic causes, raiding by the nomads (Ute or Navajo) on the
province, an d extended east into th e Rio Grande outlying settlements, pillaging of crops, an d so forth,
VaIley, as their- contemporary descendants (the 'lbs a force contrihuting to aggravate the unrest, and
Hopi, Zuñi, and Rio Grande Indians) still do. t'intervillage rivalry, intravillage feuding; an d perhaps
The climate of that area ranges from arid (with malnutrition and disease, as secondary or derivative
less than 250 mm. of mean annual precipitatioIi an d factors, is the explanation favoured bxBrew [13, p. 298]
mean annual temperatures above 11° C.) below'> and KeIley [39, p. 384], and th e o n e ~ e present1,800 m., to semi-arid between 1,800 m. and 2,200 m. wri te0eing an outsider, finds intellectually more
(precipitation between 250 mm. and 450 mm., tempe- satisfactory.
ratures averaging hetween 110 C. and 70 C), an d The hearth of the Anasazi tradition líes in the San
humid and cool at higher elevations. The vcgetation Jlian River drainage, in north-eastern Arizona, south
changes with tlie altitude, from desert scrub in th e eastern Utah, south-western Colorado, and north
lower climatic zone, to sage-brush, grass, and juniper western New Mexico. There is no evidence of agricul
an d pinyon trees in th e intermediate elevations, ture in that area until th e beginnings oí th e Christian
an d to pine and Douglas fu, or spruce in th e highest era (Basketmaker 11 phase), at first with maize and
mountains. Cucurbita moschata, to which Phaseolus vulgaris
Below about 2,100 m. rainfa ll is generally too seanty were added in Basketmaker nI imes (c. A.D. 400·700),to mature crops, and in the moist zone at high altitudes while P. lunatus and P. acutifolius varo latifolius
th e growing 5ea50n is too short;l crops are grown seem to have entered stilllater [16, p. 66 and p. 72];
most successfully in climates between th e extremes on archaeological evidence for beans in some sites
[14, p. 445; 31, p.8]. oí the San Juan area see Appendix C, by VoloeyThe climate doea net seem to have changed mate· H. Jones, in [13]. Although cotton was traded
rially during th e period which concerns us here (see into th e Anasazi countxy already in Pueblo 1 times
Brew [13, p. 5], for arehaeologieal evidenee), although (700-900), eotton growing seems to have begun in th e
occasional spells of prolonged drought, recorded north central Arizona section of t1 e territory by about
by the tree-rings, have worsened temporarily th e 1000, and spread from there to other parts in late
normally preearious situation throughout th e semi·arid Pueblo n and In phases [16, p. 80; 38, p. 76].country. Cucurbita pepo was introduced to that area at about
r- In one period however the devastating effects of the same date, and also spread from there over th e
;'recurrent droughts seem to have produced lasting Pueblo country [16, p. 20].'results. The succession of consecutive dr y years With regard to farming systems, we shall turn
which occuned over the plateau area from A.D. 1276 to ethnographical information suppleménted, when
to 1299-and ha s been recorded, too, in th e Rio possible, by archaeological evidence, to see them in
Grande Valley, with dates 1269-1296-appears to have bistorieal perspective.been merely t1 e climax of a long period of defective In th e zones of higher rainfall ordinary rain farming
precipitation. This ahnormally prolonged dr y period is practised, although contingent to t1e killing frosts,has heen correlated with an epicycle of erosion, attested owing to t1 e ahort growing season. Hack [31, p. 34]
in diverse localities within the Anasazi country. mentions dr y farms in the zone of yellow-pine of
Stream cutting would have been particularly destruc- !he Defiance Plateau of north-eastern Arizona,
tiveto the narrow flood plains ofthe San Juan drainage,
thus destroying the ground basis oí Anasazi agriculture
an d forcing th e desertion of the settlements in that
260
l. ID present da y Hopi country. th e average leDgth oí the growwg scaSOD isabout 130 days. the bulk oíthe Hopi crops need over 120 daya ío r maturatiOD{31. p. 20 ; 16. p. 101].
heing left dry or even by diverting w ater to individual
planta. In severa! instances, according to Rack;
th e ratio hetween th e areas of cultivated ak·chinfields an d th08e oC the arroyo watersheds which
supply them with water varied from 3 : 100 to 6 : 100;th e variation between these limita is explained by
th e differences in the amount oC run·off froro on e
drainage ba8m to another, depending on the nature
of th e ground [31, p. 31].The eJFeets ofthe late epieycle of erosion (sinee 1880)
have increased th e importance oC th e ak-chin fields,
hy reducing th e amount oC farmland availahle in th e
lIoodplains of large streams [31].A second class, with a few subdivisions, is that of
fields located on the Hoodplains oí th e main valleya,th e low Hood tenaces oC large arroyos, or in th e
dry heds of small intermittent streams, in places
where th e water will spread over a wide area during a
Hood and yet will not he destructive to the crop. This
practice is limited nowadays hy · th e deepening an d
widening of the stream channels in recent times.
An example of such valley-hottom fields cultivated by
th e Navajo-who, incldentally, learnt agriculturefrom th e Puehlo Indians-is given hy Bryan [14,p. 451]. The Hopi, and less frequently th e Navajo,
sometimes fuect th e Hoods by constructing temporary
earthen diversion dams, up to 30 cm. or more in
height [30, p. 104]; these are renewed every season.
A third class, fields watered by hillside wash, was
deserihed hy Bryan [14, p. 445] as praetised today
by th e Spanish-speaking inhabitants of small valleys
in th e Sandía Mountains of New Mexico. Since th e
settlement of that area dates only from 1815 and the
procedure has not been reported for the Indian country
[31, p. 30], is roight he colonial rather than ahoriginaI.
Anyway, th e distinction between these fields an d
those of th e second class does no t seem to be cIear
cut, to judge from Bryan's pictures.
- Still another technique-which involves works of a
more permanent nature-is th e building of check
dams in small wash hottoms to form terraces. Th e
dams are made of loosely píled stones an d brush,
an d serve to hold soíl an d moisture, by slowing th e
run-off, and to obstruct gully erosiono The crop planted
in those terraces is nearly always early corn, which
is associated with ceremonial celebration [31, p. 30].
Aeeording to Gregory [30, p. 104], remains of sueh
check dams on th e terraced Hoor of washes may be
seen near many of th e anclent ruins. In addition,
prehistoric check dams across th e run-off an d on
rock ledges are found at Mesa Verde (south-western
Colorado), an d it is likeIy that more would he foundifthe San Juan area were searched for them [13, p. 10].
Finally, canal irrigation was practised, at least in
the San Juan area, during Puehlo III times (lIOO·1300)as evidenced by th e ancient reservoirs an d ditches
found in the Mesa Verde region. The main ditch
262
on Mesa Verde is about 6 km . long, over 9 m. wide, [an d 28 cm. deep, an d falls more than 200 m. with a
very regular gradient [13, p. 10]. Modern Hopi gardens
irrigated hy diversion are deserihed hy Haek [31, p. 34],an d in the Navajo eountry hy Gregory [30, p. 105]However, th e amount oC land watered in this way 1Iis very small, an d it wa s apparently th e same in thCj(
past [13, p. lI].
IRRIGATION AGRICULTURE
OF THE MEXICAN BORDER
At th e close of th e seventeenth century, when th e
fust Christian missionaries entered th e section of
northernmost Sonora an d southern Axizona, later
known as th e Pimería Alta, th e agriculture of th e
Pima Indians depended upon canal irrigation [18, p. 4].
