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

By Carol Goland, 1993.

Chapter 6 - Inputs to Agriculture Labor and Its Scheduling

(page 211)

or 127 days/ha. For papa panq'o, planting is conducted much as it is in the estancia, with couples using a footplow.

WEEDING

Weeding is the most time consuming task in growing maize. Most fields are weeded twice (the terms llank'ay and hallmay are used here as well). Some are given a third weeding (termed masiy). In total the maizales required 264 days for these various weedings, or 227 days/ha.

For the papa panq'o fields, the weedings and hillings are given the same names as for fields of wata papa (k'ujray and llank'ay). This chore is the most labor intense of all tasks associated with this crop. The three papa panq'o fields required a total of 12 worker-days of labor, or 276 days per hectare for weeding.

HARVEST

The corn fields are the last to be harvested. Whether they are located in Aripo, or somewhere between Aripo and the community of Ura Ayllu, the harvest may begin in May and continue well into June. Toward the end of this period the papa panq'o fields are also harvested. Thus by the time all of the harvests are completed in Ura Ayllu--the end of June--the planting of papa milli in the lowest elevation fields of the community is only two weeks away. Agricultural activities in Ura Ayllu last nearly year around.

The main harvest of the maizales in June is fairly simple. The ears are simply snapped off the stalk and carried home. The stalks are left in the field for grazing animals, or they are cut at the base and used to feed to the cuys (guinea pigs). In 1986-87, this work consumed 48 days of labor for Ura Aylleños, or 59 days per hectare. For papa panq'o fields, harvested at about the same time, the three fields required 9 days of work, or 176 days/ha.

One of the key features of the non-estancia lands of Ura Ayllu--especially the papa milli fields--is that they may be used opportunistically for small harvests prior to the major one. Maize fields are also subject to an ankacha, albeit a very special one. The timing of the ankacha de maiz is set by Holy Week and Easter. On Good Friday one can enter someone else's corn field and remove the ripening ears without penalty because to do so is "to liberate Christ." Theft of corn, whether of this or more secular inspiration, is a common problem. When I accompanied Abraham on his ankacha de maiz (on the day before Good Friday), he made a point of covering the residue of our ankacha (uprooted stalks, etc.) with weeds and sods. This was a sign to his wife that it was he rather than a thief who had been in the chakra.

The ankacha of corn provides Ura Aylleños with a special culinary treat. When I met Abraham on the road the day before he and his family were to go to Aripo for the ankacha, he was insistent that I cancel my plans for the day and go with them. We left Ura Ayllu at 5:00 in the morning. The walk to Aripo took about two hours. As soon as we arrived at their house in Aripo (inherited from Abraham's grandfather), he went to the nearest maize field and brought back some 15 to 20 stalks of corn. We immediately removed the ears and these were boiled and served with fresh cheese and coffee for breakfast.

After breakfast we split up. Sra. Elisa and a son who accompanied us went to one set of fields, while Abraham and I went to others. Abraham and I descended to the lower of their maize fields, located in an area called Yaqaqori. The path to Yaqaqori was steep and was slippery with the dew of the morning. Sra. Elisa specifically asked me if I wasn't going to be

(page 212)

afraid to walk on the path, and she clearly did not want to be the one to go to the Yaqaqori fields, indicating that her p'istu (home-spun skirt) would get all wet.

As we entered the fields, Abraham explained that he chooses the best ears for the ankacha. These are the smallest ears (ones which are apparently "stunted" in growth but otherwise among the most mature, based on the complete formation of the kernels). The mechanics of the harvest were simple: we either uprooted whole stalks or cut them at the base with a machete. After we had cut a pile we removed the ears (usually each stalk produced only one ear). We generally removed all but a few of the layers of the husk from around the cob, although with some of the larger ears we left all of the husk (the reason to be explained shortly). Portions of the stalks which had not yet dried out were set aside. These are wirus, the core of which are a sweet and special treat.

