Marginality Chiribaya - Zaro

Today, the intervalley coast between theTambo and Ilo rivers is a desiccated land-scape. Agriculture, largely supported bysmall desert springs, is restricted to only a few sur-viving farmsteads, while vegetation among inlandhills has diminished considerably. Much of thisintervalley coast is completely depopulated, withthe vast majority of people currently living withinthe adjacent Tambo and Ilo river valleys. The area

is widely viewed to be an industrialized landscapewith little value beyond a desert wasteland. In the1960s, Southern Peru Copper Corporation (SPCC),Perus largest mining operation, positioned its cop-per smelter along this stretch of coastline, where itwas thought to pose the least possible detriment tohuman health and livelihood in the region (Hall1992:55). Furthermore, since the port of Ilo largelyexpanded in the 1950s and 1960s as a company

TIERRAS OLVIDADAS: CHIRIBAYA LANDSCAPE ENGINEERING ANDMARGINALITY IN SOUTHERN PERU

Gregory Zaro, Kenneth C. Nystrom, Alfredo Bar, Adn Umire Alvarez, and Ana Miranda

The Peruvian south coast between the Tambo and Ilo rivers is a deserted wasteland. Yet tracts of abandoned farmland andexpanses of desiccated lomas vegetation indicate that it was once vibrant and productive. Scattered habitations and ceme-teries also indicate a pronounced resident Chiribaya population between A.D. 1200 and 1400. While river drainages of thewestern Andes and their canalized extensions are often treated as primary analytical units of study, our investigation of so- called peripheral intervalley regions points to a highly engineered and intensively managed landscape. When compared tothe Ilo Valley, intervalley Chiribaya were organized into smaller communities and managed smaller spring- fed irrigationsystems but were also heavily invested in adjacent maritime and lomas resource bases. Utilizing a mixed economic strat-egy, we argue, these intervalley populations were anything but marginal to the Chiribaya seoro of the Ilo region; rather,they were fully engaged in the social, political, and economic spheres of the late Intermediate period Peruvian south coastand formed a significant component of the Chiribaya cultural landscape. Our results hold implications for other similarlymarginalized areas of the Andean coast, where distinct microenvironmental parameters and human ingenuity intersectedto significantly transform the western Andes.

Aunque la costa entre los valles Tambo e Ilo en el sur del Per es hoy da un gran desierto, los campos abandonados de cul-tivo junto con las extensiones de vegetacin disecada de lomas sugieren que en tiempos antiguos era un paisaje vivo y pro-ductivo. Habitaciones y cementerios dispersos tambin indican una poblacin Chiribaya muy marcada entre 1200 y 1400 d.C.Mientras que los drenajes principales de los Andes occidentales y sus extensiones canalizadas constituyen frecuentemente lasunidades analticas primarias de estudio, nuestra investigacin de llamadas regiones perifricas entre valles sugiere un paisajemuy intensivamente manejado. En comparacin al valle de Ilo, los Chiribaya entre valles se organizaron en comunidades mspequeas y manejaron sistemas de irrigacin ms simples, pero a la vez invirtieron intensivamente en el manejo de los recur-sos adyacentes del mar y de las lomas. Utilizando una estrategia econmica mixta, sugerimos que estas poblaciones entrevalles no fueron marginadas, sino que estaban bien integradas en los crculos sociales, polticos y econmicos del perodoIntermedio Tardo de la costa sur del Per, y formaron un componente significativo del paisaje cultural Chiribaya. Nuestrosresultados tienen importancia para otras regiones marginadas de la costa Andina, donde el cruce de parmetros microambi-entales particulares y la inventiva humana transformaron significativamente a los Andes occidentales.

Gregory Zaro Department of Anthropology and the Climate Change Institute, 5773 S. Stevens Hall, University ofMaine, Orono, ME 04469-5773 ([email protected])Kenneth C. Nystrom Department of Anthropology, State University of New YorkNew Paltz, 1 Hawk Drive, WSB 124,New Paltz, NY 12561-2443 ([email protected])Alfredo Bar Universidad Nacional Mayor de San Marcos, Lima, Per ([email protected])Adn Umire Alvarez Universidad de Tarapac/Universidad Catlica del Norte, Chile ([email protected])Ana Miranda Universidad Catlica de Santa Mara, Arequipa, Per ([email protected])

Latin American Antiquity 21(4), 2010, pp. 355374Copyright 2010 by the Society for American Archaeology

355

356 LATIN AmerIcAN ANTIquITy [Vol. 21, No. 4, 2010

town, few have historical connections to the areaprior to the twentieth century, rendering the historyof this distinctive landscape forgotten.While in many respects the intervalley land-

scape may be considered marginal today, ourresearch in these forgotten lands points to animpressive human component more than five cen-turies ago. Recent archaeological survey has iden-tified a pronounced Chiribaya occupation beyondthe hydrological limits of the principal river val-leys. Indeed, scattered habitations, cemeteries, andthe intensive management of coastal, agricultural,and lomas (winter fog /mist- fed vegetation)resources point to a highly engineered landscapebetween six and eight centuries ago. While earlierwork confirms that Chiribaya populations weregreatest along the Ilo River (e.g., Lozada and Buik-stra 2002; Owen 1993, 1994, 2005; Umire Alvarezand Miranda 2001), we suggest that intervalleycoastal peoples formed a significant component of Chiribaya fully engaged in social, economic, andpolitical spheres of the time and were anything

but marginal. Studies of human populations pastand present along the western Andean watershedhave tended to rely upon primary river drainagesand their canalized extensions as principal analyt-ical units of inquiry. In this article, we propose thatthe investigation of so- called peripheral interval-ley regions has the potential to significantly alterperspectives of ancient societies, ecological man-agement, and contemporary landscapes of the aridAndean coast.

The TamboIlo Coast

The Tambo and Ilo rivers constitute two of thesouthernmost valleys of Peru (Figure 1). Bothcourse their way through hyperarid desert, trans-porting glacial meltwater and rainfall from the adja-cent highlands to the Pacific coast. Farmingcurrently dominates the landscapes of both valleys,though they differ substantially in production.Satellite imagery indicates that the Tambo Valley,with a much wider floodplain and greater annual

Figure 1. Map of the study region showing principal locations mentioned in the text.

discharge, supports upward of 7,000 ha of irrigatedbottomlands. On the other hand, the Ilo River flowsthrough a narrower, steep- sided canyon that rarelyreaches 450 m in width. In the late twentieth cen-tury, only 390 ha of land were under production(ONERN 1976), while satellite imagery indicatesthat cultivation still does not exceed about 400 hatoday.The rugged intervalley landscape separating

these rivers is characterized by steep escarpmentsand deeply incised canyons, many of whichdescend from the adjacent Clemes Desert (1,200m asl). Oceanic and topographic elements along thecoast have maintained a relatively arid climatethroughout the late Holocene, though punctuatedby periodic flooding episodes associated with ElNioSouthern Oscillation (Fontugne et al. 1999;Keefer et al. 2003; Sandweiss et al. 2001). Marinerecords off the coast of central Peru indicate theoccurrence of a Medieval Climatic Anomaly (A.D.8001250) with persistently weak El Nios fol-lowed by a period of moister conditions (Rein etal. 2004). In the current study area, however, Mag-illigan et al. (2008) have assembled a 20,000-yearrecord of El Nio frequency from mid- valley trib-utaries of the Moquegua River.1They conclude thatEl Nio events increased in frequency and magni-tude from about A.D. 700 to 1610 (when comparedto the previous 700800 years), with mega- Niosoccurring around A.D. 1330 and 1650. Water geo-chemistry and isotopic data also suggest the impor-tant role El Nio plays in local aquifer recharge inthis hyperarid environment (Magilligan et al.2008:2627).Along the intervalley coast, the density of the

archaeological record, and particularly as it pertainsto agriculture between 800 and 600 years ago, sug-gests that there was considerably greater moistureand spring discharge during the late Chiribayaperiod (A.D. 12001400) than today, possiblyresulting from centuries of elevated ElNioinduced aquifer recharge during and prior tothat time (see Magilligan et al. 2008). Historicalarchives also indicate that the lomas among theinland hills were much lusher during the late pre-hispanic and early Spanish colonial periods thanthey appear today. Ethnohistorian Mara Rost-worowski (1981), for instance, notes that the lomasaround Ilo were some of the most productive in lateprehispanic and early Spanish colonial Peru, while

eighteenth- century French explorer AmadeoFrezier (1982:156 [1713]) comments that the hillssurrounding Ilo were once covered with trees butthat these had receded because of deforestation.Elsewhere, Zaro (2005a, 2005b) suggests that thisstretch of coast may have experienced increasingaridity since Chiribaya. Spring systems likeWawakiki witnessed a reduction in land use alongwith increasingly shorter and steeper canals overthe past half millennium, while similar incremen-tal reductions in land use are reported for neigh-boring springs like Carrizal (Clement and Moseley1991). Ultimately, historic aerial photographs indi-cate that farming had ceased at Wawakiki prior to1951 and well before the construction of the SPCCsmelter (Zaro 2007:Figure 3).