The climatic features of that area are a precipita
tion low in average an d extremely unreliable owing to
annual Huctuations, a very high rate of evaporation,
an d a long growing season. Of th e rivers which
supplied th e water for irrigation, th e Gíla and the
Santa Cruz ar e discontinuous streams, and the Sa nPedro How, although perennial, Huctuates greatly
[18 p. 13]. Piman settlements an d irrigation were
found along th e stretches of continuous flow. South
of th e present international boundary they lived
hy th e upper reaches of th e Sonora an d Sa n Miguel
rivers, an d down th e Sa n Ignacio an d Altar, where
rich irrigated land was found as fa r west as Caborca
[18, p. 5, quoting Mange's Luz de tierra incógnita, based
on his visit in 1694].As noted by eighteenth-and nineteenth-century
observers of th e Gila Pimas practice, temporary
diversion dams were bullt oC upright poles, driven
into th e river be d in zigzag lines, as a supporting
frame for a h a ~ r a g e of branches an d bundles of
bush; rocks and tree trunks were also used sometimes
to reinforce th e construction. These harrages rarely
stood longer than ayear , being washed away by
sudden floods. The networks of conveyance an d
distribution canals were very extensive; from th e
main canals many lateraIs branched out on both
sides of them. The main ditches were as much as 3 m.
deep at the intake an d generally 1.20-1.80 m. in
width. The irrigated fields were suhdivided hy earth
ridges into reetangles of ahout 70 m. hy 35 m. The
main canals were communally owned, and they
were deaned an d repaired every spring under th e
direction of elected officials, who also regulated th e
distrihution oC water. Often th e men of several
communities worked reciprocally for th e maintenanceof their irrigation systems [18, p. 156].
With regard to land ownership, elongated cobble
stones se t vertically in th e ground, suggestive of
study of their pattern of arrangement, hut i t wa s
ohliterated by recIamation works before it s signi
:ficance was appreciated hy archaeologists [35, p. 8].In addition to canal irrigation, sorne floodwater
farming wa s practised in a few favourable places.
Also, i t is noteworthy that food-gathering was an
important activity, intricately interwoven with gar
dening in Piman economy. Even in normal years th e
products of wild plants, mainly mesquite beans(Prosopis sp.) and cactus fruits an d seeds (Carnegieagigantea, LemairBocerBus thurberi, Opuntia fulgida,O. echinocarpa), constituted an important part of
th e dieto When crops were washed ou t hy overflows
or they falled for other reasons, the relative importance
of farming, food-gathering an d hunting as th e founda
tion of subsistence became reversed [18, p. 28 an d p. 63].I t must be noted, however, that increased reliance
on food-gathering an d hunting as important comple
m e n ~ s of th e suhsistence basis was possihly a reversion
which occurred at th e time when th e prehistoric
irrigation systems dwindled-in th e early fifteenth
century.l
The prehistoric sedentary culture of southernArizona is called Hohokam. Apparendy i t evolved
out of th e food-gathering Cochise Desert culture
with th e introduction of agriculture shordy before
th e heginning of the Christian era. Although there
is a dark period in th e development of this tradition,
from t1 e early fifteenth century to th e late seventeenth,
there is no reason to douht that th e Hohokam people
were th e ancestors of th e Upper Pima Indians of
later days.
Since no evidence of canal irrigation is yet known
for most of th e fust thousand years, sorne form of
Hoodwater farming is postulated for that periodo
River floodplain farming might have been feasible
in th e vicinity of th e archaeological site of Snaketown,
a setdement close to the Gila River whose origins are
much earlier than the attested antiquity (c. A.D. 800)of th e local canal system [33, p. 57]. Later, th e
development of th e canal networks permitted setde
ment away from th e rivers; an example of tbis was
th e town whose ruins are now called El Puehlo de los
Muertos, that flourished in th e fourteenth century
as th e centre of a large population cluster, 10 km .
south of th e Salt River [34, figures 1 and 24, an d
p. 14-43; 35, p. 9].Th e remains of prehistoric networks of irrigation
canals are very extensive in south central Arizona;
in th e Salt River valley th e aggregate lengtb of
conveyance ditches has been calculated to he more
than 300 km., of .which 120 km. correspond to th eLos Muertos system alone [35, p. 8; and 34, p. 41,quoting Turney'5 PrehistoTie ITTigation in ATizona].
Some of th e ridged depressions revealing th e course
of an aneient canal can he traccd for more than
30 km .
Land use in PTe-Columbian America
An excavated section oí one oí th e Los Muertos
ditches, 10 km. !roro th e intake, was found to he 9 m.
in width at th e top and over 2 m. in depth, with a large
groove at th e bottom, apparently deaigned to carry
th e small flow of water during perioda of scarcity,
a feature also found in other broad channels. In a
section of th e Snaketown canal, th e actual water con .
duit is abont 2 m. in width and 1 m. in depth, although
from crest to crest of th e side ridges i t is abont 10 m.
in average. Haury [33, p. 56] pointedly notes that th ewidths given for th e Hohokam canals on th e basis
oí surface indications do not necessarily correspond
to the actual size of th e channels, which has to he
aseertained through excavation oí their cross-seetions.
Tbe study of th e Snaketown canal [32, 33] revealed
an interesting story oí deposition oí alkaline cmsts,
silting an d re-excavation of channels, eovering a
time span of at least 500 years, beginning at about
A.D. 800.The Hohokam canal systems reaohed their maxiroum
development from about 1200 to 1400 [32, p. 50].
Their planning, construction, and maintenance would
have required a considerable degree of community
control by a political organism; th e fact that in someinstanees several villages evidently drew water from
th e same canal, indieates co-operation an d co-ordina
tion at th e intervillage level [35, p. 8]. The layout of
settlements an d canals in th e Los Muertos area,
suggests some degree oí political centralization oí th e
kind that, in larger an d more compaet irrigation dis
tricts in other parts oí th e world, produced th e rise
of city-States. This phenomenon, in th e Hohokam
area, was apparently limited to a very incipient stage
hy the seattering of irrigable zones, at least under th e
resources-technology equation operative in th e
Hohokam culture.
THE SONORAN DESERT
Further south, along th e middle and upper Yaqui
River an d th e mid-course oí th e Sonora River (i n
th e zone oíBShw climate), th e agriculture ofthe Lower
Pimas (Nebome and Ures groups) depended on
well-developed canal irrigation, according to th e
reports of the early Spanish explorers.
Th e desert's Papago division of th e Pima people,
living in the very arid section west of th e Santa
Cruz River, forcibly depended more on gathering
of wild plants and/or hunting than on agriculture.
Th e average annual precipitation over that area
is less than 125 mm . Residence was seasona!, near
the scanty permanent water sources during th e
l. Th e causes of tbis are still undetermwed. In a personal communicatioDdated 8 April 1959, Dr. Haury writes that lowering Qf the main-streamc h a n l l e l ~ or alkalinization have beeo mentioned as possible explanations,bu t Done iB directly attested by archaeological evidcnce. Personally.
he Cavours ratber a break-dowD oC th e 80ciety_by efl'ect oC COrees no t ye tsurely recognized, which were operntive over a Inrge part oC th e NortbAmerican south_wcst_as tbe main caUlie of the cultural recession. '
willter months and in th e places favourable foragriculture after th e summer rains started. Cropswere raised mostly hy thunderstorm floodwaterf a r m i n g ~ A k ~ c h i n fields ar e reported hy the earliestchroniclers (end of !he seventeenth century); lowemhankments, hrush harriers, an d shallow ditcheswere used to capture and spread th e water. Also, ina few places they irrigated ,Yi!h water derived fromsprings or th e stretches of permanent How of the
Sonoita River [18, p. 40 and p. 168].The Yuman trihes of th e lower Colorado Rivervalley-in th e western section of the Sonoran Desertprovince-depended on natural1 river Hoodplain. farming, planting as soon as th e water receded in th eground moistened an d fertilized hy th e annual summeroverHow (see !h e thorough study hy Castetter an dBe n [19]). The erratic ways of th e river-variahilityan d even' failure of seasonal Hoods an d unseasonaldestructive ones-precluded the maintenance of permanent frelds [19, p. 38 and p. 69 et seq.] and made th eYuman economy unstahle. In spite of th e eventualpossihility of a hounteous harvest of cultivated crops(maize, tepary heans, an d pumpkins), suhsistencedepended to a large extent (50·70 pe r cent) uponthe products of wild plants-mainly Prosopis julijlora,P. odorata-and frshing [19, p. 66 et seq. and p. 238].
River Hoodplain farming, hased on natural inundation, was likewise th e source of suhsistence of th epre-Spanish nations of southern Sonora an d th eSinaloa coastal plain, in north-western Mexico.There, dependahle summer Hoods an d less so, hutstill ordinary, winter ones-which originated in th eupper drainage of th e rivers in th e high WesternSierra Madre-and th e all-year-round growing season(climate BWh'w in th e north, BSh'w in th e south)alIowed two an d even three crops ayea r on th econtinuously re-fertilized fields. I t is quite significantfrom !h e nspect of sociological theory !hat, although
i t sustained a dense an d prosperous population,tm s natural irrigation did no t produce an y appreciahledegree of urhanization no r a complex political structure.
