Impact of a fourteenth century El Niño flood on an indigenous population near Ilo, Peru

MISSING IMAGE

Material Information

Title:
Impact of a fourteenth century El Niño flood on an indigenous population near Ilo, Peru
Physical Description:
xiii, 400 leaves : ill. ; 29 cm.
Language:
English
Creator:
Satterlee, Dennis Ray
Publication Date:

Subjects

Subjects / Keywords:
Floods -- Peru -- Ilo Region   ( lcsh )
El Niño Current   ( lcsh )
Ilo Region (Peru)   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 362-398).
Statement of Responsibility:
by Dennis Ray Satterlee.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001942932
notis - AKB9170
oclc - 31054316
System ID:
AA00003259:00001

Full Text










IMPACT OF A FOURTEENTH CENTURY EL NINO FLOOD
ON AN INDIGENOUS POPULATION NEAR ILO, PERU










BY

DENNIS RAY SATTERLEE


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1993


UNIVERSITY OF FLORIDA LIBRARIES





























Copyright 1993

by

Dennis Ray Satterlee






























To Carole (D.L.)--The Wind Beneath My Wings









ACKNOWLEDGMENTS


A dissertation is large undertaking which can never be

accomplished by a single individual. Many important contributions
were made to this study by a number of people, and I wish to extend

to each of them my most heartfelt thanks and deepest appreciation.

I wish to thank all my Peruvian friends who either directly
helped me in the field or who graciously imparted knowledge about

Peruvian prehistory. In particular I am indebted to some of my

fellow Programa Contisuyu members, Shawn Penmann, Rick Reycraft,
Bruce Owen, David Jessup, Elva Torres, Nene Lozada, and Jorge Tapia,
who help to make fieldwork pleasurable, most of the time. Special

thanks go to Dr. Karen Wise, curator at the Los Angeles Natural

History Museum, not only for the use of her slides of Miraflores

Quebrada in making some figures used in Chapter 5, but also because
she is a very good archaeologist and helpful friend. I also wish to
thank the other Peruvians, Walter, Helbert, and Felipe, who helped
me in the field, and who were always kind to the large Gringo who

speaks the not-so-good Spanish.

I would be remiss if I did not generously thank the personnel
of S.P.C.C., such as Ernie, Wayne, Ralph, Eduardo at Campamentos,
and Rick, and a myriad of other unknown personnel who do the
fabrication of anything that we need, and who also perform the
maintenance on our venerable field vehicles.
I wish to thank a few of my fellow UF graduate students who
helped me, and who took the time to listen to me. In particular, I
iv









want to thank Tom Eubanks, who was always a pleasure to be

associated with, and who helped me immensely in learning the
Illustrator computer program which I use to make my maps, etc.

Also some helpful suggestions were made from time to time by Steve

Kryzton. I wish to thank Greg Smith for being one of the most
friendly graduate students in the UF anthropology department. I

wish to thank George Avery and Ryan Wheeler, who were always
willing to help when any was needed. Special thanks go to the
"Maestro," Prof. Michael E. Moseley, for being my major professor for

these last four years and helping me to realize my life-time dream of

becoming a real archaeologist.

A number of figures were adapted from the works of others
including the following: Figure 1-5 was adapted from Clement and
Moseley 1991; Figure 2-2 was adapted from Dillon 1985; Figure 2-3
was adapted from S.P.C.C. 1985; Figure 3-2 was adapted from

Moseley 1992; Figure 3-5 was adapted from Silgado 1978; Figures 5-
14 and 7-26 were adapted from original drawings made by Nikki

Clark, with additional interpretations by Michael E. Moseley and
Jorge Tapia. Figures 7-11 and 7-12 were adapted from original field
drawings made by Karen Wise and myself in 1990. Of course, any
errors in translations and interpretations are solely my

responsibility.
Lastly, I wish to thank profusely Carole, my wife and chief field
assistant, for enduring the hardships and the loneliness while I was
in Florida and in Peru doing field work. It was not an easy task for
either of us.









TABLE OF CONTENTS



ACKNOW LEDGM ENTS ........................................................................................iv

A BSTRA CT ............................................................................................... xii

CHAPTERS

1 INTRODUCTION............................................ ........................ 1

Purpose of Study ................................................................ 1
Physical Setting............................................... ................5
Far-Southern Peru ................................................... .............. 5
The Coastal Quebradas ........................................ ........... .. 9
Introduction ................................... ................................ 9
Catchm ent A rea .................................................................... 1 1
Carrizal Quebrada............................................................. 12
Total Agricultural Area ................................ ............ 12
Settlement Size.............................................................. 1 3
Miraflores Quebrada ........................ ............................. 13
Agricultural Area........................................................... 13
Settlement Size.............................................................. 1 4
Pocoma Quebrada............................................................... 1 4
Agricultural Area........................................................... 14
Settlement Size............................ ................................ 1 5
Chapter Summaries....................... .................................... 15

2 EL NINO-SOUTHERN OSCILLATION................................. 15

Introduction .......................................................... ............... 20
Background Information...................................................... 23
The Cause of El Nifios........................ .................................24
The Peruvian (Humboldt) Current...................................27
Southern Oscillation .......................................................... 29
Climatic Changes Associated with ENSOs..................... 30
Exceptional Rainfall .......................................... .............. 30
Drought Conditions ............................................ ............. 3
Correlations Between Volcanic Activity and
El Nifio Phenomena..............................................................









Correlation Between Global Warming and
the Frequency of El Nifio................................. ............. 6
Effects of Strong El Nifios................................ ............ 37
F flooding ..................................................................................... 3 8
Disease and Pestilence ..................................... ............. 1
Impact on Coastal Agriculture.......................... ............. 3
Impact on Highland Agriculture.......... ................................ 4
Domesticated Animals......................................................4 5
M arine L ife ...................................... ...........................................
Guano Birds...........................................................................4 9
Econom ic Im pact .................................................................... 50
Positive Consequences of an El Nifio Event............5....5 1
Lom as....................................................... ............................ 5 1
Applying Modern El Nifio Data to
Prehistoric Settings ............................................ ............... 3

3 ARCHAEOLOGICAL BACKGROUND .......................................5 6

Initial Period.................................. ......................................5 6
Early Horizon........................................................................ 5 7
Early Intermediate Period............................. ........... 9
Middle Horizon .................................................................... 59
Late Intermediate Period ..................... ........................ 60
Cultural History of the Ilo Region .....................................6 5
Lithic Period .......................................................... ............... 6 5
Preceram ic Period......................................................................
Initial Period........................................................ ................6 8
Early Horizon........................................................................6 9
Early Intermediate Period ...............................................
Middle Horizon ................................................ ................7 1
Late Intermediate Period................................. .............. 2
Intensification and Development of
Irrigated Agriculture............................... ............. ................ 7 6
Introduction .......................................................... ............... 7 6
Water Management...........................................................8 3
Development of Agriculture................................. ........... 83
Social Change Associated with Agriculture...............8...8 5
Motivating Factors for the Development
of A agriculture .................................................. ................ 88
Advantages of Agricultural Terracing ........................8 9
Use of Fertilizers................................................................. 90
Risk Management........................................ ....................... 9 1
vii









Contribution of Agriculture to the Prehistoric Diet.....9 5
Religion and Agriculture....................................................
Background ........................... ......................................... 9 6
Role of the Gods in Agriculture.......................................9 7
Adoration of Huacas......................................... .............. 8
Oracles and Religious Centers ............................................ 102
Environmental Stress .......................................................... 109
Constant Stress.................................................................... 109
Tectonics ............................................................................. 11 0
Earthquakes....................................................................... 11
Tectonic Uplift ....................................................................... 11 6
Volcanic Eruptions....................... ................................ 11 7
El Nifio Rains and Floods .............................................. 120
Flood Studies Conducted in Peru.......................................125
Introduction ............................................................................ 12 5
Previous Flood Studies...................................................... 25
The Prehistoric Flood Record in Northern Peru.....1......127
The Prehistoric Flood Record in the Ilo Valley and
in the Coastal Quebradas Near Ilo,
Far-Southern Peru ................................................................ 130
C onclusions..................................................................................... 13 7

4 M ETH O D S........................................................................................ 14 0

Introduction ............................................................................ 14 0
Field Survey ............................................................................ 141
Unit Excavations ..................................................................... 143
Trenches................................................................................... 144
Shovel Testing ........................................................................... 146
Unit Profiles and Floor Plans............................................... 148
Quebrada Geologic Columns.................................. ..... 150
Mapping ................................................................................... 152
Laboratory Analysis ............................................................... 153
Recovery of Carbon .............................................................. 153
Computer Methods .................................................................. 154
Creating Computer Maps, Profiles, and
Illustrations ........ ....................................................... 154
Producing a Three Dimensional Model of the
Ilo V alley .............................................................................. ... 1 5 6
D discussion ................................................................................. 16 0


viii









5 SITE EXCAVATIONS.......................................................... .. 1 6 3

Introduction ............................................................................... 163
Choosing the Locations of Units.........................................164
Carrizal Quebrada.................................................................... 164
Miraflores Quebrada ...................... ................................ 169
Pocoma Quebrada........................................................... 17 1
Excavations at Carrizal Quebrada......................................173
Introduction ............................................................................ 17 3
Location and Descriptions of Units.................................. 173
Location and Descriptions of Geologic Columns.......... 78
Location and Description of the Prehistoric Canal.... 182
Agricultural Terraces ................................... ............... 183
Shovel Testing at Carrizal Quebrada............ ............. 85
Introduction ............................................................................... 185
Location and Description of the Shovel Tests.............185
Cultural Area North of the Carrizal Quebrada .............187
Excavations at Miraflores Quebrada ................................188
Introduction ............................................................................... 188
Sunken Features at Miraflores Quebrada...................188
Location and Description of Units and
Geologic Columns.......................................................... 193
Survey of the Upper Miraflores Quebrada...................1 97
Excavations at Pocoma Quebrada...................................... 199
Introduction ............................................................................... 199
Location and Description of Units and Profiles......... 199
Location and Description of Shovel Tests ......................203
Prehistoric Terraces..............................................................205
1982-83 Run-off Channel..................................................205
Irrigation C anals................................ .......................................206
Geologic Columns ................................................................ 207
Location and Description of Shovel Tests ..................... 208
Investigations in the Ilo Valley.........................................2 11
Introduction ..................................... .......................................... 1 1
The Ilo Valley Flood Sequence........................................211
Agricultural Terraces ....................................... ............. 1 2
What do Excavations Indicate about the
Flood Severity? ................................................ ........... ..... 13
Introduction .............................................................................. 13
Impact at Carrizal Quebrada............................................ 1 3
Impact at Miraflores Quebrada....................................1 6
Impact at Pocoma Quebrada ..............................................2 17
ix









Impact in the lo Valley......................................... ............. 1 8
Evidence of the Survival or the Demise of
The C hiribaya ............................................................................ 2 2 0
Post-Miraflores Cultural Activity.................................... 220
Carrizal Quebrada................................................................. 220
Pocoma Quebrada......................................... ................ 22 1
The Ilo Valley ...................................................................... 222
Irrigated Agricultural in the Study Area ...................224
Introduction ............................................................................. 2 2 4
Types of Terraces Used in the Study Area ...............224
Types of Canals Used in the Study Area.................227
Irrigation Reservoirs ...................................................... 228
D discussion .......................................................................... 229

6 EXCAVATED DATA............................................................... 191

Introduction ............................................................................ 2 3 1
Types and Quantities of Material Expected from
E ach L ocality ....................................................................... ... 2 3 1
Excavated Data from the Carrizal Quebrada................... 233
Excavated Data from the Miraflores...............................243
Excavated Data from Pocoma Quebrada..........................269
What Recovered Artifacts Indicate about the
Strength of the Flood ..........................................................277
What Recovered Artifacts Indicate about a
Cultural Response or Change Caused by the Impact
of the Miraflores Flood .................................................... 285
Why were the Agricultural Terraces near Ilo
A bandoned......................................................................... .... 2 8 6

7 PROFILE AND COLUMN DATA ...........................................292

Introduction ............................................................................ 292
Carrizal Quebrada.................................................................. 292
U nit Profiles.......................................................................... 292
Quebrada Geologic Columns................................................299
Miraflores Quebrada........................ ..................................0 1
Unit Profiles....................... ............................................. 0 1
Quebrada Geologic Columns.......................................... 1 8
Pocoma Quebrada.................................................................... 323
U nit Profiles............................... ............................. 32 3
Canal Profiles ...................................... 327
x









Quebrada Geologic Columns................................................ 335
Ilo V alley ...................................................................................... 3 3 7
Tomb Site Stratigraphic Profile....................................... 3 7
Planting Surface #1 ..........................................................340
Canal Trench.............................................................................. 340
Geologic Column #1 ...........................................................................343
D discussion ............................................................................ 34 6

8 SUMMARY, CONCLUSIONS, AND
RECOMMENDATIONS ...................................................... 349

Calculating the Volume and Speed of the Miraflores
Flood at the Miraflores Quebrada ....................................349
Dating the Miraflores Flood..................................... ....35 1
Flood Impact on the Agricultural System .....................353
Declining Demographics ...........................................................354
Impact on the Chiribaya Culture......................................... 355
Cultural Responses to the Miraflores Flood ...................357
Emigration into the Ilo Area after
the Miraflores Flood.......................................................... 5 8
Cultural Change Resulting from Natural Disaster.........359
Recommendations for Future Investigations.................360

REFERENCES............................................................................ 362

BIOGRAPHICAL SKETCH........ .......... ..................................... 399








Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IMPACT OF A FOURTEENTH CENTURY EL NINO FLOOD
ON AN INDIGENOUS POPULATION NEAR ILO, PERU

By

DENNIS RAY SATTERLEE

December, 1993

Chairman: Professor Michael Moseley
Major Department: Anthropology



A strong El Nifio-Southern Oscillation (ENSO) is the only natural
phenomenon which can slow the earth's rotation and disrupt global

climate and rainfall patterns for a year or more, potentially altering a
culture's subsistence base. Some of these El Nifios are Pan-Andean

catastrophes that have stochastically occurred for approximately
5,000 years. If a strong El Nifio flood wrecks havoc upon the modern
population along the Peruvian littoral, the possible consequences of
an even stronger event could have been devastating to the

prehistoric indigenous populations of southern Peru.
This dissertation is based on my investigations of the largest,
late prehistoric El Nifio flood yet identified in the southern Andes of
Peru. The research was conducted in three coastal quebradas--
normally dry drainages--and in the Ilo Valley near the modern
fishing port of Ilo, Peru (170 S. Lat.) during the summers of 1990-
1992.