The Concept of Marginality

By nearly any definition of the term, the TamboIlointervalley coast has become marginalized, a termapplied regularly to human societies, fractions ofhuman societies, or landscapes. As a cultural term,it is most often associated with the social, politi-cal, and/or economic exclusion of a particular groupof people within a more broadly defined humandemographic (Beattie et al. 1981; Binns 2007; Pol-lard 1997; Potter 2001). In this context it is syn-onymous with terminology like undeveloped,unconventional, minority, impoverished,disenfranchised, or even third world. In otherinstances, it relates to relative poverty or locationsthat are distant from centers of power and wealth(Pollard 1997). Marginality has also been appliedto physical landscapes, referring to rural, distant,or peripheral regions. In other contexts, it reflectspeoples conceptualization of harsh landscapes,evoking images of mountains, deserts, swamps,and even forests, depending on idiosyncratic val-ues placed on such environments (Pollard 1997). Agro- ecological marginality, another physicalapplication of the term, usually refers to the suit-ability of a landscape for farming (Beattie et al.1981; Doolittle 1988; Waldhardt et al. 2001; Wu etal. 2001). In this regard, it is tantamount to con-cepts like unpredictable, undesirable, andadverse conditions or even land that requiresintensive investment to achieve similar productiv-ity to optimal lands. Referring to cultivated land-scapes in eastern Sonora, Mexico, Doolittle

Zaro et al.] cHIrIBAyA LANDScAPe eNGINeerING AND mArGINALITy IN SOuTHerN Peru 357

(1988:255) notes that marginality is often associ-ated with intermittently farmed terrain. He alsopoints out that marginal lands are routinelydichotomized with those thought to be optimal foragriculture. However, as Doolittle (1984) has alsodemonstrated, intermittent farming can graduallylead to permanent cultivation with the accumula-tion of incremental changes over time.A number of studies also hint at what we term

scholarly marginality, or the lack of scholarlyinquiry pertaining to either sociocultural or agro- ecological concerns due to the perceived unim-portance of a people or landscape. In geography,Potter (2001) describes underdeveloped nations tobe marginal in the sense that they receive littleattention and are afforded far less importance whencompared to urbanrural landscapes of NorthAmerica and Europe. Similarly, Doolittle (1988)attributes the lack of scholarly attention to inter-mittently cultivated fields in Sonora, Mexico, totheir perceived unimportance. Elsewhere, Binns(2007) suggests that marginality can even beapplied to geographical trends in scholarly publi-cation. He notes that journals published in so- calledmarginalized regions like Africa or Latin America,for example, receive far fewer submissions for pub-lication than do those published in less marginal-ized Europe and North America, even when thescientific focus of those manuscripts is centered onthose marginal regions. Consequently, these areasremain underrepresented with respect to the distri-bution and control of a knowledge base constructedfrom their lands and people.Whether applied to socioeconomic status, to

physical landscapes, or to trends in scholarlyinquiry and publication, marginality is clearly nota fixed condition but, rather, may be applied sub-jectively in a range of contexts and vary as socioen-vironmental conditions change. The current stateof the TamboIlo coast would certainly qualify itto be marginal according to most definitions. Inmany cases, freshwater springs that once irrigatedprehispanic and Spanish colonial farmsteads havedried up completely, while others have significantlydiminished, reducing the areas overall biologicalpotential. The southern half of the coastline is dom-inated by the SPCC copper smelter, though severalolive groves established during the Spanish colo-nial period still cling to the margins of a few springchannels. Only a handful of people continue to live

perennially along this coastline today, maintainingsubsistence primarily from small- scale spring- fedirrigation agriculture, fishing, and pastoralism. Theinland lomas are also highly desiccated, which maybe linked to wider processes of desertification(including human agency) over the past 600 to1,000 years, though this continues to be the focusof an ongoing investigation (Zaro 2008a; Zaro etal. 2008). Agro- ecologically, the intervalley regionoffers little to the farming economy of the Peru-vian south coast, and with little infrastructurebeyond a rocky, pockmarked road that parallels thecoastline, the few farmsteads that do survive in thearea have little opportunity to integrate withregional economies.

The Chiribaya Seoro and the Peruvian South Coast

This bleak picture of the intervalley coast, and par-ticularly its lack of infrastructure and biologicalpotential, has left it relatively absent from archae-ological inquiry when compared to work carriedout along the Ilo River. In the 1980s and 1990s,archaeological research performed under the aus-pices of the Programa Contisuyo in Moquegua,Peru, focused primarily upon human occupation ofthe Osmore drainage (e.g., Watanabe et al. 1990)but with collaboration among colleagues through-out southern Peru, Bolivia, and northern Chile toprovide a broader cultural context. Central to thisarticle, the research program and its collaboratorsdocumented a moderate distribution of late Inter-mediate period (ca. A.D. 9001400) Chiribayaceramic styles in the region, ranging from limitedquantities as far inland as the Upper Osmoredrainage (Stanish 1992) to as far north as the TamboRiver and as far south as northern Chile (Jessup1990). However, Owens (1994) systematic surveyof the Lower Osmore drainage identified by far themost abundant Chiribaya remains, indicating thatit was primarily a coastal valley phenomenon.Chiribaya settlement stretched inland 25 km fromthe rivers mouth, reflected in habitation sectors,midden scatters, mortuary remains, and agricul-tural fields scattered along the margins of the IloRiver. By A.D. 1000, the socioenvironmental land-scape of the lower drainage largely consisted ofChiribaya farming villages situated along the flood-plain of the river. Chiribaya also engineered a 9-


km- long canal along the north canyon wall to irri-gate a series of hanging river terraces (Reycraft2000:99). Closer to the Pacific Ocean, Chiribayafishing villages like San Gernimo exploited richfisheries off the Andean coast (Jessup 1990).Following ethnohistorian Mara Rost-

worowskis model of horizontality, Lozada andBuikstra (2002, 2005) have interpreted the Chiri -baya to be a seoro, composed of specializedgroups of labradores and pescadoreswho operatedunder the rule of a supreme authority, perhaps sit-uated at Chiribaya Alta. They draw upon archaeo-logical and bioanthropological data sets generatedprincipally from Ilo Valley populations to supporttheir conclusions. Elsewhere, research centered onsmall coastal villages immediately north and southof Ilo has also provided glimpses of the complex-ity and variability of late prehispanic socioeco-nomic organization (Bawden 1989; Clement andMoseley 1991; Covey 2000; Miranda and UmireAlvarez 2007; Reycraft 1998; Satterlee 1993; Sat-terlee et al. 2000; Umire Alvarez 1994, 1996; Zaro2005b, 2007; Zaro and Umire Alvarez 2005). Inparticular, an emphasis on diversified economicstrategies may have been the norm among inter-valley coastal villages, while river valley popula-tions may have favored the moderately specializedapproach argued by others (i.e., Lozada and Buik-stra 2002). At Wawakiki, which lies on a rocky seacliff with access to a sandy beachfront, Zaro andcolleagues identified intensive stone- faced agri-cultural terraces that were once irrigated by multi-ple spring systems (Zaro 2007; Zaro and UmireAlvarez 2005). Agricultural fields were associatedwith Chiribaya habitation terraces and domesticmiddens dense with marine shell. The presence ofcamelid dung may also indicate that the nearbylomas were of some importance to this small vil-lage. Zaro (2007) argues that the proximity of mar-itime, agricultural, and lomas resources to a singlecommunity may have encouraged this explicitlydiversified subsistence strategy.In general, however, scholarly understanding of

the strength of Chiribaya cultural expression amongintervalley arenas has sorely lagged behind that ofriver valley populations and been limited to thosefreshwater springs in proximity to Ilo. Conse-quently, the human component among intervalleylandscapes has thus far been poorly understood andappeared to be insignificant when compared to that

of primary river drainages of the Peruvian southcoast.