MESO AME RICA
The term 'Mesoamerica' is used nowadays to referto th e culture-area comprising th e ancient 'Mexican'an d 'Mayan' civilizations (in their many varietiesas manifested in time and space), which in th e pasthave heen geI).erally considered as rather distinctdevelopments.2 It s fundamental unity is now heyond
question, an d th e differences in s o c i o ~ e c o n o m i c structures an d settlement pattern ca n well he explainedas ecological adaptations to contrasting environments.
The climate of the eastern lowlands, facing th eGulf of Mexico and the Carihhean Sea, where th eclassic Mayan and other cultures of related type
264
Hourished, is for th e most part of th e t r o p i c a l ~ r a i n forest (Kiippen's Af) and monsoon forest (Amw,Amw') types, th e main exception heing th e tropicalsavanna (Aw) area of the n o r t h ~ w e s t e r n part of thePeninsula of Yucatan-a limestone formation with nosurface running water an d where, noteworthily, adistinctive variety of Mayan culture existed sinceancÍent times.
On th e other hand, notwithstanding th e great
climatic contrasts effected hy th e relief, most of th ehighlands, th e interior valleys, an d the Pacific coastshare a common characteristic-well-marked periodsof seasona! drought lasting over half of th e year(generally, seven to eight months). Over most of th eplateau and range regions, aside from !h e high mountains, th e dominant climate is th e mesothermalsavanna (Cw) type, and in the interior depressionsan d along most of th e Pacific coast th e tropicalsavanna (Atv) type is found. The temperate forest(C!) only occurs in the mountains. In th e zone ofrain shadow west of th e Eastern Sierra Madre highpeaks, there ar e tracts of cold steppes (BSk), and inth e depths of th e inland valIeys tropical steppes(BSh) and even cactus desert (BWh).
In th e light of the extreme environmental diversitywitmn Mesoamerica an d what is known ahout th elevel of civilization attained hy it s people, i t is hardIysurprising that, contrary to an opinion widely an drather uncritically accepted until a few years ago,th e diversity of agricultural practiccs found in thatarea corresponds, as might he expected, to the v a r i a ~ hility of. climate, reHef, hydrography and soil.
We are concerned here only with those sections ofth e culture-area where unreliahility or extremeseasonal limitation of rains, an d excessive e v a p o ~ transpiration an d run-off losses,3 ca n he, an d were,advantageously compensated for hy means ofirrigationan d where land productivity was sustained hy th e
application of techniques for th e conservation ofsoil an d water resources, allowing intensive e x p l o i t a ~ tion. These regions correlate with th e Cw and Aw zones,which constitute th e greatest part of central, western,
l. Although Il mid-eighteenth eentury eye-witncss credits the Indiana withIl method to divcrt th e overflow, building diagonal dikes of 10gB Bupportedby poles driven iuto the river bed, ruad spreading th e water mt o swales bymeans of eartb dams disposed to form an incipient system of bw;in irrigationU9, p. 1331.
2. 'l'be nortbern froutier of lIIesoameriea can he defined in various ways, eitbcron Illl arcbaeological or ethnological hasÍB, hu t in any case it l1uctuatedtbrough time. Th e 220 paralIel eonstitutes a convenient geographiealrefcrencc, aItbough cultural influences can he traced tar north of that line,to tb e north-west Ilnd north-east, and the ltIesoamericau cultmal patterntrcspussed it at various pIaces an d periods; eonversely, tbe farming froutiersettlementshadretracted southwardsin tbe central seetion before th e alTivalof tb e Spaniards in the sixteenth century. Th e northena neighbours of tb csedentary Mesoamcriean people in the semi_nrid zone of north central
Hexico were tb e nomadic hnnting-gathering Chichimees. Th e southernlimit of Mesoamcriea does uo t coneern th e purposes of tbis papero3. Although I huye no t seen statistics on tbis sub¡'eet, anyhody familiar with
thcse regions knows tbat e v a p o _ t r a n ~ p i r a t i o n 085( '8 must he inordinlltelyhigh. owing to thc rate of insolation an d low reIativc humidity. Witb regardto run_off over tb e whole arca nnder consideration, excluding tb e negligihleCJzone of th c high mountains. rain almost always falls in torrentilll sbowers,thus tailing to soak sloping ground. Th s i8 oC speeial significance slnce,owing to tbe relative scarcity of l1at lands, much oC tb e cnltivlltion is doneon th e hiUsides al' rolliog gronnd.
The subsistence hasis of civilization in these regions,
although resting to a large extent on th e rain-farming
fallowing system2 cannot he properIy understood i f -as
has heen done to o often-the importance attained
there in pre-Columhian times hy improvement an d
conservation methods, an d consequent influence
on population density an d clustering, settlement
patterns, an d some fundamental aspects of th e socialand political sutructures, is minimized or ignored.
There has heen an awakened interest in th e distri
hution of native irrigation agriculture by means. of
canals an d ditches at th e time of th e Spanish conquest
or shortly thereafter and, in a few concrete cases,
in th e historical information referring to late pre
Columbian times. A few results have heen published
since the late forties, covering th e area north of t1 e
Isthmus of Tehuantepec : Sauer's investigations
on th e role of irrigation in the preconquest economy
of the' zone of Colima an d neighhouring sections of
Jalisco and Michoacan, in western Mexico [69];
my own research fo r most of the area that has heen
pnhlished in detail only for the part referring to th eBalsas River drainage, which eomprises, however,
a large part of southern central Mexico an d incIuded
th e prized irrigation district of southern Puehla an d
eastern Morelos [2]; an d Palerm's comprehensive
paper for th e whole area under consideration [62].
No similar studies have heen made for some sections
of Mesoamerica southward from the Isthmus of
Tehuantepec where enviromental conditions might
have encouraged analogous practices. However, a
late sixteenth-century reference indicates irrigatíon
for cacao-tree plantations at Santa An a (western El
Salvador), and it is implied for places in Xoconochco
(the coast of Chiapas, Mexico), and farther south
in Nicaragua and Costa Rica (to th e península of
Nicoya), where native cacao culture is recorded and
th e long dry season would have made watering of
l. See also Snnders [67, p. 74; 68, p. 116J.2. With reference to its effects on the type of civilization,th e uplnnds Cnllowing
system is, anywny, twice as productive in proportion to th e amount ofcultivable land as the shifting-fieldsprocedure typical oC tIJe humid lowlnnds[67,63].
annual rainfall [58, p. 66, p. 73, p. 76 and p. lU].
Th e studies mentioned in th e preceding paragraph
are based on sixteenth century historical or adminis
trative documents, some of which are transcriptions
of pre-Spanish traditional history, or pleas for confu .
mation of water rights by colonial authorities. Unfor
tunately, very little archaeological research has been
done to date to suhstantiate th e documentary evi
dence an d to investigate th e prehistory of irrigation
works. A noteworthy exception is Wolf aq.d Palerm's
field survey of the extensive remains of aqueducts
and canals in th e piedmont of the Sierra east of the
former Acolhua capital, Tetzcoco [90]. Th e longest
aqueduct found is carned over an emhankment
1,000 m. long and 12 m. high at the highest point,
another although shorter (ahout 300 m.) reaches 20 m.of maximum height over th e saddle of th e ground
[90, p. 269 and p. 272]. In those 1 have seen the
channel is stone bullt, revetted with lime plaster
and, in a shon stretch over th e hill at th e entrance of
th e emhankment leading to the royal gardens, th e
water conduit is cu t into solid rock.
