Cultivating Diversity: Field Scattering as Agricultural Risk Management in Cuyo Cuyo, Department of Puno, Peru

By Carol Goland, 1993.

Chapter 4 - An Introduction to the Communities of Cuyo Cuyo

(page 70)

Shortly after the conquest, the Spanish learned of the gold deposits in Carabaya. The province became a valuable encomienda,⁷passing among various prominent Spaniards (including the chronicler Garcilaso de la Vega) until it was returned to the Spanish crown for the final time in 1569 (Recharte 1990). Early records pertaining to these encomiendas identify an ethnic group named "Cuyo Cuyo" residing in the area of Sandia (ibid.). A later visita (1614) indicates the presence of three indigenous ayllus⁸in the region: Cuyo Cuyo, Lacaique, and Queneque, along with ayllus pertaining to the Qolla around Lake Titicaca (ibid.). Between the late 16th century and early 18th century, large numbers of immigrants swelled the population of the Sandia region.⁹In two hundred years' time, the tributary-paying population increased eightfold: in 1771, three-quarters were identified as foreign to the region (Recharte 1990). By at least the late 1700's, the present-day communities of the District of Cuyo Cuyo were minor ayllus (segments) of the maximal ayllu of Cuyo Cuyo (see Recharte 1990:29-45). In the process of gaining official recognition as comunidades campesinas (see footnote 1), the maximal ayllus dissolved and gave way to the present configuration of communities by the late 1960's (Recharte 1990:52-56).

ENVIRONMENT AND CLIMATE IN CUYO CUYO

The PSE project maintained a weather station near our home in Ura Ayllu. Each day at noon data were recorded on current temperature, 24-hour maximum and minimum temperatures, precipitation for the last 24 hours, cloud cover (percent and type), barometric pressure, and relative humidity. In addition, a mechanical pyranograph continually recorded incident solar radiation, and temperature was recorded with a continuous 7-day thermometer. A detailed analysis of these data is in progress (Winterhalder n.d.); here I present some preliminary results.¹⁰

Temperature and Precipitation Regimes in Cuyo Cuyo

(page 71)

As an eastern escarpment site at latitude approximately 14^o 30' S, Cuyo Cuyo is subject to relatively abundant rainfall, concentrated into a distinct wet season. SINAMHI¹¹ has recorded basic climatic data in Cuyo Cuyo since October 1963. Data collected from SINAMHI offices and by PSE together provide a 24 year sequence (up to October 1987) (Winterhalder 1988).

Sixteen years of precipitation data are available. During these years, the average yearly precipitation recorded in Cuyo Cuyo was 823.8 mm. The wettest year received 1696.5 mm of precipitation; the driest year 536.0 mm. During the first study year, precipitation was somewhat higher than average, with a total for 1985 of 998.3 mm; in 1986 total precipitation was a near-average 826.6 mm. Figure 4.5 illustrates monthly precipitation for the 31 months that data were collected by the PSE project (August 1985 - February 1988).

The first year of the study, 1985-86, was a particularly wet year, not only in Cuyo Cuyo but over much of the Department of Puno. The Lake Titicaca area was inundated by rain. Much of the lake basin agricultural crop was lost, roads were washed out, and families were displaced due to flooding. The PSE data show prolonged heavy rains (in excess of 115 mm/month) from November 1985 to March 1986. In contrast, in the following year the rains began to diminish in February.

Figure 4.5 demonstrates clearly the seasonal distribution of rainfall. May - August are the driest months of the year. Less than 31 mm of precipitation fell during each of these months of the PSE sample. In both 1985 and 1986, the rainy season commenced with light rains in September (less than 100 mm), which were then reduced in October. Several Cuyo Cuyo farmers remarked that this was a typical pattern, naming the September rains as lluvias del sembrio (rains for the planting).

Most rain falls between October and March. From October 1985 to March 1986, 769.3 mm of precipitation fell in Cuyo Cuyo, an average of 128.2 mm/month. During the remainder of the year (April to September 1986), only 215 mm fell (35.8 mm/month). In the subsequent agricultural season, from October 1986 to March 1987, 686.4 mm of precipitation fell (114.4 mm/month); from April to September 1987, only 142.6 mm fell (23.8 mm/month). The total precipitation for six dry months of 1986 combined was less than that for the previous December or January alone (166.4 and 207.4 mm, respectively).