Engineering a Landscape:The TamboIlo Intervalley Coast

Following informal reports of extensive archaeo-logical remains in the region, we recently carriedout a systematic archaeological survey of theTamboIlo intervalley coast. The project wasdesigned to document all human activities in thematerial record, generally from the mid- Holocenethrough the twentieth century. The survey followedthe coastline northward from Wawakiki to Que-brada Iane, extending Umire Alvarezs (1994,1996) previous work between Wawakiki and Ilo(Figure 1). Our investigation also reached inlandto elevations of approximately 800 m asl. WhileSpanish colonial, republican- era, and twentiethcentury activities are certainly notable across thelandscape, the most pronounced record over thepast several millennia appears to relate to lateChiribaya (A.D. 12001400) settlement locatedpredominantly within coastal canyons below 250m asl. In addition, a much more ephemeral late pre-hispanic archaeological record is found among the now- desiccated lomas of the inland hills.Given the paucity of excavation data from inter-

valley contexts, we recognize that we cannot rejectoutright the possibility of seasonal occupation ofthis zone or perhaps interannual movements of peo-ple from the valley (during El Niorelated rain-fall events, e.g., when lomas often expand).However, the regular association of formallybounded cemeteries with agricultural fields andhabitation sectors among intervalley springs wouldstrongly suggest a degree of territoriality and likelypermanent, rather than seasonal or interannual,occupation. Furthermore, while Knudson and col-leagues (2007) have demonstrated seasonal varia-tion in diet among some Ilo Valley Chiribayaregarding the consumption of marine products andC4 plants like maize, the potential for both resourcebases existed along the intervalley coast and wouldnot have necessitated the seasonal movement ofpeople (see Zaro 2007). Therefore, we suggest thatthe material expression of landscape engineeringalong the TamboIlo coast reflects a relatively per-manent occupation during the late Chiribayaperiod. Our discussion focuses on agricultural and


mortuary components identified in these areas andtheir pertinence to a reinterpretation and greaterappreciation of these often- overlooked segments ofcultural landscapes in the western Andes.

Intervalley Chiribaya AgricultureIntensive Chiribaya canal and terrace agriculturaltechnologies were identified in 23 intervalleycoastal canyons (Figure 2). The chronological affil-iation of agricultural remains is assessed using avariety of absolute and relative measures across theregion, which include radiocarbon age ranges pro-duced directly from field systems and associateddomestic sectors (Figure 3), superposition of Huay-naputina volcanic ash from A.D. 1600 (see Thouretet al. 1999), and the presence of Chiribaya ceramicforms and styles identified in related domestic andmortuary contexts (following Jessup 1990, 1991;Owen 1993). A summary of chronological infor-mation for intervalley Chiribaya sites is providedin Table 1.Agricultural Strategies. Terraced field systems

were engineered along canyon margins and acrossdesert pampas, and they were irrigated with canalsthat transported water to field systems from inlandsprings, anywhere from several hundred meters toupward of 2 km away. Terrace construction varied

in form and size, largely reflecting the influence oflocal topography. Most terraces, however, rangefrom .5 to 1 m in height and are faced with stackedblocks of pink granite, the most abundant rock typealong the intervalley coast for such constructions(Fortin 2008). In some cases, terraces were con-


Figure 2. Three dimensional model of a portion of the Tambo-Ilo intervalley coast showing the local topography and thedistribution of the agricultural and mortuary landscapes. Oblique satellite image produced from NASA World Wind 1.4.Vertical elevation is exaggerated 3x.

Figure 3. Calibrated radiocarbon ages. Probability distri-butions (with 2 brackets) were produced with OxCal4.1.1 (Bronk Ramsey et al. 2009) using a southern hemi-sphere atmospheric curve (McCormac et al. 2004). SeeTable 1 for sources of radiocarbon dates.


Table 1. Basis for Late Prehispanic Chiribaya Chronology of Intervalley Coastal Springs.

Documented Cases of Superimposed Huaynaputina Ash and Reported Radiocarbon Ages Stem

from Those Sites Where Archaeological and/or Geological Work Has Been Performed.

* Reported Radiocarbon Ages.

Basis of Chronology

Location

Superimposed

Huaynaputina

Ash (A.D. 1600)

Uncalibrated

Radiocarbon

Ages (B.P.)

Additional Observations References

Chuza Documented Chiribaya ceramic forms/styles from

domestic and mortuary contexts

Miranda and Umire Alvarez 2007

Umire Alvarez 1996

Carrizal Documented AA37179, 561 51

AA37181, 785 48

Stone terraces; Chiribaya ceramic

forms/styles from domestic contexts

Clement and Moseley 1991

Owen 2005*

Reycraft 1998

Satterlee 1993

Alastaya Chiribaya ceramic forms/styles from

domestic contexts Umire Alvarez 1994

Miraflores Documented PITT0948, 600 45 Chiribaya ceramic forms/styles from

domestic contexts

Satterlee 1993*

Umire Alvarez 1994

Pocoma Documented Chiribaya ceramic forms/styles from

domestic contexts

Satterlee 1993

Umire Alvarez 1994

Wawakiki Documented

AA56774, 480 30

AA56760, 575 31

AA56771, 592 35

AA56772, 627 30

GL30460, 640 50

AA56761, 651 32

AA56762, 685 33

GL30458, 700 60

GL30461, 720 90

AA56773, 727 30


forms/styles from domestic contexts

Zaro 2007*

Zaro and Umire Alvarez 2005*

Alfaro


forms/styles from domestic and

mortuary contexts

Umire Alvarez 2006

Chololo Documented Stone terraces; superimposed debris

flow Umire Alvarez 2006

Higocinto



mortuary contexts

Umire Alvarez 2006

Platanal

Stone terraces; Chiribaya decorated

ceramics from adjacent domestic

contexts

Umire Alvarez 2006

Callango



mortuary contexts

Umire Alvarez 2006

Molles



mortuary contexts

Umire Alvarez 2006

Apacheta


forms/styles from domestic contexts;

One Inca decorated ceramic fragment

Umire Alvarez 2006

Yerba Buena Chiribaya ceramic forms/styles from

domestic and mortuary contexts Umire Alvarez 2006

Concepcin Stone terraces Umire Alvarez 2006

Alpechn Stone terraces Umire Alvarez 2006

Amoquinto



mortuary contexts

Umire Alvarez 2006

Honda Stone terraces; Chiribaya ceramic

forms/styles from domestic contexts Umire Alvarez 2006

Cola de Zorro Documented

AA78828, 560 36

AA78831, 569 36

AA78830, 617 36

AA78829, 735 44



mortuary contexts

Umire Alvarez 2006

Zaro 2008b*

Dispensilla/TI-188 Stone terraces; Chiribaya ceramic

forms/styles from domestic contexts Umire Alvarez 2006

Jess Chiribaya ceramic forms/styles from

domestic contexts Umire Alvarez 2006

Pacae/Cocotea



mortuary contexts

Umire Alvarez 2006

structed along the margins of drainage channels(Figure 4) in similar fashion to what have beentermed valley floor terraces, or low- walled benchterraces found perpendicular to stream channels onvalley floor areas (Treacy and Denevan1994:100101). In other cases, small discontigu-ous sets of terraces are situated across stretches ofdesert pampa between coastal drainages. Canals are

lengthiest in such cases, where it became neces-sary to first channel water out of deeply incisedcanyons and subsequently transport it across thedesert to distal field systems. Whether along themargins of drainage channels or dispersed acrossthe desert, Chiribaya farmers often included an ele-ment of what may be termed grid terracing setsof roughly equilateral quadrangles that follow sub-


Figure 4. Photograph and surface map produced from a section of late prehispanic channel margin terraces at the Colade Zorro archaeological site along the Tambo-Ilo coast.

tle changes in topography (Figure 5). Comment-ing on morphologically similar (albeit much moreextensive) fields along the Ilo River, Williams(1997:114) notes that this type of field organiza-tion is unique in the Osmore region and may reflectsome degree of central planning by Chiribaya elite.While central organization does not appear to havebeen necessary to construct the much smaller ter-raced field systems of the intervalley coast, theirpresence indicates that coastal and main valley pop-ulations shared similar engineering principles.Extent of Agrarian Production. In general, inter-

valley Chiribaya irrigation systems were relativelysmall scale when compared to the construction andmanagement of a contemporaneous 9- km- longcanal and its associated field systems along the IloRiver (Reycraft 2000:99). Over the past fewdecades, the Ilo Valley has supported about 390 haof irrigated floodplain (ONERN 1976:276; andsatellite imagery), while during the Chiribayaperiod, we generously estimate from satellite

images that an additional 75 ha of irrigated farm-land could have been supported by the canal. Whencompared to the Middle Moquegua Valley of thelate twentieth century (2,810 ha) or the Tambo Val-ley (estimated 7,000 ha from satellite imagery), theIlo River has never irrigated more than just a mod-est amount of terrain, yet it was home to the largestof Chiribaya populations known to date andarguably central to Chiribaya sociopolitical author-ity. In contrast, the most extensive intervalley springsystem was Pocoma, which encompassed a maxi-mum of about 30 ha of cultivated terrain (Satterlee1993). Most intervalley systems, however, operatedon an even smaller scale, ranging anywhere from1 to 15 ha in size. While these intervalley agro- engineering projects may seem insignificant ontheir own, combined they produced nearly 200 haof canal and terrace farmland (Table 2, Figure 6).Furthermore, our calculations do not include poten-tially cultivated land from intervalley spring sys-tems north of our survey area between Quebrada


Figure 5. Grid terraces at the Apacheta archaeological site along the Tambo-Ilo coast. Each quadrangle measuresapproximately 12 m per side.