These authors did a splendid job of relating theanclent vestiges to sixteenth centm'y documentary
sources, and to present-day native irrigation in th e
area. Being familiar with these places, l fully agree
with their conclusions that available evidence suggests
a mid-fifteenth-century reclamation project, under
taken in Nezahualcoyotl's reign. Although hitherto
usually referred to as th e royal gardens of that king,
it is clear in th e field and supported by documentary
evidence, that th e regal domaine was only a part
of a more extensive endeavour designed to benefit
many communities. Mid-fifteenth century was a
period of Acolhua preponderance in central Mexico,
with King Nezahualcoyotl at the summit of power
as political and mili.tary leader of th e Triple Alliance,
just before it s decline and th e accession of the kings
of Tenochtitlan to pre-eminence within th e frame
work of th e expanding Aztec empire. However,
l believe that extensive use of aerial photography an d
spade work, that has no t been done, would not fail
to reveal considerahly older hydraulic works in that
general section of th e Valley of Mexico,l
The only excavations made in Mesoamerica to
uncover and date vestiges of a prehistoric irrigation
system are those of th e dams and canal in a small
valIey between th e Maravilla and Altatongo hilIs,
in th e neighbourhood of the ancient metropolis of
Teotihuacan. The indications leading to the discovery
were found by examination of aerial photographs,
and the preliminary survey on th e ground reported
by Armillas, Palerm, and Wolf [3]. The subsequent
excavations have been very ahly conducted by MilIon.
His conclusions [59] are that this system may have
originated in the ToItec period (e. A.D. 800-1200)
or perhaps in' older Teotihuacan times.2 Millon has
266
also made considerable research, most of which is
still unpublished, to relate prehistoric, colonial, an d
modern practice of irrigation in th e valley of Teoti
huacan. Th e major canals which ru n through th e
valley today existed in 1580, an d an abandoned canal
has been discovered in th e Hat lands of it s central
section that looks a promising site for an investigation
of the very origins of irrigation in th e district of
Teotihuacan.
In th e Basin of Mexico, th e shallow fresh water
lakes oí Zumpango, Xaltocan, Xochimilco, Chalco
and parts of Lake Tetzcoco,3 were conquered for
chinampa agriculture, th e so-called 'floating' gardens
[74,84]. Th e chinampas are, and were, artificial islands
bullt in shallow waters by piling up layers oí aquatic
plants and silt from the hottom oí th e lakes.4 In order
to keep th e porous soil of the chinampa perpetually
moist, by infiltration from th e surrounding waters,
an d to íacilitate supplementary manual irrigation,
th e islands· are built in comparatively narrow, e l o n ~ gated strips. Also, th e to p plane is regulated-by
adding or scraping to p soíl as required-in relation
to the water level, so that moisture might reach th e
root-leveI. Additional moisture is supplied directlyto the individual plants by lifting water from th e
surrounding canals hy means of simple hand tools,
such as cloth huckets mounted in a twig ring with
a long handle or by wooden scoops. As the water is
muddy and rich in organic nutrients, this amounts
to th e addition of new soil.
Besides th e building of th e land, infiltration of
moisture, and manual irrigation, th e chinampa far
ming system includes today, an d seems to have
included in the past, techniques such as th e use of
hotbeds-and consequently transplanting-and ferti
lizer. lndian seedheds are depicted in Fray Bernardino
de Sahagún's Florentine Codex (third quartel' oí th e
sixteenth century). FrayAlonso
de Molina'sVocabu-lario en lengua castellana y mexicana (printed in
Mexico City in 1571) provides some linguistic evidence
ahout th e fertilizers used-at least for th e central
Mexican area of Nahuatl speech, which includes th e
chinampas region. According to Molina, th e verb
coquipachoa nitla, is translated ' to dung the soíl in
some way', coquiatl is mud, bog, hence th e verb
means ' to muck', which describes welI th e practice
1, Perhaps related, an d 1 mention it just as a possibility to be investigated,
to th e ruills oC Cerro Portezuelo, an important settlement íartber south in th esame piedmont area.
2. In my opinion, the possibility oí the system originating in the Teotihuacanperiod seems rnther Iikely, although admittedly based on the íaint evidenceoí th e diversion oC th e stream írom ¡t s natural watercourse rS9, p. 164].
3, High salinity was apparently eonfincd to th e ellatern parta oí Lnke Tetzcoco,bu t the fiow and re60w oC braekish water, depending on th e í1uctuatioos in
water level withln the intercommurucated system oí lakes, afFected itawestern section, around th e island where th e twin dties Tenochtitlan andTlatelolco rose. Tb e construction oC the 'albarrndón de Nezahualcoyotl·
-a dike 16 km. long, built by the Azteca under the guidance oC Tetzcocauteclw.iciana in 1449-besides protecting th e capital írom flooding musthave improved the possibilitiea oCland reclamation, by means oí chinampas.
iu the marshes oC tbat arell,4. The c/¡jnampa system can thus be classHied as drainage an d irrigntion. The
building oC th e islands ia a very effective mcaos oC reclamation and tbeircultivation iB a specialized an d sophisticated method oC gardcning,
prize irrigation districts, forming a ke y eeonomicarea. The famolls religiolls an d commercial centreof Chollolan was located on th e upper end of thiscompact irrigation zone, that extended downwardsto Atlixco an d Itzocan, which were also importanturban centres at th e time of th e Spanish conquest.In th e lower parts of this irrigation district th e lesser
altitude permitted th e cultivation of cotton-animportant trade commodity in pre-Spanish t imes
which was no t possible around Chollolan.
The relative importance of pre-Spanish irrigatedlands and rain farming in the zone of Colima an dsurrounding sections of Jalisco an d Miehoacan ha sheen investigated hy Sauer [69]. His study shows
how th e contrast hetween th e aIluvial soils of theriver valleys, which constituted districts of concen
trated irrigation, an d th e clayey soils of the surroundin g uplands, that were no t too well suÍted for nativemethods of cultivation [69, p. 63], affected th e distrihution of population an d th e degree of- urbanization.There were at least five aggregations of urban importance in the lowlands an d thl'ee in the highlands,each one l'anging from pel'haps five to te n thousand
people. While. th e capitals of th e valieys were th ecentres of irrigation distl'iets, those in th e uplands werecommercial towns concel'ned in different degrees
with mining operations [69, p. 81], although small-scaleirrigation was also practised in that zone.
But by fal' th e largest an d most compact districtof intensive agriculture-with canal irrigation, chinam-pas an d terracing-was th e Valley of Mexico. Thiswas, apparently, th e only seetion where th e term'hydraulic agricuIture' as defined by Wittfogel[89, p. 153] does apply. Some State enterprises for
th e construction an d maintenanee of productivean d protective water works have been mentionedalready, more are known from historical sources.
Sanders' detailed studies of population for th eValley of Mexieo at th e time of th e Spanish conquestindieate a minimum of one million for 8,000 km .2 ofterritory. The total population of the Áztec empireat that time ca n be calculated, on th e basis of differentstudies by various authors, at between five an d sixmillions, for an area of about 200,000 km.2 Thus, inth e Valley of Mexico with an extension of only onetwenty-fifth of that of th e empire, about 20 pe r centof th e total population was concentrated, notwithstanding th e fact that th e empire included sorne ofth e most densely populated areas of central Mexico
outside of th e valley. In relation to th e whole areaof Mesoameriean civilization an d it s estimated population th e proportion is even more startling, for th eVaIley of Mexico, 'with an area of abont a hundredand-twentieth ofthe total, contained about 10 pe r centof the population:
In th e ValIey of Mexico, and in th e southern Pueblairrigation distriet as well, th e number of inhahitants ofa few nucleated urban centres ra n into tens of thousands
268
- and well passed fifty thousand in the empire'scapital, Tenoehtitlan-Tlatelolco. Many rural distrietcentres were towns-with part-urhan funetionsof several thonsand people.
It was on th e foundation of tlle politieal unificationof tbis nucleus that th e Áztec empire' was built,an d archaeology reveals the paramount role playedby that area in Mesoamerican prehistory, before th ebeginning of th e Christian era; this suggests that a
comparable situation, with regard to density andconcentration of population an d inferentially agrieul
tural practices, existed during Teotihuaean times(from ahout 200 B.C. to A.D. 700).