The focus here is to assess the impact of this inordinately large

El Nifio flood (ca. 1350 A.D.) on the irrigated agrarian systems and

settlements located within the study area and to analyze the stress

upon the resident Chiribaya Culture (ca. 1000-1350 A.D.). The

purpose was to test the hypothesis that the Miraflores Flood had
ultimately led to the demise of this culture.
Analysis of the frequencies of the recovered data support my
original hypothesis. Investigations also led to the identification of a

new phenomenon, which I have called a "SCDE" (Synergistically

Coupled Destructive Event), which combines the forces of seismic
events and El Nifio floods into a whole, that far exceeds the sum of its
parts in destructive power. In this case, a prehistoric SCDE must

have formed a massive wall of mud five to six meters high, which
roared down the mountains at 113 k.p.h., totally obliterating the

Chiribaya Culture at the Miraflores Quebrada as it swept every trace
of their village into the Pacific Ocean. In other areas, it so crippled

their infrastructure of agricultural terraces and irrigation canals that
the Chiribaya who managed to survive the SCDE disaster, sank into a
permanent cultural decline.


xiii














CHAPTER 1

INTRODUCTION


Purpose of Study
The purpose of this study is to assess the impact of a 14th
century flood event on the agricultural infrastructure of the

prehistoric indigenous population of the Chiribaya Culture and to test

the hypothesis that this flood ultimately led to the demise of the

Chiribaya Culture of the Osmore Drainage in extreme southern Peru
(Figure 1-1). This group occupied and farmed three coastal

quebradas (normally dry valleys) of Carrizal, Miraflores, and Pocoma,
and the Ilo Valley from about 1000-1350 A.D. In order to discover

evidence of this enormous flood event, which occurred about 1350
A.D. (PITT 0948), research was conducted from 1990-1992. The

study area in southern Peru included the 3 quebradas, located North
of the modern fishing port of Ilo and the Ilo Valley, located 5 km

North of this same city at 170 S. Lat. (Figure 1-2)
The Chiribaya were an autonomous group (Jessup 1991) which

used available river and spring flow to irrigate terraced fields in the
research area. This same group constructed a 9 km-long canal which
was used to irrigate the largest agricultural system ever built in the
lower Ilo Valley (Satterlee 1991; Figure 1-3). Both of these areas
demonstrate the dramatic impact of the same prehistoric flood event
on the Chiribaya Culture.





































C.


0 100 200 300
Kilometers


O /-
I----~----

Brazil
t-J
fI


% 1
_ -



1
rr ^
P I-


Chinck


C
c,


Y
s


- -- International Boundaries


80' W


Figure 1-1: Osmore Drainage in Extreme Southern Peru


12'S





















N


S
C" (#,k.

s


I0 1 2 3
I Kilometers
IL D..S. 1993
D.R.S. 1993


Figure 1-2: The Ilo Coastline


















































01 45Kilom
Kilometers


River



N

W ,-E

s





0 250 500 1000
Meters


D.R.S. 1993



Figure 1-3: Chiribaya Irrigated Agricultural System--Ilo Valley







5
Physical Setting
Extreme Southern Peru

Peru is located on the coast of South America, and it shares
common borders with Columbia and Ecuador to the north, with Brazil

and Bolivia to the east, and with Argentina and Chile to the south
(Figure 1-1). Although the Pacific littoral of Peru is the driest desert

in the New World (Lettau and Lettau 1978), it was home to some of
the most advanced civilizations in the Western Hemisphere, including

the Moche, Chimu, and the Inka.

In the dry altiplano of the southern Andes, the Tiwanaku
Empire held sway over most of the region for almost an entire
millennium. About 70 km from the study area, near Moquegua,
several outlying settlements with Tiwanaku affiliations have been
identified and studied (Goldstein 1989).

Peru is a country with ecological extremes. Much of it can be
divided into three distinct zones, i.e. the western dry coastal desert

with intermittent oases; the lush, humid tropical lowlands to the east;
and the intensely cold, high altiplano ringed by the Andean
Cordillera. These areas were occupied by scattered, mostly

autonomous, indigenous communities, many of which interacted
while practicing Ecological Complementarity (verticality), i.e. the
exchange of products between the different zones where only certain
crops could be grown because of a difference in the altitude. (Murra
1978; Stanish 1992).
Extreme southern Peru is generally known as the coasts of the
Departments of Moquequa and Tacna, which lie between the Tambo
and Izapa Valleys--16030'-1830' South Latitude and 69030'-72o East







6
Longitude. It is an area limited to the north by the highlands of

Arequipa and to the south by the Chilean border. It is bounded by
the Pacific Ocean on the west and by the high eastern Cordillera of
the Andes on the east. Archaeologically, this area has been one of
the least known in all of Peru (Vescelius 1960).

Much of this region was home to the Chiribaya Culture between

the years of 1000-1350 A.D. Through a bit of serendipity, some of
the Chiribaya remains were uncovered by the Tsunami of 1868, but
it was not until 1956 that the Chiribaya Culture was first identified

by Humberto Ghersi Berrera (1956). Ghersi excavated seven
unlooted tombs in the cemetery that is now called "Chiribaya Baja,"

and he located several Chiribaya sites between the valley and the
mouth of the Ilo River (1956:90). Typical grave goods found were
decorated ceramics, woven sacks containing grain, food stuffs, herbs,
and coca leaves, panpipes, and some textiles (1956:107-110).

The territory from 500 m to 2,500 m of the Chiribaya
homeland is characterized, by some, as being unhealthy because

maladies, such as Uta (a form of leishmaniasis), Verruga (weeping
skin ulcers), goiter, and malaria are often found at these elevations

(Belan 1981). Despite the presence of these diseases, cultural
occupation extended from the edge of the Pacific coast up to an
elevation of 2,500 m, but the majority of Chiribaya artifacts are
found from sea level up to 1,000 m (1981:23; Stanish 1992).
Situated in the southern Andes, the Osmore Drainage (Figure 1-
4), which includes the upvalley Moquegua River and the downvalley
Ilo River, is a long, narrow valley (125 x 25 km) that covers slightly




























S


8080


0 5 10 20
Kilometers


(Note: Universal Transverse Mercator coordinates in meters)
Figure 1-4: Average Annual Precipitation--Osmore Drainage







8
less that 3,500 km2 with a population of about 70,000 (Rice and

Watanabe 1989). The geology of the drainage, like other Peruvian
coastal valleys, is composed of sedimentary, metamorphic, and

igneous rocks dating from the Cenozoic eras (McCreary and Koretsky

1966). The drainage can be conveniently divided into three sections
according to the highly variable precipitation influenced by a change

in altitude. While most of the central and northern coasts of Peru
receive an annual rainfall varying from 15-40 mm (Moore 1991:29),

the coast at Ilo is somewhat drier and receives annual rainfall of only

5 mm (McCreary and Koretsky 1966). In the Moquegua region (1400

m), the average rainfall increases to 100 mm. It is only at 3,900 m
and above that the precipitation is 250 mm or more (ONERN 1976;
S.P.C.C. 1985; Figure 1-4).
The Osmore River begins its flow as the Rio Asana at 5,100 m.

At Moquegua other smaller tributaries join to form the larger Rio

Moquegua. The lower 20 km of the river system is the spring-fed
Rio Ilo that is separated from the upper Moquegua Drainage by 35
km of dry river channel because the upper river flow disappears
underground at a higher elevation of about 1200 m. Only during the
years of particularly heavy highland precipitation does the lower Ilo

channel experience river flow, which helps to recharge the local

aquifer and to source seeps and springs. The last 20 km of the
drainage system has sparse natural vegetation and some feral cotton,
but the coastline is mostly devoid of any natural vegetation.

Scientific research began in Peru in the early 1900s (Uhle
1910), but until the early 1980s, when the Programa Contisuyu
began its interdisciplinary, multi-institutional research program in







9
the environs surrounding Ilo, little was known about far-southern
Peru. Perhaps this area was neglected because it lacks the

sometimes spectacular monumental architecture that is so

prominently visible along much of the central and northern coasts of

Peru. Although lacking in free-standing prehistoric architecture, this
region is rife with archaeological sites.


The Coastal Quebradas
Introduction
In any desert environment, a reliable water source is the
primary element needed to support agriculture. However, in the

study area, the Carrizal springs supply the only regular freshwater
source in the desert until one reaches the Tambo Valley, 8 km to the
north (Bawden 1990). Today the quebrada contains a spring source
that is 20+ meters lower in elevation than it was during the early

Spanish Colonial Period (Figure 1-5; Clement and Moseley 1991). At

the Miraflores Quebrada, a limited amount of irrigation water is
pumped from a small reservoir. At the Pocoma Quebrada, the
intermittent trickle of spring water is collected in a shallow concrete
tank. Since the water supply at any of the coastal quebradas is
meager, in order to provide sufficient irrigation water, it now must
be stored in a reservoir in order to create enough volume and
pressure to reach the scant remaining olive trees located downslope.
































































D.R.S. 1993


Figure 1-5: Water Table Drop at Carrizal Springs








Catchment Area

The "catchment" of a valley or quebrada is the total area of the
watershed which collects rainfall and funnels it into the lower

reaches of the valley or quebrada. The area of the catchment needs
to be at least 15 times greater than the area which the water

irrigates. Even then, this areal proportion is barely large enough to

supply ample runoff for agrarian surfaces (Doolittle 1990:34). Unlike
some of the northern valleys of Peru, which have permanent flowing
rivers that were traditionally diverted from one river valley to

another to increase the amount of water for agricultural purposes
(Ortloff et al. 1982), far-southern Peru lacks such resources.
Fortunately in some areas of the Ilo Valley and in the coastal
quebradas there are springs, which can be used to augment
whatever ephemeral flow came from the higher regions around
Moquegua or above.
After some investigating, it was soon discovered that there
were insufficient aerial photographs locally available to the author
which could provide the coverage needed to calculate the total
catchment area for the individual quebradas. It was hypothesized
that these figures might show some correlation between the depth of
the flood deposits found at a quebrada and the total catchment for
the quebrada. However, this idea could not be tested with the
accessible aerial photographs. Beginning with the Pocoma Quebrada,
Figure 1-2 shows that the approximate catchment for each quebrada
is about one-half of its neighbor to the south. Although the Carrizal
Quebrada has the largest catchment, the spring flow is none too great







12
despite the fact that the aquifer should be adequately re-charged

during each rainy season.


Carrizal Quebrada
Total Agricultural Area

The measurements of the agricultural and settlement areas
were made using a planimeter, which is an instrument that can be
used to calculate the area of a polygon of any shape. The numbers
from the vernier scale of the planimeter are squared and then

multiplied by a factor which is constant for each particular photo
scale, yielding the number of hectares within a measured area. The
accuracy of such measurements is dependent upon how closely one
follows the outline of the polygon. Any slight deviation from a path
will produce a different reading each time. Thus, several readings

are usually taken and then averaged for a more accurate result.
Since there are computer programs which will calculate both the

perimeter and the area of a polygon in microseconds, using a
planimeter is a somewhat antiquated method of measuring areas,
but, nevertheless, the results are usually quite acceptable.
Based on the analysis of aerial photography, the total land used
prehistorically for agriculture at the Carrizal Quebrada was
approximately 25.86 ha, while the modern agricultural land used for
olive cultivation accounted for 2.48 ha. Therefore, the percentage of
land used by modern agriculture is 9.6% of the prehistoric land
usage, which compares favorably with the results of a field study
conducted by Clement and Moseley (1990a and 1990b, 1991), who
concluded that the late modern agriculture accounted for only 6% of







13
the total land under cultivation compared prehistorical farming

activities at the Carrizal Quebrada.


Settlement Size

It was difficult to make an exact determination of the location
of the domestic areas at Carrizal because the quebrada has been
farmed by the Spanish Colonialists as well as by modern farmers.
Deciding on the extent of domestic areas was further complicated

because at least one area may have been used before and after the
Miraflores Flood. Lacking the data from house foundations or from

mortuary estimates, it is almost impossible to estimate accurately the

prehistoric native population which lived at Carrizal. Population
figures from the ethnohistorical records such as the visitsas" (e.g. San
Miguel 1567; Toledo 1570-75; Zufiiga 1562) are not too useful
because it is a known fact that the native population declined by as
much as 70-80% after the smallpox pandemic beginning in 1521

(Moseley 1992). Therefore, 1.65 ha, based on the measurements of
two separate suspected domestic areas, is probably a reasonable
estimate of the settlement size at the Carrizal Quebrada.


Miraflores Quebrada
Agricultural Area

At the Miraflores Quebrada, prehistoric agriculture
encompassed an estimated 18.49 ha, while the late modern
agriculture now covers only 8.83 ha. These figures mean that the
decrease in total agricultural usage at this quebrada is 47.8%. This
decline would seem to indicate that the total food production for the







14
Ilo region may have suffered substantially in the last few centuries.

However, since most of the agricultural land in these three
quebradas has been devoted entirely to olive production for over
400 years, much of the comestibles would have had to come from the

highlands, as they do today.


Settlement Size

Although the domestic terraces of the village at the Miraflores

locale are totally blanketed by flood deposits, the terraces are still

easily delineated using stereo viewing of an aerial photograph.
Furthermore, since the land was covered so completely by sediments,
this settlement is one of the few places that neither colonial nor
modern agriculture disturbed the domestic area. The settlement
here measures approximately 140 m by 140 m or 1.96 ha, which is

only slightly larger than the domestic areas at the Carrizal Quebrada.




Pocoma Quebrada
Agricultural Area
Pocoma Quebrada has the largest active olive grove for 20 km
along the coast immediately north of Ilo. There are still 15 ha of
olive trees growing here. This agricultural area still does not
compare favorably to the 29.7 ha which was used for prehistoric
agriculture. Thus, similar to the Miraflores Quebrada, the total area
under cultivation at Pocoma Quebrada has decreased by slightly over
50% since the Chiribaya people farmed this area. The most alarming
fact discovered while doing the analysis of the prehistoric and







15
modern cultivated areas at each quebrada was the vast difference in

the total hectares under cultivation today compared to when the
Chiribaya Culture occupied these quebradas. The most glaring
example is the very small amount of modern agriculture at the

Carrizal Quebrada, but this same pattern of less land being cultivated
through time also holds true for the other two quebradas
investigated.




Settlement Size

If all of what appears to be domestic terraces were once
occupied, then the Pocoma Quebrada would have had the largest

Chiribaya settlement, covering a total of 2.61 ha. Although there is
evidence of the Miraflores Flood even at this highly elevated area
(compared to the other two quebradas), the impact of the Miraflores
Flood may not have affected the Chiribaya living here as directly or
severely as it did elsewhere.




Chapter Summaries


Chapter 2 begins by discussing the causes of the global-
impacting weather phenomenon known as the El Nifio-Southern
Oscillation (ENSO) and the associated changes in climatic regimes,
such as torrential rains and exceptional drought. Further discussion
includes the possible correlations between volcanic eruptions and the
onset of an ENSO and between global warming and the frequency of







16
ENSOs. Also considered are the often severe effects of El Nifio
flooding on the Peruvian people, coastal and highland agriculture,
domestic herding activities, and marine life. In conjunction with

these adverse effects, the economic impact on Peru's predominant
capital-generating enterprise, commercial fishing, and the related
industry of fishmeal production is discussed. In an effort to offset
some of the intense negativism, the positive consequences of a strong

El Nifio are also mentioned. The final topic discussed is the
application of modern data in a prehistoric milieu.
Chapter 3 begins with a brief synopsis of each of the Periods
and Horizons that are associated with significant cultural changes and
achievements in Peruvian Prehistory. A good portion of this chapter

is devoted to the important and often debated topic of the

intensification and development of irrigated agriculture. Social
changes associated with agriculture are also discussed. Some of the
motivating factors which are credited with giving impetus to the
development of agriculture are examined. The advantages of
agricultural terracing, the use of fertilizers, and risk management, all
of which are pertinent topics related to farming in the often hostile

Peruvian environment, are also discussed. Discussion likewise
includes how the agriculture-based prehistoric diet of Peru compares
to the modern Peruvian diet. The difference between traditional and
modern methods of water management are explored. A number of
topics closely aligned with the relationship between religion and
agriculture, such as the role of the gods in agriculture and oracles
and religious centers, are investigated at length. A delineation of
important environmental factors, including constant stresses and







17
periodic stresses, which adversely affect agriculture, is given. The
final topics considered in this chapter are the previous flood studies

conducted in Peru, the general prehistoric flood record in Peru, and
the specific prehistoric flood record in the study areas of the
Osmore/Ilo Valley and the coastal quebradas.

