Document Sample

                             Edited by

                    M E L I N D A A. Z E D E R

                    D A N I E L G. B R A D L E Y

                     EVE E M S H W I L L E R

                      B R U C E D. S M I T H

          UNIVERSITY O F CALIFORNIA                PRESS
                 Berkeley Los Arzgeles London
                                                             CHAPTER 4

                        A Morphological Approach to Documenting
                     the Domestication of Chenopodium in the Andes
                                                          MARIA C. B R U N O

Introduction                                                              Changes in testa thickness have also been shown to
                                                                       have occurred during the domestication of t h e Andean
con lie no pods wcre grown as domesticated crop, in many
                                                                       chenopods (Nordstrorn 1990; Eisentraut 1998; Bruno 2001).
 regions of the Americas prior to l i ~ ~ r o p e a n
                                                    contact, and they
                                                                       In the Andes, however, distinguishing between wild and
are known to have been independently domesticated in the
                                                                       domesticated chenouodh based on testa thickness is corn-
Andes of South America, i r i Mexico, ancl in tllc eastern United      I,licatcd ],). vjric;!                         ci,ecies
States. 'l'oday, howelrer, while they remain an important              were utilized.
cor-nponent of the Andean diet, chenopods survive only as
a rninor crop in Mexico and are no longer cornrnercially
growrl north of Mexico. Ironically. less is known about the                Chenopods of the Andes
domestication of the Andean chenopods. (~.'l~etropodi~trnI3~11nans                     interact with a range of different chenopod taxa
qrrirrou and Cl~erropodirrtrr pullidic-ultle, than about the dornes-       in the Andes, including two domesticates (C. q~rirrou     Willd.
tication of their extinct eastern North American counter-                  and C. pallidicu~rlrAellen), several varieties of agricultural
part, (,~.hcrlarndirri ssp. jorr~siar~r~rrr.the Andes, the genetic
                                          In                               weed (C. quirrou ssp. millenrr~rnr  Aellen, also referred to as
and archaeobotanical research necessary to identify wild                   C. yrtinou var. rnelurrosperrr~~rnl
                                                                                                            Hunziker), and wild species
progenitor populations and to document where and when                      such as C. arnhrosioides Aellen, and C. hiriirr~rnr Schrader
initial chenopod domestication occurred is just beginning.                 (see Table 4.1).l C. quinoa, or quinoa, is the most widely
Narrowing down the time and place of Andean Clret~opodirrrn recognized Andean chenopod and is stiH grown by indigenous
domestication is not only essential to advancing our general               farmers throughout the Andes. The other, less well-known,
knowledge of agricultural origins in South America, but also               domesticated ct~enopodis C. pallidicazrle, or kafiawu. It is
necessary for evaluating alternativr developmental nlodels                 categorized as a "rustic" domesticate because it maintains
regarding the transition from Archaic (ca. 8000-1500 RC:) to               several wild characteristics such as self-seeding, differential
Formative (1500 BC-AD 500) period socioeconomic systems                    maturation, and easily shattering seeds (Gade 1970). These
in the south centraI Andean highlands.                                     characteristics, however, make it a reIiable crop in the high,
   In this case study, I employ approaches developed by                    dry regions of the Andean altiplano. C. qllinoa ssp. millean~trn,
researchers working o n eastern North American chenopods                   C. quinoa var. rnelar~osperrnurn,and C. hircinurn are all
to examine the seed morphology of four modern Cheno-                       spontaneous, weedy chenopods that are recognized by their
podiurn taxa encountered in the southern LakeTiticaca Basin,               black seeds and are commonly referred to as quinoa negra or
Bolivia. A set of inorphological characteristics is identified that        ajara (Hunziker 1943, 1952; Nelson 1968; Wilson 1988b,
permits the identification of these species in the archaeolo-              1990). The first two are agricultural weeds, conspecific with
gical record, and which car1 be employed in distinguishing                 quinoa, that grow throughout the Andes. C. hircinutn is a
the seeds of wild chenopods from those of domesticated crop                wild chenopod comnlon to the eastern Andean slopes and the
plants.                                                                    plains ot Argentma, Uruguay, and Paraguay (Wilson 1988a,
   In the eastern United States, investigators have shown that             1988c, 1990). Though not specifically an agricultural weed,
sceds of tlie doinesticate (,'. berlutrriit~ri  ssp. jorrrsiizrlrtrrr have C . Irircinum often occupies habitats created by h u m a n
a significantly thinner seed coat (testa) than those of its                disturbance. C. arnbrosioid~s, paiko, is a wild chenopod
wild counterpart C. Oerlandieri s s ~berlnnriieri (Wilson 1981;
                                           .                               encountered throughout the Andes and is used as a vermi-
Smith 1984, 1985a, 1985b). Accurate measurement of testa                   fuge (La Rarre 1948, 1959; Bastien 1987; Franquemont et al.
thickness with scanning electron microscopy (SEM), along                   1990).
with direct accelerator mass spectrometer (AMS)radiocarbon

                                                                                 inn   rhenopodium Domestication:
                                                                                         ~dGenetic Approaches
tion of this crop plant in this region (Smith 1984, 1985a,             Chenopods are weedy, colonizing plants that are adapted to
1985b, 1988; Smith and Cowan 1987; Fritz and Smith 1988;               disturbed habitats (Raker 1972; Holzner and Numata 1982).
Gremillion 1993a, 1993b).                                              As a result, theories of weed domestication proposed by
Section                                                                                                                                :\11111rin:1

                                                                                   Quinoa negra,                                       I'ai h o
                                                                                    ajara, ayara

Relation t o                                                                       Agrirult~~ral                                       I-hrivcs in
 Humans                                                                             weed                       disturt)cd               disturl)ecl
                                                                                                               1iat)itnts.              hi~l~it:~ts,
                                                                                                               includirig               collected f o r
                                                                                                               agricultural             medicilic

                         Southern                    Peruvian and                  Southern                  fastern ,lopes
                          C:olombia to                Bolivian altipli~no           Colombia t o              ot t h e Andes;
                          northern Chile                                            northern Chile            valleys of
                          a n d western                                             a n d western             Paraguay,
                          Argentina                                                 Argentina                 Uruguay, anti

Physical                 Tall a n d little           Short, bushy a n d            Tall with little          'Tall with little        Short, buctiy,
 Characteristics          branched with               branched with                 t o some                   t o some                and branched
                          a n exserted leafy          many small                    branching a n d            branching               with ebracteate
                          panicle                     inflorescences                diffuse t o                a n d diffuqe t o       inflorescence
                                                      concealed in                  compact                    compact
                                                      foliage                       inflorescence              inflorescence

  S O U R L E : Rased o n Aellen and J u s t 194.3; Hunziker 1943, 1952; Sirnrnondc 1965; Gdrldarillas 1074; Gade 1970; Tapia et al. 1979; Wilson 1980. 1981.
1990; Ndtional Research Council 1989.