MESOAMERICAN FRONTIER ZONE
Th e northern Mesoameriean frontier zone in northerncentral Mexico is an exceedingly important areain which to investigate th e interrelationships of natUl'al
an d cultural factors in the history of land use.Following th e inland piedmont (from about 2,250 m.
to 1,850 m. in altitude) of th e Western Sierra Madre,
at the edge oí th e juniper-oak-pine vegetation zone,an d along th e savanna (Cw) belt to th e limit of th esteppe (BSkw), th e Mesoameriean cultural patternbecame established, either as a resulto of diffusion
aeculturation or eonquering colonization, after perhapsA.D. 700 (to judge from th e earliest carbon-14 dateobtained from th e ruins of La Quemada, at th e southernentrance to tbis corridor [24, p. 1104]). The northernmost extension of tbis pattern reaehed th e presentDurango-Cbihuahua horder [40, p. 132 and p. 138].
In th e semi-arid zone of th e central section of th efrontier-a region oí transition to the ChihuahuaDesert province-the nOl'thern neighhours of th esedentary Mesoamerican farmers were th e nomadic
hunting-gathering' Cbichimees' (the Zacatecs, Guachichiles, an d Pames). The historie limit between sedentarians and nomads roughly coincides with th e presenthorderline between Cw on th e one hand, and BSk
(in ih e west) 01' BSh (in th e east) climates, on th eother. In view ofthe knownretraction ofthe agriculturalfrontier-attested archaeologically an d by traditionalhistory in territories formerly occupied by sedentaryeultivators an d taken over by th e nomads at thetime of th e fall of Tollan (towards th e close of th etwelfth century, perhaps in th e fust "half of thethirteenth)-it would be of great interest to investigatewhether some minor climatic Huctuation triggeredecological changes, 01' impl'ovident clearing of marginallands for th e expansion of cultivation indueed aridityin that area.
The fact that similar an d more or less contemporaneous troubles are recorded in th e south-west of th eUnited States (see previous remarks on Anasazi an dHohokam and hihliography [38, p. 120]), in the iulandpiedmont of the Western Sierra Madre [40, p. 139], and
in the eastern margin ofthe Great Plains, tends to favour
th e hypothesis of climatic deterioration. I t Iooks as i fth e North American arid zonc was expanding in al l
directíons between th e late twelfth an d earIy fifteenth
centurics. Since local environmental an d cultural
factors intervene in widely separateplaces, it is not to he
expected that th e effects would havc heen strictly syn
chronous; twelve score years or so is no t an inordinately
large allowance.The apparent coincidence might be a mirage,
however, an d certainly much more factual information
in different licIds of knowledge is necded to decide
th e qucstion; ncw interdisciplinary approaches will
have to he developed., to gather and digest all th e
pertinent data on this complex problem.
PERU
The rise an d progress of th e ancient coastal Peruvian
civilization ar e unequivocalIy related to the improve
ment of th e techniques for full exploitation of th e
riverine 03ses hy Ill:eans of canal irrigation. Only inth e Virú Valley has th e histOly of the development
of hydraulic engineering an d it s relation to population
density an d settlement patterns and, by inference,
to the origins of the State, been studied in some detail
[86, passim, specially, p. 361-.371]. Generalizationson th e history of irrigation works for th e remainder
of th e zone are based on that sequence, an d p r e s u m ~ tively dated illustrative examples from other places.
W have already seen that agriculturc-as distinct
from cultivation supplementary to an economy hased
on wild life-became established aftcr th e introduetion
of maize, an d that it wa s expanding in th e early
eenturies of the first millennium B.C. There are no
remains of canals or vestiges of irrigated plots that ca nbe attrihuted to this period, hut th e attested expansion
of settlement away from th e shore-line, to hottom
lands or on th e valley margins, is presumed to be
related to agricultura! practices, possihly Hoodwater
farming [22, p. 19].In an y event the tapping of the streams for irrigation
may have hegun hefóre th e middIe of th e millennium
in the upper narrows of th e rivers [8, p. 142; 86, p. 31,p. 361, p. 392, and in fignre 82, note the clustering ofsites at th e Quehrada de Huaeapongo]. The latter
half ofthe millennium-archaeologists' Early Gallinazo
period-saw th e heginnings of large-scale irrigation
works, with th e related phenomena of marked increase
in population an d its spread into al l parts ofthe valIeys,
and the formation of th e first large agglomerated
communities of several thousands of persons living
within an area of· two or three square kilometres
[86, p. 396]. This indicates the attainment of a mastery
of water control tha1; must have required a period,
however hrief, of experimentation.
By Late Gallinazo times (early eenturies of th e
Landluse in pre.Columbian America
Christian era), th e Virú Valley canal system extended
from the water intakes in th e narrows (Quebrada de
Huaeapongo an d Upper Virú) at th e foot of the
mountains to the sea. Th e old canal lines an d distri .
bution of settlements suggest that at least 40 per cent
more land was under irrigation then than nowadays
(9,800 ha. against 7,000 ha. maximum today) [86,
p. 394, see also p. 20]. Th e difference prohahly does not
indicate a change in t le total volume of water dis
charged hy th e gorges; i t may welI be due to higher
consumption pe r acre required by the commercial
crops grown today, sugarcane in the upper valley an d
cotton in th e lower section [5, p. 19].
Old irrigation furrows seen in th e neighhourhood
of Huaca Gallinazo, th e site naming th e period, that
seems to have heen occupied uninterruptedIy for
ahout one thousand years (from about 500 B.C. to
A.D. 500), are depicted hy Strong and Evans [78]and reproduced hy Willey [86].
In th e ensuing Huancaco period, during th e middle
eenturies of the first millennium, th e layout of tlle
canal network and the distrihution of population
clusters changed somewhat, without noticeably
affecting th e total acreage of irrigated lands or th e
number of people; the irrigation an d popuIation maxi
ma attained in former times seem to have been
maintained. However, it oecurs to me that the deerease
in land usage in th e sections intensively exploited in
the previous period (according to Willey (86, p. 393]),might he very significant as regards possible causes
of th e general decline of th e Virú ValIey economy
evident in later times.
In th e neighhouring valley·oasis of Moche an d
Chicama, to th e north, impressive remains of
irrigation works are seen. Extensive canal systems
were heyond reasonahle doubt in operation-although
direct dating of the vestiges has not yet been done sofa r as I know-during th e Mochiea period ofthe a r e h a e ~ ological nomenclature; according to CoIlier's an d
Willey's synchronologies i t overlaps Middle and Late
Gallinazo an d th e following Huaneaco periods of th e
Virú VaIley, having lasted for about on e thousand
years, from hefore th e beginning of the Christian
era to one or two eenturies prior to the close of its
first millenary [22, 87].Th e magnitude of th e engineering work involved in
th e building of the irrigation systems ea n be best
evaluated hy th e example of th e Ascope aqucduct,
admittedly one of th e greatest accomplishments ofthe
aneient Peruvian lndian engineers. Th e Ascope canal,
generalIy helieved to date from this period, follows
th e basal eontour of the.hills on the north side of th e
Chicama VaIley, heing conducted aeross a hay of th e
plain over an embankment huilt of earth and adobe,
1,400 m. long and 15 m. high ahoye the adjacent valley
Hoor, an estimated volume of 785,000 m.a of earthwork
[8, p. 157; 43, p. 71; 46, p. 162; 86, p. 411].Also, inter-valley irrigation systems-implying the
constitution of multi-valley political entltles-weresupposedly built in Moehiea times [22, p. 21], an d acanal over 120 km. long deviated water from th eChicama River into th e lower Moche Valley, to th esouth [46; 65, p. 58].
Willey [86, p. 365] makes the interesting remarkthat th e main irrigated section of th e Virú Valley
through _ he Gallinazo period was avoided in subsequent times. In my opinion, as pointed ou t aboye,this might have a bearing on th e apparent weakeningof th e coastal States in later days, when they becamean easy prey to th e invading highlanders, coupledwith th e astonishingly disastrous e:ffects of th e Spanishconquest which followed, an d also considering t h a t -
contrary to what happened in Mexico-no majorepidemic waS responsible for th e decimation of th eIndian population within half a generation afterPizarro's arrival. The social chaos brought about byth e conquest, and the ravages caused by th e factionwars between th e conquistadores, were certainlycontrihuting factors, but th e deep causes of thedecline of th e coastal civilizations might go fa r backinto times earlier than th e coming of th e beardedwhite meno
The trend to IDove away froID th e older fields,already started during Huancaco times, continuedduring the later Tomaval, La Plata, an d Estero periods(from about A.D. 800 to the Spauish eonquest). A newzone of settlement became occupied, since Tomavaltimes, along th e coastal dune belt. This probablyindicates th e beginning of th e practice of utilizingground moisture by means of th e pukio basins (seebelow) to raise erops [28, p. 34; 86, p. 368]. Fromabout 1200 onwards, diminution of population isevident; most of the valley-bottom dwelling siteswere abandoned, habitation clustered in th e upper
narrows of th e valley an d along th e dune zone. Notwithstanding that sorne canal irrigation was still operativein 1548, at th e time of Cieza's visit, th e pre-Columbianrecession had been evidently aggravated by th e effectsof th e Spanish conquest [21, chapter LXX, in whichth e Virú Valley is mentioned under th e name Guañape].