Iane and the Tambo River (an additional 25 kmof coastline). Nor do they include potentially cul-tivated fields among the inland hills, where thereis substantial evidence for twentiethcentury agri-cultural furrows indicative of dry farming between500 and 800 m asl. A local farmer and longtimeresident of the intervalley coast confirmed that3040 years ago he would expediently plant pota-toes and other cultigens among the inland hills,noting that the seasonal gara (winter fog and mist)was at times dense enough to support such activi-ties. While we did not identify any direct surfaceevidence of Chiribaya dry farming, the ephemeralnature of the late prehispanic and Spanish colonialarchaeological records among the inland hills mim-ics the equally scant modern material record, sug-gesting that similar activities may have transpiredperiodically in antiquity. When joined with terracedlandscapes of lower canyon lands and potentiallyirrigated farmland in unsurveyed regions north ofour study area, 250 ha of cultivable terrain during

the late Chiribaya period is not an unreasonableestimate for the intervalley region. This would haveconstituted more than 50 percent of the maximumextent of contemporaneous land use along the IloRiver. Add this to a lomas base described to be oneof the most productive in late prehispanic and colo-


Table 2. Areal Extent of Chiribaya and Modern Agricultural Land Use in the Ilo Region.

Intervalley Ilo Valley, Ilo Valley, Location Chiribaya (ha) Chiribaya (ha) Modern (ha) ReferencesChuza 3 Umire Alvarez 1996Carrizal 26 Satterlee 1993Alastaya 15 estimate from satellite imagery Miraflores 18 Satterlee 1993Pocoma 30 Satterlee 1993Wawakiki 11 Zaro and Umire Alvarez 2005 Alfaro 5 Umire Alvarez 2006Chololo 2 Umire Alvarez 2006Higocinto 6 Umire Alvarez 2006Platanal 1 Umire Alvarez 2006Callango 6 Umire Alvarez 2006Molles 2 Umire Alvarez 2006Apacheta 8 Umire Alvarez 2006Yerba Buena 8 Umire Alvarez 2006Concepcion 2 Umire Alvarez 2006Alpechin 4 Umire Alvarez 2006Amoquinto 17 Umire Alvarez 2006Honda 5 Umire Alvarez 2006Cola de Zorro 12 Umire Alvarez 2006Dispensilla 3 Umire Alvarez 2006Jesus 10 Umire Alvarez 2006Pacae 1 Umire Alvarez 2006Ilo Valley (Chiribaya) 465 ONERN 1976, plus estimate from

satellite imageryIlo Valley (Modern) 390 ONERN 1976Total (ha) 193 465 390Note: Ilo Valley chiribaya land use is estimated from satellite imagery based on relic fields associated with 9-km long canalalong the north canyon wall and modern irrigation of floodplain.

Figure 6. Comparison of estimated cultivation (ha) amongintervalley Chiribaya, Ilo Valley Chiribaya, and Ilo Valleyagricultural landscapes of 1976.

nial Peru (see Rostworowski 1981) and to one ofthe worlds richest fisheries, and a picture of ahighly engineered landscape emerges far fromthe marginalized industrial landscape observedtoday.

Chiribaya Mortuary RemainsWhile the archaeological record points to a signif-icant agricultural landscape along the intervalleycoast during the thirteenth and fourteenth centuries,numerous middens, habitation terraces, and mor-tuary remains also indicate a fairly pronounced res-ident Chiribaya population associated with themanagement of field systems and marine resources(see also Zaro 2007). Our discussion here focuseson the mortuary component of the intervalley land-scape, where a total of 24 discrete Chiribaya ceme-teries have been identified from surface remains(see Figure 2).2 Comparing the mortuary land-scapes of the intervalley region with those of theIlo Valley permits some assessment of relative pop-ulation sizes and the socioeconomic identity ofintervalley populations.Chiribaya Cemeteries and Relative Population

Size. All intervalley cemeteries discussed here wereat least partially looted and identified from surfacescatters of human bone and fragments of ceramics,textiles, slab stones, and other mortuary remains.In some cases, tomb constructions are visible withinlooters excavations and indicate the presence of stone- lined rectangular and circular pits (Figure 7).Intervalley cemeteries are positioned along the mar-gins of drainage channels or on immediately adja-cent desert pampas or small crests, and all aresituated near habitation features (terraces or struc-tures) and extensive scatters of domestic debris thatinclude fragments of ceramics, lithics, and marineshell. In several instances, multiple discrete ceme-teries are located in proximity to each other alonga single drainage. Cemetery areas are measured bythe visible extent of looting, with most encom-passing very modest areas ranging from less than100 m2 to about 2,000 m2. However, the two largestcemeteries along the intervalley coast measure8,576 m2 and 3,630 m2, both of which are found atthe site of Callango near Quebrada Molles (seeFigure 2).In contrast to the intervalley coast, Bruce

Owens (1994) archaeological survey of the LowerIlo Valley documented slightly fewer Chiribaya

cemeteries overall,3 but they were much moreextensive on average, ranging from less than 800m2 to over 14,000 m2. In total, 18 demonstrablyChiribaya cemeteries were identified in the lowervalley, with two additional Chiribaya cemeteriesidentified at El Yaral in the lower portion of the mid-dle drainage. Chiribaya were most numerous alongthe Ilo River, but when combined, intervalleycoastal populations (represented by bounded orclustered cemetery areas) constituted about 32 per-cent of the total measured in the lower valley, sug-gesting that these communities formed apronounced (though somewhat dispersed) compo-nent of the Chiribaya cultural landscape (Table 3,Figure 8).The Chuza Site and Expression of Socioeco-

nomic Identity. To date, only one intervalley Chirib-aya cemetery has been excavated. Located atQuebrada Chuza behind the SPCC smelter (seeFigure 2), it was excavated in its entirety as a sal-vage project in the face of modern constructionactivity (Miranda and Umire Alvarez 2007). Thesite is predominately Chiribaya and suitably posi-tioned to have managed maritime, agricultural, andlomas resources. Overall, 108 intact or disturbedtombs were excavated, with 97 single intermentsand 11 double interments. Tomb construction var-ied but included rectangular, circular, and irregu-lar stone- lined and unlined pits. Burialaccompaniments included ceramic vessels, textiles,and other artifacts crafted of metal, ceramic, wood,stone, and bone. Our discussion is based on Chuzacemetery data initially published by Miranda andUmire Alvarez (2007).While the mortuary analysis presented in this

article is preliminary, it is interesting to consider thecontext (residential vs. midden), construction form,and grave contents of tombs at Chuza in light of theestablished relationship between these variables andethnic group membership among Chiri baya inhab-iting the Ilo Valley. Evidence from a number of dif-ferent attributes, including differential distributionof tomb forms (Buikstra 1995), cranial modifica-tion styles (Hoshower et al. 1995; Lozada and Buik-stra 2005), and biochemical analyses of skeletalmaterial (Tomczak 2003), points to a degree of eco-nomically specialized ethnic groups (Buikstra1995). More broadly, the context in which the ances-tors are interred speaks to resource control and, byextension, regional sociopolitical interactions.