Abraham and I gathered about 85 ears of corn, while the others returned to the house with about 50. Once we were gathered, Abraham and I set about removing the kernels while his son worked on grinding these in a hand-cranked grinder. It produced a thick paste. From the larger ears which we had brought back with the husks intact, Abraham began to remove the husks by cutting them around the base of the cob. This preserved large sheets of husks. The paste was spread into these and then folded. We were making humintas, not unlike tamales, which are more familiar in North America.

The raw humintas were baked atop a stone slab resting on two uprights between which a fire blazed. An additional stone slab was placed on top of the humintas. When we ran out of husks in which to bake the humintas, the remaining corn meal paste was used to make mazamorra. Mazamorra is a thin porridge with slices of potatoes thrown in. We feasted on this and the fresh humintas for lunch.

We departed from Aripo at about 4:00 in the afternoon, loaded down with humintas, wirus, and the ears of corn that hadn't been used. Each of us had a wiru as a walking stick. I had the sense that we carried these as staffs to announce to the community and those we met on the path that we had made our maize ankacha. When we rested briefly along the way, we sucked the sweet moisture from wirus. The family presented me with a substantial quantity of humintas before I left them to return to my own home. In the days that followed, I found that many others sought me out to give me gifts of humintas, ears of corn, and wirus.

Abraham commented to me that now he could return to the jungle contentedly, because he had tasted choqllos. Indeed, he had made the trip to Ura Ayllu from the jungle below Sandia (where he engages in gold-mining and gathers incense) just for this occasion. The ankacha de maiz is not so much a religious holiday (despite its coincident timing with Easter) as it is something like American Thanksgiving: a secular event consisting primarily of harvest celebration, festive food preparation, and eating.

Comparison of Labor in the Transitional Fields and Maizales

Although the calculation of labor intensities for the papa panq'o fields of the maizales zone is suspect due to problems with area measurement for one field, comparison of these fields to panq'o fields in the transitional zone may still be worthwhile. Field preparations in the panq'o fields of the maizales are much reduced when compared to those in the transitional zone. On the other hand, labor intensities for planting and weeding in the two zones are roughly similar. Labor for harvest is reduced in the maizales zone. In sum, total labor intensity for papa panq'o production in the maizales zone is considerably less than that for the

(page 213)

transitional zone, 754 days/ha versus 1031 days/ha. Again, due to suspect field area measurements, it is difficult to assess the significance of this information.

The sample of maize fields in the transitional zone consists of only two examples. This also makes comparing labor between this production zone and the maizales difficult. Maize production in the transitional fields is less time consuming per unit area than maize production in the maizales (296 days/ha vs. 392) (Figure 6.4). In both areas, the most labor demanding task associated with maize production is weeding (371 days/ha in the transitional fields; 227 days/ha in the maizales). In fact, in both cases, weeding represents between 65% and 70% of all labor inputs to these fields (Table 6.15). In general terms, the labor intensity of maize production in these two lower elevation zones is similar to that for habas production in the estancia manda. In both places, tuber production (potatoes) is more time demanding.

SUMMARY - LABOR AND LABOR INTENSITY IN CUYO CUYO

Comparison among the Different Zones and Crops of Cuyo Cuyo

I begin with the last point made: growing tubers in Cuyo Cuyo appears to take much more time than growing grains. Note that this is not an assessment of net production. Tuber production may take longer, but the net returns to labor may be relatively high. This is a topic which will be closely examined in Chapter 8.

Another factor impedes comparison of labor intensities for various crops. Field preparation prior to planting has been allocated ("charged to") the first crop planted after fallow, typically potatoes. But there is nothing inherent in potato production that requires this extra work effort. Any crop which broke the fallow would appear to be especially time demanding. It may be more accurate to view field preparation after fallow as an investment to be amortized over all of the subsequent crops and cropping years. Alternatively, rather than distributing it in equal parts to each crop in the rotation we can ignore it entirely. This will produce a more accurate comparison of labor intensities across the various crops.