The population decrease mentioned in th e precedingparagraph has been tentatively explained as an e:ffectof social an d politieal disturbanees [86, p. 421], butth e alternative hypothesis of local failure of th e Írri-gation system fo r purely technical reasons is worthyof consideration, since th e abandonment of th e ancÍentirrigated sections of th e Virú seems to have begunbefore th e major upheavals apparently brought
about by the Tiahuanaco expansion an d what followed.Soil salinization and· excessive rlsing of th e watertable due to prolonged use of irrigation with inadequate drainage may have occurred. That salts mayhave given trouble to th e native farmers in the VirúValley an d elsewhere in th e Peruvian coast is borneou t by Willey's remarks [86, p. 16] an d also by Ford andWilley [28, p. 26], quoting Lareo Hoyle. I t is worth
270
while to note that owing to mismanagement of waterresources, salinization is a:ffecting nowadays th edistriets of Piura and Pisco [65, p. 20 and p. 45].
The soldier an d chronicler Pedro Cieza de León-wrlt ing of the Chilca Valley, 40 miles south fromLima, in about th e middle of th e sixteenth centuryexpresses amazemcnt that, in a rainless country and
with no surface running streams available for tapping,th e place was verdant with cultivated fields. Togrow maize, vegetables, and fruit trees, th e Indiansplanted in wide an d deep basins purposely dug inth e soil, to capture underground moisture at root levelof the plants [21, ehapter LXXIII].
Willey [86, p. 16] reports that there are arehaeolo-gical remains of this practice in th e dune helt frontingth e ocean shore at th e lower Virú Valley : basins-sometimes rectangular, sometimes irregular, varyingin size from 100 X 50 m. to 30 X 30 m., sunk about1 ID . below th e surrounding ground surface an d separated by ridges 2 to 4 m. high, built up with th e soilremoved froID th e excavations-dug by man, serveto retain seepage water which lies closest to th e surfacenear the beach [cf. 5, p. 19] on th e groundwaterlevel in this valley). These basins are known locallyas pukios.
The only indications of age are th e proximity ofsuch excavated hollows to a ruin certainly datingfrom th e Tomaval perlod (beginning towards th e en dof the fust millennium A.D.) an d surface potsherdsfrom th e same an d subsequent periods, up to Colonialtimes, collected in another area of pukio cribs, bothplaces being situated in the lower valley, no t fa rfrom th e beach. Th e concentration of habitationsites dating from th e Tomaval and La Plata perlods(up to the late fifteenth century) in the dune belt,where th e groundwater level is higher, suggests a
relationship hetween th e shifting of populationto that zone an d th e beginning of th e practice [86,p. 368]; whether i t might be older in other places isno t known.
Horkheimer [37, p. 78] also mentions this ancienttechnique for obtaining undersurface watering, an ddescribes two such great basins, one of them about500 ID . long an d 10m. deep, surrounded on three sideshy ridges an d open on th e fourth side facing th e sea.This basin was dug a short distance from th e beachnear th e ruins of th e former Chimu capital of Chanchan(fourteenth an d fifteenth centuries), on th e northcoast of Moche VáIley. The same author quotes Regal[64, p. 104] for similar ancient hollows in Chala,
Mala, Atiquipa, Atico, an d Pica. Reparaz [65]illustrates modern cultivation of cotton utílizingunderground moisture in th e desert Pampa de Villa-curí, hetween Pisco an d Ica.
An excellent report on the state of natíve irrigationin the coastal valleys in 1 5 4 8 ~ 1 5 5 0 , after th e c a t a s ~ trophic depopulation brought about hy th e factíonwars among th e Spanish conquistadores, is given
more, valley fields are liable to he washed over hy th e
Hood from th e precipitous slopes in the rainy season.
Th e second, hut not th e least, achievement of th e
highIanders to master their environment was th e
reclamation for agriculture of th e steep mountainsides
by means of stone-faeed irrigated terraees.
Incredihly steep slopes were eonquered for culti
vation hy terraeing an d sometimes whole -valleys
were reshaped and regraded. Rowe [66, p. 210]
mcntions as outstanding examples thosc at Yucay
an d Ollantaytamho, an d Cieza de León [21,
ehapter XCI) refers with admiration to those at the
Valley of Xaquixaguana. A series of amazing air
and ground views of telTaced steep slopes in the
Colea an d Andagua valleys, north-north-west from
Arequipa, were pubJished by Shippee [75]. Detailed
descriptions, maps, an d photographs of aneient
terraces in th e Vileabamha mountain range are
found in th e report of the Viking Fund's explorations
[27].Irrigation canals often ra n for miles along th e side
of a valley to irrigate a comparatively smalI terraced
area [66, p. 233].Not very much is known of th e origins and deve
lopment of th e teehniques of reclamation hy meanS of
tenace construction, although many ofthe complicated
tenace arrangements could he assigned to th e first
millennium of th e Christian era, aceording to Bennett's
remarks [6, p. 21 ; 8, p. 157]. Sorne of th e ancient
terraces are stilI cultivated but many are ahandoned.
THE CHILEAN DESERT
The resourees that th e desert extending along th e
Pacific from latitude 18° S. to ahout 30° S. offers
for th e suivival ofpopulations with simple technologiesar e severely limited. As along the Peruvian coast to
th e north, rainfall is sporadic-no rain may fall for
years, but here, hecause of th e composite effects of th e
Humboldt cold current parallcl to th e shore and th e
prevailing winds, even th e slopes of the Andes are
dr y an d bare; no permanent stream reaches th e ocean
hetween Pisagua an d Copiapó. Nevertheless, th e
dryness is lessened at times by sea fogs, especially in
winter (elimate BWkn); from just north of Taltal
to the south, th e desert eoast is fringed hy a heIt
of loma vegetation (annual grasses an d herhs) whieh
increases in width southward [9, 10, passim].
Similar eeological conditions may have prevailed
for a long time in th e past, for, in his excavations at
several plaees along this coast, Bird [9, p. 314] found
no evidence for any signifieant change in precipitation
or wild life within th e known period of human occu
pation, extending hack, presumahly, over 4,000 years
[see 87, figure 11, sequence chart]. Nor- does Bennett
[6, p. 600] see the need to explain by cHmatic change
th e fact that vestiges of irrigation ditches are found
272
in some now waterless sections or of old algarroho
trees partially burled by sand; such occurrences,
he writes, might he accounted for in other ways.
However, a more general distrihution of springs
an d groundwater is implied hy th e location of sites,
and is horne ou t hy some of Bird's own remarks.1
This is pointed out hy Schaedel, who also makcs an
interesting reference to the observed changing dip
of the wator table [73, p. 34].Prior to the introduction of agriculture th e early
inhahitants of the driest sections ha d no alternative
hut marine suhsistence, since game must have heen
always to o scarce to he even oí sccondary economic
importance. The choice oí camp locations for th e
s h o r e ~ d w c I l i n g people wa s further restricted hy th e
prevalence of heavy seas, scarcity oí sheItered places,
and irregular availahility of drinking water. South
oí , th e 25° parallel, land game, supported in limited
numbers by th e coastal vegetation, allowed them to
supplement fishing with hunting; also, a water supply
is less difficult to secure. Agricultural expansion was
restricted hy th e limited volume of water availahle
for irrigation and the small size oíthe valley hottoms;along much of th e coast hetween Arica an d Coquimho,
farming never replaced entirely th e mode of life hased
on fishing and hunting.