These wet and dry seasons are also distinguished by temperature. The dry season is called chiraw tiempo (the cold season). Table 4.5 presents data on absolute and average seasonal maximum and minimum temperatures, as well as average seasonal mean. These data are also illustrated in Figure 4.6, aggregated by the two divisions of the year noted above. In 1986, the average temperature was 48.8^o F; the highest maximum recorded was 67^o (in December); the lowest minimum was 26^o F (both June and July). In 1987, the average temperature was 50^o F; the highest maximum was 69^o F, while the lowest recorded temperature was again 26^o F (July).

Neither average monthly mean nor maximum temperature varies much throughout the year. During the 32 months of temperature recording, the average monthly mean varied only slightly more than 10^o F, from 43.3^o F to 53.5^o F. The average monthly maximum also varied only by about 10 degrees, from a low of 55.6^o F to a high of 65.7^o F. The absolute maximum

(page 75)

ranged from 62^o F to 70^o F, only 8 degrees. In contrast, the absolute minimum varied from 26^o F to 40^o F, a full 14 degrees difference. This pattern was noted in Chapter 3: the greatest temperature differences in tropical mountain environments are not seasonal, but diurnal. And, variation in the minimum temperature accounts for a great portion of whatever seasonal differences in mean temperature do exist (Figure 4.6).

Differences in monthly minimum temperatures in large part can be accounted for by cloud cover. During the rainy season, heavier cloud cover prevents the escape of warm air and maintains relatively warm minima. In contrast, the clear skies of the dry season permit the warmth to escape as radiation during the night, producing relatively low minimum temperatures. Figure 4.7 superimposes monthly minimum temperature (line) on average monthly cloud cover (bars) from August 1985 to February 1988.

Geography and Vegetation

The landscape of Cuyo Cuyo is high, steep, and heavily dissected.¹²Data on the distribution of lands by slope in Puna Ayllu and Ura Ayllu are presented in Table 4.6. In Ura Ayllu 73% of the land falls on slopes of between 20 and 45 degrees. In Puna Ayllu, 45% of lands occur on slopes less than 5 degrees. This figure underscores the vast pampa encompassed by the Puna Ayllu territory. In contrast, Ura Ayllu's territory is confined to the Cuyo Cuyo Valley slopes. Virtually the only flat land within Ura Ayllu is the river floodplain.

The steep terrain of Cuyo Cuyo inevitably gives rise to a landscape characterized by diverse aspect. Table 4.7 presents data on aspect classes (in 45 degree increments) for the study communities. Puna Ayllu lands are fairly evenly distributed across all aspects, with some bias towards north-, west-, and northwest-facing slopes. This diversity illustrates the spread of Puna Ayllu lands across the western wall of the Cuyo Cuyo Valley and the eastern wall of the Awi Awi Valley. In contrast, the lands of Ura Ayllu are spread linearly along the eastern side of the Cuyo Cuyo Valley. In this community, 86% of all land faces either north, west, or northwest. Differences in aspect are significant especially in altering receipt of solar radiation, thus modifying growing climate for crops.

These geographic features of Cuyo Cuyo--extreme contrasts and diversity in elevation, slope, and aspect--give rise to parallel diversity in plant communities, further altered by human influence.¹³ Table 4.8 presents data on landforms and associated plant communities in the region. Within the Cuyo Cuyo Valley, puna vegetation occurs in small areas above each community at elevations of 4200 m and greater. The characteristic elements include cushion-plant life forms and bunch grasses, especially Stipa and Festuca (Bennett, n.d.). Grass-shrub communities predominate on three of the landforms defined by Bennett: exposed ridges, mountain slopes, and fallowed and abandoned terraces. Grasses and herbs predominate on slopes which have been heavily grazed; shrubs grow on steeper and thus more inaccessible slopes. In a glacial valley which lies above Ura Ayllu, plants are similar to those which occur

(page 79)

in temperate alpine regions. Dominant families include Asteraceae, Gentianaceae, Juncaceae, Poaceae, and Violaceae (ibid.).

On the floodplain terraces near Ura Ayllu, weedy species dominate, as they do along the disturbed habitat of roadsides. Along river banks a few remnants of once extensive stream side forests survive, but most of the riparian vegetation has been destroyed. There are a few ponds in the region as well. Associated plants are aquatics, including Scirpus (tortura) and Juncus (ibid.). Bennett (ibid.) defines a rock outcrop plant community in which bromeliads dominate. The stone-faced terraced walls of Cuyo Cuyo share some plant species with the rock outcrops. On both, ferns and begonias grow. Finally, Eucalyptus trees have been introduced in various spots throughout the Cuyo Cuyo Valley.