Chapter 4 discusses the methods used during the three field
seasons in Peru. Field survey was used to search for potential sites

and to analyze the flood impact to the domestic and agricultural
areas. Excavation techniques, such as unit excavations, trenches, and

shovel testing are discussed. Methods used to prepare the geologic
columns, unit profiles, and canal cross sections for the transfer of

pertinent features and information to graph paper are discussed.

The methods for creating maps which contained the locations of
units, trenches, and geologic columns are outlined. Laboratory
techniques used to process materials from the field and their

analysis are also explained. Methods used to recover carbon for 14C
dating are discussed. The final topics considered are the techniques

used in creating computer maps, profiles, and illustrations from field
drawings for inclusion in this dissertation.
Chapter 5 discusses a number of important aspects of my field

work influencing decision making and interpretations. For example,
the criteria for choosing the location of units and geologic columns is
outlined. A brief description of the location of units, columns, and
trenches and other pertinent data are given. What the excavations
indicate about the severity of the Miraflores and Chuza Floods is
discussed. Also discussed is what the archaeological record suggests







18
concerning the post-flood survival or demise of the Chiribaya people
who occupied the study area.

Chapter 6 begins by discussing the types and quantities of
cultural materials which could be expected to be found at each

investigated locality. The main focus of this chapter are the data
contained in the tables listing the artifactual materials recovered
from each level of a particular unit, feature, geologic column, or
midden. The total sherd distribution from each quebrada, the sherd
distribution from the individual floods, and the sherd weight

distribution per natural strata are analyzed in order to infer the
relative strengths of the two flood episodes and the relative flood
impact at each location. Several hypotheses are offered as possible
explanations for the existing discrepancies in the sherd distributions.
Finally, a discussion of the possible reasons for why the agricultural
terraces near Ilo, Peru, were abandoned is presented.
Chapter 7 presents an analysis of the unit profiles and geologic
columns. These data are analyzed to determine the composition of
the Chuza and Miraflores Floods. The flood record and the
stratigraphy are scrutinized to assess the consistency of the flood

deposits found in the individual quebradas. Finally a discussion is
presented concerning the depths of the flood deposition at each
quebrada and at the specific locations where units and geologic

columns are located in an effort to determine whether the deposits
were found at a uniform depth at each investigated location.
Chapter 8 presents the summaries, syntheses, and conclusions
of all the data gathered during the course of three field seasons in
far-southern Peru. These data are analyzed to estimate the overall







19
impact of the Miraflores flood on the Chiribaya Culture. Further

analysis includes the determination of whether or not the evidence
supports the original thesis that the impact of the Miraflores Flood
was of sufficient magnitude to destroy totally or partially the

irrigated agricultural system which was the Chiribaya subsistence

base. A scenario of a possible response by the Chiribaya people in
the months following the flood devastation is presented. Possible
recommendations are made which could help to improve or modify

future flood investigations so that maximum information can be

obtained, while at the same time using the minimum of amount of

field time and spending the least amount of money on field
assistants.














CHAPTER 2
EL NINO-SOUTHERN OSCILLATION

Introduction

Few, if any, natural phenomena have the global impact of an

El Nifio-Southern Oscillation (ENSO). A strong event can slow the

rotation of the earth, alter the length of day, impact global climate,
and create worldwide disasters (Salstein and Rosen 1983). Strong

ENSOs can displace normal climatic regimes, especially in the

Tropics, for periods lasting from a few months to a few years
(Rasmusson 1984:5). The very strong 1982-83 El Nifio caused

record rainfall in California, severe spring flooding in the northern
United States, record droughts in Africa (the worst of the century)

and Australia, unusual wintertime conditions as far apart as the
U.S.A. and New Zealand (Rasmusson 1984), and devastating rains

and flooding along the western coast of South America (Caviedes
1984; Glantz, 1984; Tapley and Waylen 1989; Waylen and Caviedes

1986). This extraordinary event was the most prolonged and
catastrophic El Niiio ever recorded, surpassing the great El Nifios of
1925 and 1891 (Rasmusson 1984:11; Glantz 1984). Although a
strong El Nifio affects humanity worldwide, the devastation is often
most apparent along the Peruvian littoral (Figure 2-1). Here the
torrential rains cross one of the driest deserts in the world (Lettau
and Lettau 1978) and descend one of the steepest watersheds
found anywhere--falling from 6,000 plus meters to sea level in a

20























Brazil




I.
J




11
'-- I -- '



^- ~'-
V^" xxI


0 100 200 300
Kilometers


C
(V
4,


N

W -E
Y
S


- International Boundaries


Figure 2-1: The Peruvian Littoral


12"S


C.
/

C


80' W







22
distance of a few hundred kilometers. Thus, severe flooding, flash
floods, and immense mud slides are commonplace along the

Peruvian littoral during strong ENSO events (Glantz 1984; Arntz
1984).

The purpose of this chapter is to explore the following: the
causes of the stochastic weather phenomenon that has become

known as the El Nifio-Southern Oscillation; climatic changes--
exceptional rainfall and drought--associated with strong El Niiios;

possible correlations between volcanic eruptions and the El Nifio

phenomenon; possible correlations between global warming and
the frequency of El Nifios; the effects of the ENSO-induced torrential
rains and flooding; the subsequent impact on the Peruvian

population, agriculture (both coastal and highland), domesticated
herd animals, and marine life; economic impact; positive
consequences of an El Nifio; and, finally, the application of modern
data to prehistoric settings.
Since the 1991-92 El Nifio is continuing into this year (Sonja

Guillen, personal communication 1993), this study of an extremely
large Prehistoric event may help us better understand that some
large 20th century ENSOs, which are viewed as very severe, are
possibly but a harbinger of what may lie ahead for the inhabitants

of the Peruvian coastline since there seems to be an increase in
Strong to Very Strong events in the last two centuries based on
historical records (Table 2-1).








Background Information

Because of the Colonial Spanish's penchant for keeping
administration and litigation records, we have accounts of extreme

rainfall and flooding in Peru as early as 1541 (Quinn et al. 1986).
The selected El Nifios presented in Table 2-1 are only those strong

events that have a high confidence rating as determined by Quinn,
et al., through a literature search of Spanish Colonial documents,

early Spanish chroniclers and clerics, and other non-Hispanic,

historical sources concerning Peru.



Table 2-1: Strong Historical El Nifio Events


YEAR STRENGTH YEAR STRENGTH YEAR STRENGTH

1541 S 1728 VS 1899-00 S

1552 S 1791 VS 1911-12 S

1567-68 S 1803-04 S+ 1917 S
1578 VS 1828 VS 1925-26 VS

1607 S 1844-45 VS+ 1932 S

1624 S+ 1871 S+ 1940-41 S
1652 S+ 1877-78 VS 1957-58 S
1701 S+ 1884 S+ 1972-73 S
1720 S+ 1891 VS 1982-83 VS


(Note: S=Strong; VS=Very Strong)







24
Although there has been information available for centuries

concerning devastating Nifio floods along the Peruvian coast, and

good records exist for the 1891 event (Murphy 1926), it was not
until the 1970s that evidence for strong events and their effects
was widely presented to the public. Prior to this date, media

coverage of the El Nifio phenomenon, in any form, was virtually
non-existent because the 1957-58 event was little known outside

of Peruvian newspaper accounts, but the 1972-73 El Nifio was the

first to receive a great deal of worldwide attention because of the

widespread droughts in West and East Africa, Ethiopia, the Soviet
Union, Australia, and Central America that were associated with
this perturbation (Glantz 1984:15-16). Because of the interest in

this event, the recent media coverage given to the 1982-83 El Nifio

was so extensive that it might be believed by some that El Nifios
are 20th century weather phenomena. However, since 1979 (Nials
et al.) El Niiio flooding has been studied archaeologically, and recent

geoarchaeological evidence suggests that the Peruvian desert coast
has experienced massive flooding from cataclysmic El Nifio rains for

about 5,000 years (Sandweiss 1986; Rollins et al. 1986).


The Cause of El Nifios
"The term El Nifio (Little Christ Child) was coined long ago by

Peruvian fishermen who witnessed the annual warming of the
coastal waters just after Christmas" (Dillon 1985:6). An El Nifio is,
among other things, the appearance of uncommonly warm water
along the coasts of Ecuador and Peru, which causes disastrous
ecological and economic consequences. Although its effects have







25
been traced at least as far as the western equatorial Pacific, the
manifestations of El Nifios are especially dramatic along the

Peruvian littoral (Smith 1983). This phenomenon, which can
persist for 6 to 18 months (Thayer and Barber 1984), is much more

involved than the simple occurrence of unusually warm water
along the South American strand.
Before the actual onset of an El Nifio, there are stronger than
average easterlies in the western equatorial Pacific for at least 18

months. These winds tend to move water from the eastern Pacific
toward the West as indicated by (a) in Figure 2-2, and
consequently the sea level is usually higher in the West than in the

East (Cane 1983). In September or October the easterlies begin to
diminish along the equator west of the International Date Line, and
subsequently this "dome" of warm water in the central Pacific,

which is 2-3 meters higher than in the eastern Pacific, "sloshes
back east colliding with Peru and overriding the cold Peruvian
(Humboldt) Current" (also known as the South Equatorial Current)

(Feldman 1983:17; (b) in Figure 2-2). At the same time, the
thermocline in the West is depressed and becomes deeper than
average by as much as 50 meters or more and the equatorial

upwelling is, also, reduced (Smith 1983). Both local and remote
responses to the wind contribute to rises in both sea level and sea-
surface temperatures (SST) that are characteristic of El Nifio (Cane
and Zebiak 1985).
In the fall of the year preceding the 1983 Nifio, SST

anomalies varied from 3.5C to 80C above normal along the coast of




























) -.T' ". -u 3>, C 0



/ 2
I




I z
0
Lj 3 | *




IIO

3 a.




-, 0 u

13 o ="

I' '6:^
IUI
'U 3U
' 0 v


SV0
WI V

i1 r7 I 1 3
"! S
"-- v^







27
Peru. For example, at Callao the SST increased 20C per month for

the last third of 1982 (Cane 1983). These extremely warm waters
are carried south by the South Equatorial Current (Reverse

Humboldt Current) when its flow reverses directions during an
ENSO event as indicated by (c) in Figure 2-2.
After an actual Nifio event, the anomalies slowly return to a

normal condition, but SST anomalies lag behind their atmospheric

counterparts. "This is consistent with the ocean's 'thermal inertia'
and immense capacity for the storage of energy" (Ropelewski

1984:592). Nevertheless, after another slight warming beginning
in the following December and continuing into the next year, the

SST falls and often is cooler than normal, and the waters again

resume their normal westerly flow.


The Peruvian (Humboldt) Current
Interacting with the El Nifio Current, also known as the
Equatorial Countercurrent, is the Peruvian (Humboldt) Current, i.e.

South Equatorial Current, which is a layer of water driven by the

wind northward toward the equator as indicated by (c) and (b) in

Figure 2-2. Averaging 2.80C cooler than other waters of the same
latitude (Mason 1957), the Peruvian Current is an integral part of
the weather system along the Peruvian coast because it cools the
air and causes rain to fall off-shore, thus perpetuating the desert
conditions on land. It also creates the persistent garua (fog and
mist) along the coast during the winter in the Southern
Hemisphere. As early as 1555 A.D., Augustin de Zarate (1965)
noted the great coolness of the sea along the Peruvian coast.







28
However, it was the prolific author and great naturalist Alexander

von Humboldt, after whom the current is named, who brought vital

attention to the current in his early 19th century writings. He

noted that the coastal waters of Peru were cooler than the air, and,

therefore, the water must cool the air and not vice versa as was

suggested earlier (Humboldt 1818; Merriman 1955).

Interacting with the Humboldt Current is the Peruvian

Undercurrent, which flows just below the Humboldt Current along

the coast southeastward toward the South Pole [see (d) in Figure 2-

2]. This action supplies the cooler water and nutrients that upwell

along the coast of central Peru (Smith 1983), which, in normal

years (Table 2-2), is carried North and then to the West where it


Table 2-2: Classification of Oceanic-Atmospheric Conditions


El Nifio Years Normal Years Anti-El Nifio Years

1925 1957 1927 1940 1956 1930 1964

1926 1965 1928 1942 1959 1937 1966

1932 1972 1929 1944 1969 1947 1967

1933 1973 1931 1945 1971 1948 1968

1939 1977 1934 1946 1975 1950 1970

1941 1978 1935 1949 1976 1951 1974

1943 1983 1936 1952 1981 1954 1979

1953 1938 1955 1982 1963 1980


(After Waylen and Caviedes 1985)







29
mixes with the warmer equatorial water. During the anti-El Nifio
years, the waters are even cooler than normal. However, during El

Niio years, these processes reverse and anomalies transpire.


Southern Oscillation

Always accompanying a strong El Niflo is the Southern

Oscillation, which is "a coherent pattern of pressure, temperature,
and rainfall fluctuations discovered and named by Sir Gilbert
Walker more than a half-century ago" (Rassmusson and Wallace

1983:1195). The primary manifestation of the Southern Oscillation

is an alternating change in atmospheric pressure at sea level

between the southern Pacific subtropical high and the region of low

pressure stretching from Africa to northern Australia. Other
manifestations involve fluctuations in sea-surface temperatures in

Africa, Indonesia, and northern Australia (Rassmusson and Wallace

1983).
The connection between an El Nifio and the Southern

Oscillation was not identified until the 1960s by Bjerknes, who
discovered that a cycle exists where there is a positive feedback
between the ocean and the atmosphere. Stronger equatorial
easterlies increase upwelling in the West and, therefore, an east-

west temperature contrast. This difference in temperature, in turn,
increases the thermal driving of the atmosphere, thus creating
stronger easterlies. The negative phase of this cycle is an El Nifio
Event.
The vacillating barometric pressures of the Southern
Oscillation, the westerly winds, the warmer than normal SST, and







30
the ocean currents all interact helping to create the anomalies of an
El Nifio. The "driving force" behind these interactions is the ocean's

circulation, which plays the role of a flywheel in the climate
system, and this circulation is responsible for the extraordinary

persistence of the atmospheric anomalies from month to month or
even sometimes from season to season (Rassmusson and Wallace

1983).


