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.
L.) Achenes from Mexico and Eastern North America 1
SOMAYEH S. T ARIGHAT2, D AVID L. LENTZ*,2, STEPHEN F. M ATTER
2, AND
R OBERT B YE3
2Department of Biological Sciences, University of Cincinnati, Cincinnati, OH 45221 USA 3 Jardín Botánico, Instituto de Biología, Universidad Nacional Autónoma de México, 04511 México,DF, México
Morphometric Analysis of Sunflower (Helianthus annuus L.) Achenes from Mexico andEastern North America. Sunflower (Helianthus annuus L.) has played a major role in theevolution of agricultural systems in the Americas. The discovery of ancient domesticatedremains from archaeological deposits in pre-Columbian Mexico offers new dimensionsto widely accepted viewpoints on the domestication pattern of H. annuus. AlthoughAmerican sunflower populations north of Mexico have been examined extensively, Mexicanindigenous domesticated landraces have not been studied in any detail. In this study,we morphologically assessed wild and domesticated sunflower achenes from Mexicoand compared them to similar datasets from eastern North America. Additionally, weevaluated the utility of four computer-assisted shape measurements in discriminatingbetween wild and domesticated sunflower achenes (fruits) and compared variation inachene size among modern wild and cultivated populations from both Mexico and the U.S. We
found that, of the shape parameters tested, none were informative in distinguishing wild achenesfrom domesticated varieties. Subsequent size analysis, using conventional parameters of length,width, and thickness, showed that modern wild populations from Mexico had smaller achenescompared to modern populations from eastern North America. Domesticated achenes unearthedfrom Mexican archaeological sites, however, were significantly larger than the early domesticatedspecimens recovered from eastern North America. Our methodological results indicatethat variation in archaeological sunflower achenes is better described by conventionalsize parameters rather than computerized shape analysis.
Análisis Morfométrico de Aquenios de Girasol (Helianthus annuus L.) de México y Este delos Estados Unidos de América. El girasol (Helianthus annuus L.) ha jugado un parte importanteen la evolución de los sistemas agrícolas en las Américas. El descubrimiento de restos domésticosantiguos de yacimientos arqueológicos en el México precolombino ofrece nuevas dimensiones
a los puntos de vista ampliamente aceptados en el patrón de domesticación deH.
annuus. Aunque las poblaciones de girasol de México y Estados Unidos de América (EUA)han sido examinadas ampliamente, los restos arqueológicos domésticos de las razas indí-genas de México no han sido estudiados en detalle. En este estudio, se evaluaron morfo-lógicamente aquenios de girasol silvestre y domesticado de México y se compararon conconjuntos de datos similares del este de EUA. Además, se evaluó la utilidad de cuatromediciones de forma asistidas por computador en la discriminación entre los aquenios(frutos) de girasol silvestre y domesticado además la variación en comparación de tamaño entrelos aquenios modernos de poblaciones silvestres y cultivadas de México y los EUA. Se encontróque de los parámetros probados ninguno fue determinante para distinguir entre los aqueniossilvestres y las variedades domesticadas. El análisis de tamaño, utilizando por los parámetrosconvencionales de la longitud, anchura y espesor, mostró que las poblaciones modernasde México presentan aquenios más pequeños, en comparación con las poblaciones modernasdel Este de EUA. Sin embargo, los aquenios domesticados desenterrados de los sitios arqueoló-
1 Received 19 April 2010; accepted 21 July 2011;published online 13 August 2011.
gicos de México fueron significativamente más grandes que los ejemplares domesticadosy recuperados del Este de EUA. Nuestros resultados indican que la variación metodo-lógica para aquenios de girasol arqueológicos está mejor descrita por los parámetros detamaño convencional en lugar del análisis de la forma informática.
Key Words: San Andres, Cueva del Gallo, shape factors, sunflower, Helianthus annus.