STRATEGIES FOR COPING WITH VARIABILITY

In preceding chapters, common risk buffering mechanisms in the Andes and elsewhere were reviewed. These include transmission of knowledge, food storage, mobility, exchange, and diversification. In Cuyo Cuyo, all of these practices are found, though not all are intended to cope with variability, nor are all equally effective in this task. Diversification is the most important strategy for reducing risk in Cuyo Cuyo, and is practiced in the communities of Puna Ayllu and Ura Ayllu.

Tales of Famine

Older Cuyo Cuyeños (60 and older) recount a story of famine and plague told to them by their parents, who experienced it as youngsters.¹⁴ According to this tale, the famine (hambruna [Sp.]; muchuy [Q.]) followed a year of especially good production. During or before the harvest of that year, the human population was hit with a plague (peste), and many people sickened and died within days. Because of the disease, they say, crops were left in the field; people hadn't the strength to harvest the crop or carry it back home. Some who remained healthy and attempted to steal the remaining crops were struck down with the disease and died quickly. In another version of this tale, under super-abundance, people were careless with the crop, and cavalierly left much of it in the fields. This angered the apus, and in response they brought the plague.

Whether the result of epidemic disease or failure of human responsibility, this initial calamity (plague) was followed by three successive years of crop failure due to drought. In the first year, stores of food saw most of the population through. In the second year, the stockpiles ran out for all but the richest families (ricachones). The poor and hungry begged for work from the ricachones. At first they were paid a q'ullu (a soup-sized bowl) of food for a day's work; later they were given only a "chua pukana" (about the size of a jigger).¹⁵In the second year of the famine, when only the richest still had seed with which to plant their fields, people entered after the planting and dug up the seed to eat. Not even the richest had food to give in payment for other types of work (weaving, terrace repairs); people began to die of hunger.

(page 80)

In the third year of the famine, people scoured the hillsides for edible wild foods: a grass that could be soaked and pulverized to create a gruel; the aquatic plant murmuntu,¹⁶processed like ch'uño and eaten; a wild root like racacha called kapaso, growing at high elevation; papa corika (a semi-wild potato); a plant called haq'aqo which grows near Sandia. One man recounted that eventually people ate their doors, which were made from the hides of animals, and that people began to steal each other's doors for this reason. Some families left altogether, descending to Sandia and below in search of work and food.

Whatever the historical accuracy or inaccuracies of this tale, it provides important information about appropriate behaviors and response to stress. The plague which initiates the cycle of hunger befalls Cuyo Cuyo because of acts which angered the gods (apus). People were wasteful of the crop (leaving it in the field), or acted in socially inappropriate ways (stealing from fields), and were ravaged by a plague.¹⁷Further information imparted in this tale concerns alternative food sources (where to find them, how to process them), and alternative strategies to gain food (activities to be bartered in exchange for food, migration to areas that may not be affected). The story also highlights the importance of stored foods and seed.

The famine as such was attributed to drought. Although not part of the story recounted above, other Cuyo Cuyeños told me how in times of drought the community gathers to waraq'ar: playing music on traditional panpipes and beating drums, carrying long steel bars, members of the community dance their way to lakes above the Cuyo Cuyo Valley. There they dislodge large rocks and send them tumbling into the lake. The sprays of water which jump from the lake are said to form into clouds and bring rain to Cuyo Cuyo.¹⁸

Risk Buffering Strategies

As described in Chapter 3, food processing and storage were dominant elements of prehispanic Andean economies, and its importance has been described in several contemporary settings (Rhoades et al. 1988; Werge 1979). I expected to find the same in Cuyo Cuyo.¹⁹In fact, processing tubers into freeze-dried products is prevalent in the study communities. Potatoes are processed into ch'uño and tunta; oca are processed into kaya. However, I found that in most cases, stores of both fresh and processed food barely lasted until the following harvest; in some cases, they were heavily depleted months before the harvest (see also Graham 1991). This was not due to unusually poor production, since by all accounts the study years were quite bountiful.