Climatic Changes Associated with Strong El Nifios
Exceptional Rainfall
One of the more unusual anomalies associated with an El Nifio

is the inordinate amounts of rainfall that occur along the normally

hyperarid Peruvian Coast. "The result in the particular strong ENSO
event of 1982-83 was a 40- to 60-fold increase in precipitation in

the region" (Tapley and Waylen 1989:62). Some areas of northern
Peru and Ecuador received rains 1,000 per cent greater than the

15-year monthly average (Arntz 1984:36). Although the 1972-73

El Nifio was not as strong, nonetheless, increases of 15- to 30-fold
in precipitation are recorded for that year.
An increase in maximum daily river runoff is, normally, a
prime indicator for the amount of precipitation an area receives.
All the rivers of the northern Peruvian coast exhibit a dramatic
increase in runoff during strong El Nifio perturbations (Table 2-3).
In particular, the Piura River had a runoff 400 times its 30-year
mean in January of 1983. It should be noted that precipitation
patterns do vary somewhat from valley to valley and region to
region (Waylen and Caviedes 1984), and large river discharges can







31
Table 2-3: Maximum Daily Runoff of Selected Rivers in


Peru, December, 1982 to May,
in Cubic Meters per Second.


1983.


Thirty Year Means measured


River Dec. Jan. Feb. Mar. Apr. May
Chira
Max. Runoff 288.2 1197.2 1641.9 2282.0 2437.1 2375.4
30-Yr. Mean 36.6 86.6 220.4 309.3 323.5 134.8
Piura
Max. Runoff 320.0 1314.6 1418.0 2428.4 2064.0 2473.0
30-Yr. Mean 0.5 3.6 59.8 108.2 89.7 29.9
Chicama
Max. Runoff 66.2 112.3 81.8 900.0 600.0 400.2
30-Yr. Mean 8.9 33.4 66.6 101.7 78.2 29.7
Moche
Max. Runoff 90.0 120.0 24.0 240.0 280.3 28.8
30-Yr. Mean 3.9 9.8 17.0 34.2 29.9 10.2
Viru
Max. Runoff 14.4 80.0 9.0 70.3 120.0 10.0
30-Yr. Mean 1.3 4.3 10.1 14.8 10.0 4.0


(After Caviedes 1984)


be misleading unless accompanied by rainfall on the coast (Quinn et

al. 1986). Nonetheless, when such massive runoffs occur,
devastating concomitant floods usually follow.
This variation in precipitation can be seen in southern Peru

where the 1993 El Nifio precipitation is more than it was for 1991-

92 event. The normally dry, lower Osmore River generally flows
only briefly for one or two days in March because of highland
precipitation (Figure 2-3; Manuel Pacheco, personal communication
1990). However, the river flowed in December, 1992, and again in

January and March of 1993. The January flow was exceptionally
strong and severely damaged some of the coastal highway which


Northern




































Sp


2,C..J m


100 mm
s


0 5 10 20
SKlometers
Kilometers


D.RS. 1993


Figure 2-3: Average Precipitation--Osmore Drainage







33
runs through the town of Ilo in southern Peru (Sonja Guillen,
personal communication 1993).

As Table 2-3 clearly shows, the daily run-off from each of
these rivers swelled dramatically above the 30-year average
during the 1982-83 El Nifio. When rivers increase their flow as

much as 3,600+ per cent above average in a span of a month of so,

the impact on the surrounding terrain must be enormous. In light
of the evidence, the massive coastal floods produced by such a

torrent should have caused severe damage.


Drought Conditions

While much precipitation falls along the north coast of Peru

during an El Nifio, normally little rain falls along the southern coast
and little snow or rains occurs in the high altitudes of the southern
highlands. Currently, some of the most reliable data concerning

both prehistoric and modern climatic conditions come from deep,
glacial ice cores. Studies of the 1,500-year-record contained in the

ice cores from the Quelccaya Glacier in southern Peru (Figure 2-1)
demonstrate that a major dry season persisted from 920 to 1050
A.D. (Thompson et al. 1984). It was about this time that the ridged
fields around Lake Titicaca were abandoned because of severe

drought conditions (Thompson et al. 1984). The data also indicate
approximately a 30 percent reduction in precipitation in the
highlands during 1972-73 and 1982-83 events (Thompson et al.
1985). These were years with strong and very strong El Nifios,
respectively (Table 2-1). Even though the 1991-92 El Nifio was not
nearly as severe as the 1982-83 event, the drought conditions







34
were even more pronounced in some areas, especially in the valley
surrounding Arequipa (Lorenza Carpe Diez, personal

communication 1993).
These events seem to indicate that the pattern of copious

rainfall along the coast during ENSOs, with a corresponding dearth
of precipitation in the highlands, held true in prehistoric times as

surely as it does in modern times. This pattern should not be too
surprising since the modern climatic regime has remained
fundamentally the same for approximately 4,500 years. According

to recent evidence, there has been stochastic major flooding along
the Peruvian littoral during this time (Rollins et al. 1986).


Correlations Between Volcanic Activity and El Nifio Phenomena

Strong volcanic eruptions have been proposed by some as

contributors to El Nifio events--not so much as the singular cause of
an El Nifio, but as an adjunct to the atmospheric and oceanographic

conditions that produce such an event. Handler (1984) has shown
that a positive correlation exists between large magnitude, low

latitude volcanic eruptions and the onset of El Nifios (a negative
correlation for high latitude volcanic eruptions exists). In 1600
A.D., Huayna Putina, in the southern highlands of Peru, erupted
continually from February 19 until March 6, eclipsing the sunshine
for seven days in Arequipa (Thompson et al. 1986). Therefore, if
Handler is correct, then very possibly the disruption of the
atmosphere by the Huayna Putina eruption could have contributed
to the strong El Nifio in 1607 A.D. (Table 2-1), whose flood deposits
are found throughout the study area.










There are other instances of El Nifio occurrences during
or following a year with major volcanic eruptions in
tropical areas, the most recent being El Chichon during
April, 1982, in Mexico. Notably the eruption of
Krakatoa in 1883 was followed during the summer of
1884 by torrential rainfall in northern Peru, a typical
manifestation of an El Nifio event. Evidence supports
that same association of El Nifio phenomena during
1721, 1728, and 1804 with reported volcanic eruptions
in the Pacific basin. In contrast, the occurrences of El
Niiio in 1911, 1925, and 1957 were independent of
volcanism, but concurrent with large meteorological
anomalies elsewhere (Caviedes 1984:290).
There seems to be a strong correlation between large

magnitude volcanic eruptions and the El Nifio phenomenon. The

recent 1991-92 El Nifio, rather than diminishing, is continuing into
1993, which is rather unusual because this is the first time in over
four decades that an ENSO has persisted longer than one season

(Quinn et al. 1979). It seems plausible that the 1991-92 Nifio is
continuing for another year because of the inordinate amount of

volcanic tephra that was spewed into the atmosphere by the
Pinatubo volcano in the Philippines in 1992.
Volcanic eruptions often do affect the climate, but the results
seem to vary worldwide depending on the latitude of the volcano

and the size of the eruption. The 1982-83 El Niiio, the strongest
event of the 20th century, was preceded by the eruption of Mt. St.
Helens in 1980 and the eruption of El Chichon in 1982. Both of
these events impacted the climate and may have affected the
intensity of the 1982-83 event.
Although ENSOs can occur independently of volcanic
eruptions, there is a correlation between large magnitude eruptions







36
and strong ENSOs. Therefore, it appears as if additional research
concerning the link between climatic change and volcanic activities
would be very useful. Fortunately, in the last decade, researchers
have been investigating information, such as precipitation amounts,
airborne dust, and volcanic ash, which are trapped within the
glaciers of Iceland (Hammer et al. 1980), Peru, (Thompson at al.
1984, 1985, 1986, and 1988), and more recently in China
((Thompson et al. 1989; Ellen Mosley-Thompson, personal
communication 1993).


Correlation between Global Warming and the Frequency of El Niiios
There may also be a connection between global warming and
its effects on the frequency and, perhaps, the magnitude of El
Nifios. Although early Spanish Colonial records may be lacking
somewhat in detail, there seems to be a pattern of an increase in
stronger events since the Industrial Revolution (Table 2-1).
However, tree ring data from Chile and Argentina present contrary
evidence that show no trend toward warming since the start of the
Industrial Revolution (The Gainesville Sun [TGS], 31 May 1993).
The mean temperature in this region has risen and fallen many
times in the past few millennia. Nevertheless, ice core sampling
from Greenland's ice sheet suggests that air pollution could lead to
dramatic shifts in the climate over the next 100 years (The Tampa
Tribune [TTT], 18 July 1993), which might affect the frequency of
major ENSO anomalies. Innovative studies of fossil temperature
changes retained in the earth, recovered by deep boreholes, show a
general warming trend in a number of areas of the United States







37
and Alaska. Models predict that the warming trend should be most
vigorous at the high latitudes. Nearly everywhere the warming has
20th century onset. Some areas evince temperatures that exceed

global average, while others have cooled. Significant gaps in the

borehole data exist for the Amazon Basin and other regions (Pollack
and Chapman 1993:46-50). Currently, there are no data for the
coast of Peru. Thus, the controversy continues, and it is only
through such studies as this one that this connection may

eventually be dismissed or established with certainty.


Effects of Strong El Nifios
Introduction

When assessing the effects of a strong El Niflo on a prehistoric
culture with no written records, we must rely heavily on current

data and draw conclusions from the impact of modern flood events
on human populations and their infrastructures. Furthermore, we

must assume that similar results could have occurred easily in
prehistoric times as well. Only then can we test hypotheses against
the geoarchaeological and archaeological records.
Climatic anomalies seem to be the norm for the years when a

strong El Nifio occurs. "During the summer of 1983 in the U.S., the
climate was such that June was the 6th coldest on record, while

August of that year was the hottest on record" (Ropelewski
1984:591). Europe and the British isles experienced an extremely
dry and warm summer (Chen 1983), and July was the hottest ever
in England where weather records have been kept for almost 350
years (Radcliffe 1983).










Flooding

The most immediate and dramatic consequence of a major El
Nifio is the catastrophic flooding that cripples the Peruvian littoral.

These areas are normally inordinately dry since many of them
receive only 45 mm of rainfall in a quarter of a century preceding
an El Niiio (Nials et al. 1979). "As might be expected, the pattern of

annual flood size is dominated by elevation in normal and Anti-El

Nifio years (Table 2-2). However, in El Niflo years, the pattern
becomes more strongly influenced by latitude as the equatorial air

masses related to the InterTropical Convergence Zone and the
Equatorial Countercurrent move farther south" (Waylen and
Caviedes 1986:151).

Devastating El Nifio flooding is not exclusively a modern

phenomenon because geoarchaeologists have determined that

major flooding occurred in the Moche Valley as early as 500 B.C
(Rollins et al. 1986). Another great flood--2 to 4 times greater than

the 1925 flood--inundated, Chan Chan, the powerful capital of the
Chimu Empire, in the early 12th century A.D. (Quinn et al. 1986).

This flood disrupted the culture, and eventually the capital was
abandoned.

About 1000-1100 A.D., the infamous "Chimu Flood" probably
occurred (Nials et al. 1979). This monstrous flood may be the
source of the legend about the great deluge which led to the demise
of the king of Chot (Chotuna), called Fempellac (Donnan 1990).
After he was tricked into sleeping with a demon, disguised as a
beautiful woman, it rained continually for 30 days, which caused







39
much hunger in the area. The priests of the temple and other

leaders became so angry because of Fempellac's offense against the

gods that they bound his hands and feet and threw him into the
sea ("lo hecharon en el profundo de el mar") (Cabello Valboa
1955:327-329 [1586]).

Early historical records from the 16th through the 18th
centuries are often lacking in details concerning flooding since such

an event was commonly viewed as a punishment for transgressions
against a wrathful god (Quinn et al. 1986). Fortunately, 19th
century observations were more scientific, and, thus, accurate
records of the rainfall and the consequences of devastating flooding

were kept.
The strong El Nifio of 1891 was documented by a number of

individuals in Peru, one of whom proposed a tentative theory

concerning the inordinate amounts of rainfall. "The most
reasonable explanations of these rains is that they have something

to do with the Corriente del Ninio or Reverse Humboldt Current"

(Murphy 1926:35). "The last rains were in February, 1891, and
they were certainly torrential It seemed to come down in
sheets, like a cloudburst, but was by no means local" (Murphy

1926:36). The flat plain at Talara in northern Peru was covered by
three feet of water, and the resultant mess was like quicksand. As

one author exclaimed, "In 1891, the situation was abnormal in the
extreme" (Merriman 1955:70).
The flooding from the 1925 Niiio was even more catastrophic.
A third of the town of Huanchaco was totally obliterated as adobe
homes were dissolved and washed away (Nials et al. 1979).







40
Hundreds of people died and thousands of homes were destroyed.

Since "gutters and tin roofs were a luxury rather than a necessity,"

(Caviedes 1975:501) the incessant rains soaked the adobe walls,
ceilings, and foundations to such an extent that they disintegrated.
Severe flooding has also destroyed much of Peru's patrimony.
The impressive Huaca del Sol, which was already severely damaged

in the early 1600s by the Spanish hydraulic mining efforts to

retrieve precious metals from the tombs, experienced further
damage during the 1925 El Niiio. Yet, the damage was not nearly

so much as that done by an ancient flood, ca. 500 A.D., which left a
high-water mark a full 8 meters above the 1925 flood level (Nials
et al. 1979). Evidence that both the Huacas del Sol and de la Luna
were damaged, by this ancient flood, lies in the fact that the adobe
bricks were saturated and "glued" together by the torrential rains
and the flooding.

Located in the La Leche Valley, the Baton Grande-Poma
archaeological Complex, with some three millennia of history, was
not badly damaged by the 1925 event. Nevertheless, ca. 1100 A.D.

a mammoth flood overwhelmed the citizens of the area forcing
their relocation (Craig and Shimada 1986). Fortunately, these
MegaNiiios only happen at a rate of about two per millennium, but
they are significantly larger and much more damaging than the
1982-83 event (Sandweiss 1986). Apparently the 14th century El
Nifio flood studied near Ilo is one of these rare events since it left
widespread, deep deposits that would easily qualify it as a
MegaNifio. It is also possible that this flood could have been







41
associated with previous, strong tectonic activity which would have

sufficiently loosened materials for easy transport by flood waters.
There were a number of minor El Nifios from the 1930s

through the 1960s, but almost 50 years elapsed before the next
major event. In 1972-73 the country of Peru was again pummeled

by a strong El Nifio. This time more people died and more property
was damaged or destroyed, but this 1972-73 event merely

foreshadowed the widespread destruction that would arrive in
1982-83. For months, the coastlines of Ecuador, Peru, and even
Chile suffered relentless rainfall and flooding.


Disease and Pestilence
Flooding is an obvious and immediate consequence of a

strong El Nifio, but it is the disease and pestilence in the ensuing
weeks and months following such an event that is often neglected

by the news media in favor of the more sensational flooding

sequences. It was reported, for example, that the floods
immediately destroyed 12,500 homes and damaged 28,000 more in
the Tumbes and Piura River valleys alone, during the 1982-83
event (Jackson 1984), but such massive flooding frequently causes
many belated problems for humans that are often ignored.
Thus, although 800 people died in Peru during the 1982-83
El Nifio (Thayer and Barber 1984), the survivors did not escape
unscathed. In the north along the Rio Tumbes, "constant rains and
humid atmospheric conditions compounded the picture of misery in
which mosquito swarms and outbreaks of malaria, typhoid, and







42
skin diseases haunted the inhabitants of the valley throughout the

summer, fall, and winter of 1983" (Caviedes 1984:277).