Tomb Context. In the Ilo Valley, El Yaral, Chirib-aya Alta, and Chiribaya Baja have formallybounded cemeteries adjacent to residential terraces.The burials at San Gernimo, on the other hand,were recovered largely from midden contexts. Thisdifference may indicate that the ancestors at the lat-ter site were not being utilized as a corporate man-ifestation of resource control (Buikstra 1995:258).

It has been argued that the population at San Ger -nimo had an economic focus on marine resourcesbased upon grave goods (Buikstra 1995) anddietary reconstruction (Tomczak 2003). It is pos-sible that the rich littoral resources at the coast obvi-ated the need to signal corporate ownership orcontrol of those resources. Arable land, however,is more circumscribed and requires a significant


Figure 7. Exposed stone-lined tombs from looted cemeteries along the Tambo-Ilo coast. A rectangular tomb is visible atthe site of Callango (top) while a circular tomb is exposed at the site of Alfaro (bottom).

investment of time and energy, leading to the cre-ation of bounded cemeteries as indicators of con-trol (Goldstein 1976).Among intervalley coastal springs like Chuza,

Zaro (2007) has argued that the proximity of tra-ditional resource zones (e.g., lomas, arable land,

marine resources) may not have necessitated eco-nomic specialization among those Chiribaya com-munities but, rather, encouraged intensivemanagement of a mixed subsistence economy. Ifwe are to accept the relationship between burialcontext and corporate resource control mentioned


Table 3. Areal Extent of Chiribaya Cemeteries in the Ilo Region.

Intervalley Location Coast (m2) Ilo Valley (m2) ReferencesChuza 600 Umire Alvarez 1996Carrizal 1100 estimate from average size of all other coastal springsAlastaya 1100 estimate from average size of all other coastal springsPocoma 1100 estimate from average size of all other coastal springsPocoma 1100 estimate from average size of all other coastal springsPocoma 1100 estimate from average size of all other coastal springsAlfaro 690 Umire Alvarez 2006Higocinto 1017 Umire Alvarez 2006Callango 8576 Umire Alvarez 2006Callango 3630 Umire Alvarez 2006Yerba Buena 1250 Umire Alvarez 2006Amoquinto 600 Umire Alvarez 2006Cola de Zorro 314 Umire Alvarez 2006Cola de Zorro 79 Umire Alvarez 2006Cola de Zorro 600 Umire Alvarez 2006TI-188 170 Umire Alvarez 2006TI-188 120 Umire Alvarez 2006TI-195 880 Umire Alvarez 2006Cocotea 1256 Umire Alvarez 2006Cocotea 600 Umire Alvarez 2006Cocotea 225 Umire Alvarez 2006Cocotea 160 Umire Alvarez 2006Cocotea 225 Umire Alvarez 2006Pacae 1962 Umire Alvarez 2006San Gernimo II 3493 Owen 1993San Gernimo 1046 Owen 1993Chiribaya Baja 2952 Owen 1993Chiribaya Baja 5481 Owen 1993Chiribaya Baja 1043 Owen 1993Algodonal 1925 estimated from map in Owen 1993Site 222 2844 Owen 1993Loreto Viejo 1816 Owen 1993Loreto Viejo 3217 Owen 1993Site 237 2000 estimated from map in Owen 1993Chiribaya Alta 7967 Owen 1993Chiribaya Alta 9754 Owen 1993Chiribaya Alta 5564 Owen 1993Chiribaya Alta 4977 Owen 1993Chiribaya Alta 9972 Owen 1993Chiribaya Alta 14436 Owen 1993Chiribaya Alta 7631 estimated from average of all Chiribaya Alta cemeteriesChiribaya Alta 749 Owen 1993El Yaral 2200 estimate from map in Buikstra 1995El Yaral 900 estimate from map in Buikstra 1995Total 28454 89967Note: Estimates were generated for those cemeteries whose areas were not originally reported, or for those whose areaswere combined with much larger domestic areas.

above, it is interesting to note that Chuza burialswere found within both small midden contexts anddomestic contexts (adjacent to, within, or beneathremnants of domestic architecture). Hence, itappears that the economic diversity among inter-valley Chiribaya may have relaxed the need forstrict material expressions of corporate resourcecontrol, resulting in the placement of tombs inmixed domestic and midden contexts within a sin-gle cemetery.There is also a spatial pattern to the distribution

of lined vs. unlined tombs within the Chuza ceme-tery. Those interspersed with small domestic mid-den features (found exclusively on the westernextent of the excavated area) have stone- lined walls,

while those associated with architectural features(all but one of these features are found in the east-ern half of the excavated area) are unlined. This pat-tern may relate to the stability of the surroundingmatrix: Midden deposits are notoriously porousand unstable, perhaps rendering stone- lined tombsmost effective when creating burial pits in theseareas. Regardless, all tomb forms discussed below(i.e., circular, rectangular, irregular) are representedin each of these areas.Tomb Form. Tomb form serves as an important

mortuary attribute when making comparisonsamong Ilo Valley Chiribaya, as there are significantdifferences in tomb form distributions betweenthese populations (Buikstra 1995). In the lower


Figure 8. Comparison of looted cemeteries (m2) among intervalley Chiribaya and Ilo Valley Chiribaya mortuary land-scapes.

Figure 9. Comparison of Chiribaya tomb form among intervalley (Chuza), lower valley (San Gernimo), and lower por-tion of the middle valley (El Yaral) mortuary landscapes.

portion of the middle valley, circular to oval cistspredominate at the site of El Yaral, with only eightexamples of rectangular tombs. This distributionleads Buikstra (1995:258) to suggest that the mor-tuary behavior at El Yaral resembles earlier mid- valley tomb forms. On the other hand, rectangulartombs are common in coastal valley Chiribayacemeteries (Buikstra 1995:258). At San Gernimo,for instance, just a few hundred meters from thecoast, all 92 excavated tombs were rectangular.Contrary to what is observed among these Ilo Val-ley sites, there is no dominant tomb type at Chuza.Rectangular tombs are more common (n = 47, or44 percent), but circular tombs are also well rep-resented (n = 27, or 25 percent). There are an addi-tional 34 irregular tombs (31 percent) that couldnot be easily classified into these categories (Fig-ure 9).4 This variability within the Chuza cemeterymay reflect the mixed strategies inherent amongintervalley coastal villages that embraced bothfarming (typically valley- centric) and maritime(typically littoral /coastal- centric) activities.Grave Accompaniments. There is significant

regional variability in the quantity of grave goodsamong Ilo Valley Chiribaya sites. Tombs at El Yaralwere not particularly rich (i.e., quantity of mater-ial inclusions), while at San Gernimo theyincluded an average of more than 30 grave itemsper tomb (Buikstra 1995:259). At Chiribaya Alta,many tombs contained numerous ceramics, tools,and ornaments. Most notably, Burial #419, whichcontained one adult male and two adult females,included a total of 32 ceramic vessels, 20 baskets,25 textiles, and six metal items (Buikstra1995:260). Strontium isotope analyses of the maleproduced a local Ilo Valley signature (Knudson andBuikstra 2007:576), and he has been described asa potential paramount lord (Lozada and Buikstra2005).While not quite as lavish, the richness and diver-

sity of mortuary assemblages at Chuza are com-parable to those of Ilo Valley centers. For instance,Burial #3, a rectangular tomb with a double inter-ment of an adult and child, contained 20 ceramicvessels in addition to bone and lithic artifacts. Sim-ilarly, Burial #6, a rectangular tomb containing twochildren, included 17 ceramic vessels, one minia-ture balsa raft, and several baskets (Figure 10).Elsewhere, the irregularly shaped tomb of Burial#39 contained a single adult with 13 ceramic ves-

sels, two textiles, a wind instrument, a small raft,animal bones, and marine shells, among otheritems. A number of Chuza burials also containedexamples of four- cornered hats. The individual inBurial #39, for instance, was buried with a red four- cornered hat decorated with three small copperplates. Four- cornered hats have also been recoveredfrom Chiribaya Alta in the Ilo Valley, a settlementargued to be the center of Chiribaya political author-ity. Farther inland along the Moquegua River, thereappears to be a correlation between four- corneredhats and tabular forms of cranial modificationamong earlier Tiwanaku settlers (Blom 2005;Janusek 2004, 2008). Among Chiribaya settle-ments, Lozada and Buikstra (2005:216) argue thattabular forms of cranial modification are most com-monly associated with valley agriculturalists.The material artifacts recovered from the Chuza


Figure 10. Burial #6 at Chuza showing richness of graveassemblage.

cemetery underscore two points relative to our argu-ment. First, the richness and diversity of gravegoods demonstrate that intervalley coastal popula-tions during the late Chiribaya period were neitherperipheral nor marginalized. Although Chuza lieson the small end of the intervalley spectrum ofcemetery sizes (i.e., 600 m2), its mortuary assem-blage suggests that these populations, througheither direct production or exchange, had access tothe quantity and quality of polychrome ceramicvessels and other materials typically invested in IloValley tombs. Second, material indicators of eco-nomic activity point toward diversification andmixed subsistence strategies. Evidence of eco-nomic diversity includes the recovery of artifactstraditionally associated with an emphasis on mar-itime resources (e.g., harpoons, miniature balsarafts, and marine shells), while an abundance ofcamelid bones may suggest exploitation of the sur-rounding lomas for grazing. The presence of stone- faced terraces associated with the Chuza site alsopoints to farming as a significant activity as well.Finally, if items like four- cornered hats correlate tovalley agriculturalists, then the presence of four- cornered hats like that of Burial #39 at Chuza rep-resents a degree of socioeconomic integrationbetween intervalley and main valley people.