In th e ía r north, incipient cultivation appears
perhaps at ahout th e heginning of th e Christian era,2
with maize, gourds, and cotton, the latter possihly
cultivated but perhaps wild [9, p. 248]. Howevcr,th e pattern oí weIl-established agriculture found
in Inca times in th e valleys oí th e Arica hinterland
th e year-round Howing Lluta River and the Azapa,
of intermittent streams now running only with o c c a ~ sional :fioods-does not appear to have developed
until th e cIose of th e first milIennium, coinciding
with th e Tiahuanaco expansion from th e centralAndes that also inHuenced this zone.
Farther south, th e incipient farming s ~ a g e is repre
sented by maize, gourds, and cotton, found in th e
trash deposits at th e ancient fishing station of Punta
Pichalo, hy Pisagua, showing evidence of cultivation
in th e vicinity oC that site heginning, apparently,
very early in th e Christian era. Nevertheless, th e
economy of these people hasicalIy depended on th e
products of th e sea : fish, s h e I l ~ f i s h , and sea mammals
[9, p. 273 an d p. 276; 8, p. 92].
In th e Atacama hinterland, population centred
in th e Calama oasis, on th e Lo a River. Irrigated
agriculture wa s th e hasis of Atacameño suhsistence,
and aqueducts are known that conveyed water íor
considerahle distances; however, the very limited
productivity of th e area impeded an y growth. Herding
wa s also of great importance; th e llamas an d alpacas
l. At Quiani 'perhaplI water was formerly seeured from a well or spring in tbebottom of tb e guUy'; at Punta Pichalo 'perbaps, in former timell, n springwall nvai1able' [9, p. 232 and p. 253J.
2. In th e phase Quinni 2, possibly eontemporaneous with Pichnlo 1 ler. 9,p. 277; 87, p. 3701.
as well as wool an d other raw materials. Llamas also
served for transportation, as pack animals, and thera
is evidence that th e Atacameños acted as middlemen
hetween th e coast and the interior, to north-western
Argentina [6, p. 38; 7, p. 600, and p. 606; 8, p. 90].·The Atacameño culture of th e Chilean oasis looks
like a pcripheral extension of a large zone of puna
adapted mixed economy, i.e., combined, limitedoasis irrigated agriculture an d herding, as mutually
complementary activities for th e exploitation of th e
meagre resources of th e enmonment (Schaede! an d
Munizaga [73, p. 33] mentioning Bowman). The anti
quity of this pattern in th e Chilean section doesno t seem to he earlier than th e latter half of the first
millennium of th e Christian era.
Th e southern Andean arid zone agriculture-and .
grazing pattern of land use extended over aU north
western Argentina-the cold desert high punas,
th e intermontane dry valleys, an d th e piedmont
steppes an d savannas to the Pampean Sierras-and
~ h e ZOne of transition hetween th e northern desert
an d th e centra! Chilean 'mediterranean' climates,in th e present da y provinces of Atacama an d Coquimbo.
The remains of agricultura! terraces are found in th e
high punas, th e Quehrada de Humahuaca, and the
Santa Victoria range in th e north.1 Their frequency
appears to diminish to the south, although th e Calchaquí Diagnita that inhahited the southern part of
th e zOne are credited with irrigation an d stone-faced
terraces for farming [8, p. 88].2 Irrigation was impor
tant in the Quebrada de Humahuaca. Speaking
in general, for th e whole area, hoth irrigated an d rain
farming tenaces are. mentioned in the availahle
summaries, hut their respective distrihution does not
seem to have heen mapped in detail.The extension of th e limited intensive agriculture
an d .grazing pattern of land use over north-western
Land use in pre-Columbian America
Argentina does not, however, seem older that th e
latter part of th e first millennium of th e Christian
era, as in northern Chile. I t was evidentIy derived
from inHuences emanating from th e central Andes,
that also brought into these regions elements of th e
Tiahuanaco style. Intensity diminishes towards th e
periphery.Although incipient farming (with maize, at least),
possihly depending on heavy summer rainfall, appearsearlier than that-perhaps in th e early centuries of th e
Christian era-in La Candelaria section (climate Cwa)
of th e Province of Salta, no details are known of this
culture's pattern of land use [85, p. 663] [for dating
see 87, figure 12, sequence chart].
At th e time of th e S p a n i ~ h exploration, nver Hood .
plain farming was practised in the tierra bañada alongthe Dulce and the Salado rivers of Santiago delEstero; native economy depended upon cultivation
of maize, quinoa, heans, and pumpkins, gathering of
wild plants, hunting and fishing, and llama herds
[55, p. 657]. Further south, the Comechingón Indians
of th e provinces of Córdoha an d San Luis, whosedwelling sites were found near strearos or small springs,
are also c:redited with cultivation an d herding hy th e
earlySpanishreports [1, p. 676]. This marks the extreme
south-eastern limit of native American cultivation.
Beyond, the Pampas an d Patagonia were inhahited
hy nomadic hunting hands.
l. For an example oC :irrigoted terracell in tb e high punllll lIee Krapovickns[42, p. ll ] on Tehenchique or Tebenqniche. an ancient settlement at3,500 m. altitudc, in tb e blenk Brea north-wellt oC Antofagasta de la Sierro.As typical examplell of terracell in the Puna, Cosanova [17, p. 620] mentionsstepped billBidell at sarate and Cnsabindo; it is no t clcar from this referencewhether those werll irngated or dry. He also m e n t i ~ o s Coctnca an d Alfarcitoin tile Quebroda de Humahuaco. In tbe Santa Victoria area, Mkrquez}liranda [56, p. 24] mentions Higuerllll and Huoira.huasi.
2. However, Márquez Miranda [51, p. 640] states tbat in the whcle Dioguitaarea ogrieulture depended on roin, an d that terraces (apparently foundonly or mostly in tbe Calchaqui suh.oreo) were a meons oí incrensing rainefficiency, while Bennett [6, p. 40] minimizes th e importanee of Diaguitaterracing.
BIBLIOGRAPHY
1. APARICIO, F. DE. Th e Comechingón an d their neighborsof the Sierras de Córdoba. In : Handbook of South.Ameri
can Indians. Vol. 2: the Andean civilizations. Washington,1946. p. 673-685. (Smithsonian Institution, Bureau of
American Ethnology, bulletin. no. 143.)
2. ARMILLAS, P. Notas sobre sistemas de cultivo en Mesoamérica : cultivos de riego y humedad en la cuenca del
río de las Balsas. Anales del Instituto Nacional de Antropología e Historia, vol. II I (1947-1948). México, 1949.p. 85-113.
3. --; PALERM, A.; WOLF, E. R. A small irrigation
system in the Valley of Teotihuacan. American Antiquity,
vol. XXI, no. 4,1956. p. 396·399.
4. BARGHOORN, E. S. ; WOLFE, M. K. ; CLISBY, K. H.
Fossil maize from the Valley oí Mexico. Botanical
Museum Leaflets, Harvard University. vol. 16, no. 9,1954. p. 229-240. .
5. BENNETT, W. C. Archaeology of the North Coast of Peru :
an account of exploration and excavatwn in Viru and
Lambayeque Valleys. New York, 1939. (The American
Meuseum of Natural History. anthropological papers,
vol. XXXVII, pt. I. )
6. --o The Andean Highlands: an introduction. In :
Handbook of South American Indians, Vol. 2 : The
Andean Civilizations. Washington, 1946. p. 1-60. (Smith-
. sonian Institutwn, Bureau of American Ethnology,
8. --; BIRD, J. B. Andean culture history. New York, oí Mexico. Botanical Museum Leaflets, Harvard Univer-
1949. (The American Museum of Natural History, sity, vol. 16, no. 6, 1953. p. 115-172.
handbook series, no. 15.) 26. ENGEL, F. Sites et établissements sans céramique de la
9. BIRD, J. B. Excavations in northern Chile. New York, cate pérnvienne. Journal de la Société des Américanistes,
1943. p. 169-318. (The American Museum of Natural Paris, vol. XLVI, 1957. p. 67-155.
History, anthropological papers, vol. XXXVIII, pt . IV.) 27. FEJOS, P. Archaeological explorations in the Cordillera
10. --o The cultural sequence oí the -North Chilean coast. Vilcabamba, south-eastern Peru. New York, 1944.