At the peak of El Niiio, rainfall, poor diet, and constant
high humidity favored the spread of tuberculosis: a
health report issued in June 1983 stated that 60
percent of the population of Chulucanas had contracted
the disease .. .in the village of Canchaque .. an
epidemic of Uta, a leishmaniasis of the skin, spread,
particularly among children, and caused terror among
the rural population...[usually transmitted by biting
flies, this disease produces single or multiple lesions
with proliferating weeping ulcers and sloughing of skin]
(Beck and Davies 1976).


Were this not enough, the valley was invaded by tropical bugs,

among which was the catigaza, whose sting produced swelling,
ulceration, and even paralysis of the affected limb, accompanied by
fever (Caviedes 1984).
Health conditions were atrocious in many areas of Peru

during 1983, but they were even worse in 1925. In addition to the
millions of dragonflies, caballitos del diablo, there were millions
more mosquitoes which carried the malaria virus. Besides the
rampant malaria, dengue fever was a common malady. This

infectious fever, which is usually epidemic, causes excruciating pain

in the joints and muscles of its victims--hence, its nickname of
"breakbone fever."
Because many railroad beds/bridges and roads were
destroyed by flooding, dozens of towns and cities were isolated
from food supplies. The poor diet contributed not only to







43
gastroenteritis, which claimed the lives of hundreds of children,

but, also, to beriberi which caused the death of scores more

because of the lack of vegetables (Murphy 1926). Presented with

the facts concerning the suffering following a modern event, one
could easily imagine the misery and decimation that must have run

amuck along the coast of southern Peru following the gargantuan

14th century El Nifio event.


Impact on Coastal Agriculture

Humans are inextricably bound to the products of agriculture,

and, when this valuable resource is disrupted by whatever means,

we usually suffer. The 1925 flooding washed away, not only crops-

-i.e. rice, sugarcane, and cotton--but the none too plentiful top soil

as well. To add to the misery of the winter winds and sand storms,

hordes of crickets--grillos--feasted upon the scant vegetables

(Murphy 1926). Burros and llamas replaced the railroad to carry

the desperately needed foodstuffs slowly from the sierra to the

coast.

In 1972-73, the rivers were so clogged with debris that they

overflowed and completely drowned the riparian orchards and

fields. In the Jequetepeque river basin, irrigation channels for the

rice fields were destroyed (Caviedes 1975). Since the volumes of
water associated with the 1982-83 event were so much more than
other historical El Nifios, the damage to agriculture and to related

support facilities was even greater. Of course, crop loss was not
restricted only to Peru. In two Ecuadorian provinces--Los Rios and







44
Machala--54,000 hectares of cropland were destroyed by flooding

(Caviedes 1984).
Since the coast of Peru is a hyperarid desert, agriculture is

totally dependent on the technology of irrigation. The rains of the

1982-83 El Nifio either severely damaged or completely destroyed
many irrigation canals. The Moche River washed away 500 meters
of two main irrigation canals, while it filled some other sections of

these open canals completely with sand. Also destroyed was
almost a full kilometer of highway which is needed to transport the
agricultural products to the markets and processing facilities
(Feldman 1983).
Again, many of the El Nifio related problems, such as damage
to agricultural infrastructure and diseases that plague modern Peru

following a major perturbation, should hold true for prehistoric
times as well. Also, the impact of the 14th century A.D. megaflood
must have been even greater owing to the lack of adequate
technology and a sufficient labor force with which to recover from
such grievous disasters.


Impact on Highland Agriculture
While the coast suffers from devastating rains and floods

during a strong El Nifio, the highlands suffer from catastrophic
drought. In the southern highlands around Lake Titicaca, in 1983,
70 percent of the major staple crop, potatoes, was destroyed. This
crop failure was quite serious because the indigenous population
mainly survives the long winter months by eating freeze-dried
potatoes--Chuhio--which can be easily stored for the entire winter







45
season. Unfortunately, because of the severe drought and the
extreme food shortage, even the seed potatoes for next year's

planting had to be eaten (Jackson 1984). Even though there was
some precipitation in the highlands around Lake Titicaca, "the

scattered rainfall of February and March 1983 was insufficient to

grow potatoes, quinoa, maize, and alfalfa during the peak of
summer" (Caviedes 1984:288).
Famine is normally alleviated by kin groups sharing food

with each other. This native "disaster relief" is usually sufficient to

cope with the frequent droughts and subsequent low crop yields,

but in 1983 the situation was so dire that this traditional "risk

management" system was inadequate. To compound an already
grievous situation, some highland people were incapable of helping

their kin since entire villages, along with all the potentially helpful
relatives, were completely buried by mud slides. A full one-third
of the remaining villages were also badly affected by this
particular ENSO event (Feldman 1983).


Domesticated Animals

Not only humans endure great suffering during strong ENSO

events, but domesticated animals, mainly camelids, such as the
Llama and Alpaca, also do not fare well during these times. During
both the 1925 (Murphy 1926) and 1982-83 events (Jackson 1984),
as pasturage became scarce because of the drought, the sheep and
camelids in the highlands began to starve. At the other extreme
along the north coast, especially near Trujillo, vast tracts of







46
valuable pasture were totally obliterated by flash floods, further
impacting domesticated grazing animals.
The most seriously drought-affected areas were the

highlands of southern Peru and Bolivia. By the end of April, 1983,
the drought had spread from the Titicaca basin to the entire
altiplano and had wreaked hardship among traditional cattle and
llama herders. As was the case during previous El Nifio-related

droughts (1891, 1941-42, 1957, and 1972), the level of Lake
Titicaca dropped to record lows. The herds of cattle and llama
become so thin that social problems resulted for many poor
peasant families because they were accused of not properly caring
for their livestock. Unable to maintain the weakened animals,
some peasants sold them at extremely low prices, while other
families pushed to the limits of need and despair, even sold their
own children to wealthier families (Caviedes 1984:288-289).


Marine Life
In normal years, the main upwelling centers of the Pacific
Ocean, extending as far South as the Paracas Peninsula, coincide
with the greatest carbon-producing areas and with the richest
anchovy grounds along the Peruvian coast. An El Nifio triggers a
devastating crisis in the ecosystem. The upwelling of the cool
water weakens, and the less salty and oxygen-poor warm tropical
waters overlap the cold waters of the Peru Current, resulting in a
sizable decrease in phytoplankton and zooplankton, which feed the
anchoveta (Caviedes 1975).







47
A strong El Nifio event severely disrupts the ocean's
production of both chlorophyll and phytoplankton by as much as

75 percent. The larval anchovy feed on phytoplankton, but, as fate
would have it, the zooplankton do also. So the failing food supply is
now even more diminished for the young fish and, as a result of
the competing zooplankton, millions of the larval anchovy do not
survive (Walsh et al. 1980). As if this destruction were not enough
misfortune, even those anchovy which do survive decrease in

absolute growth and reproductive success (Barber and Chavez
1983).
Even the weak El Nifio of 1975 caused an 80 percent
reduction in the productivity of nutrients, and this production

occurred in a narrower zone than in normal years (Cowles et al.
1977). "When the warm surface water invades the Peruvian

littoral, the sensitive anchovy dives down some twenty meters to
colder waters or migrates southward ." (Caviedes 1975:498).
"The major El Niiio, in 1972, combined with overfishing,

almost wiped out the vast school of anchoveta (Engraulis Ringens)
that had propelled Peru into first place among fishing nations (with

a catch greater than the combined total of both North and Central
Americas)" (Feldman 1983:17). Before the 1972 event, Peru
accounted for one-fifth of the world's total fish production
(Caviedes 1975). This loss of the anchoveta catch, of course,
affected the fishmeal production and, consequently, the price of
beef since cattle eat the fish meal as a protein source. Without the
fish meal, cattle owners had to turn to more costly soybeans as a
substitute.







48
The anchovies died off or were so severely depleted after
1977 that pilchard (a member of the herring family) replaced

anchovy as the mainstay of the fishing industry. These small fish

can survive in warmer waters between 23C and 25C, while
anchovy cannot. Unfortunately, the horse mackerel, which preys

on both pilchard and anchovy, also prefers these same warm
waters. Therefore, the number of the fish available to commercial

fishermen steadily declined (Caviedes 1984). For example, the
anchovy catch declined from 1.7 million tons in 1982 to 100,000
tons in 1983 (Arntz and Tarazana 1990). Of all the marine life it is
the middle of the normal food web that is disrupted the most by

the anomalies of an El Niio (Arntz and Tarazana 1990:333).

Yet another threat to the fishing industry in Peru, caused by
El Nifios, is called a red tide, which has caused the death of

countless fish along the Peruvian coast. "Red tides are spectacular
dinoflagellate blooms that occur in oceans and often lead to mass
mortality of marine fishes and invertebrates" (Krebs 1978:540).
Because of the warm weather anomaly, when the water

temperature, salinity and/or nutrients are in certain proportions,
the marine protozoans dinoflagellatess) rapidly increase. The
Biological Oxygen Demand (BOD) required by aquatic bacteria to
decompose the accumulated metabolic waste from the billions of
protozoans is such that there is not enough oxygen for the fish to
survive (Owen 1980). In March of 1990, 24 tons of fish washed
ashore at Chimbote (Francisco Mamani, personal communication
1990). Although it is uncertain at this time, it appears that a red
tide may have caused their demise.







49
Murphy (1926) relates that, in 1891, during the day the sea
was covered with blood-like patches many acres wide. The final

effect of this freak of nature was that so many fish and other

marine life died that the subsequent hydrogen sulfide produced
actually blackened the white paint on the ships in the Callao Harbor

(Merriman 1955). In 1925 another red-tide appeared. "As

testimony to the loss of life, the gruesome phenomenon is a
harbinger of death appropriately known as El Pintor (the painter)
(Nials et al. 1979:7). The loss of human and animal life in total

numbers, as a consequence of a strong El Nifio event, seems small
when one assesses the total impact on marine life. Yet,
understandably, this loss may not be of great significance to
anyone--except, perhaps, the fishermen.


Guano Birds

Many of the guano producing sea birds eat pilchard, anchovy,
and other smaller fish when they are driven to the surface by

larger marine predators. During an ENSO event, these larger
predators move from the extraordinarily warm waters to cooler

waters elsewhere. When this happens, some birds migrate to other
areas, but most birds actually starve to death because the smaller
fish are not readily available at the surface. For example, there are

normally 300,000 birds on Christmas Island; by the Spring of 1983,
there were none inhabiting the island (Thayer and Barber 1984).
Those birds which stay closer to the coast fare somewhat better,
but thousands still die for want of sufficient food (Barber and
Chavez 1984).







50
The guano birds are quite important economically to the

people of Peru since these guanay--cormorants, gannets, and

pelicans--produce a valuable fertilizer (Caviedes 1984). Prior to
the 1925 event, the guano bird population was estimated at 30
million. The strong 1925, and 1972-73 El Nifios have reduced their
numbers to 7.5 million (Nials et al. 1979). Murphy (1926)
describes the scene following the 1925 El Nifio, by saying, "By the
end of January sick and dead guano birds began to be numerous in
northern Peru, and the peste (plague) spread rapidly southward
until countless thousands of carcasses lined the whole shoreline of

the country" (1926:32). Surely it will take years for nature to
rectify the decimation of the millions of birds along the Peruvian

littoral caused by the El Nifios of this century.


Economic Impact
It is almost unbelievable that something as innocuous as an

influx of warm water could so drastically disrupt the ecosystem
along the western coast of South America, but the evidence speaks
for itself. During the 1982-83 El Nifio, the worldwide combined
loss of human life from floods, polluted water supplies, and drought

has been estimated at 10,000 or more deaths. Property damage
has been estimated at $10 billion (Dillon 1985:6-7). In addition to
the death of 800 humans in Peru, the total destruction from the
1983 event alone is estimated at over 3 billion dollars (Thayer and
Barber 1984). Included in this staggering figure is the loss of
hundreds of millions of dollars for the Peruvian fishing industry,
which has for years been a major source of income for the country.







51
The fishing industry first felt a financial crunch in the 1920s when

the local fisheries along the coast failed as the common schooling
species, i.e. anchovy, etc., died or departed (Murphy 1926).
The 1972-73 El Nifio's disruption in the ecosystem started an

immutable chain reaction of events for the 128 fishmeal factories
then in existence. Burdened by debts, low productivity,
maintenance costs, and wages, sixty-five small fishing plants, most
locally owned, faced bankruptcy, and the owners were willing to

sell their plants to the Peruvian government for whatever was
offered. The remaining plants had financial backing because they
were owned by foreign countries or by 10 Peruvian magnates who
could absorb the losses and continue operations at a reduced level.

Finally, to preserve the fishing industry, on May 7, 1973, it was
nationalized in an effort to control the growth of the industry, the
welfare of the fishermen, and the marketing of the sea products
(Caviedes 1975). Following the 1982-83 event, the fishing

situation became so desperate that the entire fishing fleet
continually tried a variety of new techniques to catch whatever it
could (Barber and Chavez 1983).


Positive Consequences of an El Nifo Event
Lomas
Fortunately, there are some positive consequences of an El
Nifio. After encountering so much suffering, destruction, and death,
any small glimpse of beauty is a welcome respite. Ironically, while
the human disaster continued, a vast area of the desert in northern
Peru was in "full bloom with green shrubs everywhere" (Jackson







52
1984:33). In some places, wet fog drifts inland from the sea, and

when there is enough fog, a seasonal flora develops in what are
called Lomas--these areas are not a continuous band, but more like
an "island archipelago" of verdant patches of vegetation (Dillon

1985).
"Current tabulation of the entire lomas flora shows it to
contain ca. 1,000 species of angiosperms and ferns" (Dillon and

Rundel 1990). Included in these various families, genera, and
species are rare plants, which are being carefully studied in an

effort to understand their origins and the formation of these idyllic
"islands" (Pare 1984). It is believed that the composition of the

plants and flowers reflect past climatic and geologic events (Dillon
1985) and, thus, can be useful in modern attempts to reconstruct
and to understand these ancient events.
For years, people have been observing and recording the

beauty of the desert when it awakens from its long, barren sleep.
It has been suggested that Pizarro's march through the Piura area,

as he and his men trekked toward Cuzco in 1532, "was possible
only because he chanced to land upon the desert shores during one
of the rare "aios de abundancia or years of abundant water and
vegetation" (Murphy 1926:54). In 1891, during another one of the
"years of abundance," when the rains occurred, one person noted,

"the desert soil is soaked by the heavy downpour, and within a few
weeks the whole country is covered by abundant pasture" (Murphy
1926:35). Although this pasture is ephemeral, when it dries it
affords goats a natural hay for a year or so. Often seasonal lakes

and ponds are created by these monsoon-like rains and the plant







53
life proliferates, attracting flocks of ducks from as far away as the
Guayaquil region. In some places, for the first time in forty years,
flowering plants reached the coast (Murphy 1926)
During the rain of 1972, as well as of 1925, in many places

the buried grass seed germinated providing ample grass which
more than met the grazing needs of the ranchers in northern Peru.
In fact, this new-found abundance encouraged some local ranchers
to import many thousands of feeder cattle and breeding stock.
Goat herds were seen roaming around various areas taking
advantage of the unusual fresh vegetation during the fall and
winter months of 1972 (Caviedes 1975).
In 1983, there was a particularly rich bloom in the Lomas
along the coast of Peru and as far south as Chile. Typically the
unusual outburst by the dormant vegetation appears some two or
three months after the maximum rainfall. The Peruvian
government has even designated one area as a national park--
Lomas de Lachay--which is located about two hours drive north of
Lima (Caviedes 1984). Even in the far south of Peru, there is lush
pasture in the normally desert-like Sama river valley during such
events, and the local herders bring their cattle into the valley to
enjoy this rare serendipity (Francisco Mamani, personal
communication 1990).