Discussion and Conclusion

While most research along the Andean coast isrestricted to primary river drainages and their canal-ized extensions, our work suggests that Chiribayaliving along the TamboIlo intervalley coast werenumerous during the thirteenth and fourteenth cen-turies and anything but marginal. These regularlydispersed populations were organized into smallervillages and managed smaller farmsteads, but com-bined, they formed a significant segment of theChiribaya cultural landscape. Today, the area mayindeed be identified as marginal, since residentpopulations are all but absent, agricultural produc-tion is considerably diminished, and the inland hillshave become highly desiccated. We believe that thisbarren state of the region has unfortunately shapedcontemporary human perspectives on its value andimportance in the more ancient past. The combi-nation of a rich marine resource base (both offshoreand littoral), a productive inland lomas, and anintensively agro- engineered intervalley landscape

was likely more self- sustaining than the Ilo Valleyby providing resident populations with an oppor-tunistic mixed economic strategy. The strength ofthe archaeological expression of Chiribaya acrossthe intervalley landscape also points to a complexdegree of socioeconomic organization within theChiribaya seoro. These results suggest thatancient populations were not simply restricted toprimary river drainages along the western Andeanwatershed but, rather, significant segments of apopulation could intensively engineer intervalleylandscapes into regional subsistence economies ina manner that was substantially different from thatof their river valley neighbors.These sorts of expressions of human ingenuity

are found beyond the canalized extensions of otherwestern Andean river valleys, though they oftenreceive less attention by scholars. One exception isthe Lomas de Atiquipa region, situated about 350km north of the TamboIlo area between Quebradade Chala and the Yauca River of southern Peru. Sim-ilar to the TamboIlo coast, inland hills there reachheights above 1,200 m asl in a short horizontal dis-tance from the sea, while the general slope andaspect of the terrain create favorable conditions forlomas vegetation, where winds predominately carrymoisture from the south to intersect with groundsurfaces above 250 m asl. Condensation of thismoisture supports the growth of herbaceous plantsand trees, which in turn compounds favorable con-ditions for vegetation by providing ground coverand shade, further reducing evapotranspiration andenhancing soil moisture retention (Canziani 1998).While the lomas today support about 2,190 ha offorested area, spring- fed canals irrigate about 350ha of cultivated terrain that includes the productionof fruit trees, alfalfa, and maize (CIZA- ONERN- SENAMHI 1989). It has been argued that the deser-tification of these lomas occurred in large part dueto deforestation and overgrazing since the Spanishcolonial period, but they have been partly regener-ated over the past four decades with the develop-ment of atrapanieblas, or fog catchers, to provideirrigation water for the initial reestablishment ofendemic vegetation. Once mature, these standingforests serve as living atrapanieblas in these coastalecosystems (Canziani 1998; Talavera et al. 2005).While the Lomas de Atiquipa and the TamboIlo

coast continue along very different trajectoriestoday and with different cultural values placed upon


them, they do share comparable human compo-nents and land- use histories during the late prehis-panic and Spanish colonial periods, varying onlyin degree rather than kind. The Lomas de Atiquipaexhibits similar ecological zones as the TamboIloregion in the arid littoral, alluvial plains, and inlandlomas. The archaeological record of the regionpoints to numerous late prehispanic villages in thesevarious zones, such as Quebrada de la Vaca and LaCaleta along the littoral and Cahuamarca amongthe inland hills. Furthermore, extensive tracts ofabandoned stone- faced terraces situated belowabout 300 to 400 m asl testify to the regions inten-sive agricultural past. From preliminary observa-tion, Canziani (1998:177) estimates that spring- fedcanals and stone- faced terraces transformed about2,600 ha of arid terrain into a productive agricul-tural landscape in antiquity. In this case, the inter-section of microenvironmental conditions andhuman ingenuity produced a highly engineeredlandscape that rendered it a focal point of a mixedsubsistence strategy among late prehispanic coastalpopulations situated beyond the hydrological lim-its of primary river drainages.We believe that, while admittedly on a much

smaller scale, the TamboIlo coast represents asimilar configuration that provided late prehispaniccoastal populations a rewarding alternative to sim-ply valley- focused or coastal- focused subsistencestrategies. In many ways, the management oflomas, agricultural, and marine resource bases byintervalley coastal people is more reminiscent ofhighland strategies embedded within the conceptof ecological complementarity than socioeconomicmodels of horizontal exchange that have typicallycharacterized the Andean coast. The upland lomasmay have provided grasslands for grazing and con-siderable winter moisture for dry farming, whilefreshwater springs situated among lower alluvialplains offered the potential for intensive canal andterrace agricultural strategies. Finally, the rockycoastline with intermittent sandy beaches providedan ideal setting for managing the rich littoral andoffshore resource bases. Similar to highland com-munities that managed field plots among a varietyof temperature and precipitation gradients in theAndes, intervalley coastal Chiribaya may havereduced risk and secured a degree of self- sufficiency by engineering these vertically stackedintervalley resource potentials into a complex sub-

sistence economy.A comparison of these ancient landscapes to

their contemporary conditions may also point tosignificant environmental change on timescalesranging from centuries to millennia. As demon-strated here, the mixed resource base of theTamboIlo coast was intensively managed duringthe thirteenth and fourteenth centuries and sup-ported a significant resident population. Today it isan industrialized wasteland with apparently littlebiological potential. While the rate, timing, andagents of ecological change remain unclear, it isevident that significant changes have occurred overthe past five or six centuries. Clearly, moisture wasmore available in the past, where freshwater springssustained enough discharge to be directed throughlong irrigation canals and into terraced field sys-tems. Local aquifers during the late prehispanicperiod likely benefited substantially from 900 yearsof increased El Nio activity prior to the sixteenthcentury, as Magilligan et al. (2008) have indicatedfrom their Moquegua Valley study. A few farm-steads today continue to be irrigated by freshwatersprings, but other terraced field systems lie aban-doned as springwater continues to flow nearby. Stillother springs have diminished completely. In thecontext of global change, archaeologists can thusplay a central role in investigations of ecologicaldynamics on multiple temporal and spatial scales(see also Beresford- Jones et al. 2009). In southernPeru, we hope that our continued focus on humanagency and environmental change will clarify thatprocess as it unfolded over the past millennium andprovide a much needed historical ecological con-text for the marginality that characterizes the regiontoday.We believe that the desiccated and industrial-

ized nature of the TamboIlo intervalley landscapehas led not only to its marginality and devaluationon a scholarly level but also to its marginalityamong contemporary communities in the region.Few residents in the Ilo area identify with the his-torical significance of this landscape. The archae-ological record of the lower valley and coast pointsto a significant population collapse after the lateChiribaya period, where numbers remaineddeflated until the striking growth of Ilo in the lasthalf of the twentieth century with the establish-ment of the Southern Peru Copper Corporation(Hall 1992; Zaro 2005b). Importantly, a compari-


son of historical documents of the early twentiethcentury (Paernio 1908) to historic aerial pho-tographs of the midtwentieth century indicatesthat many spring systems near Ilo were alreadyexperiencing a sharp decline (and therefore adecline in agricultural importance) well before thepopulation boom catalyzed in part by SPCC (Zaro2005b:201). Consequently, only a small percent-age of Ilo residents has historic ties to the area priorto the twentieth century, when the regions agro- ecological past was more pronounced.Our reference to tierras olvidadas, or forgotten

lands, thus has dual meanings. It refers to the lackof historical connection to these intervalley land-scapes by contemporary peoples of the Ilo regionbut also to their perceived unimportance withinscholarly communities, at least relative to adjacent,more populated areas. This study illuminates thepotential for some of these marginalized environ-ments to significantly alter our perceptions andunderstanding of the complexity, breadth, anddiversity of human ingenuity in resource manage-ment and the processes of socioenvironmentalchange beyond the hydrological limits of arid west-ern Andean river valleys.