Anderson, Harlan, d e Wet, a n d others are clearly applicable                  origin along t h e same geographic a n d phylogenetic path. It
t o C . quillon a n d C . palliriicaule (Anderson 1952; Harlan                  is therefore difficult t o place ajnrn in a basal, intermediate, or
a n d d e Wet 1965; d e Wet 1973; Harlan et al. 1973; d e Wet                   derived phyletic position relative t o q~rinoi~, its phyletic
a n d Harlan 1975; Kuznar 1993; Pickersgill 1977; Pearsall                      distance from the progenitor taxon is equivalent t o that of
1989, 1993; Smith 1992). Most regional specialists believe                      qrrir~oo"(Wilson 1990: 99). Analysis of random amplified
that the Andean domesticates arose from wild populations                        polymorphic DNA (RAPD) of C. qrrir~oaa n d C . qtritloiz var.
that colonized anthropogenic habitats (Sauer 1952; I'earsall                    rr~rlanosperm~rm Ruas et al. supports Wilson's assessment
1980, 1 9 8 9 , 1992; Kuznar 1993). Researchers have also                       that the croplweed complex is a "monophyletic, possibly
suggested t h a t t h e process of domestication resulted iri                   co-evolving unit" (Ruas et al. 1999: 26).
croplweed complexes. Genetic research has focused o n under-                       Identifying t h e wild progenitor of t h e q~ririoc2/cliclri11ez/ilic1rrz
standing t h e phylogenetic relationships of the croplweed                      crop/weed complexes has been more difficult. Wilson (1990)
complexes, as well as o n identifying their wild progenitors.                   hypothesizes that a wild North American tctraploid progeni-
                                                                                tor (C. berlnndicri) traveled t o btexico and South America via
   T h e quinon croplweed complex, consisting of the domes-
                                                                                h u m a n migrations or by long-distance bird dispersals. The
ticate C. clrtirloa a n d the black-seeded types C. quinoa var.
                                                                                p o p ~ ~ l a t i o ins each region were subsequently domesticated
                   and (I. griinoa ssp. miller~ri~r~rr, been defined
rn~~l~~nospern~rirr~                               has
                                                                                independently: C. Derlnr~rlirri jor~esiorlrttrlin eastern North
by several botanists (Nelson 1968; Heiser a n d Nelson 1974;                    ~Imerica, bcrlnnrlirri s s p i~trttullicr~~
                                                                                             C.                             in Mexico, and (J. rl~rirlocl
Wilson a n d Ileiser 1979; Heiser 1976; Pickersgill 1977; Wilson                in the Andes. Since the greatest modern diversity of qriirlon
1988b, 1990). Wilson's allozyme studies suggest that the crop                   occurs in the south central Andean highlands, researchers
a n d its weedy relatives (all of which he refers to as ujelra),"have           following a Vavilovian model have posited that domestica-
moved together from a temporal a n d geographic point of                        tion occurred in this region (Gandarillas 1974; Wilson 19882,
                                                                     11;1\ i r ~ ~ l l ~ ~ i ~(clearl!. ;I s i ~ r ~ l i~icr-c;~\c ~
  1990). L\'ll\on'\ alloi.!.nlc r C ' \ C ' d r t l i . Iloi\.t'!.~'r.                              ~l                              i r \c.c.d \ I / V tioc\ not rcprcscnt
 that an earlier forrrl ot (,'. Irir(-irr~rtt~ttic, strongexstcanclid;~tc ;In aticcluate ~ n o r ~ ) l l c ~ l o ~ ~ ~ ~ :r~01 c domestication for
                                                                                                                                       g~~n k s       I
 f c ~ rIlaving been the, wilci progenitor o! ( ' . r/lliflOrl. I.:lcY'trc)-                (,'lrc~rrc~~rodirrrir thc :21)dc,s. I<rowrn'~n
                                                                                                                in                                      ancl I'rarsall also
 1)lloretic \.ariation places (J. Irirc-itlrrrrr in a \I;IS;II ~Iositinn u g g c \ t tllat tlic t)imo(lal pattcrn ol~served sccd diameter                 in
 relative to tlic other Andc;ln tetral)loid che~lrq)ods,                      making        rne;~surc~ncllts     might retlect two (:/rt~rlopodi~rrtr perhaps   taxa,
 the prew~it-ci;~y    geographical range of (:. /?in-irrrrrrr(eastern                       (:. clrrirror~ and (.. [~lrllitlic-trrrlc, u n o f o I l o \ \ ' - ~ ~ ~ ) to test
                                                                                                                                    l~ l                        studies
 Andean slopes a n d plains) a likely locatic)rl for tlic ir1iti;ll                         this hypothesis further have I'een carried out.
domestication of (,'. qoitror~(Wilson l988c, 1990).                                              111 the snrnc time franlr., rewarchers interested in chenopod
     In contrast to rlrritrorl, very little is known allout kntiuwrl                        ciome.;tication i11 the eastern IJnitcd States also found
dornesticdtion. [lased on its weedy character, a "durn\) heap"
                                                                                            negligible differences in scr2clsize I>et\veeriwild and domestic
scenario of domestication is plausible. tleiscr ( 1 97h) suggests
                                                                                           chenopods (Asch a n d Asch 1977; Smith 1985a). In an effort
that the wild forrn of kufiuwu could have been a w e d in cl~rirro<r
                                                                                            to find other possil~le      morphological means of distinguishing
fields. Early farmers probably recognized its tolerance tv
                                                                                            between wild ,lrlcl dor~~esticatecl              chenopods, researchers
drought and freezing temperatures and began managing it.
                                                                                            in eastern North America turned t o scanning electron
Taxonomic studies have yet to verify whether a weed evolved
                                                                                            microbcopy to examine tlic n~icro-n~orphological                          features
alongside the crop. Hlinziker (1943, 1952) classified a dark-
                                                                                           of (,'lrerroporli~rrr~ seeds morc l~recisely          (Smith 1988).
                             as                            var.
seeded form of koili~wrr (:. prrllitiico~rlc~ t n e l t r t ~ o . \ p c ~ r r r ~ r ~ ~ t ~ .
                                                                                                'l'he (,'h~noporlirrrtrfruit has : round, starchy perisperrri with
[.eon ( 1964) described a "wild," dark-seeded form of kr~firrwr~
                                                                                           an ernbryo that wr;~ps          around the pcrirncter. A membranous
growing in the Lake Titicaca region, which local people called
                                                                                            layer called the testa, or seed coat, encapsulates the perisperm
                        While the "wild" f o r n ~
cl!,aru or rlliitac-l~wrj.                                     described by I.eon
                                                                                           and embryo and controls gernlination dormancy (Figure 4.1).
may represent a sympatric weed, 110subsequent studies have
                                                                                           111 1981, Hugh Wilson useel scanning electron microscopy
investigated this possibility, making it difficult to determine
                                                                                           (SEM) to cornpare the testa thickness of modern Mexican
whether a companion weed evolved along with kuilnwu.
                                                                                           domesticates h~runrontle c-llin(varieties of (. t1erlandieri ssp.
                                                                                                                            and                            :
Simmonds (1965) indicates that krztiuwa is predominantly
inbreeding and that there is only limited evidence for intro-                              t~lrttc~lliue  (Safford) Wilson and Heiser), a modern wild eastern
gression with wild or weedy taxa. To date there have not been                              North American chenopod ((J,brisl~iat~r~nr), pale-seeded        and
any genetic studies focusing o n the domestication of krlriuwn;                            archaeological chenopods from the Holman rock shelters in
however, the RAPD analysis by Ruas et al. (1999) shows that                                northwestern Arkansas. 'l'he SEM micrographs revealed
C. pallidicuule is a distinct group from the croptweed complex                             that both the Arkansas and Mexican chenopods had thin
of C. quinoa.                                                                              tcstas, whereas t h e wild seeds had thick testas. Wilson
                                                                                           suggested that humans selected seeds with thin testas because
                                                                                           they had reduced germination dormancy, another trait
Chenopodiurn Domestication: A Morphological                                                desirable in seed crops (Harlan 1975). Analysis of seeds
Approach                                                                                   from other eastern North American sites clearly established
Researchers interested in finding evidence for the initial                                 that low testa thickness (less than about 20 microns) was
domestication of Chenopodiutn in the archaeological record                                 the most reliable indicator of Chetropodiurn domestication
of the Andes have looked primarily for the kinds of changes                                in the region (Smith 1984, 1985a, 1985b, 1988; Fritz and
i n seed morphology that might be expected. The first                                      Smith 1988).
morphological analyses of Chet~opoditrtt~                      domestication in                Other morphological differences were noted between
the Andes focused o n seed size (seed diameter) because a n                                wild and domesticated chenopods, many of which are
increase in seed size is a common change associated with                                   correlated with a reduction in testa thickness. Many domes-
the domestication of seed plants (Harlan et al. 1973; Harlan                               ticated chenopods are lighter in color than are their wild coun-
1975). Pearsall's (1980, 1989) analysis of chenopod seeds                                  terparts. This is attributed to the thin, nearly transparent
from Pachamachay and Panaulauca Caves, Peru, and                                           testa, which leaves the white perisperm visible (Wilson 1981).
Browman's (1986) similar study of archaeological che~iopod Because some dark-colored seeds have thin testas (Fritz and
seeds from Chiripa, Bolivia, both recognized an increase                                   Smith 1988) and archaeological seeds are usually charred, this
in seed diameter over time, and a clear bimodal distribution                               characteristic is of limited usefulness in the analysis
of seed size (Browman 1986: 145; Pearsall 1989: 322). Both                                 of seeds recovered from a n archaeological context. Margin
Pearsall and Browman propose that the documented increase                                  configuration also differs between wild and domesticated
in seed size over time reflected selection for larger seeds under                         che~iopods:        domesticated seeds are truncate, whereas wild
domestication, but both researchers also acknowledge                                       seeds are round to biconvex (Asch and Asch 1977; Wilson
difficulties with this conclusion. None of the archaeological                              1981; Smith 1985a; Figure 1.) . Wilso~iI               (1981, 1988a) suggests
seeds in t h e assemblages they studied were as large as                                   that the reduced testa allows "fruit morphology to be
seeds of modern do~nesticatedpopulations, and the range                                    influenced by structural features of the embryo" (1988a:
of seed diameter documented for the modern domesticate,                                   487). The flat (in cross-section) cotyledons of the embryo
C. pallidicaule (1.G1.5 m m ) overlaps other wild chenopods                               create a truncate margin in t h e fruit. Finally, seed coat
(0.8-1.2 mm). (See Table 4.2.)                                                             texture varies: domesticates tend to have smooth seed coats