Applying Modern El Niiio Data to Prehistoric Settings
The Peruvian landscape is one of the most dynamic in the
New World, if not in the entire world. Peruvian history is rife with
examples of major seismic events, including the most devastating







54
historical earthquake in the Western Hemisphere, i.e. May 31, 1970

(Silgado 1978; Webb and Fernandez Baca de Valdez 1991). Further
proof is provided by the impressive number of El Nifios events that
have re-sculpted the Peruvian countryside during the last 450
years (Quinn et al. 1986). Both earthquakes and strong El Nifios
have a great impact on humanity, and it seems reasonable that
"phenomena that alter subsistence systems and disrupt means of

agricultural production are likely candidates for triggering change
or ethnic movement during the course of Andean civilizations"
(Moseley 1987:7).

Using an ethnohistorical approach, one should be able to use
the available modern data concerning the impact of El Nifio flooding
on modern humans and to apply these data to prehistoric
populations. We can compare the modern accounts (e.g. Murphy
1925; Caviedes 1975, 1984, among others) to the early historical
accounts of floods in Peru to extrapolate and to determine the
validity of these written records concerning the consequences of
these events for humans. Vivid descriptions by Alcocer (1987
[1582]) of the 1578 A.D. catastrophic rains and flooding that
severely affected the north coast of Peru are but one example of
the early works which can be used to infer what the impact of the
earlier 14th century A.D. Miraflores Flood might have been for the
Chiribaya Culture of extreme southern Peru.
Since the occupants of the coastal quebradas studied in this
dissertation were dependent upon a highly developed,
sophisticated irrigated agricultural system, there is little doubt
that, following the destruction of this system by the 14th century El







55
Nifo, the indigenous population, which had occupied the area for

almost 400 years, would have been severely reduced by starvation

and disease. Without an adequate workforce to rebuild, repair, and

maintain the agricultural subsistence base, the culture could not

have regained its pre-flood eminence. Therefore, it would have

lost its hegemony, and it would have been in a very vulnerable
position with regards to invasion or simply an in-migration by a

non-resident culture. (For a detailed discussion of the impact of an

early El Nifio on the Chiribaya Culture refer to Chapter 8).
As previously noted, archaeologists must, by necessity, draw

upon modern flood data when interpreting the effects of El Nifios

on prehistoric populations. Flooding of cultivated land, along with
diminished sea resources, as a result of a particularly intense El

Nifio, results in drastic food shortages today. The prehistoric

consequences must have been even more disastrous (Jackson and
Stocker 1982:22). Currently, there is no contrary evidence that the
devastation and the diseases associated with contemporary events
should not apply equally to a prehistoric population. In fact,

because of the lack of modern medicine and technology, the
disruption of an autochthonous prehistoric society should be even

greater than that suffered by a modern society.















CHAPTER 3

ARCHAEOLOGICAL BACKGROUND

Introduction

As with all other countries, the history of Peru is a fascinating

story of the gradual development of unique cultures throughout
many centuries. This history is divided in various ways, but it is

almost universally agreed that the most significant milestones of
cultural development that fall within the scope of this study can be

demarcated by the titles of "Initial Period" (IP), "Early Intermediate
Period" (EIP), and "Late Intermediate Period" (LIP). Therefore, it is
important to examine each of these periods in turn to understand

the vital role that irrigated agriculture played at each step along the
Peruvian journey through time.


Initial Period

A number of cultural changes occurred during this period.
Early coastal sites were abandoned; monumental architecture

shifted inland; and there was an increased consumption of
cultivated plants (Moseley and Deeds 1982). At the beginning of
the Initial Period, around 1800 B.C., agriculture began to spread into
new habitats, which included coastal valleys. "This must imply
some mastery of irrigation techniques, since without these,
agriculture would scarcely have been possible in such arid







57
surroundings" (Fung 1988:82). Orlove (1985) suggests that maize
was a relatively new domesticate that was preceded by potatoes
(ca. 2000 B.C.), peanuts (ca. 2500 B.C.) and the "industrial cultigens"
cotton and gourd (ca. 3000 B.C.) during the Preceramic Period (:46).
Some, however, would argue that maize has been cultivated as
early as 4000 B.C. in Colombia (Bray et al. 1987).
At this same time, population increased and the mountain and

coastal settlements began to crystallize into "poles" of influence
(Moseley 1992). Spectacular developments, such as mound
building, free-standing architecture, and weaving during this period
are largely attributable to the technology of irrigation agriculture,
which provided people with more leisure time (Pozorski and
Pozorski 1987:125). Also, pottery, associated with storage, cooking,
and, perhaps even more importantly, with the brewing of chicha (a
native corn beer used as an integral part of religious ceremonies)
appears on the central coast of Peru around 1800 B.C.

Early Horizon
Building upon the incipient agricultural base created during
the Initial Period, early Peruvians increased their dependence upon
maize agriculture during the Early Horizon. In some of the coastal
valleys, irrigated agriculture was undoubtedly expanded. In the
highlands, the use of ridged fields and cochas (sunken fields)
increased agricultural yields and extended the farming season
(Moseley 1992). New cultigens, such as manioc (Manihot esculenta),
tree tomatoes (Solanum muricatum; Cyphomandra splendens),
Jiquima (Pachyrryzus tuberosus), and probably sweet potatoes
(Ipomoea batatas) were added to the ever-growing inventory of







58
domestic plants. However, the Peruvians still placed an emphasis
on marine resources and further supplemented their diet with
hunting and gathering. As a consequence of the diversification in
subsistence resources and an improved diet, population increased
considerably and nearly all of the coastal and highland valleys were
settled (Lumbreras 1969).
Religious centers, with new art and iconography dictated by
the Chavin cult, appeared in the highlands and along the Peruvian
coast from the Moche River to the Mala River valley. With the new
art came changes in textile production, such as textile painting,
dying of camelid hair, and the heddle loom. Pottery styles now
included stirrup-spouted bottles, stamped decorations, and the
neckless olla with motifs stressing super-natural beings, especially
the fanged anthropomorphic "staff god". Using such innovations,
the spread of corporate art and iconography probably helped
precipitate a coalescence of a multitude of distinctive local and
regional cultures into what is commonly referred to as the Chavin
Horizon, a significant, but short-live phenomenon (Burger 1988).
Metallurgy also experienced meaningful modifications with the
introduction of three dimensional forms produced by soldering pre-
shaped metal sheets, silver/gold alloying, and repouss6 decorations.
Accompanying these changes, were marked social stratification and
an increase in the demand for exotic goods exchanged between
ecologically complementary zones (Burger 1988).









Early Intermediate Period
A change in the settlement patterns occurs with the
residential centers now outnumbering the ceremonial centers

(Moseley 1992). There were other significant developments during
this period, which included an even greater increase in population

accompanied by construction of large irrigation systems on the
desert coast. Hints of a possible increase in warfare or a greater
need for defense are suggested by the many fortifications built at

this time. Nucleated settlements, with populations as large as
10,000, were present at Moche, Nazca, and Pukara, and an
estimated two million people inhabited the coastal areas (Lanning

1967).


Middle Horizon
This was a time of severe droughts which adversely affected

both agriculture and the great complex cultures which depended so
heavily upon it. A great drought lasted from 562-594 A.D.

(Thompson et al. 1988), that greatly curtailed agriculture since
evidence indicates that the precipitation was 30 percent less than
normal. This drought was probably a contributing factor in great
ethnic movements and conflict of the Middle Horizon (Moseley

1992).
There were not only severe droughts during the Middle
Horizon, but devastating floods as well. In the Moche Valley,
flooding stripped away many meters of precious soil and severely
damaged fields and irrigation canals, forcing the abandonment of
the Mochica capital, Huaca del Sol (Moseley 1992). A new capital







60
was built at Cerro Galindo, which was located at the valley neck

near the irrigation canal intake-perhaps to protect them from
outsiders (Bawden 1985).

New irrigation techniques were introduced, and the Huari
(Wari) were among the first to irrigate slopes using short, low

canals to farm limited areas (Moseley 1992). At Huari, high water
sources fed long, primary canals which, in turn, fed into secondary
canals that watered the extensive terraces. The Huari used this

new technique to develop the unfarmed slopes in the Moquegua

Valley, 60 km upvalley from the study area (Goldstein 1989; Figure
3-1; Table 3-1).
Table 3-1: Middle Horizon Cultures


Time N. Coast Cen. Coast S. Coast Moquegua Titicaca

1000 A.D. Sican Chancay Ica Tumilaca Tiwanaku

900 A.D. Sican Ica Tumilaca Tiwanaku

800 A.D. Sican Ica ChenChen Tiwanaku

700 A.D. Sican Huari Ica ChenChen Tiwanaku

600 A.D. Moche Huari _Omo iTiwanaku

(After Moseley 1992)



Late Intermediate Period
The Late Intermediate Period (L.I.P) began in about 1000

A.D., approximately concurrent with the decline of the Tiwanaku
Empire, and lasted until the rise of the Inca Empire, ca. 1476 A.D., at
the beginning of the Late Horizon (Table 3-2; after Rowe 1962;







































C?
/

0


0 100 200 300
Kilometers


C
c


61





S \-

















-Lnha '- .Cuz








Naer
Aca rl
fi\3~ / AC /,




Yauca,
Ocan

Nl Moquegua
S? Drainage

N

w E

S


- International Boundaries
76'W 72'


razil


Figure 3-1: Moquegua in Southern Peru


- -I
1
1







62
Lumbreras 1974). The capitals at both Huari and Tiwanaku had
already been abandoned toward the end of the Middle Horizon

(Rowe 1963). Therefore, this period was a time of change because
"the decline of the Wari (Huari) Empire disrupted the unity that had

been imposed on the Central Andes and permitted the resurgence
of local or regional political organizations" (Lumbreras 1974:179).

The Late Intermediate Period was similar to the Early
Intermediate Period because it was also a time with increasing

population, small political units sometimes associated with larger
confederations, and warfare (Lumbreras 1969). Some coastal areas


Table 3-2: Periods and Horizons of Peruvian Prehistory


Periods/Horizons Time Scale Also Known as

Late Horizon 1476-1532 A.D. Imperialist

Late Inter. Period 1000-1476 A.D. Urbanist

Middle Horizon 600-1000 A.D. Expansionist

Early Inter. Period 200 B.C.-600 A.D. Experimental

Early Horizon 900-200 B.C. Cultist

Initial Period 1800-900 B.C. Formative

Preceramic 2500-1800 B.C. Archaic

Lithic 8000-2500 B.C. Hunter/Gatherer


were depopulated and, within a few centuries, small regional states
appeared in various parts of the former Huari Empire (Patterson







63
1973:101). It was a time of marked local cultural diversity and
relative isolation, contrasting with the Middle Horizon which had
been dominated by the two rather large, interactive empires of

Huari and Tiwanaku. According to Inca tradition, there was "a
situation of extreme political fractionation in the sierra at the

beginning of the Inca Dynasty" (Rowe 1963:16).

With the exception of perhaps Chan Chan and Pachacamac,
most of the cities were abandoned by the end of the Middle
Horizon, and the prevailing pattern of settlement became one of

small, urban centers with dispersed dwellings. Perhaps this new
pattern was a reaction against large cities which exerted too much

control over all the inhabitants--much like the rural population

revolting against the late Roman cities. This backlash occurred in
areas where city traditions were old, but not on the north coast of
Peru (Rowe 1963:19-20). This Middle Horizon settlement pattern
would remain unchanged until the Incas placed people in nucleated
settlements where they could better watch and control the people's

activities (Patterson 1973:69).
The only regional state in existence during the L.I.P. was
Chimor, the great Chimu Empire that extended for a 1,000 km along
the northern and central coasts of Peru (Conrad 1981; Moore, 1991;
Figure 3-2). In other areas of Peru, there were polities that

developed a social organization based on small towns and villages;
however, they did not achieve large scale political integration, like
that of Chimor. These polities included the following: the Chancay
on the Central Coast; the Lake Kingdoms, Colla and Lupaqa, around
Lake Titicaca; Pisco, Ica, and Nazca Valleys of the south coast ruled









64


















ter
ama
-I















PaLI
Sg
Ca' e te


8 WChnc .


a Ica Ic
Naz, ~
Acar
O kYauca
0 300 At
ters 0
t s0 Chiribaya -



N

W


S


--- International B
80" W 76' W


I


ra
j


%1 ,
I J
''- -\ \


I
Z I
.UZCO


oundaries


Figure 3-2: Cultures of The Late Intermediate Period


zil


-.

S0I


0 100 20
Kilome


/ I







65
by the "Lord of Chincha"; and the Chiribaya on the extreme

southern coast in the department of Moquegua (Table 3-3).


Table 3-3: Late Intermediate Period Cultures


Time N. Coast Cen. Coast S. Coast Moquegua Titicaca

1476 A.D. Inca Inca Inca Inca Aymara

1400 A.D. Chimu Chancay Ica Estuquifia Kingdoms

1300 A.D. Chimu Chancay Ica Chiribaya Aymara

1200 A.D. Chimu Chancay Ica Chiribaya Kingdoms

1100 A.D. Chimu Chancay Ica Chiribaya Aymara

1000 A.D. Chimu Chancay Ica Chiribaya Kingdoms


(After Moseley 1992)




Cultural History of the Ilo Region


Lithic Period
The Ilo coast has a lengthy human history which extends back

into the Lithic (Hunter/Gatherer) Period. Located 7.5 km South of
the city of Ilo and dating to ca. 8000 B.C., the Ring Site is the oldest
littoral site in Peru (Sandweiss et al. 1989; Table 3-4). This
maritime site consists of an almost perfect 26 m-wide circle of
discarded mollusks shells (Richardson et al. 1990). In addition to
over one hundred stone tools, 7 different types of bone implements,
including a bone harpoon were recovered at the Ring Site. Faunal









66
Table 3-4: Ilo Time Chart

Time Lower Io Ilo Coast Moquegua Highlands Arica, Natural
Frame Valley Chile Events
In it i a 1 Period
900 B.C. Wawakiki/
Carrizal
1000 B.C. Wawakiki/
Carrizal
1100 B.C. Wawakiki/
Carrizal
1200 B.C. Wawakiki/
Carrizal
1300 B.C. Wawakiki/
Carrizal
1400 B.C. Wawakiki/
Carrizal
1500 B.C. Wawakiki/
Carrizal
1600 B.C. Carrizal

1700 B.C. Carrizal/
K-4
1800 B.C. Carrizal/
K-4
Preceramic Period
1800 B.C. Carrizal/ Chinchorros
K-4
1900 B.C. Carrizal/ Chinchorros
K-4
2000 B.C. Carrizal/ Chinchorros
K-4
2100 B.C. Carrizal/ Chinchorros
K-4
2300 B.C. Carrizal/ Chinchorros
K-4
2500 B.C. Carrizal/ Chinchorros
K-4
Lithic Period
2500 B.C. Chinchorros

3000 B.C. Villa del Toquepala Chinchorros
Mar Asana
4000 B.C. Villa del Toquepala Chinchorros
Mar Asana
5000 B.C. Villa del Toquepala Chinchorros
Mar Asana
7000 B.C. Ring Site Toquepala
Asana
8000 B.C. Ring Site Asana

9000 B.C. Asana







67
analysis shows that the diet was based almost exclusively on
marine resources, such as gastropods, bivalves, and chitons.