Acknowledgments. The Heinz Foundation for LatinAmerican Archaeology graciously provided support for theinitial archaeological survey of the TamboIlo coast, whilethe University of Maine and the State University of NewYorkNew Paltz provided additional funds. Archaeologicalpermits to conduct the survey and subsequent excavationswere granted by the National Institute of Culture in Lima,Peru. The Museo Contisuyo and its staff were also very kindto offer logistical support throughout the duration of the fieldportions of our work in southern Peru, and we particularlyrecognize Antonio Oquiche Hernani, Ryan Williams, andMike Moseley for their continued encouragement of ourresearch along the coast.

We thank the participants of the 27th NortheastConference on Andean Archaeology and Ethnohistory heldat the University of Maine (October 2008), where some ofthese ideas were first presented. We also thank HelaineSilverman, Chris Pool, Gabriela Uruuela y Ladrn deGuevara, Dan Sandweiss, Jane Buikstra, and the LatinAmerican Antiquity staff for their constructive comments andeditorial assistance. Finally, we recognize Lic. Daniel IvnDvila Manrique and students Antoine Carranza, LouisFortin, Laura Labbe, Heather Slivko- Bathurst, and AlissaDubois for their contributions in various phases of thisresearch. While we are grateful for the contributions of oth-ers, any shortcomings of this article remain strictly our own.

References CitedBawden, Garth1989 Settlement Survey and Ecological Dynamics on thePeruvian South Coast. Andean Past 2:3967.

Beattie, Kathleen G., Wayne K. Bond, and Edward W. Man-ning

1981 The Agricultural Use of Marginal Lands: A Reviewand Bibliography. Lands Directorate, EnvironmentCanada, Working Paper No. 13.

Beresford- Jones, David G., Susana Arce T., Oliver Q. Whaley,and Alex J. Chepstow- Lusty

2009 The Role of Prosopis in Ecological and LandscapeChange in the Samaca Basin, Lower Ica Valley, SouthCoast Peru from the Early Horizon to the Late Intermedi-ate Period. Latin American Antiquity 20:303332.

Binns, Tony2007 Marginal Lands, Marginal Geographies. Progress inHuman Geography 31:587591.

Blom, Deborah E.2005 Embodying Borders: Human Body Modification andDiversity in Tiwanaku Society. Journal of Anthropologi-cal Archaeology 24:134.

Bronk Ramsey, C., T. F. G. Higham, F. Brock, D. Baker, and P.Ditchfield

2009 Radiocarbon Dates from the Oxford AMS System:Archaeometry Datelist 33. Archaeometry 51:323349.

Buikstra, Jane E.1995 Tombs for the Living . . . or . . . for the Dead: The OsmoreAncestors. In Tombs for the Living: Andean Mortuary Prac-tices, edited by Tom D. Dillehay, pp. 229280. DumbartonOaks Research Library and Collection, Washington, D.C.

Canziani, Jos1998 Las Lomas de Atiquipa: Un Caso de Paisaje Culturalen la Costa Desrtica del Sur del Per. In Paisajes Cul-turales en los Andes, edited by E. Mujica Barreda, pp.169190. Patrimonio Mundial, Arequipa.

CIZA- ONERN- SENAMHI1989 Aprovechamiento de nieblas costeras en las zonasridas de la costa, Lomas de Atiquipa, Provincia de Car-avel, Departamento de Arequipa, Peru. Consejo Nacionalde Ciencia y Tecnologa, Lima.

Clement, Christopher O., and Michael E. Moseley1991 The Spring- Fed Irrigation System of Carrizal, Peru A Case- Study of the Hypothesis of Agrarian Collapse.Journal of Field Archaeology 18:425443.

Covey, R. Alan2000 Inka Administration of the Far South Coast of Peru.Latin American Antiquity 11:119138.

Doolittle, William E.1984 Agricultural Change as an Incremental Process.Annals of the Association of American Geographers74:124137.

1988 Intermittent Use and Agricultural Change on Marginal Lands The Case of Smallholders in Eastern Sonora,Mexico. Geografiska Annaler Series B Human Geogra-phy 70:255266.

Fontugne, Michel, Pierre Usselmann, Danile Lavalle, MichleJulien, and Christine Hatt

1999 El Nio Variability in the Coastal Desert of SouthernPeru During the Mid- Holocene. Quaternary Research52:171179.

Fortin, Louis W.2008 Geoarchaeological Investigations Along theTamboIlo Coast of Southern Peru. Unpublished Mastersthesis, Climate Change Institute, University of Maine,Orono.


Frezier, Amadeo1982 [1713] Relacin del viaje por el Mar del Sur. Bib-lioteca Ayacucho, Ayacucho.

Goldstein, Lynne1976 Spatial Structure and Social Organization: RegionalManifestations of Mississippian Society. UnpublishedPh.D. dissertation, Department of Anthropology, North-western University, Evanston.

Hall, Arthur C.1992 Nothing Down. Sundance Press, Tucson.

Hoshower, L. M., Jane E. Buikstra, Paul S. Goldstein, and A.D. Webster

1995 Artificial Cranial Deformation at the Omo 10 Site: ATiwanaku Complex from the Moquegua Valley, Peru. LatinAmerican Antiquity 6:145164.

Janusek, John Wayne2004 Identity and Power in the Ancient Andes: TiwanakuCities Through Time. Routledge, London.

2008 Ancient Tiwanaku. Cambridge University Press, NewYork.

Jessup, David A.1990 Desarrollos generales en el Intermedio Tardo en elValle de Ilo, Per. Field report on file at Programa Con-tisuyo, Moquegua, Peru.

1991 General Trends in the Development of the ChiribayaCulture, South- Coastal Per. Paper presented at the 56thAnnual Meeting of the Society for American Archaeology,New Orleans.

Keefer, David K., Michael E. Moseley, and Susan D. deFrance2003 A 38,000-Year Record of Floods and Debris Flows inthe Ilo Region of Southern Peru and Its Relation to El NioEvents and Great Earthquakes. Palaeogeography, Palaeo-climatology, Palaeoecology 194:4177.

Knudson, Kelly J., Arthur E. Aufderheide, and Jane E. Buikstra2007 Seasonality and Paleodiet in the Chiribaya Polity of South-ern Peru. Journal of Archaeological Science 34:451462.

Knudson, Kelly J., and Jane E. Buikstra2007 Residential Mobility and Resource Use in the Chiri -baya Polity of Southern Peru: Strontium Isotope Analysisof Archaeological Tooth Enamel and Bone. InternationalJournal of Osteoarchaeology 17:563580.

Lozada, Mara Cecilia, and Jane E. Buikstra2002 El seoro de Chiribaya en la costa sur del Per. Insti-tuto de Estudios Peruanos, Lima.

2005 Pescadores and Labradores Among the Seoro ofChiribaya in Southern Peru. In Us and Them: Archaeol-ogy and Ethnicity in the Andes, edited by Richard M. Rey-craft, pp. 206225. Cotsen Institute of Archaeology,University of California, Los Angeles.

Magilligan, F. J., Paul S. Goldstein, G. B. Fisher, B. C. Bostick,and R. B. Manners

2008 Late Quaternary Hydroclimatology of a Hyper- AridAndean Watershed: Climate Change, Floods, and Hydro-logic Responses to the El NioSouthern Oscillation inthe Atacama Desert. Geomorphology 101:1432.

McCormac, F. G., A. G. Hogg, Paul G. Blackwell, Caitlin E.Buck, T. F. G. Higham, and Paula J. Reimer

2004 SHCAL04 Southern Hemisphere Calibration, 011Cal kyr BP. Radiocarbon 46:10871092.

Miranda, Ana, and Adn Umire Alvarez2007 Costumbres sepulcrales y contextos funerarios. Gr-fica G&C, Arequipa.

ONERN1976 Inventario, evaluacin y uso racional de los recursosnaturales de la costa: Cuencas de los Ros Moquegua,Locumba, Sama y Caplina. Oficina Nacional de Eval -uacin de Recursos Naturales, Lima.

