The central Peruvian desert coast extends southward through several valley systems. The major rivers that descend to the Pacific and break the desert are the Pativilca, Huaura, Chancay, Rimac and Chillon, Malay, Caiiete, Chincha, Pisco, lea, and Nazca. At the time of the conquest, most residents of the central coast subsisted on the basis of irrigation agriculture, with food supplemented by marine resources. Most of the area was brought into the Inca Empire through conquests between 1460 and 1480. Pachacamac was the most important center in that sector of the Peruvian coastal region. It contained a relatively dense urban population and, as the chief ceremonial shrine along the entire coast, attracted a large number of temporary residents.

The central Peruvian highland corregimientos are in the heart of the Andean chain. The elevation of the area is high. The frigid, windswept pun a appears inhospitable but provides fodder for extensive flocks of llamas and alpacas, which are used for wool. Upper sections can also produce ichu and quinoa, a native cereal. Potatoes and other tuberous crops are grown in cool zones, and in the more temperate valleys wheat, corn, and other crops are cultivated. The central highlands are a major mining region of the Peruvian Andes. The modern Cerro de Pasco mining district is in the north; minor silver deposits were worked throughout the central highlands; famous mercury mines are located at Huancavelica in Chocorvos; and there was a major silver-mining center at Castrovirreyna in the southern part of the same corregimiento. In this chapter we shall examine the relationship between mining, which was so important in the central sierra, and population change.

In the preceding pages we attempted to gauge the magnitude of Peru's preconquest population. By any standards the Inca population was large. Perhaps 9 million people resided within the limits delineated by Peru's contemporary boundaries. The number of inhabitants remaining a century after contact was roughly a tenth of those that were there when the Europeans invaded the Andean world. In the following chapters we shall examine the transformations that were taking place as a result of the demographic collapse of native America. The impact of population change varied from one place to another within Peru. Some Indian communities were wiped out. In a few cases numbers actually increased, in spite of the general downward tendency. We shall examine these developments in as great detail as possible. Analysis is within the framework of major geographical regions. We have already seen the differences between coastal and highland change. More subtle variations are discernible as one moves southward, down the Peruvian coast. Although on occasion such a structure becomes too rigid, the results are useful for comparative analysis.

The northern Peruvian Andes are distinct from the central and southern sections of the chain. The Andean mountains are both lower and narrower in the north than elsewhere. This fact accounts for the different ecological structure of the north and also helps explain the unique demographic development of the region. The highest point on the Ecuadorian border is only 3,726 meters. The passes between the coast and the upper Amazon basin are also lowest in the north. The cut of Porculla connecting the Olmos and Marañón valleys is a mere 2,144 meters in elevation. The transverse cuts in Figure 9 illustrate the low and broken nature of the Peruvian north. The whole of the northern sierra is also characterized by the Marañón River system. Running from south to north along the area, it divides the region into the more highland west, and the montaña to upper rain forests in the east.

The demographic transformation of Peru was almost complete by the end of the first century after the Spanish arrived at coastal Tumbez. “Collapse” most aptly describes the process along the Pacific desert strip. With few exceptions, a once dense native population was wiped out and partially replaced by Europeans, Africans, and migrants from other regions. Remaining Indians were quickly integrated into the lower strata of colonial society. Collapse also occurred in the low-lying areas of the northern highlands. In the north and northeast, fewer outsiders immigrated to establish permanent settlements. That area therefore remained economically and socially outside the mainstream of the historical development of Peru during the colonial perod. The highlands proper, from Cajamarca, Conchucos, and Huailas southward to Lake Titicaca, also entered into the downward spiral, but in a less pronounced fashion. The highland population was more evenly spread over vast distances in contrast with the coastal populations, which were concentrated in narrow valleys. Population decline of the central and southern sierra was considerable; but not catastrophic as elsewhere. Large populations allowed the highland Indian to maintain social and economic institutions long after they were lost on the coast and in the far north.

Modern medical researchers are able to project the time, place of appearance, rate of infection, and mortality for disease epidemics. The projections are not completely accurate; nonetheless, predictions of impending outbreaks of potentially dangerous epidemics are taken seriously by public health officials. Indeed, modern disease control is based on the premise of predictability. In this chapter we shall attempt to project a preconquest population figure on the basis of estimated mortality for known epidemics that swept Peru in the sixteenth and early seventeenth centuries.

Numerous complicating factors confront the researcher of early colonial epidemics in America. First, descriptions of epidemics are conflicting. It is often difficult to ascertain the correct disease agent that is called various names by colonial eyewitnesses. Second, the American Indians apparently lacked immunity to common European diseases. Measles, for example, which had a low mortality rate for Old World residents, had very high rates for the virgin American Indian population. Third, general health and diet affected the mortality rate for those who became ill. Warfare, famine, and overwork contributed to the death rate for those who contracted the disease. Fourth, it appears that elevation affected the course and outcome of several types of epidemics. This factor is especially important in the Andes. Fifth, either diseases or humans or both evolved during the century after the first contact.

Human populations cannot expand forever. Densities are ultimately limited by the environment's ability to sustain them. One of the principal limits to population is the supply of essential foodstuffs. The population of Europe in the years prior to the Black Death in the mid-fourteenth century appears to have surpassed the limit and was suffering from a series of inadequate harvests and famine before the devastating epidemics of 1348. The population of Central Mexico under the Aztecs also seems to have exceeded the carrying capacity of its agricultural economy. Sherburne F. Cook has suggested that extensive practice of human sacrifice by the Aztecs was an implicit societal attempt to redress the balance between inhabitants and land. Study of the maximum carrying capacity of an ecological system can assist the researcher in establishing population bounds, but, as we shall see, there are many variables in the equation, and the estimated limits are far from precise.