                                                                           1 \It1 I   -1 L
  yumnlary of pul~lishcclIlata o n :\rc~lacological(:/ic'rlo/~o,iilrm
                                                                    5cc.d \irts ,111d I.c\ta ' l ' l l i c . h ~ l c ~ \t\r o ~ nI,.;lrl>- \it?\    111   lllc .lntlc\
                                                 dnd Eastern Nortli :imc.rita

                                                                                                                  Sirt2( I I I I I I ~           Tt.\til 77ii(-krl(,\\
source                                                      .Sit ~~/l'/l(~>
                                                                       c,                              llange                     Mean                      P
                                                                                                                                                  K ~ T I S or Meti11

Andean Highlands

pearsall 1980                            I'acha~nachay,I'ertr

Rrowman 1986                             Chiripa, Bolivia
                                          300 BC:-AL) 50
                                          600-300 BC
                                          850-600 U  C
                                          1350-1000 RC

Nordstrom 1990                           I'anaulaca Cave, Peru
                                           ['ancan, I'eru (combined)
                                           AD 365-1300
                                           1620-700 BC
                                           3000-1620 BC
                                           3000 RC

Eisentraut 1998                          Quelcatani
                                          WXX: Middle Formative
                                          WXXI: Early Formative
                                          WXXII: Early Formative
                                          WXXIII: Early Formative/
                                           Late Archaic
                                          WXXVII: Archaica

Eastern North America
                                         - -

Smith and Fritz 1998                     ca. 2000 BP
                                          Russell Cave, AL
                                          Ash Cave, OH
                                          Edens Bluff, AR
                                          White Bluff, AR

Gremillion 1993a,1993b                   Late Archaic (1500-1000 BC)
                                          Newt Kash (16420)
                                          Newt Kash (16446)
                                         Early Woodland (1C)OO-300 RC)
                                          Big Bone, TN cultigens
                                          Big Bone, TN seeds
                                          Salts Cave, K (vestibuIe) cultigens
                                          Salts Cave, K (vestibule) weeds

   a Stratigraphically level WXXVII is thought to be Archaic because it falls below a stratum dated to 3600 BP. However,                 Ch~r~opo~lrrirrr from this
level were AMS dated a n d vield a date of 2740 i 50 R!
                    Seed Morphology                                    I.lsVntr;iut ( 101)H) z;irricd o11t thc jcc-ond \l.hl a11;ll21\of
                                                                    ,Andcan c h c n o l ~ o c l o n arc-hac~ologicalsprcirnens from
                                                                   Quclcdtani ( h e and C:;umata in I'eru. Occupation o f these two
                                                                   sites s p ~ n n e d thc transition betwrcn the Archaic I'eriod
                                                                   ( 7000-3500 BI') 21nd Forlnative I'rrioil (3500-2000 RP).

                                                                   \tic t o u n d seeds with tejta t h i c k n c s values of les, than
                                                                   75 lliicro~isin each level she examined. She described the
                                                                   thin-testa seed\ as generally truncate and t h e thick-testa
                                                                   heeds as biconvex, rounded, or equatorially banded. 'The
                                                                   surfaces of the thin-testa seeds were smooth or smooth and
                                                                   dimpled, while the thick-testa seeds were either smooth
                                                                   and wavy or r e t i ~ u l ~ ~ t e - i ~ v e oin t surface appearance.
                                                                                                                la e
                                                                   Scrd size was q ~ ~ i cuntistent over time (range of means
                   Margin Configuration                            O.K.3-0.92 mm with a standard deviation of only 0.90 mrn).
                                                                   .An AMS radiocarbon date of 2740k 50 RP on a thin-testa seed
                                                                   from Quelcatani Cave was the earliest date obtained in her
                                                                      'i'hese landmark studies demonstrated that the methods
                                                                   uscd to identify do~nesticatedchenopods in the eastern
                                                                   IJnited States could also be employed in documenting
        Biconvex                         Equatorially Banded
                                                                   thc. initial domestication of Clzerloporiizirn in the Andes.
                                                                   Although thesr studies have clearly shown that testa

                                                                   for different chenopod taxa in the Andes, the current study
        Rounded                                Truncate            of chenopods from the southern Lake Titicaca Basin, Bolivia
    uR E 4 , ~   h          ~seed m o r p ~ o ~ o~~ B~~~~
                                  ~       h     (after ~       ~   considers various potential, morphological indicators of dome-
                                                                    ~      i     ~      ~           ~
and Whitehead 2003).                                               stication, such as testa thickness, within a larger context of
                                                                   variability across taxa.

as opposed to the reticulate seed coats o f the wild and weedy
chenopods. Employing these features, researchers have been         Case Study: Chenopods of the Southern Lake
able to determine how, when, and where North American              Titicaca Basin, Bolivia
chenopods were domesticated (Smith 1985a, 1988, 1990,
                                                                   The southern Lake Titicaca Basin is located in the southern
1995; Smith and Cowan 1987; Fritz and Smith 1988;
                                                                   portion of the Andean nltiplano, a high ( 2 5 0 M 0 0 0 m) plateau
Gremillion 1993a, 1993b).
                                                                   hetween the eastern and western Andean cordilleras that
   A similar examination of Andean chenopods in search of
                                                                   extends from Peru t o southern Bolivia. The southern Lake
differences in testa thickness and associated morphological
                                                                   Titicaca region is dry (750 mni annual precipitation) and
changes was first carried out by Nordstrom (1990), who
                                                                   cold (9" C annual temperature) because of a combination of
considered both archaeological and modern samples from
                                                                   elevation and inverse weather patterns (Boulange and Aquize
central highland Peru. Her study demonstrated that there are
                                                                   Jaen 1981; Binford and Kolata 1996).
distinct differences in testa thickness of modern Andean
                                                                      hZodern Cl~enoporii~i~nspecies of the southern Lake Titicaca
Clienopodirirn taxa. Domesticated populations were found t o
                                                                   Basin provided a baseline of comparison for my analysis
have relatively thin seed coats (C. qziinoa 0-20 microns and       of archaeological chenopods recovered by the Taraco
C . palliriicnrrle 5-25 microns), while wild taxa had relatively   Archaeological Project (I'AP) at Chiripa, Bolivia (1500 BC-
thick testas (C. saliniurn 25-80 micror~s   and C. nrnbrosioidcs   AD 100) (Figure 4.2). Chiripa is one of the most important
15-80 micron^).^ Acknowledging some overlap in the wild            Formative Period sites in the southern Andean highlands
and domesticate populations, Nordstrom suggests that this          (Browman 1981; Mohr-Chavez 1988; Hastorf 1999, 2003),
might be due to introgression. In her analysis of chenopod         dnd it exhibits many of the key cultural developments
assemblages from two early Peruvian sites, Panaulaca and           associated with this period, including sedentary villages,
Pancan (3000 BC-AD 1300), she found that 89 percent of             long-distance trade networks, small civic-ceremonial
specimens had seed coats between 5 and 25 microns thick.           centers, and agriculture.
The earliest of the thin-coated seeds came from a level at           TAP recovered large quantities of charred Chrnopodiurn
Panaulauca Cave dating to approximately 3000 BC.                   seeds from their excavations (Whitehead 1999). Although