IntroductionRecent reports of pre-Columbian sunflower
(Helianthus annuus L.) in Mexico (Bye et al.2009; Lentz et al. 2001, 2008a , b; Pope et al.2000) have focused attention on sunflower populations from that country. Unfortunately,
very little is understood about sunflower in
Mexico, either wild or domesticated. This study is an effort to provide comparable morphometricdata for assessment of modern and archaeologicalsunflower populations from Mexico and easternNorth America. Aside from its importance as oneof the world’s major oilseed crops (Steffansson2007), the study of sunflower bears an evolu-tionary significance to our understanding of ancient cultures of the pre-Columbian Americas.Moreover, the study of germplasm from warmer climes takes on an additional significance as
global warming and other aspects of climatechange become manifest. Archaeological records of H. annuus consist
mainly of achenes because these are the parts of the plant most likely to be preserved. Convention-ally, achene size has been the key indicator of domestication in H. annuus and, through thestudy of modern wild and domesticated popula-tions, achenes longer than 7 mm have beencategorized as structures derived from domesticatedplants (Heiser 1985). Determining whether anunearthed achene is domesticated or wild anddating these remains play a key role in reconstructing the domestication trajectory of H. annuus .
Seed identification and classification coupled withthe application of novel methods, which are faster and more reliable, have significant economic andtechnical importance in studies of seed populations(Granitto et al. 2000). In this context, moderncomputerized shape analysis methods have addedto the knowledge base developed through conven-tional size analysis. Image processing methodsapplied to plant morphology have been useful in
distinguishing wild from domesticated varieties as well as crop seeds from contaminants (Zayas et al.1989). Russ and Rovner (1989) utilized computer-assisted image analysis of plant phytoliths to
discriminate between wild Zea varieties and domes-ticated maize. The main measured shape factors,including formfactor, roundness, solidity, curl, andaspect ratio, proved to be informative in distin-guishing wild and domesticated maize populations.
Despite the growing application of computers inmorphometric approaches, there are not many
examples of computer-assisted studies involving seed populations (Rovner and Gyulai 2006). Travisand Draper (1985) used an image processing system to distinguish crop seeds from contaminant
weed species by measuring the size of the seeds andone shape parameter (formfactor). Their resultsindicated that fi ve of seven crop species could beseparated from weed species based on thesecharacteristics and that most of the weed seedscould be distinguished from each other using thissystem of measurements. Granitto and colleagues
(2000, 2002, 2005) also evaluated computer-assisted image processing methods examining mor-phometric (size and shape) characteristics of seedsand these techniques have been shown to be highly effective in automated weed seed identification.
Rovner and Gyulai (2006) developed andapplied new seed taxonomic approaches using computer-assisted image processing. These freshapproaches include measuring shape parameterssuch as formfactor, curl, convexity, and round-ness, which cannot be obtained through manualmethods. Heretofore, standard classification of seeds in sunflower has been based on sizeparameters usually represented by a composite
value called a “size index ” (length x width).Computer-analyzed shape features of seeds, whichmay be informative in distinguishing wild seedsfrom domesticated varieties, have not beenutilized in examining H. annuus . Developing new means of seed differentiation would enhanceclassification and identification endeavors inarchaeological contexts as well as in modern seedtaxonomy. Fast and reliable computer-facilitated
approaches could be of great benefit in this fieldof scientific research.
We sought to test the utility of modern imageprocessing approaches by measuring and assessing
261TARIGHAT ET AL.: MORPHOMETRIC ANALYSIS OF SUNFLOWER 2011]
four shape parameters (i.e., formfactor, curl,convexity, and roundness) to discriminatebetween wild and domesticated sunflower achenes. We used a dataset that included
numerous samples of Mexican indigenous culti- vars and wild populations (Figs. 1 and 2)collected in Mexico during the course of a previous study (Lentz et al. 2008a , b). Further-more, we compared size information frommodern data to records of archaeological evi-dence from Mexico and eastern North America (ENA) in search of further insights into domes-tication patterns of this important crop plant.
Materials and Methods
Groups of seeds collected from Mexico andENA, both wild and domesticated, were used for this study. ENA indigenous domesticated land-races were obtained from the United StatesDepartment of Agriculture (USDA). Mexican
wild and indigenous domesticated landraces andENA wild populations were collected by Lentz
and Bye during field studies conducted from2001 through 2005. To minimize the potentialeffect of introgression between cultivated sun-flowers with conspecific wild H. annuus popula-
tions on achene size (Rieseberg et al. 1999),diligent sampling was conducted by collectors,
who selected only plants exhibiting wild pheno-type characteristics (i.e., multiple slender pediclesand small flower heads; Doebley et al. 2006).Overall, seeds from 27 wild and domesticatedpopulations including 14 wild (10 Mexican and 4eastern North American) and 13 domesticatedlines (8 Mexican and 5 eastern North American)
were examined in this study (Fig. 1). A completedescription of the Mexican sunflower domesti-
cates can be found in Tarighat (2010). For mostof the populations, 100 seeds were measured, butfor fi ve populations, only 40 seeds were availablefor measurement.