The human factor adds a major complication in assessing the ecological potential of a region. No two groups of people are exactly the same.

There are three distinct ecological systems in Peru: the coast, the highlands, and the montaña with the Amazon basin. Each of the regions has its own set of unique characteristics and is subject to its peculiar internal divisions. Common to all the Peruvian coast, and the factor that contributes to its special climate, is the Humboldt current. That broad river, flowing in a counterclockwise fashion in the eastern South Pacific, sweeps cold Antarctic water along the South American coast. The prevailing westerly winds push the air, cooled by the water, against the unbroken Andean chain, which in places surpasses 6,000 meters in elevation. The result is an area of high barometric pressure. Almost no rain falls along the mid-Chilean to north Peruvian coasts. The consequence is a narrow desert strip some 3,000 kilometers in length. As the air rises along the western Andean slopes, moisture condenses. In some areas, at roughly 600–1,500 meters in elevation, the moisture forms a fog or garua. At higher altitudes there is seasonal rainfall and at the highest points snow, which provide the water for coastal rivers and irrigation systems.

The basic source for historical demography is the census. The Incas of Peru made periodic counts of their subjects and recorded the numbers on knotted-string mnemonic devices, which they called quipus. Pedro Cieza de Leon, writing in Cuzco in 1550, clearly explained why the native rulers wanted statistical information on the inhabitants and how the count was administered: “the principal men and their delegates in all towns and provinces of Peru had to keep track of the men and women who died and were born each year, for the purposes of tribute, and in order to know the number of warriors and defenders of the community.” At the appropriate time the Inca sent an official called a lunaquipo, or quipocamayo, to each province. The leaders of each valley met the imperial representative at a central place. The quipu records of the previous census were brought, note was made of deaths and births, and a new division of the population was made into several age categories.

As one continues down the Peruvian coast, the desert becomes more pronounced and the river valleys fewer and farther between. Less rainfall descends on the western slopes of the Andes as the Chilean border is approached. Scarce water results in little land under cultivation and a small population. Even in pre-Spanish times the south coastal population was lower than in the other coastal sectors. Another negative factor affecting population of the area is geological instability. A series of volcanos extend from roughly 150 kilometers northwest of Arequipa to 100 kilometers southeast of the city. There are frequent and severe seismic movements in the region, and these have historically limited the rate of development of the afflicted area.

Demographers have for a long period of time estimated populations for dates on which no census was taken. Even if national censuses were available for 1920, 1930, and 1940, for the purpose of analysis it might be important to know the total population at intervening dates. If accurate and complete vital registration were available, estimating the total population at a point in time would be a simple process of addition and subtraction; but an adequate system of vital registration is rare, even in the twentieth century, especially in underdeveloped portions of the globe. Therefore, for the purpose of estimating population in the past or future the standard formula for population change becomes a necessary analytical tool. The formula P2 = P1ert represents change as a continuous process, which is indeed true of population movement. The curve of population growth, positive or negative, is an exponential curve, not a straight line. The process reflected by the formula is the same as that for the continuous compounding of interest.

In the previous chapters we examined a variety of approaches to making population estimates of the native population of Peru before first direct contact with the Old World invaders or the diseases that preceded the Europeans. Some of the avenues are less useful than others, and we had to discard them outright.

The ecological model, or a study of the carrying capacity of pre-Columbian Peru, may best suggest the limits of population growth. There is certainly a maximum population that may be sustained by an ecosystem. Due to the nature of Peru's coastal agriculture, it is easier to establish theoretical maximums for coastal valleys than for the highlands. The maximum 6.5 million that we reached (Table 3) for the coastal population is, in my view, a valid high point, but the total productivity of the highlands is not well enough known to establish a limit for that area. At best, we might apply seven people per square hectare as the carrying capacity of Peru's total cultivated land in 1961 to arrive at a figure of 13.3 million. This highly tentative number is about as good as any other figure one is able currently to derive from the method, given the large number of variables and the insufficient evidence. The question of whether the Empire reached the limits of the agricultural base is not yet fully answered. Evidence that it did comes from the extension of cultivated lands into marginal areas: ridged fields in floodplains, sunken fields along the coast, extensive terracing of the hillsides. Warfare did not seem to be a significant check on growth, however, nor did famines.

Little did I realize when I first read Woodrow Borah's New Spain's Century of Depression in 1962 that I would begin a search for similar material for the Andean area of South America that would continue for almost two decades. Borah's description of the relationship between population and economy in Central Mexico stimulated my investigation for parallel patterns in Peru. While preparing my master's thesis, under the direction of Lyle N. McAlister at the University of Florida, I became aware of the limited knowledge of the dynamics of population change in colonial Peru. By 1969 I had completed much of the basic research on Peru's Indian population from 1570 to 1620 for my dissertation at the University of Texas with Thomas F. McGann and James Lockhart, but teaching duties, and revisions, led to a delay in completion until 1973. I have continued the search for supplementary materials since that date, conducting new research in Peru in 1973, 1974, and finally 1977. I have now reached the point where I feel continued investigation would provide only diminishing returns. Most of the population data for Indian Peru for the period prior to 1650 have already been collected. Census results, at least the totals, for almost 585 repartimientos – the primary local unit of Indian administration – for various dates provide the basis for the presentstudy. Undoubtedly, other information will be found, but it is unlikely that the data base will be substantially modified. Many people have assisted me during the course of research and writing. Woodrow Borah and William Denevan have read a draft of the present volume. I owe them special thanks for their comments and support.