                                                                                               ---   Rivers
                                                                                               - Roads
                                                                                               - 200 m contour interval
                                                                                               ---               boundary

      FlG U R E 4.2   Map of the southern Lake Titicaca Basin (from Hastorf 1999)

scholars had long assumed that t h e people at Chiripa           for eneral comparison. With the possible exception of
were agriculturalists, this conclusion was based primarily on    C. an~brosioides, taxa conforrn to the patterns observed by
indirect data (Whitehead 2000). Morphological examination        Smith and Nordstrom. In the case of paiko, the uncharred
of chenopod seeds recovered from Chiripa provided the            testa thickness measurement (23.5 microns) is much larger
first direct test for domesticated chenopods and a food          than average charred measurement (range 11-14.5 microns).
production economy being present at Chiripa during the           The possible significance of this result requires investigation
Formative I'eriod (Bruno 2001; Bruno and Whitehead 2003).        beyond the scope of the project reported here.

Materials                                                         Morphological Analysis of Chenopodium Seeds

The modern Chenopodium taxa examined were C. quinoa,              Measuring Seed Diameter
C. quinoa var. melant,spermurn, C. pallidicaule, and C. ambro-   I obtained seed diameter data from two different sets of
sioides. Dr. Christine Hastorf and her students collected        specimens using two different measuring techniques. The
most of the samples in this study during previous field          two sets of specimens were (1) several large samples of
seasons at Chiripa and Tiwanaku. I also collected specimens      modern charred seeds (N = 2605) used to test the utility
from Chiripa, Tiwanaku, and Achuta Grande (a community           of seed diameter in distinguishing wild from domesticated
5 km east of Tiwanaku) in July 2000. Although most of            populations, and (2) smaller samples of modern charred and
the wild and weed seed samples were taken from whole             uncharred seeds (N = 46) for subsequent examination of testa
plants, many were contributions from the harvests of local       thickness using SEM. The digital imaging system of the
farmers and from collections they had made of the weeds          Washington University Ethnobotany Laboratory was used
that grow in and around their fields. Therefore, each sample     to determine seed diameter measurements for the first set of
represents seeds from various plants within the same species     specimens.Vhe second set of seeds was measured using a
(Table 4.3).                                                     Wild M3 dissecting microscope with an ocular micrometer.
   All the modern Cher~opodiumsamples were charred in a             In his analysis of North American assemblages, Smith
muffle furnace for con~parison    with the charred archaeo-      found that the "beak diameter" (the tip of the beak to the
logical specimens. Smith (1985a) and Nordstrom (1990)both        opposite edge) most often represented the maximum fruit
recorded a 5 percent decrease in seed diameter with charring.    diameter, as opposed to the perpendicular measurement,
Smith (1985a) also found that charring did not affect            which is the line perpendicular to the beak (Smith 1985a:
seed-coat thickness. Because these values had been estab-        120; see Figure 4.1). Because most of the seeds 1 analyzed
lished previously, I did not quantify changes in my own          tended to be more oval than circular and the "beak diameter"
charring trials. I did, however, measure seed diameter and       was not always the maximum seed diameter, the location of
testa thickness for one uncharred specimen of each taxon         my "maximum diameter" measurements varied.

                                                                     MORI'HOLOGY      OF A N D E A N C H E N O P O D I U M   37
                                             1 \I{] I      4.   {

                            \l<jlivrll ( /il~illl/lO,/;llill , l l l l [ ) I ~ ~ \
 ~       -                                            .~

 ?/Ji'c;I,\./ iY,i/ ( '111111110ll       ,\'111111'                                   1 0 1 11~iOli

(    ,(/rriii~~c~     c111i1io~il ; ~ r i c ~
                                li                                                \cliuti~Or:~ii<lv
                      cliiinoa I)lanc;i
(.'. I / I I J I I ~ ~ I                                                           .li~ril);~
( . I / I I I I Iq u i ~ i o a
                        I I ~ ~ hlanca                                          'I i ~ v ~ i n ~ i k ~ i ~ '
(:. i/r~irio,~        iluinoa l ~ l a n c a                                     1.a 11;17.mar kc.^,'
(:. ~/rri~roi~        quinoa arniirillo                                         ,\ctiuta <;rand?
(.'. q~rirroir u l ) u m a r i l l o                                            (:li~ripa"
( :. i/rri~~otr        jupa amarco                                              (;Iiiripi~<'
<,.. [ / I I ; I I (\,ar. ~ ~ I ~ , / I I I I O ~ / ~ ~ , I . I ~ I I ~ I ~ ~   <:liiripa,'
(,'. i/lii11011       var, ii10/(1t1~15/~t,r1tir1it1                            <;Iiiri[)a~'
   ~ L I [ I ; I aiiirci

         I/I~;IIO[Z        t t r c ~ / ( ~ t ~ o . ~ [ ~ t ~ i ~ i t ~ ~ ~ ('hiril)""
  ajara ncgra
(1. qrritlc~,~ ~ r i r l n r l o c [ ~ c , r ~ i r c r , ~ ~ ~                  CIiirij)aL'
  jupa ajaru
(,'. [~~~//it/ii-rir~/c~                                                        Tiwariaku
(~,./ J I I / / ~ ~ / ~ ~ - ~ ~ I I / ~ ~                                                     .
                                                                                l . a l l a ~Ir~gavi.
(I. iz1rl1)ro.t
             ioiricc paiko                                                      Chiripa

                                                                                                                  F I G U R E 4 . 3 Measuring testa thickness from a scanning
                                                                                                                  electror~micrograph. The testa is the layer of tissue o n
                                                                                                                  the left ot the micrograph. The a\;erage width is
Measuring Testa Thickness                                                                                         meds~red    with a ruler a n d recorded in centimeters. The
                                                                                                                  measurement in microns is obtained by multiplying the
A Hitachi S-450 Scanning Electron Microscope was employed                                                         centimeter value by 10,000 a n d then dividing this value
t o measure seed coat thickness (Figure 4.3).4 For each taxon,                                                    by the degree of magnification.
the metric measurements included maximum seed diameter
a n d testa thickness. Ratios of testa thickness to seed diameter                                              Testa Thickness
were also calculated in order t o provide a relative measure of
testa thickness that takes seed size into account. Nonmetric                                                    A very clear pattern emerges for testa thickness. The cultigens
traits recorded (when present) included pericarp patterning,                                                    C. I I ~ I ~ I I L J a n d (:. p~lllidil-rlrrlr a v e t h i n testas, 1 .25-3.75
                                                                                                                                     L~                       h
seed-coat texture, a n d margin configuration. These traits                                                     microns (rr = 11; x = 2.39 microns; 511 = 0.719) a n d 4.25-7.50
were best observed with t h e SEbt, b u t could also be seen                                                    microns ( r ~ 12; s = 5.95 microns; SL) = 1.11), respectively
                                                                                                                (Table 4.5, Figure 4.5). Conversely, t h e weed C. yrrinoa var.
under a dissecting microscope at magnifications of 10-50x.
                                                                                                                                              has a
                                                                                                                r~~t~lnr~usperrnrr~rr thicker seed coat of 22-51 microns
                                                                                                                (n = 16; x = 39.50 microns; 511 = 8.74). C. ambrvsioides is
Results                                                                                                         problematic because, according t o t h e charred seed coat
                                                                                                                measurements, it has a relatively small testa thickness range
Distinct patterns were observed a m o n g t h e four modern
                                                                                                                of 11-14.5 microris (11 = 6; x = 12.58; S1>= 1.59), yet t h e o n e
Clic,tlopo~ii~rr~ e x a m i n e d , especially w h e n all of t h e
                                                                                                                uncharred specimen is 23.5 microns. As mentioned above, the
               attributes were considered together.
                                                                                                                discrepancy between charred a n d uncharred testa thickness
                                                                                                                requires further inquiry. 'l'he differences in testa thickness
Seed Size
                                                                                                                among taxa are statistically significant (F= 139.23, df = 3, 41,
A total of 1.50 t o 200 charred seeds were ineasured from each                                                 /I < 0.001).