In : Handbook of South American Indians, Vol. 2: the (Viking Fund publications in anthropology, no. 3.)
Andean civilizations. Washington, 1946. p. 587-594. 28. FORD, J . A.; WILLEY, G. R. Surface survey ofthe Vira
(Smithsonian Institurion, Bureau of American Ethnology, Valley, Peru. New York, 1949. (TheAmerican Museum ofbulletin, no. 143.) Natural History, anthropological papers, vol. 43, pt . 1.)
11. --o Preceramic cultures in Chicama and Virú. A 29. FORDE, C. D. Hopi agriculture and land ownership.reappraisal of Peruvian archaeology. Menasha, 1948. Journal of the Royal Anthropological Instilute of Greatp. 21-28. (Memoirs of the Society for American Archaeo- Britain and Ireland, vol. 61, 1931. p. 357-405.
logy, no. 4.) 30. GREGORY, H. E. The Navajo country : a geographic and
12. BIÍAmWOOD, R. J.; REED, C. A. The achievement and hydrographic reconnaissance of pares of Arizona, New
early consequences of íood-production : a consideration Mexico, and Utah. Washington, 1916. (Department
of the archaealogical and natural-historical evidence. of the Interior, Geological Survey, water-supply paper,
Cold Spring Harbor Symposia on Quantitative Biology, 380.)vol. XXII, 1957. p. 19-31. 31. HACK, J. T. The changing physical environment of the
13. BREw, J. -O. Archaeology of Alkali Ridge, southeastern Hopi Indians of Arizona. Cambridge, 1942. (PeabodyUtah. Cambridge, 1946. Peabody Museum, Harvard Museum, Harvard University,papers, vol. XXXV, no. 1.)University, papers, vol. XXI.) 32. HAURY, E. W. The Snaketown canal. Symposium on
14. BRYAN, K. Flood-water farming. Geographical Review, prehistoric agriculture. Albuquerque, 1936. p. 48-50.New York, vol. XIX, no. 3, 1929. p. 444-456. (University of New Mexico, anthropological series.)
33. --o The Snaketown canal. In : G l a d y ~ . j n , H. S.; Haury,15. --o Pre-Columbian agriculture in th e south-west,
as conditioned hy perlods of alluviation. Annals of the E. W. ; Sayles, E. B. ; Gladwin, N. Excavations at Snake-
Assocíation of American Geographers, vol. XXXI, no. 4, town, l . Globe, 1937. p. 50-58. (Medallion papers, no. 25.)34. --o The excavation of Los Muertos and neighboring
1941. p. 219-242. . . ! S IR ' 11 h A C druzns ln t le a t tVer va ey, sout ern rizo na. ambri ge,16. CARTER, G. F. Plant geography and culture history in 1945. (Peabody Museum, Harvard University, papers,
the American South-tvest. New York, 1945. (Viking XXIVvol. , no . 1.)Fund publications in anthropology, no . 5.) 35. _ -o Speculations on prehistoric settlement patterns
17. CASANOVA, E. The cultures oí he Puna and th e Quebrada in th e s o u t h ~ w e s t . Prehistoric settlement patterns in the
ofHumahuaca. In: Handbook of South American Indians, New World. New York, 1956. p. 3-10. (Viking Fund
Vol. 2 : the Andean civilizations. Washington, 1946. publications in anthropology, nO. 23.)p. 619-631. (Smithsonian Institution, Bureau of American 36. HOOVER, J. W. Tusayan : th e Hopi lndian country
Ethnology, bulletin, no. 143.) of Atizona. Geographical Review, New York, vol. XX,
18. CASTETTER, E. F. ; BELL, W. H. Pima and Papago no. 3, 1930. p. 425-444.
Indian agriculture. Alhuquerque, University of New 37. HORKHEWER, H. La alimentación en el Perúprehispánico
Mexico Press, 1942. y su interdependencia con la agricultura. Lima, 1958.
19. - - ; --o Yuman Indian agriculture : primitive sub- Mimeographed. (Unesco, Programa de estudios de la zonasistence on the lower Colorado and Güa rivers. Alhuquerque, árida peruana.)
University oí New Mexico Press, 1951. 38. JENNINGS, J. D. el al. The American south-west : a
20. CHARD, C. S. New World origina: a reappraisal. Anti- problem in cultural isolation. Seminars in archaeology:
qui'y, vol. XXXIII, no. 129, 1959. p. 44-49. 1955. Salt Lake City, 1956. p. 61-127. (Memoirs of ,he21. CmZA D E LEON, P. DE. Primera parte de la crónica de,,-l__ ~ w c i e t y for American Archaeology, no. 11.)
Perú. (Several printings, edition quoted : La crónic KELLEY, J. C. Factors involved in the ahandonment oí
del Perú ' Colección Austral', no. 507. Buenos Aires, certain perlpheral south-western settlements. American
México, Espasa-Calpe Argentina, 1945. English trana- Anthropologist, vol. 54, no . 3, 1952. p. 356-387.
lation: The travels of Pedro de Cieza de León, A.D. --o Settlement patterns in north-central Mexico.1532-1550, Landon, Halduyt Society, 1864.) Prehistoric settlement patterns in the New World. New
22. COLLIER. D. Development of civilization on the coast York, 1956. p. 128-139. (Viking Fund publications in
oí Pern. Irrigation civiliza tions: a comparative study. anthropology, no. 23.)Washington, 1955. p. 19-27. (Pan American Union, 41. --o The desert cultures and th e Balcones phase :social science monographs, l. ) archaic manifestations in th e south-west and Texas.
23. COLTON, H. S. Sunset Crater : th e effect oí a volcamc American Antiquity, Salt Lake City, vol. XXIV, no. 3,
ernption on an ancient Pueblo people. Geographical 1959. p. 276-288. ..-
Review, New York, vol. XXII, no. 4, 1932. p. 5 8 2 ~ 42. KRAPOVICKAS, P. El yacimiento de Tebenchique (Puna
590. de Atacama). Buenos Aires, 1955. (Universidad· de
24. CHANE, H. R. ; GRIFFIN, J. B. University oí Michigan Buenos Aires, Instituto de Arqueologfa, Publicaciones,
Central Plains. American Anthropologist, vol. LV, no. 4,1953. p. 499·514.
81. --o Changing settlement patterns in th e Great Plains.Prehistoric seulement patterns in the New World. NewYork, 1956. p. 81-92. (Viking Fund publications in
anthropology, no. 23.)82. --o Las Praderas Centrales de Norteamérica ¿naturales
u obra humana? Estudios sobre ecología humana. Washington, 1958. p. 39-69. (Uni6n Panamericana, estudios
monográficos, 111.)
83. WEST, R. C. Cultural geography 01 the modern Tarascanarea. Washington, 1948. (Smithsonian Institution,
Institute 01 Social Anthropology, publication, no. 7.)84. -- ; ARMILLAS, P. Las chinampas de México. Cuadernos
Americanos, México, Ano IX , no. 2, 1950. p. 165-182.85. WILLEY, G. R. The culture of La Candelaria. In :
276
Handbook of South American Indians, Vol. 2 : the Andean
civilizatrons. Washington, 1946. p. 661-672. (Smithsonian
Institution, Bureau of American Ethnology, bulletin,
no. 143.)
86. --o Prehistoric settlement patterna in the Virú VaUey,
Perú. Washington, 1953. (Smithsonian Institution,Bureau of American Ethnology, bulletin, no. 155.)
87. --o Estimated correlations and dating of South
and Central American culture sequences. American
Antiquity, Salt Lake City, vol. XXIII, no. 4, 1958.p. 353·378.
88. -- ; CORBETT, J. M. Early Anc6n and early Supe
culture: Chavfn horizon aites of the Central Peruvian
Coast. New York, Columbia University Press, 1954.(Columbia studies in archaeology and etknology,
vol. III.)
89. WITTFOGEL, K. A. The hydraulic civilizations. In :Man's role in changing the face of the Earth. Chicago,University of Chicago Press, 1956. p. 152-164.
90. WOLF, E. R.; PALERM, A. Irrigation in the oldAcolhua
domain, Mexico. Southwestern Journal of Anthropology,
Alhuquerque, vol. U, no. 3, 1955. p. 265-281.91. WORMINGTON, H. M. Ancient Man in North America.