Three principal size dimensions, namely length, width, and thickness, were measuredusing a Vernier caliper (accurate to 0.01 mm)
Fig. 1. Map of North America showing modern distribution of wild sunflower (Lentz et al. 2008a ; Heiser et al.
1969). Closed circles show the collection areas of modern wild H.annuus in Mexico and ENA. Stars indicatearchaeological sites where domesticated sunflower achenes were discovered. Three well-preserved achenes (330–250 B.C.E.) were unearthed at the Cuevo del Gallo site and a seed and achene (2875–2482 B.C.E.) werediscovered at the San Andrés site. Two of the oldest ENA sites with clearly domesticated sunflower remains werethe Higgs site (1259–829 B.C.E., Brewer 1973), and Marble Bluff (1264–912 B.C.E., Fritz 1997).
for each individual seed. Due to the irregular shape of sunflower seeds, the widest points for both length and width were recorded. A single
value, size index = length (mm) x width (mm), was calculated for each seed.
Shape factors were measured using an image-processing program (MATLAB version 7.4).Modern seeds were placed in groups of 50 seedsfor domesticated and 100 seeds for wild popula-
tions per image and randomly numbered by computer. Images of seeds of all populations,taken using a digital camera, were analyzed for four shape factors. Definitions of the shape
features follow Russ (2007) and are provided inTable 1.
Computer-assisted shape measurements wereobtained using digital images of more than oneobject in an image. Color digital images wereconverted to binary (black and white) images, whilean edge detection algorithm traced the exterior boundaries of each object using a threshold value.
We optimized the threshold value for each image toget the best boundary detection. Parameters werethen simultaneously measured in a two-dimensional
binary image for each individual object. TheMATLAB script developed for image analysis isavailable upon request from the corresponding author.
We initiated the computer-assisted measure-ment process using four selected shape parame-ters: formfactor, roundness, convexity, and curl.Formfactor is the configuration of the perimeter relative to the area of an object. A perfect circlehas the smallest ratio of perimeter to area and a formfactor value of 1. Departure from a true
circle will result in a smaller formfactor value(Rovner and Gyulai 2006). Roundness is similar to formfactor but it is measured using the length(longest chord) of the objects rather than their perimeters. This property of roundness makes itmore sensitive to how elongated the object is,
whereas formfactor is more responsive to irregu-larity of the outline of the object. There are many examples where formfactor and roundness of a single object vary significantly (Russ and Rovner 1989), suggesting that both parameters areimportant for a full evaluation of an object.Convexity is used to measure the degree of irregularity of an object, quantified by the ratio
Fig. 2. Sample digital images of domesticated and
wild Mexican sunfl
owers set for computer-assisted imageanalysis. Images of multiple achenes were uploadedto the MATLAB image analysis program to obtain shapemeasurements.
T ABLE 1. DESCRIPTION AND FORMULA OF SHAPE FACTORS ASSESSED IN THIS STUDY (RUSS 2007).
Variable Description Formula
Formfactor Departure of an object’s perimeter from a smooth circle
4π area/perimeter 2; it is 1.0 for a perfectcircle and diminishes for irregular shapes.
Convexity The ratio of a convex perimeter to perimeter
Convex perimeter/perimeter. It diminishesif there are surface indentions.
Curl Degree of departure of anobject from a straight line Length/*fi
ber length
Roundness Departure of an object from a circle 4 area/ π x length2
*Fiber or skeleton of a shape is obtained by erosion (sequential removing) of edge pixels of that shape.
263TARIGHAT ET AL.: MORPHOMETRIC ANALYSIS OF SUNFLOWER 2011]
of the true perimeter of the object to its fittedconvex hull. Curl is used to quantify thedeparture of an object from a straight line. Dueto the significantly high systematic error of
measurement in sunflower achenes when using the curl parameter and confident in the fact thatcurl provides the least discriminating power, wecompleted the analysis using only the remaining three feasible shape parameters: formfactor, con-
vexity, and roundness. Assessing the utility of these three factors in discriminating between wildand domesticated seeds was one of our principalobjectives in this study.