In the highlands above Moquegua, hunter/gatherers of the

Lithic Period occupied natural rock shelters and caves. People used
laurel-leaf shaped stone projectile points to hunt camelids and
other small animals of the puna (Lumbreras 1969). The Toquepala

Cave contains wall paintings of humans hunting camelids (Table 3-
4). Situated at about 2000 m.s.l., this cave was occupied, mostly

during the wet period from October to April, for many years,

beginning ca. 9500 B.C. (Moseley 1992).
Another wet season site, dating from about 9600 B.C. to 3600
B.C., is the open-air settlement of Asana located at 3450 m on the

highland puna above Moquegua. The inhabitants of Asana

exploited the local fauna all year, and, during the dry season,
humans and animals alike depended on the ground seeps-
Bofedales-as freshwater sources (Aldenderfer 1989). This site
contains one of the earliest ceremonial structures in all of Peru.

Measuring almost 12 m by 9 m, this rectangular ceremonial
complex has a wide clay floor, altars constructed of rock and clay,
and walls made of mud and clay (Aldenderfer 1990).


Preceramic Period
Preceramic Period sites along the Ilo coastline include K-4,
which dates to as early as 4620 B.P. +/- 90 (BETA) (Wise et al.
1994), with occupation continuing into the Initial Period (Table 3-
4). This site consists of a large shell midden and domestic terraces,
which are located less than 100 m from the Pacific Ocean. Early







68
sites at the Carrizal Quebrada date from about this same time

period, i.e. 4690 B.P. +/- 100 years (BETA) (Wise 1990). Human
occupation at the Carrizal Quebrada is almost continual until
modern times (Bawden 1990) probably because of the freshwater
springs and the fact that this quebrada has easy access to the
marine resources, which were exploited by the early residents, just
as they still are today.

Another important coastal site dating to this period is Villa
del Mar, first investigated in 1986 because of a double burial
uncovered during a construction project (Torres et al. 1990).

Several Chinchorros-like burials were excavated here in 1990 by
Dr. Karen Wise. The Chinchorros were fisherfolk who occupied the
northern Chilean coast from ca. 5000 B.C. and used a wide variety of

fishing tackle including nets, harpoons, and fishhooks made of shell
or cactus thorns. The Chinchorros people are perhaps most noted

for their elaborate mummification techniques, which involved the
preservation of the viscera and also the bracing of the vertebral
column and the long bones of the deceased with cane supports
(Allison et al. 1984).


Initial Period
The preceramic sites, which were located along the coast

adjacent to the springs at the Carrizal Quebrada, were still occupied
during this period, with the inhabitants clearly exploiting all of the
micro-ecological zones. Marine resources were augmented by
terrestrial hunting and gathering (Bawden 1990), but there could
have been some early agricultural fields which were subsequently







69
destroyed by the later Chiribaya occupation of this quebrada. This

subsistence strategy was the beginning of an adaptive system
which would be incorporated into the later ceramic periods and

would endure until the Colonial Spanish Period.
The primitive Carrizal pottery sequence, which includes the

neckless olla, is related to other early undecorated ceramics, such as
Faldas del Morro from Chile and the Huaracane style from an early
agricultural culture in the upper Moquegua Valley (Feldman 1990).
The Huaracane pottery, which also includes a neckless olla style,

belongs to a pottery horizon which begins around 850 B.C. towards
the end of the Initial Period and lasts until 300 A.D. of the Early
Intermediate Period. This same type of pottery is also found in the
Bolivian highlands during the middle Wankarani phase and at
Chiripa (Feldman 1990; Table 3-5).


Early Horizon
The Neckless Olla (Olla sin cuello) ceramic tradition begins as

early as ca. 800 B.C. in the Ilo area (Table 3-5). The style of this
large, wide-mouthed cooking vessel persisted considerably longer
than a millennia. By about 200 B.C., the neckless olla was produced

at the Early Horizon sites at the Carrizal Quebrada (Bawden 1989,
1990). The importance of this domestic vessel is clearly established
by the fact that 80% of the pottery sherds found at Carrizal were of
this type. Neckless olla also accounted for the largest ceramic
component at the Ilo Valley sites of El Algodonal and Loreto Viejo
(Owen 1992a.).









70





Table 3-5: Ilo Time Chart


Early Intermediate Period


Although located 300 km from the Tiwanaku heartland, there

was a strong Tiwanaku presence in the upper valley around the city


Time Lower Ilo Ilo Coast Moquegua Highlands Arica, Natural
Frame Valley I I Chile Events
Early Intermediate Period
600 A.D. Neckless Neckless Omo Huari Cabuza
Olla Olla Tiwanaku ua
500 A.D. Neckless Neckless Omo Hauri Cabuza
Olla Olla Tiwanaku IV
400 A.D. Neckless Neckless Omo Tiwanaku IV
Olla Olla
300 A.D. Neckless Neckless Huaracane
Olla Olla
200 A.D. Neckless Neckless Huaracane
Olla Olla
100 A.D. Neckless Neckless Huaracane
Olla Olla
0 Neckless Neckless Huaracane
A.D./B.C. Olla Olla
100 B.C. Neckless Neckless Huaracane Wankarani Faldas del
Olla Olla Chiripa Morro
200 B.C. Neckless Neckless Huaracane Wankarani Faldas del
Olla Olla Chiripa Morro
Early Horizon
200 B.C. Neckless Neckless Huaracane Wankarani Faldas del
Olla Olla Chiripa Morro
300 B.C. Neckless Huaracane Wankarani
Olla Chiripa
400 B.C. Neckless Huaracane Wankarani
Olla Chiripa
500 B.C. Neckless Wankarani
Olla Chiripa
600 B.C. Neckless Wankarani
Olla Chiripa
700 B.C. Neckless Wankarani
Olla Chiripa
800 B.C. Neckless Wankarani
Olla Chiripa
900 B.C.







71
of Moquegua. As a result of the studies conducted since 1983 by
archaeologists of the Programa Contisuyu, the local Tiwanaku

sequence is now well-understood. Based on the data from
excavations at domestic sites and cemeteries, the local sequence is

now divided into the following phases. The earliest occupation is
called Omo and corresponds to the "Classic" Tiwanaku IV phase

(Table 3-5). Ceramics from the Omo phase are characterized by
banded keros, angular bowls, and incense burners with 2 wings.

The strong affect of the Tiwanaku State on Omo pottery is indicated
by the painted decorations rendered in black, orange, and white on
red, which is a common tradition for all of the pottery from
Tiwanaku (Goldstein 1990).


Middle Horizon
The Omo Phase of the Tiwanaku influence in the Moquegua

Valley continued into the early Middle Horizon, but it was replaced,
ca. 700 A.D., by the Chen Chen phase which continued until about

950 A.D. (Goldstein 1989; Table 3-6). At about the same time that
the Chen Chen phase ends in the upper Moquegua valley, the

Ilo/Tumilaca occupation begins in the lower Ilo Valley (see below).
The ceramics from the Chen Chen Phase (Tiwanaku V) also reflect
the form and decoration of the Tiwanakan State. The common

decorated pottery forms are keros without bands or a single band,
large cups in the form of a half-kero, and jars with only one handle.
Other common utilitarian artifacts include decorated wooden
spoons. A new style of kero with a human profile appears for the
first time late in this phase. Stylized flamingos and parrots painted







72
on the vessels were among the most common motifs. Although the
neckless olla sites in the Ilo Valley had long been abandoned where
the influx of highland influence, nonetheless, the neckless olla
tradition is still well-represented in the ceramic record of the

Middle Horizon.

It was toward the end of the Middle Horizon when the
lengthy Osmore Canal was probably begun. The question of who

truly constructed this masterful irrigation system still remains a
matter of contention. At least one author, Owen (1992a), believes

that because of the proximity of their settlements to the canal, the
Ilo/Tumilaca, rather than the Chiribaya, built the canal. It must be

remembered, however, that the Ilo/Tumilaca designation used by
Owen corresponds almost exactly to what Jessup (1990, 1991) calls

the El Algarrobal Phase of the Chiribaya Culture. Therefore, it is
quite possible that even if the Chiribaya did not actually begin the

construction of the valley irrigation system, they at least expanded
it to its farthest extent.


Late Intermediate Period

At the beginning of the Late Intermediate Period, around
1000 A.D., sites in the middle Moquegua Valley were abandoned
and the upper valley and the coast were settled. Owen (1992b)

suggests that this abandonment of the upper valley sites restored
more irrigation water to the lower Ilo Valley and, therefore, could
have provided the impetus for the expansion of agricultural land.
The Ilo/Tumilaca/Cabuza (I/T/C) tradition persists until about 1250









73
A.D. when it is rather abruptly replaced by the more populous

Chiribaya Culture (Owen 1992a). Since the number of Ilo/Tumilaca



Table 3-6: Ilo Time Chart

Time Lower Ilo Ilo Valley and Moquegua Highlands Arica, Natural
Frame Valley (1) Coast (2) (3) Chile Events
Colonial Period
1607 A.D. Chuza
Flood
1604 A.D. Earthquake
Tsunami
1600 A.D. H.P.
Eruption
1550 A.D. Spanish Spanish Spanish Spanish

1532 A.D.

Late Horizon
1532 A.D. Inca Inca Inca Inca

1476 A.D. Estuquifia/ Estuquifia/ Estuquifia/ Inca
Inca Inca Inca
Late Intermediate Period
1476 A.D. Estuquifia/l Neckless Olla Estuquifia Gentilar
nca
1400 A.D. Burro Flaco Neckless Olla Estuquifia Gentilar
San Geronimo
1300 A.D. San Geronimo/ Neckless Olla Estuquifia San Miguel 1fO0 A.R
Yaral(l) Miraflores
Flood
1200 A.D. Yaral (1) Ilo/Cabuza (2) Estuquifia San Miguel
Neckless Olla
1100 A.D. Yaral Ilo/Cabuza (2) Tumilaca (3) San Miguel Chimu Flood
(I)Algarrobal Neckless Olla
1000 A.D. Algarrobal (1) Ilo/Tumilaca(2) Tumilaca (3) Maitas/
Neckless Olla Cabuza
Note: Local culture/tradition classifications according to (1) David Jessup (1991); (2)
Bruce Owen (1992a); and (3) Paul Goldstein (1989, 1990).
M iddle Horizon
1000 A.D. lIo. umiia'l- Neckles; Oila Chen Chen TIwanaku V Cabuza

900 A.D. Neckless Olla Chen Chen Tiwanaku V Cabuza

800 A.D. Neckless Olla Chen Chen Tiwanaku V Cabuza

700 A.D. Neckless Olla Chen Chen Huari Cabuza
Huari
600 A.D. Neckless Omo Huari Cabuza
Olla Huari Tiwanaku IV
Huari







74
sites had decreased by 50% by the beginning of the Ilo/Cabuza
phase, there is a strong possibility that the I/T/C simply could not
adequately compete with the Chiribaya Culture which reached its
zenith at about 1200 A.D. during the Yaral Phase (Jessup 1991;
Table 3-6). Although the Yaral Phase ceramics were identified from

the grave accompaniments at the middle valley site of La Yaral, this
pottery style is also found at the lower valley sites of Chiribaya Alta
and Chiribaya Baja, and at the coastal site of San Geronimo (Jessup
1990).
It is presently unclear as to the origin of the Chiribaya people.
It has been proposed that they may have emigrated from the
highlands into the Ilo Valley, or they may have developed

independently from an earlier Preceramic population or the
Ilo/Tumilaca people (Jessup 1990, Owen 1992b.). There is some
indication that there may have been some Chiribaya influence along
the extreme northern Chilean coast or vice-versa since the Chilean
San Miguel and Gentilar ceramics (very similar to the Chiribaya

styles) were found co-existing with the Chiribaya ceramics at the
site of Chiribaya Baja (Jessup 1987, 1990; Table 3-6). Currently
DNA studies are being conducted on some of the human remains
recovered from the cemetery at Chiribaya Baja in the Ilo Valley,
and, perhaps in the near future, the results of these investigations
may provide us with a definitive answer as to the origin of these
Chiribaya people.
The last phase of the Tiwanaku presence in Moquegua is the
Tumilaca, which actually occurs after the decline of the Tiwanakan
State. The manufacture and the decoration of the Tumilaca







75
ceramics indicate a gradual loss of contact with the altiplano. For
example, the paste now contains more sand; the color yellow is used
instead of red; and the keros are very large. Further, white dots,
similar to the common Chiribaya decoration, are used, suggesting a
trend toward the Chiribaya style of ceramics (Goldstein 1990).

The Late Intermediate Period was probably a time of hostility

because structures at the Tumilaca site include a wall and a

perimeter trench which are strong indications of a defensive
posture (Goldstein 1990). With the exception of the large Chiribaya

Alta site, no other locations in the lower Ilo Valley have defensive
moats or walls. Almost all of the major settlements by the Late
Intermediate Period were fortified, and in the Moquegua Valley, all
independent irrigation areas and their villages had walls and dry
moats (Moseley 1992). Fortified sites of this period were probably
created to protect the natural resources of the region and the
irrigated agricultural fields (Moseley 1990).

Another prominent fortified site is the Huari settlement
located on the top of Cerro Baul, a truncated cone of stone which
rises an impressive 400 m above the valley floor. The almost
impregnable fortress was likely a necessity since the Huari colony

had intruded into the Tiwanaku presence already firmly ensconced
in the Moquegua area. Centuries later, the Cerro Baul fortress

would prove to be vulnerable when it fell to the invading Incan
army after a 50 days' siege (de la Vega 1989).
Following the end of the Tiwanaku occupation in the
Moquegua Valley, the Estuquifia Phase begins around 1100 A.D.
With natural escarpments providing defense on the east and south







76
sides, the "type site" of Estuquifia was occupied from about 1100

A.D. to 1500 A.D. Late in the period, the Estuquifia were
contemporaneous with the Gentilar of Chile and the Inca (Lozada
1987). However, there is no evidence indicating that there was
habitation at Estuquifia after the Spanish Conquest (Rice et al.
1990). Estuquifia pots and bowls show few significant
characteristics other than opposing prominences around the upper
edges of the vessels. Other common Estuquiiia ceramics are double-

handled ollas, jars with one handle, and jars modeled in the shape
of a boot or a duck (Rice et al. 1990; Williams et al. 1990).
About 1100 A.D., the Chiribaya established an upvalley
enclave at La Yaral, which was occupied for about 200 years. In the

lower Ilo Valley and in the coastal quebradas, agrarian expansion
by the Chiribaya reached its height around 1000 A.D. (Moseley
1993), and then began to contract slowly because of changing
environmental conditions, and, by the time of the Inca occupation of

the Ilo area, most of the irrigated terrace systems had long since
been abandoned.