of t h e 15 samples (N= 2605) t o determine the range of seed
size for t h e four modern taxa. Of t h e four different taxa,                                                 Seed Size and Testa Thickness
C. rluitioii"s t h e largest, a n d in order of decreasing size are                                            'To assess the relationship between seed size and testa thick-
                   Z,      a
qr1i17orlI I C ~ Iknfii~wrr, n d piliko (Table 4.4, Figure 4.4). 'l'!iese                                      ness, 1 calculated a ratio of testa thickness over w e d diameter
differences are statistically significant ( I ; = 2337.18, df = 3,                                             (Table 4.5). Significant differences were recognized a m o n g
2601, p < 0.001), but t h e size ranges overlap. Therefore, iden-                                              t h e taxa (F = 353.325, df = 3, 41, p < 0.001). W h e n dia-
tification of species or do~nestication      based on seed size alone                                          grammed, this ratio provides a clear visual indication of the
is difficult.                                                                                                  obvious differences between taxa: t h e two domesticates are

 Quinoa                 1256                    1.9 159
 Quinoa negra            600                    1.4969
 Kanawa                  450                    1.1681
 Paiho                   299                    1.Oh78

    0.5   J
      N=      1256      600      450                            299
     Variety Quinoa Quinoa negra KaAawa                         Paiko
                                                                                                       Quinoa Quinoa negra
                                                                                                                             i6     12
F l G U R E 4.4 Seed diameter ( m m ) of modern Cller~opodiurn
taxa. In these box plots and those in other figures, the                                 F I G U R E 4 . 5 Testa thickness (microns) in modern
vertical lines represent the range, the dark horizontal bars                             Chenopodium taxa.
represent the mean, and the rectangles represent one
standard deviation.

                                          Small Diameter                                                    Large D~ameler
                                          Thtck Tesla                                                       Thlck Testa

                                                                                          I I
                                                                                     1 1 1
                               -     15
                               c                 E l n
                                                                  0     0
                               -                                        0            0

                                                           0      O
                                                                        0        O   0


                                                                                                       t        o

                                          Small Dlameter                                                    Large Dlameter
                                          Than Testa                                                        Thln Testa
                                                           31               32               33            34                35
                                                                             Log D~ameter

                               Fl G U R E 4 . 6 Scatterplot of log testa thickness and log diameter
                               illustrating the ratio of testa thickness to diameter for four
                               Chenopodinm taxa. Kaiiawa (C. pallidicaule) (empty circle) and
                               quinoa (C. quinoa) (black circle) are the domesticated varieties.
                               Paiko (C. ambrosioides) (empty square) and quinoa negra (C. quirrocl
                               var. melanospermrlm) (black square) are the wildlweedy varieties.
                                                         '\I" l   <.,<
          S u m ~ n ~ 1)ata t o r hlodcrn (:I/c~/~r~/~i~rli~ir/~ trorn thc Southern I.;ikc I iticac;~Bahin, Bolivia
                      ~ry                             5l1rc1rnc.n~

Modern Quinoa

                                                                                                 Smooth          'Truncate
                                                                                                 Smooth          'l'runcate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate
                                                                                                 Smooth          Truncate

Modern Quinoa negra

                                                                                                 Reticulate      Rounded
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Rounded
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex
                                                                                                 Reticulate      Biconvex

Modern Kaiiawa

                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
                                                                                                 Canaliculate    Rounded
 TtJ.‘,             lotr~llt,\ltr   1 o\q1'.lotii/                   .%,t?I                            I(c~tio   L.og Ktrtio
 T/richrrtl%%        l~/~i(~hrro.~s li.\ttr                      1 )itr!!1tTt2r     1 o<yStwi         Tt'sttt/      Tt2cttr/                               Mliryirr
 (rriic-r011.s~     /rrri~~rorr.\)" ' ~ ' I I J C ~ ~ I I ~ , % \ ( r r r r 7 r )   1~        t         t
                                                                                                    l>iurncJterc 1)iirrrrcTcr        7?xtfrrc,          (~orr/i~yrrr~rtior~
Modern I'aiL~i

 14.00                  28.00                  1.45                  .90              2.9 5             .03 1          -1.51        l'unctate           Rounded
 13.50                  27.00                  1.43                  .90              2.95              .030           - 1.52       I'unctate           Rounded
 1 1 .OO                22.00                  I .34                1.10              3.04              ,020           -1.70        I'unctate           Rounded
 I 1.00                 22.00                  1.34                 1.10              3.04              .020           I .70        Punctate            Roundetl
 1 1.50                 23.00                  1.36                 1.10              3.04              ,020           -1.68        Punctate            Rounded
 14.50                  29.00                  1.46                 1.OO              3.00              ,029           -1.54        I'unctate           Rounded

      Uecauw the w t a ~hichnessis rneasu1e~lo n onlv one side c ~ the teed, I double the tecta tl~icknessvalue 111 order to Jccount f o r the elltire area
rep~.escnted11). tile te\la when calculating the trjtddiameter r,ltio.
   1' ,A natural log was applied to transform the ralucs.

      Diameter orig~nallymeasured in mrn ancl then converted tu micron? betclrc calculating t h e ratio. To calculatr the ratio, 1 d ~ b t d rtr~taltecta
thickness (microns) hy d~nrneter( r n ~ c r o n \ ) .

                                                                    TABLE 4.6
                                            Morphological (:haracteristics of Modern C/lt7~iopoiii~rrn

                                                                                                                 C. qiiirioa var.
                               C . pallidiia~tle                                                                meianosperrn~rrrr
                                    Aellen                                C. q ~ ~ i n o a
                                                                                       Willd.                      Hunziker                     C. arnbrosioitles 1..

Common                     Kafiawa, cafiihua                            Quinoa, jupa                      Quinoa negra,                          I'ai ko
 Names                                                                                                     ajara

Status                     Domesticate                                  Domesticate                       Weed                                   Wild

Pericarp                   lrregularly punctate                         Reticulate-aveolate               Reticulate-aveolate                    Smooth

Seed Coat                  Canaliculate                                 Smooth                            Reticulate                             Foveate

Margin                     Rounded t o truncate                         Truncate                          Biconvex t o equatorially              Rounded
 Configuration                                                                                             banded

Seed Diameter              I .O-1.4 m m                                 1.6-2.2 m m                       1.4-1.6 m m                            0.9-1.1 m m

Testa Thickness            4.25-7.5 microns                             1.25-3.75 microns                 22-55 microns                          11-14.5 microns

relatively large with t h i n seed coats, whereas t h e weedy                             along t h e margins (Figure 4 . 7 ~ )Finally, C. ambrosioides has
a n d wild varieties are relatively small with thick seed coats                           a distinctive foveate o r pitted surface (Figure 4.7d).
(Figure 4.6).
                                                                                          Pericarp Patterning
Seed Coat Texture
                                                                                          W h e n the pericarp is preserved archaeologically, it t o o allows
Each taxon has a distinct seed coat texture (Figure 4.7). Like                            distinguishing a m o n g Chenopodium taxa. Both C . quirioa
its domesticated cousins in Mexico anti the eastern United                                a n d C . quinoa var. meianospermurn h a v e a reticulate peri-
States, C. quinoa has a very smooth t o slightly undulating                               c a r p (Aellen a n d Just 1943; Wilson 1980) (Figure 4.7a-b).
seed coat surface (Figure 4.7a). C . quinoil var. melano.~pt~rrnitrn                      C. pallitiicaule, however, is characterized by a s m o o t h t o
is reticulate-aveolate, again similar t o t h e North American                            ribbed pericarp (Figure 4 . 7 ~ ) Finally, C. arnbrosioides does
weedy forms (Figure 4.7b, Smith 1985a). C. pallidicrrule has a                            n o t appear t o exhibit a distinctive pattern, b u t is rather
canaliculate pattern that is most distinct near the beak a n d                            s m o o t h (Figure 4.7d).