Statistical analyses of the data were performedusing JMP analysis software (version 7). Data
points were inspected for normality and homo-geneity of variance. For non-normal data we useda log10 transformation. To determine whichcontinuous variables (shape or size features) coulddiscriminate between the main sunflower catego-ries, our data sets were subdivided into four groups: ENA wild, Mexican wild, ENA indige-nous domesticated, and Mexican indigenousdomesticated. To distinguish among the achenecategories we conducted a discriminant function
analysis which generates a set of discriminantfunctions based on linear combinations of con-tinuous predictor variables. These variables pro-
vide the most effective ways to distinguish among
the seed categories and thereby render a measureof the overall similarity among them (Rashed etal. 2008). Six measured characters includedachene length, width, thickness, achene form-factor, roundness, and convexity, all of which
were used in the analysis as predictors for membership in one of the two wild or one of the domesticated groups.
To determine if a significant difference existedamong sunflower populations of Mexico and easternNorth America, discriminant function analysis
(DFA) and analysis of variance (ANOVA) wereperformed on size measurements. Separate one-way
ANOVA tests comparing the size index variable(length x width) within wild populations anddomesticated populations were carried out. Theresults were compared to size differences betweenarchaeological specimens unearthed from Mexicoand ENA (Tables 2 and 3). In addition, weanalyzed climatic data (average annual temperature,average annual rainfall, and elevation) of wild
T ABLE 2. AVERAGE SIZE INDEX (LENGTH X WIDTH) OF ARCHAEOLOGICAL DOMESTICATED REMAINS RECOVEREDFROM MEXICO AND EASTERN NORTH AMERICA. BOTH HIGGS (BREWER 1973) AND HAYES (CRITES 1993) ARE EARLY
ENA SITES WITH SUNFLOWER EVIDENCE, BUT HAVE NO ACHENES.
Rose Island, TN Early, Middle Woodland(318 B.C.E.–C.E. 287)*
4 17.8(25.3) 1
Newt Kash Hollow, KY Late Archaic, Early Woodland(1162–369 B.C.E.)*
14 29.2 2, 3
Marble Bluff, AR (34-23-345) Late Archaic (1264–912 B.C.E.) 19 27.2(39.2) 4Marble Bluff, AR (34-23-327) Late Archaic (1032–920 B.C.E.) 14 24.5(35.2) 4Eden’s Bluff, AR Early, Middle Woodland
(170 B.C.E.–C.E. 50)4 25.9 5
Salts Cave, KY Early Woodland (654–416 B.C.E.)*57 17.4(24.4) 6,7
Salts Cave, KY (feces) Early Woodland (970–660 B.C.E.)* 1,000 16.8(23.7) 6,7Mammoth Cave, KY Early Woodland (539–239 B.C.E.)* 80 15.1(21.7) 2, 6Cueva del Gallo, Mex. Formative (330–250 B.C.E.) 3 57.5 5San Andrés, Mex. Late Archaic (2875–2575 B.C.E.) 1 36.9(51.9) 8
* Converted from conventional radiocarbon dates to calibrated dates using Fairbank ’s calibration curve (Fairbanks et al.2005).
† Numbers in parentheses represent values that were enhanced by correction factors (see Yarnell 1978). These are oftenused to compensate for achene shrinkage during carbonization.
sunflower locations using three single linear regres-sion tests to investigate the relationship betweenclimatic factors and achene size index. Climatic data at each collection site, including temperature, rain-fall, and elevation, for 14 wild populations in thespecific year of seed collection, were derived fromrecords of the Servicio Meteorological Naciónal deMéxico (SMN) and the National Climatic Data Center (NCDC) of the U.S. National Oceanic and
Atmospheric Administration (NOAA).Results
Discriminant function analysis (DFA) indi-cated that size factors (length and width) con-tributed significantly to the discriminant functionand consistently maximized the differencebetween the two wild and domesticated groups(Wilks’ Lambda= 0.06, F3, 2532=11417.8, P<
0.0001) (Fig. 3). According to the DFA of shapefactors, the morphological differences among populations were significant (Wilks’ Lambda=0.81, F3, 2532=195.5, P= 0.001); however, the
difference between wild and domesticatedachenes was much less than that seen using sizefactors (Fig. 4). Size measurements of modern
wild and domesticated populations are presentedin Table 4.