Table 6.14 presents these adjusted data. This table shows the average labor intensity for each community alone, and for both combined, for each of the major crops (potatoes, oca, habas, and maize). The figures for potatoes are presented in both unadjusted form (total labor, including field preparations as defined elsewhere in this chapter) and in adjusted form (excluding fertilizing, plowing, and k'upay labors). It is readily apparent that when field preparations are excluded, potato production is not as differentially labor intensive as first thought. For both communities combined, the adjusted intensity of potato production per hectare (591 days) is actually less than that for oca production (645 days/ha). In contrast, labor intensity for grain production (habas and maize) is considerably lower than for tubers, and quite similar to each other (386 days/ha for habas; 364 days/ha for maize). As noted above, however, the acid test for the significance of these figures will come when we convert the labor inputs to calories, and compare these to the caloric production from these fields to determine net returns to labor and land.

For each zone of production, weeding is comparatively time consuming. For many crops (in most zones of production) weeding is the most labor demanding task (Table 6.16). This is especially true for the grain crops. In Table 6.15, the amount of time invested in each task is presented as a percentage of all days worked in each field type. Again, the same pattern holds. Thus, for maize and beans, weeding is the activity that consumes the majority of the

(page 217)

time spent in production of these crops, although the intensity for all labors combined is less than that for tubers.

Even for tubers, the duration of work and percentage of all work time that weeding represents is large compared with other tasks. In the estancia mandas of both communities, weeding as a percentage of all labors increases with each successive year of the rotation: from 24% of all labor time for the potato crop to over 60% by the fourth year of the rotation. During the first rotation year in Ura Ayllu, and both first and second year of the rotation in Puna Ayllu, harvesting consumes the greatest percentage of all work time; by the third year, this relationship has reversed, so that in the third and fourth year, weeding requires the greatest percentage of all labor time. Of course, this may say more about the ease of harvest in these years than it necessarily does about difficulty or intensity of weeding. Weeding does not increase in intensity per hectare as the years progress, it simply fails to drop as sharply as the intensity of labor in harvesting.

Outside of the estancia mandas it is possible to compare labors in tuber production between a first year potato crop and a second year oca crop only in Paqhchani and the Awi Awi manda. The same pattern holds. Per hectare, the labor intensity of weeding for the oca crop is greater: 165 days/ha for potatoes versus 238 days/ha for oca in Paqhchani; 63 days/ha for potatoes versus 170 days/ha for oca in Awi Awi. And, as a percentage of all labors required for each crop, the second year also demands more weeding. In these other production zones, the oca crop requires more weeding time and intensity than does the potato crop.

Despite this, within the same crop (potato, oca), intensity of weeding drops with distance from the home community: the estancia fields of each crop are weeded more intensively than distant fields (Awi Awi fields of Puna Ayllu or the Ñacoreque fields of Ura Ayllu). Most of the Awi Awi potato fields receive only one weeding, and although the oca crop there generally is weeded twice, the intensity of labor required per hectare is much diminished from the estancia.

Of the Ura Ayllu tuber fields receiving only one aporque, both distance and crop seem to be relevant. In both years, the fields most likely to have only one aporque are those which are distant from Ura Ayllu, in Aripo or between the two communities. For example, in 1986-87, of the twelve potato fields in Ñacoreque, only one received both k'ujray and llank'ay, while the remaining fields had a single aporque. Distance seems to be the relevant factor here, and in this sense, the treatment of these distant potato fields parallels that of the Awi Awi potato fields of Puna Ayllu. The high incidence of single aporques for papa milli fields may be partially explained by distance, but another factor also may enter. Since these fields are planted prior to the beginning of the rainy season, and harvested well before its termination, it may be that the climatic conditions which favor early weed growth in other fields are not present. Thus there is reduced need for aporque.

On the other hand, crop type also enters into the equation. Despite the travel time required to visit the maizales, these are the fields with the highest incidence of three aporques (and intensity of weeding labor for the transitional maize fields is the highest of all production zones and crops). As in the estancia manda, weeding of these grain fields is the most time-demanding activity of production in this zone, be it calculated in terms of intensity per hectare or as a percentage of all days worked in the field.

Comparisons with Other Andean Communities

(page 218)

The amount of agricultural labor per unit of land area in Cuyo Cuyo is quite high. This is expected, given the review of Andean agroecology presented in Chapter 3. The figures for labor intensity in Cuyo Cuyo show the expected patterns when compared between crops and between different zones of production. But how do these findings in Cuyo Cuyo relate to what we know about other places and peoples of the Andes?