                                                                                                M O R P H O L O G Y OF A N D E A N C H E N O P O D I U M           41
                      F I G U R E 4 . 7 Scanning electron micrographs of experimentally charred modern
                      Chenopodilrrn, showing intact testa and pericarp morphology: (a) C. qlrinocz: the thin
                      testa is intact, and a patch of the pericarp remains in the center; (b) C. qlrinoa var.
                      rnelarlosperrnrrm: the thick testa is split, and some pericarp remains; (c) C. pallidica~rle:
                      the testa is intact, and pericarp remains on the margins; (d) C. nrrlbro.sioitir.s: the testa is
                      intact, and some of the pericarp remains at the bottom of the image.

Margin Configuration                                                    Clrer~o~~oiii~rrrz and in differentiating wild, weedy, and
                                                                        domesticated specimens from each other (Table 4.6). By first
Margin configuration is discernible o n seeds that have not
                                                                        assessing the natural variation present in each taxon, one can
beer) distorted by puffing. 'The variation hetween Andean
wild and domesticated taxa is similar t o what has been                 more confidently identify morphological variation caused by
observed in eastern North America (Asch and Asch 1977;                  human selection.
Srnitll 1985a)and Mexico (Wilson 1980). (,'. clrrirroa seeds are           In the Andes it is not possible to identify a domesticated
truncate, whereas (,'. pnlliiiicnzrle is generally truncate to round.   chenopod seed based o n testa thickness alone. C. arnt~ro-
            Lrar. rnelnrlosperrrrurn can be cquatorially-banded or
(1. cl~rirlon                                                           .sioidc~s, wild plant, has a charred testa thickness of less than
biconvex. (:. arnt~rosioides a very round, almost spherical
                                has                                     20 microns. Although this may be an effect of charring, most
configuration.                                                          archaeological Cl~~riopoiii~rtrr            assemblages in the .Andes are
                                                                        charred. As a resuIt, it is necessary to identify the taxon first
                                                                         sing nonmetric features. Seed coat texture, pericarp pat-
                                                                        terning, and to a lesser extent margin configuration are diag-
This case study presents a set of morphological characteris-            nostic traits for distinguishing among C. qirinon, C. i1uinon var.
tics t h a t aid in the identification of individual Andean                                        ~ m,
                                                                        n ~ e l n r ~ o . ~ p e n nC.r palliiiicurrle, and C. nrnbrosioides.
   Within a given ,pecic\, tcstil thicklle\5 and tllc testa                                      Gavle Fritz, I'att)' J o Watson,
                                                                    tile figures. [)avid Hrowm;~n,
thickuess/seed size ratio are useful for separating domestic.       Lisa Hildebrand, Christine Hastorf, and Jason Kaufman
varieties from both weedy and wild varieties. The differences       provided guidance in research and useful comments on this
in testa thickness between (,'. q~rirloaand C . rllrinoa var.       article. Michael Veith piloted thc SEM in the biology depart-
~elanosprrrnnmcan be attributed to hurnan selection for             ment at Washington University. Finally, I would like to thank
          germination dormancy. Selection for a larger secd         Bruce Smith a n d Melinda Zeder for the invitation to
size may also have occurred, because the modern C. yuino,i          contribute to this volume. The research was funded by
seeds are significantly larger than are their black-seeded          t h e Washington Urliversity in St. Louis School of Arts
                                                                    and Science's Pre-Doctoral Summer Research Grant (2000),
                                                                    Summer Writing Grant (2001), and Faculty Research Grant
                                                                    (2001 awarded to David Browman).
Conclusions and Directions of Future Research
 Comparative analysis of modern chenopod taxa and archaeo-          Notes
logical specimens resulted in the identification of a quinoa-like
form and a quinoa negra-like form in the archaeological               1. This is not a n exhaustive list of Andean chenopods,
samples from Chiripa, indicating that domesticated                       but it does represent the most commonly cited
chenopods were under cultivation at Chiripa as early as 1500             species found in arlthropogenic habitats.
BC (ca. 350CJ BPIh (Bruno 2001; Bruno and Whitehead 2003).            2. C. salir~iurnis a dark-seeded species.
This provides the earliest evidence to date of domesticated
                                                                      3. The system includes a Sony CCD black-and-white
chenopods in the Andes based on direct AMS dates from
                                                                         video camera attached to an Olympus SZ-P-f
charred chenopods.
                                                                         microscope that feeds into a Gateway PC. A live
   Eisentraut (1998) obtained a direct AMS date of 2740        +         image is captured from the microscope a n d displayed
50 BP calibrated o n a thin-testa seed identified from Quel-
                                                                         o n the computer monitor ilsing the image analysis
catani Cave, Peru. Thin-testa seeds come from levels at                  software Scion Image. The microscope and imaging
Panaulauca Cave, Peru, dating (by conventional radiocarbon               software are calibrated to a millimeter scale. a n d
dating) to as early as 3000 BC (ca. 5000 BP) (Nordstrom                  measurements can be taken o n the screen using the
 1990). Based on these independent studies, ca. 5000 BP would            software's "analysis" function. Each measurement is
appear t o be a liberal estimate of when Chenopodium was                 automatically recorded in Scion Image. I saved
first brought under domestication in the central Andean                  these data as a text file a n d transported them t o a
highlands. A more conservative estimate of around 3500 BP                statistical software package, SPSS 8.0, which I used for
could also be proposed for the south central Andean                      all statistical analyses.
highlands. Chenopodium remains from associated contexts
                                                                     4. Seed samples were mounted o n aluminum SEM
in highland Peru and coastal Chile date almost 2000 years               specimen mounts using 12-mm diameter Carbon
earlier (Pearsall 1992; Kuznar 1993; Aldenderfer 1999), but             Adhesive Tabs (Electron Microscopy Sciences,
their domesticated status has not been verified. While the              Ft. Washington, PA 19034; cat. #77825-12). After
domesticated chenopods from Chiripa and Quelcatani are                  measuring the diameter under the light microscope,
probably forms of C. q~tinoa, is not clear whether the other            specimens were cut in half using a razor blade, and
early examples are quinoa or kafiawa. Clearly, more research,           placed o n a stub with the newly exposed surface
both genetic and morphological, is needed o n the Andean                facing up. Mounted samples were coated with about
Chenopodium taxa.                                                       50 n m of gold in a Polaron ES000 Sputter Coater.
   1 recommend that future morphological studies of                     Micrographs for all specimens were taken at
chenopod domestication in the Andes follow the multiple                 magnifications of lOOOx to 4000x. Images of the
attribute comparative approach outlined here, which takes               desired area were captured using Polaroid 55 PIN
into account t h e taxonomic diversity of chenopods in                  (positivelnegative) film. Because the SEM employed
the Andes. Direct AMS dating of identified taxa from                    did not have on-screen measuring capabilities, testa
different regions and time periods will greatly advance our             thickness measurements were taken directly from t h e
understanding of Chenopodium domestication in the Andes.                micrographs (Figure 4.3).
                                                                     5. Several varieties of quinoa are available today, two of
                                                                        which were included here: q~linoa   blancu and quinoa
I would like to thank the farmers of Chiripa, Tiwanaku, and             amarilla. Although there was some variation in seed
Achuta Grande for sharing their knowledge and chenopods                 size between these populations, differences between
for this study. 1 also t h a n k Christine Hastorf, William             the testa measurements were not statistically
Whitehead, a n d t h e Taraco Archaeological Project for                significant. Therefore, 1 have included them all
giving me the opportunity to carry out this project. William            under the single category "quinoa" for this
Whitehead a n d Eduardo Machicado helped draft several of               discussion.