Analysis of variance (ANOVA) results (Table 5)showed that the achenes of modern wild sun-flower from ENA (average size index= 12.61)
were significantly larger than modern Mexican wild achenes (index size=8.6) (F1, 1338=1258.8,P<0.0001). The average index size of indigenous
domesticated seeds from Mexico ( X ¼ 64:
06) was smaller than the size of those from ENA ( X ¼ 71:62) (F1, 1190=21.4, P<0.0001). Addi-tionally, our sunflower populations showed high
within-population variation in their achene sizeindex (ANOVA; ENA wild F3, 396 =192.4, P<0.0001; Mexican wild F9, 930 =138.7, P<0.0001).
ANOVA results of archaeological sunflower speci-mens, however, showed that Mexican domesti-cated achenes were significantly larger than
domesticated samples from ENA (F1, 9 =22.53,P =0.001) (Table 5).Linear regression analysis testing the associa-
tion between climatic factors and achene sizeyielded nonsignificant results for rainfall (R 2 =0.071, P= 0.35) and elevation (R 2 =0.013, P=0.69). Average annual temperature, on the other hand, was significantly associated with size index (R 2 =0.543, P= 0.0026) in an inverse manner
Fig. 3. Discriminant function analysis (DFA) of size measurements. The left cluster represents the wild achenesgrouped together, while domesticated achene populations formed a distinct group.
265TARIGHAT ET AL.: MORPHOMETRIC ANALYSIS OF SUNFLOWER 2011]
where increasing temperature was associated withsmaller size index (Fig. 5). In the Mexicanpopulations, higher values of average annual
temperature were recorded for eastern andcentral Mexico where smaller achene size indices were observed (Table 6).
DiscussionMorphometric analysis of sunflower achene
populations provided a number of importantinsights into both modern populations and historicalinterpretations of this valuable crop plant. Thecomputer-assisted evaluations of sunflower achenesconducted using discriminant function analysis(DFA) revealed that shape analysis techniques wereless effective than techniques commonly utilized inconventional methods to distinguish between wildand domesticated populations of sunflower. Thefour shape parameters measured in this study (curl,roundness, convexity, and formfactor) did not bear suf ficient recognition robustness to recommend
their use in establishing quantifiable differencesbetween wild and domesticated achenes. In sun-flower populations, the transition from wild to
domesticated form is accompanied by dramatically altered achene size but, evidently, not the shapefeatures evaluated in this study. Computer-assistedapproaches may continue to offer cost-effective,rapid, and accurate methods for discriminating between seed populations if appropriate parametersare selected for the species or structures being evaluated, but the parameters tested in this study could not be successfully applied to sunflower achenes. Technical enhancements in differentiationand classification of sunflower achene populationsshould focus on comparative aspects that have beendemonstrated to be effective at discriminating between wild and domesticated types (i.e., achenelength or size index).
Analysis of variance of achene size indicatesthat modern wild sunflower populations fromeastern North America (ENA) have significantly
T ABLE 5. ANOVA RESULTS OF ACHENE SIZES COMPARING MODERN WILD AND DOMESTICATED ACHENES COLLECTED
IN MEXICO AND EASTERN NORTH AMERICA (ENA). THE LAST ROW SHOWS THE RESULTS COMPARING THE
MEAN SIZE INDICES OF PRE-COLUMBIAN DOMESTICATED SPECIMENS DISCOVERED FROM MEXICO AND ARCHAEO-
LOGICAL DOMESTICATED SAMPLES COLLECTED FROM ENA .
Evaluations Error DF MSE F P-value
Modern ENA Wild vs. Mexican Wild 1338 3.31 1361.84 <0.0001Modern ENA Dom. vs. Mexican Domesticated 1194 392.6 42.67 <0.0001 Ancient ENA Dom. vs. Mexican Domesticated 9 47.3 22.53 0.001
7
8
9
10
11
12
13
14
15
A c h e n e S i z e I n d e x
10 12.5 15 17.5 20 22.5 25Temperature
R2 = 0.54,P = 0.0026
Fig. 5. Graph showing the inverse relationship between temperature and average size index in modern wildpopulations sampled from Mexico and eastern North America.