Owen, B. D.1993 A Model of Multiethnicity: State Collapse, Compe-tition, and Social Complexity from Tiwanaku to Chiri bayain the Osmore Valley, Peru. Unpublished Ph.D. disserta-tion, Department of Anthropology, University of Califor-nia, Los Angeles.

1994 Inventario arqueolgico del Valle Osmore Costero:Informe del campo e informe final. Field report submittedto the Instituto Nacional de Cultura, Lima.

2005 Distant Colonies and Explosive Collapse: The TwoStages of the Tiwanaku Diaspora in the Osmore Drainage.Latin American Antiquity 16:4580.

Paernio, P.1908 Los Olivares de Moquegua. Publications of the Min-isterio de Fomento 30. Imprenta del Estado, Lima.

Pollard, Sydney1997 Marginal Europe: The Contribution of MarginalLands Since the Middle Ages. Clarendon Press, Oxford.

Potter, Rob2001 Geography and Development: Core and Periphery?Area 33:422427.

Rein, Bert, Andreas Luckge, and Frank Sirocko2004 A Major Holocene ENSO Anomaly During theMedieval Period. Geophysical Research Letters31(L17211):14.

Reycraft, Richard M.1998 The Terminal Chiribaya Project: The Archaeology ofHuman Response to Natural Disaster in South CoastalPeru. Unpublished Ph.D. dissertation, Department ofAnthropology, University of New Mexico, Albuquerque.

2000 Long- Term Human Response to El Nio in SouthCoastal Per, Circa A.D. 1400. In Environmental Disas-ter and the Archaeology of Human Response, edited byGarth Bawden and Richard M. Reycraft, pp. 99120.Anthropological Papers No. 7. Maxwell Museum ofAnthropology, Albuquerque.

Rostworowski, Mara1981 Recursos naturales renovables y pesca, siglos XVI yXVII. Instituto de Estudios Peruanos, Lima.

Sandweiss, Daniel H., Kirk A. Maasch, Richard L. Burger,James B. Richardson III, Harold B. Rollins, and AmyClement

2001 Variation in Holocene El Nio Frequencies: ClimateRecords and Cultural Consequences in Ancient Peru. Geol-ogy 29:603606.

Satterlee, Dennis R.1993 Impact of a Fourteenth Century El Nio on an Indige-nous Population near Ilo, Per. Unpublished Ph.D. dis-sertation, Department of Anthropology, University ofFlorida, Gainesville.

Satterlee, Dennis R., Michael E. Moseley, David K. Keefer, andJ. E. Tapia A.

2000 The Miraflores El Nio Disaster: Convergent Cata-strophes and Prehistoric Agrarian Change in SouthernPeru. Andean Past 6:95116.

Stanish, Charles1992 Ancient Andean Political Economy. University ofTexas Press, Austin.

Talavera, C., P. Jimnez, F. Villasante, L. Villegas, and L. Ortega2005 Atiquipas Lomas: Restoring a Unique and SingularHabitat. Paper presented at the 2nd Southern Deserts Con-ference, HumanEnvironment Interactions in SouthernHemisphere Deserts: Past, Present, and Future, Arica.

Thouret, Jean-Claude, Jasmine Davila, and Jean-Philippe Eissen1999 Largest Explosive Eruption in Historical Times in theAndes at Huaynaputina Volcano, A.D. 1600, SouthernPeru. Geology 27:435438.


Tomczak, Paula D.2003 Prehistoric Diet and Socioeconomic RelationshipsWithin the Osmore Valley of Southern Peru. Journal ofAnthropological Archaeology 22:262278.

Treacy, John M., and William M. Denevan1994 The Creation of Cultivable Land Through Terracing.In The Archaeology of Garden and Field, edited by N. F.Miller and K. L. Gleason, pp. 91110. University of Penn-sylvania Press, Philadelphia.

Umire Alvarez, Adn1994 Inventario arqueolgico de la costa norte de Ilo. Fieldreport submitted to the Instituto Nacional de Cultura, Lima.

1996 Inventario arqueolgico de la costa norte de Ilo,Segunda temporada 19951996. Field report submitted tothe Instituto Nacional de Cultura, Lima.

2006 Informe tcnico: Proyecto arqueolgico TamboIlo2006. Field report submitted to the Instituto Nacional deCultura, Lima.

Umire Alvarez, Adn, and Ana Miranda2001 Chiribaya de Ilo. Eskann, Arequipa.

Waldhardt, R., K. Fuhr- Bossdorf, and A. Otte2001 The Significance of the Seed Bank as a Potential forthe Reestablishment of Arable- Land Vegetation in a Mar-ginal Cultivated Landscape. Web Ecology 2:8387.

Watanabe, Luis K., Michael E. Moseley, and Fernando Cabieses(editors)

1990 Trabajos arqueolgicos en Moquegua, Per, 1. Pro-grama Contisuyo, Moquegua.

Williams, Patrick Ryan1997 The Role of Disaster in the Development of Agricul-ture and the Evolution of Social Complexity in the South- Central Andes. Unpublished Ph.D. dissertation,Department of Anthropology, University of Florida,Gainesville.

Wu, Gang, Jing Li, Hong-Mei Kong, Jing-Zhu Zhao, and LinGao

2001 Exploiting Pattern and Sustainable Development ofMarginal Lands in the Three Gorge Reservoir Areas. Jour-nal of Environmental Sciences China 13:173179.

Zaro, Gregory2005a Desiccated Lands: Agricultural Technology andLandscape Evolution Along the Far South Coast of Peru,A.D. 10002000. Paper presented at the 2nd SouthernDeserts Conference, HumanEnvironment Interactions inSouthern Hemisphere Deserts: Past, Present, and Future,Arica.

2005b Human Ecology and Ancient Agrarian Land Use atWawakiki Spring, South Coast Peru, A.D. 10002000.Unpublished Ph.D. dissertation, Department of Anthro-pology, University of New Mexico, Albuquerque.

2007 Diversity Specialists: Coastal Resource Managementand Historical Contingency in the Osmore Desert of South-ern Peru. Latin American Antiquity 18:161179.

2008a Agriculture, Landscape Evolution, and Desertifica-tion Along the Peruvian South Coast. Paper presented at

the 3rd Southern Deserts Conference, Climate Changeand the Peopling of the Southern Deserts, Molopo Lodge,Northern Cape, South Africa.

2008b Pampa Dispensilla Geological Project. Manuscripton file, Department of Anthropology, University of Maine,Orono.

Zaro, Gregory, Heather Builth, Claudia Rivera, Jimena Roldn,and Graciela Suvires

2008 Landscape Evolution and Human Agency: Archaeo-logical Case Studies from Drylands in Western SouthAmerica and Australia. Chungar, Revista de AntropologaChilena 40(Nmero Especial):261271.

Zaro, Gregory, and Adn Umire Alvarez2005 Late Chiribaya Agriculture and Risk ManagementAlong the Arid Andean Coast of Southern Peru, A.D.12001400. Geoarchaeology An International Journal20:717737.

Notes

1. Combined, the Moquegua and Ilo rivers (plus uppertributaries) constitute the Osmore drainage; the MoqueguaRiver flows through the middle portion, while the Ilo Riverflows through the lower and coastal portions.

2. Because of the difficulties of delineating the arealextent of habitation areas, midden deposits, and debris scat-ters from surface remains alone, we have chosen not toinclude survey data pertaining to these features in this discus-sion. These were recorded differentially among the severalarchaeological projects that have produced reconnaissancedata over the past several decades, largely stemming from theintegrity of the archaeological record itself. Therefore, weinclude only a comparative discussion of Chiribaya mortuaryfeatures.

3. Since intervalley coastal surveys generated only sur-face data, we felt it prudent to utilize only original surfacedata generated in the Ilo Valley survey for comparison (Owen1994) rather than Ilo Valley excavation data from Buikstra(1995). While excavation data provide a more precise mea-sure of the areal extent of a cemetery (and number of inter-ments), comparison of looted cemeteries to other lootedcemeteries from surface evidence alone may provide a morevalid measure of relative differences.

4. In this study, circular includes circular, oval, andsemicircular tomb forms as initially recorded in Miranda andUmire Alvarez 2007, while rectangular includes rectangu-lar, quadrangular, and semirectangular tomb forms as theywere initially recorded.

Submitted May 6, 2009; Accepted: July 29, 2009.

374 LATIN AmerIcAN ANTIquITy [Vol. 21, No. 4, 2010]

Documents

Marginality Chiribaya - Zaro