Intensification and Development of Irrigated Agriculture

Introduction
Early great civilizations, such as Mesopotamia, Egypt, and
Mexico, all reached their zeniths in arid regions under irrigation
(Kosok 1942; Mason 1957; Wittfogel 1957; Heiser 1973), and Peru
was certainly no exception. "One of the major areas of prehistoric
civilization in the New World is the stark desert landscape of coastal
Peru" (Jackson and Stocker 1982:12). "At a time when our







77
ancestors in northern Europe were still utter savages, clothed only
in skins, and living by hunting and fishing, settled agricultural

communities must have existed in the Peruvian region" (Cook

1916:474). All the major states and the independent polities, that
were extant during the preceding Middle Horizon, were highly
dependent upon intensive agriculture (Table 3-3). During the Late

Intermediate Period, the Chimu State and the smaller polities used,
as did their predecessors, some type of creative and highly

productive agricultural technology--irrigation canals (asequias),
terracing (andenes), raised fields (camellones), or sunken fields
(cochas--lagunilla in Spanish) (Guillet 1992).
There are two theories, the Drainage Theory and the
Floodwater Theory, concerning the origin of canal irrigation. As

with some other popular theories they do not apply to Peru in

general. The Drainage Theory states that canal irrigation developed
from the technology used to drain and direct excess water from
productive land. The technology was simple and minimized
downcutting, sedimentation, meandering and overflowing of
streams (Doolittle 1990:138-189). This theory might be applied by
some to northern Peru because of the presence of raised fields in
the Casma Valley. However, this technology of constructing raised

fields does not appear in the archaeological record of the north
coast until almost 1,000 years after the development of canal
irrigation. It seems more likely that valley necks and inland
locations were the setting for the new technology and economy of
irrigated agriculture (Moseley 1975b) because these inland areas







78
required shorter lead-off canals and, therefore, less labor to build
and maintain (Moseley and Deeds 1982).
The Floodwater Theory states that planters first farmed in

arid areas on alluvial deposits which were periodically inundated
by floodwaters from upland rain. Subsurface moisture would

remain for some time after the flow subsided, thereby allowing
some cultivation (Doolittle 1990:140). This theory could apply to
areas such as the Nile River Valley which, until the building of the
Aswan High Dam, had practiced such agricultural methods. It might

also be applicable to the Moche Valley where a floodplain exists
(Moseley and Day 1982). However, once again, canal irrigation in
Peru probably did not advance as a result of floodwater agriculture
because, of the 57 river valleys and drainages in Peru, only a few
have permanent rivers, which could supply sufficient spring flood
waters.
Sunken fields were late additions to the prehistoric

agricultural repertoire. They generally occur in a topographically
low area lying very near the coast, but they have been found as
much as a kilometer inland from the littoral in the Chicama, Moche,
and Pisco Valleys among others (Smith 1985:519). Sunken fields
are known by different names depending on the geographical
location and the time period in Peru. They are called Pukios or
wachaques on the north coast, mahamaes in the Chilca Valley, and
are referred to as hoyas by most of the early chroniclers (Smith
1985:603). According to Cobo, Hoyas were used by the native
Peruvians to increase farmland (Mateos Tomo II, 1956:92). Large
depressions, up to 100 m wide, were dug into the earth below the







79
water table, and crops were planted along the slopes of the sunken

fields (Cabello Valboa [1586] 1955; Donnan and Mackey 1978;

Moseley and Feldman 1984; Smith 1985).
It is only in recent times that modern engineering has been
able to surpass the irrigation projects of the PreHispanic Peruvian

agrarian societies (Horkheimer 1990).


The Chimu were probably the greatest Pre-Hispanic
agriculturalists in coastal Peru. Only within the past
decades has the amount of irrigated land in the Chicama
and Moche Valleys equaled the area cultivated by the
Chimu, and this has required the extensive use of pumps
and cement-lined canals (Kus 1972:193).


For example, North/Northwest of Chan Chan, the Chimu had

irrigated 1,600 hectares in one area alone (Kus 1972:199). During
the early Late Intermediate Period more canals were built and
extended into previously uncultivated plains (Parsons and Hastings
1988:198). These large irrigated agricultural works, occupying

suitable alluvial plains, were all state controlled by this time (Kus
1980). Rural administrative centers represented the "state
presence" even in the non-metropolitan areas of the Moche Valley

and maintained control over land, water, and labor resources
(Keatinge 1974:67). Labor was often made available as a tax

payment to the state (Conrad 1981). At its height, Chan Chan
controlled, at least, 66 percent of the irrigated coast lands (Moseley
and Deeds 1982:25).
Most early eyewitness accounts mention, with some degree of

awe, the vast terraced and irrigated agricultural systems which







80
they encountered throughout Peru (Cieza de Leon [1552], 1971;
Cobo [1653], 1979, 1990; Sarmiento de Gamboa [1572], 1967;
Guaman Poma de Ayala,[1613], 1980; Pizarro [1571], 1972; Sancho

[1550], 1917; 1972; de la Vega 1989). Cieza [1552] notes "all the
land of the valleys, where the sand does not reach, up to the
wooded areas, is one of the most fertile and abundant lands of the
world..." (1971:25). With just a little work, using irrigation water
from the rivers, it was possible to cultivate abundant crops (Cieza
1971). This type of agriculture is particularly important in the
sierra because "irrigation permits the cultivation of a broader range
of crops at higher altitudes than would be possible with natural
rainfall alone" (Mitchell 1977:38).

Since the narrow, coastal plain of Peru is one of the world's
driest deserts (Brush 1977; Lettau and Lettau 1978), where
agriculture is extremely difficult and virtually impossible without
irrigation, Cobo [1653] noted that "not a twentieth of this large
stretch of land is productive" (1979:4). According to Squier, "The
Sahara is a 'thing of beauty' and Arizona 'a joy forever' compared to
the coast of Peru" (1877:25). "It never rains, thunders, snows, nor
hails in all this coast, which is a matter worthy of admiration.
{However,} a little distance from the coast it rains and snows
terribly" (Acosta [1604] Vol. I, 1970:164). "Rain does not fall until
it reaches the cooler elevations of the Andes at altitudes above
2,500 m" (Pozorski and Pozorski 1987:1). This natural aridity
accounts for the fact that about 22 percent of the cultivable land
today in Peru is located on the coast, while the sierra has almost 62







81
percent of the arable land and produces the majority of the food for
modern Peru (Claverlas et al. 1983).
Unfortunately, the total area under cultivation today in the
Andes is from 30% to 80% less than it was the during the Pre-
Hispanic periods, depending on the region (Wright 1963; Moseley
and Deeds 1982; Masson 1986; Denevan 1987; Clement and Moseley
1991). Perhaps part of the reason for this disparity is the fact, as
one author explains, that the use of contemporary terraces is
sometime underreported (Mitchell 1985). The Titicaca region has
the largest ridged field system in the world (Kolata 1991:101),
where ridged fields have been used since 2500-3000 B.C. (Erickson
1987:374). Yet in the Titicaca Basin alone, there are 200,000 ha
(Hectares) of abandoned terraces, 90,000 ha of abandoned or
plowed up qochas (cochas) sunken fields, and 80,000 ha of
abandoned or destroyed ridged (raised) fields (Browman 1987:175).
Truly, the Titicaca area has a tremendous amount of potential
for food production because, during the Tiwanaku Empire, just the
3,500 hectare raised field agricultural system, at Pampa Koani could
have fed an estimated 60,000-100,000 people annually and still
have produced a food surplus (Kolata, 1987:40). At just 75% usage
of the ridged fields, mono-cropping could support a minimum of
285,000 people (Kolata 1991:110). Tiwanaku developed its mighty
empire partially based on the ridged field system surrounding Lake
Titicaca (Kolata 1983, 1986) and at one time dominated highland
Bolivia and the coastal portions of southern Peru and northern Chile
(Orlove 1985).







82
It is a lamentable fact that this once highly productive

agricultural system has been abandoned for centuries (Kolata

1983). Luckily, experiments in the reactivation of some of these
ancient raised fields at Puno and Pampa Koani, on the south side of

Lake Titicaca, have been conducted in recent years resulting in crop
yields that are 2-4 times the average yield from other agricultural

fields (Erickson 1987, 1988; Kolata and Ortloff 1989; Kolata 1991).
Although ridged fields are found throughout the Tropics, they
are relatively rare along the arid coast of South America. Yet in the

Casma Valley, it is estimated that the abandoned ridged field

system once encompassed as much as 3,100 hectares (Moore
1985:265; Pozorski et al. 1983). "The Casma Valley is unique in
having ridged fields, rather than sunken gardens, which are the

most common feature on the high water table regions of the coast"
(Pozorski et al. 1983:407).
Fossil ridged field systems also exist in Bolivia, as well. This

area, in general, offers important archaeological data about various
impressive projects of PreHispanic engineering. For example, in

addition to the ridged fields, there are canals, water

wells/reservoirs, and roadways. Some of the canals and roadways
are still in use today, though some have been repaired through time
(Erickson 1980). Perhaps some of the most spectacular vestiges of

ancient ridged fields are found in northern Colombia (Smith et. al
1968; Broadbent 1987), and, also, in Ecuador (Parsons and Shleman
1987), where maize cultivation, on ridged fields, dates to 500 B.C.
(Pearsall 1987:287).









Water Management

Today water rights are owned and regulated by the state
water administration (Administraci6n de Aguas) (Hatch 1976;
Guillet 1992), rather than being totally controlled by the ayllu, or
the community, as had been the case for millennia. However, the

various irrigated sections have elected water judges (Regidores),
who hold weekly meetings at which "water shares" are granted to

users (Guillet 1987c). Permission for use of water is granted in the
descending order of domestic usage, animal husbandry, agriculture,
hydroelectric power, industry, and mining (Guillet 1992). Farmers
are advised of what day their particular branch canal (rama) will be
provided with water (Hatch 1976), and they must determine their
specific needs according to the availability of water (Farrington

1985). It is a common practice for a single household to control a
branch canal, while many households control a main canal (Brush
and Guillet 1985). This type of water management is a far cry from

Pre-Hispanic days when, even during dire times, water was doled
out to everyone, regardless of rank or social status (de la Vega,

1989).


Development of Agriculture
Why did humans abandon nomadic hunting and gathering or
sedentary fishing lifestyles in favor of a new sedentary agricultural
life style? Boserup (1965) proposes that the development of
agriculture was the result of population pressure (see also Patterson
1973) which necessitated the creation of a more prolific system of
food production (see Carneiro 1970). It is generally agreed that







84
population density is a crucial factor for giving impetus to the

intensification of agriculture because it is usually the number of
people residing in a particular area that determines the necessary

level of food production. By about 1 A.D., population increase and
social demand in the Moche Valley possibly caused an expansion of

arable land (Farrington 1985:648). Based on her studies, Boserup
suggests that labor-intensive irrigation allowed for a more intensive

system of land use, "multi-cropping," which would increase the
amount of arable land and raise crop yields, thereby increasing the

carrying capacity of the land (1965:39). However, others maintain
that multi-cropping "appears to have been practiced only sparingly
in the prehistoric world" because crop varieties which allow a

traditional farmer to produce more than one crop per year have
only been recently developed (Farrington 1980:288).
Sauer (1952) thought that agriculture on the desert coast of

South America derived from elsewhere because the environment
demanded the advanced skills of irrigation. He believed that the
northwest coast of South America was the likeliest place to find the

origins of agriculture because of its abundant aquatic and riparian
life and fertile land. This region provided a sheltered basin with

the "proper balance of self-containedness and outside contact"
(1952:42). He may have been in error with these statements, but
he was, however, slightly avant garde when he stated, "sedentary
fishing people perhaps commenced the cultivation of plants and
became the first domesticators of plants and animals" (1952:103;
see also Moseley 1975a, 1992; Moseley and Feldman 1988).







85
Social Change Associated with Agriculture
With the advent of irrigation agriculture, a number important
social changes occurred. Agriculture allowed for a stable, improved
diet which in turn created a greater concentration of people,

transforming primitive societies into complex social/political
structures (Kosok 1942). These structures included a more efficient

social organization and a concomitant expanding power over people
and, especially, the control of labor.
According to Wittfogel (1957), ancient states arose and gained

despotic control of dense populations because of the control of

water in arid regions. In many hydraulic societies, the state
retained control over the private property owners by keeping them
disorganized and impotent (1957:3-4). Irrigation farming always
requires more physical labor than does dry farming because

controlling large amounts of water through channelizing requires
some type of direct authority that subjugates many people

(1957:17). Therefore, "effective management of irrigation works
requires an organizational web that covers the whole or, at least,
the dynamic core of a country's population" (1957:27).

Carneiro (1970), who advocates environmental
circumscription as the cause for the origin of the state, sees the
many, short, narrow valleys of the Peruvian coast as a classic
example of what he champions. Each valley is "backed by
mountains, fronted by the sea, and flanked on either side by desert
as dry as any in the world" (1970:735). As the size of autonomous
villages grew, the population would fission and move to other
usable land. When there was no more available land, agriculture







86
had to intensify, and previously unusable land was brought under

cultivation by means of terracing and irrigation (1970:735). With
increasing population pressure, conflicts for control ensued, and the
conquered people usually would be politically subjugated.
Wittfogel's theory and, perhaps, also, Carneiro's theory might

be true for some coastal areas of the Peru, but, in general, they may
not be true for other areas of Peru. According to Guillet (1990)

there was no despotism on the Central Coast of Peru, even though
there were centralized hydraulic systems with bureaucracies.
Further, in the Peruvian highlands, there were both small-scale
village or inter-village irrigation systems (1990:7) where water

resources, i.e. rivers, springs, and seeps, as well as land resources,
were collectively owned and operated by an ayllu--a corporate,
endogamous, hierarchical descent group (Brush 1977; Moseley
1992; Denevan 1987). There is no evidence that a more elaborate
system beyond the traditional ayllu was necessary to effectively

manage highland agricultural systems (Erickson 1987).
"Recent literature in Andean archaeology and ethnohistory
asserts the dominance of local kin groups in the organization of
agricultural production rather than supracommunity state
authority" (Kolata 1991:99). Water is distributed according to
custom by the people who assembled at the distribution points, not
according to some despotic control (Mitchell 1977:50, 57). There is
further criticism of these theories because there is a lack of studies
of contemporary highland communities in Peru and there has been
too much focus on coastal irrigated agriculture (Mitchell 1977).







87
In southern Peru, there were independent coastal populations,
such as the Ica and Chiribaya, which maintained control over their
water resources and never reached a social organization level
beyond that of a polity. Rowe (1963) explains that in the Ica
Valley, large cities came first, and only later were there major
irrigation canals. He says that it is very difficult to argue for any
relationship between irrigation and the development of cities
(1963:20). Lanning (1967) states that we cannot say that irrigation
led to the centralization of authority, but rather that once authority
was centralized, then it became possible to build and maintain

irrigation systems. Thus, irrigation was a product of civilization, not
a cause of it (1967:181-182).
Whether the society is classified as a state or polity, the social
organization and control of a society dependent on either a small-
or large-scale irrigation system remains basically the same. There
exists a controlling body that directs the construction of terraces,
fields, and irrigation canals, regulates the distribution of water, sets
production quotas, and determines the planting and harvests cycles
(Harris 1975).
Now that the stage was set for large-scale irrigation projects,
agriculture expanded into heretofore undeveloped and unoccupied
areas, such as the hyperarid coast and the steep Andean slopes
(Moseley 1992). Not only the vast tracts of abandoned agricultural
works, but also the impressive number of earthen and stone
mounds and pyramids remaining on the Peruvian landscape are all
mute testimonies to the amount of human labor that was released
from other pursuits by the accession of agriculture.