267TARIGHAT ET AL.: MORPHOMETRIC ANALYSIS OF SUNFLOWER 2011]
larger size indices than modern wild populationsfrom Mexico. Wild sunflower populations examinedin this study showed high variability in their sizeindices both within and between populations. Our
expectation was that this variation would be highly influenced by climatic factors, but our results suggesta more complicated relationship. Regression analysisrevealed no significant association between size index and two climatic factors, rainfall and elevation, butindicated an inverse relationship between size andtemperature. The largest wild achenes in Mexico
were recorded for populations collected in thenorthern state of Chihuahua, where we found thecoolest temperatures in the corresponding year.Temperature seems to be the important climatic
factor affecting sunfl
ower achene size in North America. This observation makes sense in light of the fact that sunflower is essentially a temperateplant and is well adapted to the climate of the middlelatitudes. Accordingly, our results reflect existing
variation of achenes that largely can be attributed tothe wild sunflower genome and environmentalfactors.
The observed pattern of size variation betweenmodern wild populations becomes more intriguing
when we consider them in relationship to theachene sizes from archaeological contexts. Based onavailable archaeological records of ancient domes-ticated H. annuus from both regions, ancientcultivated sunflowers from Mexico producedachenes larger than the oldest H. annuus discoveredin ENA (Lentz et al. 2008b). These observationsraise provocative questions about the trajectory of crop development in the two regions and thecultural exchanges between ENA and Mesoamerica during pre-Columbian times.
Pre-Columbian Mexico has been identified as a cradle of domestication for other important crop
plants such as squash (Cucurbita pepo), which wasgrown in the highlands of Mexico in 10,000 B.P.(Smith 2006; Sanjur et al. 2002), and maize ( Zea mays ), which is believed to have been domes-
ticated in southern Mexico more than 6,300 yearsago (Burger et al. 2008; Matsuoka et al. 2002;Piperno and Flannery 2001). It is reasonable tohypothesize, given the backdrop and long history
of previous domestication processes in Mexico,that the occupants of early Mexico had theintellectual understanding, adequate chronologi-cal space, and the necessary wild germplasmavailable to complete the domestication processby 4600 B.P., the time of the first domesticatedsunflower evidence from Mexico at the San
Andres site. This position, however, has not been without controversy. Two domesticated sunflower disseminules were recovered from San Andres: anachene and a seed. These disseminules came from
different test pits, but dated to exactly the same timeperiod. Heiser verified the identification of the San
Andres achene soon after its discovery in 2000, butrecanted his identification eight years later, long after the specimen had been consumed by theradiocarbon dating process (Heiser 2008a ). TheSan Andres seed has attracted far less discussion.Our publication of the Cueva del Gallo sunflower finds from Morelos, Mexico (Lentz et al. 2008a ,b), engendered a spirited response (Brown 2008;Heiser 2008b; Rieseberg and Burke 2008; Smith
2008), but none of these commentaries seriously questioned the basic premise of the article, that thefinds from central Mexico represented fully domesticated H. annuus achenes. These new data,therefore, reinforced our earlier contention thatsunflower was indeed a crop of the pre-ColumbianMesoamericans. It is our hope and expectation thatfuture archaeological, paleoethnobotanical, andmolecular studies in Mexico and ENA will helpto clarify the relationship of the ancient sunflower populations in these two regions.
ConclusionIn conclusion, the four shape parameters we
evaluated (formfactor, roundness, convexity, andcurl) were less effective than conventional methodsin separating wild and domesticated sunflower achenes despite their utility in other seed popula-tions. Our assessment of achene size variation inmodern wild sunflowers showed that populationsfrom Mexico had significantly smaller achenes thanpopulations from eastern North America. Observedsize diversity in modern populations probably is a
result of both genetic and environmental variation.Conversely, domesticated archaeological achenesdiscovered in Mexico were significantly larger thanearly specimens recovered from eastern North
T ABLE 6. RESULTS OF THREE SINGLE REGRESSION
ANALYSES ASSESSING ASSOCIATION OF MEAN ACHENE
SIZE INDEX WITH THREE CLIMATIC FACTORS IN MODERN
WILD POPULATIONS.
Climatic Factor R 2 P-value Relationship
Temperature 0.543 0.0026* InverseRainfall 0.071 0.35 NoneElevation 0.013 0.69 None
America. Collectively these data will provide usefulguideposts as more archaeological specimens of H.annuus from ENA and Mexico are unearthed.