Unfortunately there is not much comparative material available. What information exists is quite different from the measurements of labor investment in Cuyo Cuyo. Table 6.17 presents the findings of this chapter and similar figures derived from Brush (1977a, 1977c) for the Peruvian community of Uchucmarca (Dept. of Cajamarca, Peru) and Knapp (1984) for the Ecuadorian Andes.

Estimates of labor effort (per unit land) in Cuyo Cuyo agriculture for most crops are three to six times higher than in Uchucmarca. The disparity for growing potatoes is even greater. All of the crops except broad beans require substantially less labor per hectare in Uchucmarca than in Cuyo Cuyo. The difference between the Cuyo Cuyo data and Knapp's data for Ecuador is equally great, although Knapp presents data for only the potato and the maize crops.

Knapp does not provide information on the number of hectares cropped by the average household or person in his study. Therefore I cannot calculate what the total labor requirements for a year are. Brush calculates that the total number of days in the year that the average head of household must spend in agricultural activity is 131. Summing his data on average landholdings for each crop, we find that a total of 1.58 hectares are sown by the average Uchucmarcan family. In other words, about 83 days per year are spent working on each hectare of land. In Cuyo Cuyo, the number of days spent working per hectare is much higher, even though the total number of days spent working in agricultural tasks per year is lower. Assuming only the husband and wife work in the fields (which may be a questionable assumption as children do make contributions to agricultural labors), in Ura Ayllu the average individual spends about 99 days of the year engaged in the full-time equivalent agriculture, while in Puna Ayllu the average is 77 days of the year.

In Uchucmarca people have about three times as much land as do people in Cuyo Cuyo. Cuyo Cuyeños work their .35 hectare (Puna Ayllu) or .45 hectare (Ura Ayllu) three to six times more intensively (according to the crop and production zone) per unit of land. Whether or not this translates into increased yields--either per labor investment or per land unit--is examined in Chapter 8.

The relatively high demands of labor per land area in Cuyo Cuyo merits attention. For each of the cases presented here (Brush, Knapp, Cuyo Cuyo) there are two possible sources of error: the land estimates and the labor estimates. Brush (1977a:86) used a combination of seed planted per plot and air photo measurement to derive his estimates of average landholdings in each crop. This seems quite reasonable, though not infallible. He does not explicitly state whether these are total landholdings or the area cropped in a given year. Given a sectorial fallowing system, these can be quite different. It is less clear how Knapp derives his estimates for work per hectare. In the present study, almost all fields were measured on the ground; only a small fraction of field areas were derived either by informant report (n = 4, or 1%) or seed:land ratios (n = 17, or 4.8%). Inaccuracies in the measurements--especially on steep slopes--may occur, but by and large the errors should be relatively minor.

(page 220)

The second source of potential error is the reporting of labor time. All three studies (Brush, Knapp, and Cuyo Cuyo) relied on informant reports of labor expenditures. Brush states that he used a sample of 42 households, which were interviewed about their labor expenditures in 1971. Knapp too apparently used interviews to establish work expenditures. It is not clear whether his informants provided information on work effort on a per hectare basis, or if, like Brush and this study, final figures are the result of adjusting reported labor investments by landholdings. As described in the introduction to this chapter, in Cuyo Cuyo families recorded and were interviewed on an on-going basis throughout the agricultural year concerning their labor inputs to each field under cultivation.

An independent observation helps in evaluating the quality of the labor figures for Cuyo Cuyo. Over a two year period, the PSE project collected time allocation data for the twenty study families, using the random spot-check methodology (Winterhalder et al. n.d.). Every six days families were visited and their activities observed at the moment of contact (or the location and activity of absent members queried and later checked where possible). These data are quite reliable and provide an independent check on the informant-reported figures for agricultural labor used here.