                                                                      M O R P H O L O G Y OF A N D E A N C H E N O P O D I U M   43
     6. l'his AMS date was obtained o n f~vc,      charred                   I.ritr, C;. I. and li. 1,. Smith. lLISX.Old collectlonc and ncw
                        seeds larger t h a n 1.2 Inrn frorn a tingle            technology: I )ocumenting thc. domestication of Chc~r~c~(~odirrm
        context. In t h e seeds that remained in thi, ,ample,                   in eactern North .America. Mitt~-or7tirrc~r7lrII   /orrrrrill of /Irclr[~ro/qyy
        I identified domesticated specimens (see Whitehead                       13: 3-27.
        1999 for details o n radiocarbon dates and Bruno and                 Gade, D. W. 1970, tthnobotany of Caiiihua ((:/reno(~oriilrnr
        Whitehead 20U.i for details o n the analvtis ot the                     ()c7llicli:.~rrrl~'), seed crop of the Altiplano. Ec-orlornic Roterrry
                                                                                34: 55-61.
                                                                             Gandarillas, H. 1974. Grrreticn y oriirin dc' la qlrirlia. La Paz, Uolivia:
                                                                                lnstituto Nacional del 'l'rigo, Departmento de Estudios
References                                                                      Econornicos, Estadisticas y Comercializacion.
                                                                             Gremillion, K. J. 1993a. Crop and weed in prehistoric eastern
 Aellen, P. and T. Just. 1943. Key and synopsis of the .American
   species of the genus Chenopodi~imL. Anreric-arr Midlorr~i                    North America: The Cherropodirrnr example. American Aritiqlrity
   Nahrralist 30: 47-76.                                                        58: 496-509.
Aldenderfer, M. S. 1999. The Late Preceramic-Early Formative                .            1993b. 'The evolution of seed morphology in domesticated
   transition on the South-Central Andean Littoral. In Pacific                  Cllerroporiilrrn: An archaeological case study. /o~rrr~al                  of
   Lotirr Anlt2rica irl prehistory: Tllr cvol~ltiorr of archaic urrd            E t l r n o b i o l ~ ~ ~ 149-169.
   formative crrltrrres, Michael Blake (ed.), pp. 213-222. Pullman,          Ilarlan, J . R. 1975. Crops ilrril rrrcrrr. Madison, Wisc.: American
   Wa5h.: WSU I'ress.                                                           Society of Agronomy.
Anderson, E. 1952. l'latrtc, marl, arid life. Uoston: Little Brown.          Harlan, J . R. a n d J . M. J. de Wet. 1965. Some thoughts on weeds.
Asch, D. L. and N. R. Asch. 1977. Chenopod as cultigen: A                       Econottric Rotatly 19: 16-24.
   re-evaluation of some prehistoric collections from eastern               Harlan, J. R., J.M.J. de Wet, and E. G. Price. 1973. Comparative
                                                               ~ 3-45.
   North America. Midcontinental /oumal o f A r c l ~ a u o l o2:~ ,            evolution of cereals. Evolutiotr 27: 311-325.
Baker, H. G. 1972. The evolution of weeds. Ann~ralre vie^^ of               Hastorf, C. A. 1999. Introduction to Chiripa and the site area. In
   Ecolop arid Systenratics. 5: 1-24.                                           Early settlement a t Clriripu, Bolivia: Research of tlre Tarac.o
Bastien, J. W. 1987. Healers of the Andes: Kallowoya herl)alist.\ and           ArchaeologicalProject, C. A. Hastorf (rd.),pp. 1 4 . Contributions
   their tnedicinalplants. Salt Lake City: University of Utah I'ress.           of the Archaeological Research Facility No. 57. Berkeley:
Binford, M. W. and A. I,. Kolata. 1996. The natural and human
                                                                                Archaeological Research Facility, University of California at
                                                Archaeology arrdpaliJoe-
   setting. In Tiwonakrr rrrrd its lrintc>rlarrd:
   cology of a n Andean c-ivilization, A. L. Kolata (ed.), Vol. 1,
                                                                            Hastorf, C. A. 2003. Community with the ancestors: Ceremonies
   pp. 57-88. Washington, D.C.: Smithsonian I'ress.
                                                                                and memory in the Middle Formative at Chiripa, Bolivia.
Boulange, B. and E. Aquize Jaen. 1981. Morphologie, hydrologra-
                                                                               lorrnrnl of Antliropological A n t k r ~ p o l o ~ q 305-332.
   phie et climatologie du Lac Titicaca et de son Basin Versant.
  Revue d'Hydrobiologie Tropicale 14: 269-287.                              Heiser, C. B. 1976. Origins of some cultivated new world plants.
Browman, D. L. 1981. New light o n Andean Tiwanaku. Americarl                  Anrlual Review of Ecology and Systematics 10: 309-326.
  Sc-ientist 69: 408-419.                                                   Heiser, C. B. and D. C. Nelson. 1974. O n the origin of cultivated
.        1986. Chenopod cultivation, lacustrine resources, and fuel            chenopods (Chenopodium). Genetics 78: 503-505.
   use at Chiripa, Bolivia. In New worldpaleoetlrrrobotany: Collected      Holzner, W. and M. Numata. 1982. Biology and ecolog of weeds.
  papers in hotlor ofLeonard W Blake, E. E. Voigt and D. M. Pearsall           The Hague: Junk.
   (eds.), pp. 137-172.7heMissouriArchaeologist Vol. 47. Columbia:         Hunziker, A. 1943. Las especies alimenticias de Atnarantus y
   Missouri Archaeological Society.                                            Chenopodium cultivadas por Los lndios de America. Revista
Bruno, M. C. 2001. Formative agn'culture? The status of Cheno-                 Argentina Agronomic-a. 10: 1 4 6 1 5 4 .
  podium domestication and intensificntiorl at Clliripa, Bolivia           .            1952. Los pscwdoc-erea1e.i de la agn'mlhrra irtdt'gena de Amirin'ca.
   (1500 B.C.-100 B.C.). Master's Thesis, Department of Anthro-                Buenos Aires: Acme Agency.
  pology, Washington University in St. Louis.                              Kuznar, L. A. 1993. Mutualism between Chenopodium, herd
B ~ n o M. C. and W. T. Whitehead. 2003 Chenopodium cultiva-
          ,                                                                    animals, and herders in the south central Andes. Moutltain
  tion and Formative period agriculture at Chiripa, Bolivia. Latirl            Research arld Devrloptrretrt 13: 257-265.
  Americarr Antiquity 14: 339-355.                                         La Barre, W. 1948. The Aymara Indians of the Lake Titicaca
de Wet, J. M. J. 1973. Evolutionary dynamics of cereal domesti-
                                                                               Plateau, Bolivia. Memoir Series of the American Anthropolog-
  cation. R~rllrtirlof the Torrey Botanicul Club. 102: 307-312.
                                                                               ical Association No. 68. New York: Krai~s.
de Wet, J. M. J. and J. R. Harlan. 1975. Weeds and domesticates:
                                                                           .            1959. Materia medica of the aymara Lake Titicaca Plateau,
  Evolution in the man-made habitat. Economic Rotan). 39:
                                                                               Bolivia. Wehbia 15(1):47-49.
                                                                           Leon, J . 1964. Plantas alimetlticias Andinos. Lima, Peru: lnstituto
Eisentraut, P. J. 1998. Macrobotanical remains from southern
  Peru: A comparison of Late Archaic-Farly Formative period sites              lnteramericano de Ciencias Agricolas Zona Andina.
  from the Puna and Suni zones of the Western Titicaca basin.              Mohr-Chavez, K. L. 1988. The significance of Chiripa in Lake
  1'h.D. dissertation, Department of Anthropology, University of               Titicaca Basin developments. Expeditiotl 30: 2, 17-26.
  California, Santa Barbara. Ann Arbor, MI: University of                  National Research Council. 1989. Lost crops of the Olcas: Little-
  Michigan Microfilms.                                                         known plants of the. Andes with protnise for worldwide cultivation.
Franquemont, C., T. Plowman, E. Franquemont, S. King, C. Niezgoda,             Washington, D.C.: National Acaderny Press.
  W. Davis, and C. Sperling. 1990. The ethnobotany of Chin-                Nelson, D. C. 1968. Taxonomy and origins ofchenopodium quinoa
  chero, a n Andean community in southern Peru. Fieldiatra,                    and Chenopodium nuttalliae. PhD Dissertation, Indiana
  Botany No. 24. Chicago: Field Museum of Natural History.                     University. Ann Arbor, MI: University of Michigan Microfilms.