AcknowledgmentsThis work was supported by the National
Science Foundation (Grant Number BCS-0228049), the National Geographic Society (Grant Number 7030-01), and a Faculty Research Support Grant from the University of Cincinnati. We thank Abbas Shirinifard from thePhysics Department, Indiana University, for developing the MATLAB image processing script,an invaluable contribution to this project. Wealso thank Dr. Theresa Culley for her instrumen-
tal guidance. Kim Thompson and Edelmira Linares kindly offered editorial help.
Literature Cited Brewer, A. J. 1973. Analysis of the floral remains
from the Higgs site (40L045). In: Excavationof the Higgs and Doughty site: I–75 salvagearchaeology, eds. M.C. R. McCullough and C.
J. Falknor, 141–144. Knoxville: Tennessee Archaeological Society, Miscellaneous PapersNo. 12.
Brown, C. H. 2008. A lack of linguistic evidencefor domesticated sunflower in pre-ColumbianMesoamerica. Proceedings of the National
Academy of Sciences 105:E47.Burger, J. C., M. A. Chapman, and J. M. Burke.
2008. Molecular insights into the evolution of crop plants. American Journal of Botany 95:113–122.
Bye, R., E. Linares, and D. L. Lentz. 2009.México: Centro de origen de la domesticacióndel girasol. TIP Revista Especializada enCiencias Químico−Biológicas 12(1):5–12.
Chapman, J. and A. B. Shea. 1981. Thearchaeobotanical record: Early Archaic Periodto contact in the lower Little Tennessee River
Valley. Tennessee Anthropologist 6:61–84.Crites, G. D. 1993. Domesticated sunflower in
fifth millennium B.P. temporal context: New evidence from middle Tennessee. American
Antiquity 58:146–148.Doebley, J. F., S. G. Gaut, and B. D. Smith.
2006. The molecular genetics of crop domes-tication. Cell 127:1309–1321.
Fairbanks, R. G., R. A. Mortlock, T. C. Chiu, L.Cao, A. Kaplan, T. P. Guilderson, T. W.Fairbanks, A. L. Bloom, P. M. Grootes, andM. J. Nadeau. 2005. Marine radiocarbon
calibration curve spanning 0 to 50,000 yearsB. P. based on paired 230Th/234U/238U and14C dates on pristine corals. Quaternary Science Reviews 24:1781–1796.
Fritz, G. 1997. Crop seeds from Marble Bluff.Pages 42–62 in K. J. Gremillion, ed., People,plants and landscapes: Studies in paleoethnobo-tany. University of Alabama Press, Tuscaloosa.
Gardner, P. S. 1987. New evidence concerning thechronology and paleoethnobotany of Salts Cave,Kentucky. American Antiquity 52:358–367.
Granitto, P. M., H. D. Navone, P. F. Verdes, andH. A. Ceccatto. 2000. Automatic identificationof weed seeds by color image processing.Universidad Nacional de Rosario, Argentina.
———, ———, ———, and ———. 2002. Weed seeds identification by machine vision.Computers and Electronics in Agriculture33:91–103.
———, P. F. Verdes, and H. A. Ceccatto. 2005.Large−scale investigation of weed seed identi-fication by machine vision. Computers andElectronics in Agriculture 47:15–24.
Heiser, C. B. 1985. Some botanical considerationsof the early domesticated plants north of Mexico.In: Prehistoric Food Production in North
America, ed. R. I. Ford, 149−
203. Anthropo-logical Paper No. 75, Museum of Anthropology,University of Michigan, Ann Arbor.
——— 2008a. The sunflow er (Helianthus annuus ) in Mexico: Further evidence for a North American domestication. GeneticResources and Crop Evolution 55:9–13.
——— 2008b. How old is the sunflower inMexico? Proceedings of the National Academy of Sciences 105:E48.
Heiser, C. B, D. M. Smith, S. B. Clevenger, and W. C. Martin. 1969. The North Americansunflowers (Helianthus ). In: Memoirs of theTorrey Botanical Club, ed. T. Delevoryas,
Vol. 22, No. 3. New York Botanical Garden.Durham, North Carolina: Seeman Printary.
Lentz, D. L., M. E. D. Pohl, K. O. Pope, and A. R. Wyatt. 2001. Prehistoric sunflower (Helianthus annuus L.) domestication in Mexico. EconomicBotany 55:370–376.
———, R. Bye, and V. Sánchez-Cordero. 2008a.Ecological niche modeling and distribution of
wild sunflower (Helianthus annuus L.) in
Mexico. International Journal of Plant Scien-ces 169:541–549.