The labor reported in this chapter amounts to a total of 1972 days of work in Ura Ayllu and 1549 days of work in Puna Ayllu (Table 6.2). If we make the simplifying assumption that this represents the work of the husbands and wives of each household, then per adult, the total number of days worked in Ura Ayllu was 99 days per year, and in Puna Ayllu was 77 days. As a percentage of all days of the year, this represents 27% for Ura Ayllu and 21% in Puna Ayllu. When values for adult labor in agriculture are calculated from the time allocation data collected by the PSE project, the concordance is striking. According to that independent data set (not based on informant report) Ura Aylleño adults spend 26.3% of their time on agricultural activities; Puna Aylleño adults spend 20.3% (Winterhalder et al. n.d.). The figures provided by informants and those independently measured by the time allocation observations are nearly identical. This correspondence adds confidence to the informant-reported data used here. Moreover, if we consider the ratio of work for Ura Ayllu and Puna Ayllu based on informants' reports (56 : 44), for the independent PSE time allocation data set (56.4 : 43.6), and for the total land in each of the communities in 1986-87 (54.4 : 45.6) there is additional reason to believe that the informant reported data are accurate.

Altitude, Distance, and Agricultural Intensification

The final set of observations I wish to make concern the issue of intensification of land use with decreasing altitude. The data from Cuyo Cuyo are appropriate for examining this issue since the zones of production in the two communities span the altitudinal gradient from above 4000 m down to 2700 m. Intensification has been discussed in several ways in the Andean literature. A focus on the duration/length of fallow is the most common criterion (Godoy 1984; Orlove and Godoy 1987). Other measures of intensification include the use of greater or more costly inputs (Guillet 1987), and degree of communal control (Mayer 1985; Yamamoto 1975).

When agricultural intensity is measured as land use intensity (following Orlove and Godoy [1987], expressed as the ratio of fallow years to total years in the cycle) then it is clear that in Cuyo Cuyo intensity is positively correlated with altitude generally.¹¹ On a gross level,

(page 221)

the production zones organized in a sectorial fallowing system have higher fallow ratios than do the lowest-lying fields in Ura Ayllu, located in an ecologically distinct zone and used for producing different crops. Among the various manda systems in the two communities, the relation between elevation and fallow ratio is not entirely consistent. In the highest elevation lands examined here--the Awi Awi manda--fallow for five years is followed by two years of cropping use. In both the estancia mandas, fallow is two years in length, while fields are cropped for four years. The Paqhchani zone is a bit anomalous: fields are fallowed for four years, and cropped for three. The manda of Ñacoreque, where Ura Aylleños rent land, is cropped for one year and allowed to fallow for five. These lands are at elevations from 3500 m to 3800 m,¹² and so are comparable to the estancia zone.

By the criterion of degree of communal control, the relationship between intensification and elevation in Cuyo Cuyo is clearly inverse, as would be expected. The highest elevation lands (the Awi Awi and estancia mandas) are those in which community regulation of land use is strongest. Both timing and crop choice in these zones are highly coordinated. In contrast, in the Paqhchani fields there is a much lower uniformity in land use (see discussion in Chapter 5). Continuing downvalley, in the transitional zone and the maizales community coordination of timing and cropping uses is at a minimum.

The final criterion is that of labor intensity. In Cuyo Cuyo it is quite clear that labor intensity (the labor required per hectare of cultivation) does not uniformly increase with decreasing altitude. Labor inputs are greatest in the estancia mandas of the two communities. In both lower-lying zones (Paqhchani, transitional fields, maizales, etc.), and higher altitude zones (Awi Awi), labor inputs per hectare are reduced for most of the crops that are grown in more than one zone.

Relative elevation cannot account for the differences in labor intensity seen in Cuyo Cuyo. A more promising explanation lies in simple measures of distance between home and field. Geographers have been most active in exploring the relationship between distance, intensity of inputs, and yields in agricultural production systems. They consistently find that distance is inversely related to inputs and yields: the greater the distance, the lower the inputs and the lower the yields (both gross and net) (Blakie 1971a, 1971b; Chisolm 1982; Richardson, 1974). In this chapter, it has been shown that intensity of labor varies in this manner with distance.

In the following chapter, I continue my examination of agricultural inputs by focussing on the use of fertilizers (net yield is examined later). Distinctive styles of fertilization--both quantities and specific types--augment the definition of production zones begun in this and the preceding chapter.