(page 81)

In Table 4.9 I present aggregate data for each community on the destination of production for potatoes and oca. In 1986-87, 45.4% of the Ura Ayllu oca crop was processed, versus 15.1% of its potato crop. The high percentage of processed oca reflects in part the poor quality of the harvest that year: tubers were small and worm-infested, and thus unsuitable for consumption as wayk'u (boiled whole tubers, the most common way to prepare oca). In Puna Ayllu, 44.7% of the oca crop was processed, versus 34.7% of the potatoes. I do not believe that this proportion of processed oca was unusual in Puna Ayllu. Puna Aylleños grow a special class of oca in one of their higher production zones which they find to be bitter in taste (but hardier to the cold conditions and high altitude of this zone); a very high percentage of this is made into kaya.

I conducted a more in-depth study of the stores of four families in each community when I arrived in Cuyo Cuyo, shortly after completion of the 1986 harvest. I measured the amount of fresh and processed foods each had in storage. Later, I followed up by inquiring if and when they had depleted those stores. Five of the eight families had run out of stored food by the following November, several months before the main harvests begin.²⁰

The implication of this information is that in Cuyo Cuyo food storage cannot act as an effective risk buffering strategy on a year-to-year basis, because production is not adequate to permit accumulation of a large surplus. It is impossible to lay away stores sufficient to see a family through one year of crop failure, let alone several successive years. Why do people then process such a large portion of the crop, if it will be consumed so quickly? I believe that there are several possible responses. First, as noted above, poor quality tubers encourage processing. Tubers which are very small, diseased, or worm infested are likely to be processed rather than consumed fresh. Second, in the humid environment of Cuyo Cuyo, losses due to rotting in storage may be large. Processing, then, is important for food preservation, even on a time-scale of half a year or so. Finally, processing serves to add diversity to a somewhat monotonous diet, by altering flavor and texture. Graham's (1987) study of household diet and nutrition shows that processed foods are mixed into the diet throughout the year while available, not only during periods when supplies of fresh foods are scarce.

In contrast to storage, intra-community exchange is a more effective risk buffering behavior.²¹Obligations of reciprocity, strengthened by ties of kinship and compadrazgo, pervade the everyday life of Cuyo Cuyeños. Most exchanges are made in the context of ayni, which carries the promise of a delayed equivalent return. Ayni is particularly important in labor exchanges: a day's worth of work is repaid at a later date in kind. But labor is also a means to earn food. Comuneros who hire themselves out for the day can usually (but not always) expect to be fed at least one meal, in addition to receiving coca during breaks and

(page 83)

monetary remuneration.²²Women will hire themselves out as weavers. In this case, they weave in the home of their client, receiving food during the daytime spent there, in addition to a cash payment.

During the harvest, jornaleros (day wage-laborers) receive a portion of the crop in return for their work (see column for "payment" in Table 4.9). The measure used for payment in produce for work in the harvest is a hara, equivalent to one to one-and-a-half arrobas (25 - 38 lbs.).²³Another way to earn produce during the harvest is to rent out pack animals to transport it. In Puna Ayllu, herders descend with llamas from the pasture zone at Limatapampa to carry the harvest from distant fields. For every ten loads (cargas)²⁴they transport, they keep one.

As discussed in Chapter 2, it can be difficult to completely separate the various risk buffering strategies. In Cuyo Cuyo, this is seen clearly in consideration of mobility and diversification. The primary pattern of mobility is in the seasonal migration to mine gold. As risk buffering, however, I consider this activity more properly a form of diversification.

DIVERSIFICATION IN CUYO CUYO

There are three primary forms of diversification in Cuyo Cuyo. First, economic opportunities are diversified between income-earning activities (usually off-farm) and subsistence production. In both study communities, the primary off-farm activity is gold mining during the rainy season. Puna Aylleños communally own a placer mine in Ancoccala, on Limatapampa. In Ura Ayllu, men migrate to Puerto Maldonado where some have claims, and others work under various arrangements (Recharte 1990). In the realm of the subsistence economy, the primary forms of diversification are maintenance of a highly varied array of crops and crop varieties, and the dispersion of agricultural fields.

Income-earning Activities

During the rainy season, the main community of Puna Ayllu empties except for a few of the elderly individuals left behind as household caretakers. The comuneros of Puna Ayllu relocate themselves to Ancoccala, the placer gold mine located within its territory, some six hours walk from the Cuyo Cuyo Valley. Men are the first to depart, beginning in November, and women and children follow in December and January, once the fields have been weeded and the school year has ended. In Ancoccala each household has a claim and works a small piece of the mine, independently; however, all production is communally controlled and integrated by the technical requirements of water management (Recharte 1990).