———, M. D. Pohl, J. L. Alvarado, S. Tarighat,and R. Bye. 2008b. Sunflower (Helianthus
269TARIGHAT ET AL.: MORPHOMETRIC ANALYSIS OF SUNFLOWER 2011]
annuus L.) as a pre-Columbian domesticate inMexico. Proceedings of the National Academy of Sciences 105:6232–6237.
Matsuoka, Y., Y. Vigouroux, M. M. Goodman,
G. J. Sanchez, E. Buckler, and J. Doebley.2002. A single domestication for maize shownby multilocus microsatellite genotyping. Pro-ceedings of National Academy of Sciences30:6080–6084.
Piperno, D. R. and K. V. Flannery. 2001. Theearliest archaeological maize ( Zea mays L.) fromhighland Mexico: New accelerator mass spec-trometry and its implications. Proceedings of theNational Academy of Sciences 98:2101–2103.
Pope, K. O., M. E. D. Pohl, J. G. Jones, D. L.
Lentz, C. V. Nagy, and I. Quitmyer. 2000.Early agriculture in the lowlands of Mesoa-merica. Science 292:1370–1373.
Rashed, A., M. I. Khan, J. W. Dawson, J. E. Yack, and T. N. Sherratt. 2008. Do hover flies(Diptera: Syrphidae) sound like the Hyme-noptera they morphologically resemble?Behavioral Ecology 20:396–402.
Rieseberg, L. H. and J. M. Burke. 2008.Molecular evidence and the origin of domes-ticated sunflower. Proceedings of the National
Academy of Sciences 105:E46.———, M. J. Kim, and G. J. Seiler. 1999.Introgression between the cultivated sunflower and a sympatric wild relative, Helianthus
petiolaris (Asteraceae). International Journal of Plant Sciences 160:102–108.
Rovner, I. and F. Gyulai. 2006. Computer-assisted morphometry: A new method for assessing and distinguishing morphological
variation in wild and domestic seed popula-tions. Economic Botany 61:154–172.
Russ, J. C. 2007. The image processing hand-book, 5th edition. CRC Press, New York.
——— and I. Rovner. 1989. Stereologicalidentification of opal phytolith populationsfrom wild and cultivated Zea . American
Antiquity 54:784–792.Sanjur, O. I., D. R. Piperno, D. C. Andres, and
L. Wessel-Beaver. 2002. Phylogenetic relation-
ships among domesticated and wild species of Cucurbita (Cucurbitaceae) inferred from a mitochondrial gene: Implications for cropplant evolution and areas of origin. Proceedings
of the National Academy of Sciences 99:535–540.
Smith, B. D. 2006. Eastern North America as anindependent center of plant domestication.Proceedings of the National Academy of Sciences 103:12223–12228.
——— 2008. Winnowing the archaeologicalevidence for domesticated sunflower in pre-Columbian Mesoamerica. Proceedings of theNational Academy of Sciences 105:E45.
Steffansson, B. R. 2007. Oilseed crops. The
Canadian encyclopedia (Toronto: Historica Foundation). http://thecanadianencyclopedia.com.
Tarighat, S. S. 2010. A morphometric approachin examination of sunflower (Helianthus annuus L.) in eastern North America andMexico. Unpublished M.S. thesis, University of Cincinnati, Ohio.
Travis, A. J. and S. R. Draper. 1985. A computer based system for the recognition of seed shape.Seed Science and Technology 13:813–820.
Watson, P. J. and R. A. Yarnell. 1966. Archaeo-logical and paleoethnobotanical investigationin the Salts Cave, Mammoth Cave NationalPark, Kentucky. American Antiquity 31:842–849.
Yarnell, R. A. 1978. Domestication of sunflower and sumpweed in eastern North America. In:The nature and status of ethnobotany, ed. R.I. Ford, 289–300. Anthropology Papers 67,Museum of Anthropology, University of Michigan, Ann Arbor.
——— 1981. Inferred dating of Ozark bluff dweller occupations based on achene size of sunflower and sumpweed. Journal of Ethno-biology 1:55–60.
Zayas, I., Y. Pomeranz, and F. S. Lai. 1989.Discrimination of wheat and nonwheat com-ponents in grain samples by image analysis.Cereal Chemistry 66:233–237.