 xord\tr()nl, (,. IC)')O. I \,idcncc lor t l ~ c o r n c s l l ~ , ~ t ~ 1o (r ~ c r 1 0 -
                                                                       0       /r            -.    -    lc)02,I llc t10o~lpl~1111 t l ~ c o r ~p1d11tci(>~~lc>lic.~it~c)r~
    pot/i~rttrIn the i\ndcs. I(cl)ort to the Nation,~l lcncc toundn-
                                                           h                                      Irl                                                      I S 13.
                                                                                                      caytcrll North Arncr~c,~. Ri\8i,t> C / I I I ~ Ily ~ . ,1). Smith,
                                                                                                                                  In            of
    tion. IJnivcr,it)' of California, Rcrkcley l'nlcoethnohot,tny                                1111. 19-.(.(. Wdshington. I).('.: Sn~lttisonianInctitution I'rcss.
    1,aboratory Keportc #19.                                                                  .                                                 in
                                                                                                        1995. Seed plant doniecticat~on ?astern North America.
 pearsall, I). M. 1980. lithnobotanical report: I'lant utilization at                            In Lcrst Irrirrtc~rc-firctfarttrrr\, 7'. I ) . I'rice and A. D. Gebauer
    a hunting bate tamp. In I'rc~historic111rt~tcr.c the Iriaylril tr,lt,\.                      (cd,.). 1'1.' 193-211. Santa Fe: Schoc1l ot American I(esearcli
    ]ohn Rick (ed.),pi). 191-2.3 1. New York: Academic Pry55                                     l'ress.
 -         1989. Adaptation of prehistoric hunter-gatherers in the                            Smith, B. I). and C. W. Cowan. 1487. IDoniesticated Clrenopodirrtt~
    High Andec: The changing role of plant resources. In Forrr<yir(y                             in prehistoric eastern North :\rllerica: New accelerator dates
    atld furtnitrg, n.Harris and G. Hillman (eds.), pp. 318-332.                                 from eastern Kentucky. Attleric-(211       :ltrtiqlrip 51: 355-357.
    1.ondon: Unwirr fryman.                                                                   Tapia, M., 13. Gandarilla,, S. Alandia, A . Cardozo, A. Mujica,
           1992. Origins of plant cultivation ir1 South America. In                              K. Ortiz, V. Otazu, J. Rea, and D. Salas. 1979. Quinia y kaniwa
    n r e origitls of ngricult~rre,C. Cowan and P. J. Watson (eds.),                             cultivos Andinos. (:enfro lntert~ncionalde In\~ecti~q~rlcic~nec     para
    pp. 173-205. Washington, D.C.: Smithtonian Institution Prc~5.                                I.'/ Desarrolo ((:/ID). lnstituto Interamericano de Ciencias
-.         1993. IDomestication and agriculture in the New World.                                Agricolas (IICA). Editorial 1IC:A.
    In The lust hlrtlterc-first furrncrs, T. D. Price and A. H. Gel~auer
                                                                                              Whitehead, W. T. 1999. I1aleoethnobotanical evidence. 111
    (eds.), pp. 157-192. Santa Fe: School of American Research
                                                                                                 Early sc,ttlemet~t at Chiripci, Holiviu: Resenrch of the Tnraco
                                                                                                /2rd1ueologicalProject, C. A. Hastvrf (ed.), pp. 95-104. Contri-
I'ickersgi!l, B. 19 77. Biosysternatics of crop-weed complexes.
                                                                                                butions of the Archaeological Research Facility No. 57. Berkeley:
    Kulturpflanze 29: 377-388.
                                                                                                 Archaeological Research I:acllity, University of Califor~iaat
Ruas, P. M., A. Bonifacio, C. F. Ruas, D. J . Fairbanks, a n d
    W. R. Andersen. 1999. Genetic relationship among 19 acces-
                                                                                             -          2000. I'erspectives on agriculture from Formative and
    sions of six species of Chenopodi~rrnL., by Random amplified
                                                                                                'fiwanaku sites in the Bolivian Altiplano. Paper presented
    polymorphic LINA fragments (RAPD). Eiipl1,vtica 105: 25-32.
                                                                                                at the 65th Annual Meeting of t h e Society of American
Sauer, C . 0. 1952. Ayric~rlturaloricyit~s    nnd dispt'rsals. New York:
                                                                                                Archaeology, Philadelphia, PA, April 7.
    America11 Geographical Society.
                                                                                             Wilson, H. D. 1980. Artificial hybridization a m o n g species
Simmonds, I' W. 1965. The grain chenopods of the tropical
                                                                                                of Chenopodium sect. Chenopodilrtn. Systetnatic Botany 5:
    American highlands. Econotnic Botnt1,v 19: 223-235.
Smith, B. D. 1984. Chenopotiilrttr as a prehistoric domesticate                                 253-263.
    in eastern North America: Evidence from Russell Cave,                                    .
                                                                                             -           1981. Domesticated Chenopodium of the Ozark Bluff
    Alabama. Science 226: 165-167.                                                              Dwellers. Economic Botany 35: 233-239.
.          1985a. The role of Cherropodilrttr as a domesticate in                            .
                                                                                             -          1988a. Allozyme variation and morphological relation-
    pre-maize garden systems of t h e eastern United States.                                    ships of Chenopodiutn hircinum. Systematic Botany 13: 215-228.
   So~rtlit~ostern  Archaeology 4: 5 1-72.                                                   .
                                                                                             -          1988b. Quinua biosystematics: Domesticated popula-
.         1985b. Chrnopodiuttr berlnt~dieri jonesianum: Evidence
                                              ssp.                                              tions. Economic Botany 42: 461-477.
   for a Hopewellian domesticate from Ash Cave, Ohio. Southea-                               .
                                                                                             -          1988c. Quinua biosystematics 11: Free-living populations.
   stern Archaeology 4: 107-133.                                                                Economic Botany 42: 478-494.
.         1988. SEM and the identification of micro-morphological                            .
                                                                                             -          1990. Quinua and relatives (Chetlopodium sect. Cheno-
   indicators of domestication in seed plants. In Scantling electron                            podium subsect. Cellulata).EconomicBotany 44 (Suppl.): 92-110.
   microscopy in archaeology, S, L. Olsen (ed.), pp. 203-214. BAR                            Wilson, H. D. and C. B. Heiser. 1979. The origin and evolu-
   International Series 452. Oxford: BAR.                                                       tionary relationships of 'Huauzontle' (Chenopodium nuttalliae
-          1990. Origins of agriculture in eastern North America.                               Safford), domesticated chenopod of Mexico. American /ournal
   Science 246: 1566-1571.                                                                      of Botany 66: 198-206.

                                                                                                M O R P H O L O G Y OF A N D E A N C H E N O P O D l U M            45