A short story with some musings on human-groundwater relations along the coast of Peru.

By Carolina Domínguez [1], Maria Teresa Ore [2] and Andres Verzijl [3]

In memory of David Bayer

One of the first times when we, the authors, met, and the first time we shared a long conversation, food and drinks, was at David’s house in December of 2009. David was a retired development expert who continued, tirelessly, to defend the position of Peruvian smallholders. His house was part of a semi-gated community in the Ica Valley that was circumvented by sand dunes. Like all valleys on the coast of Peru, the climate in Ica is arid. Temperatures can rise to 38 degrees and average rainfall is less than 150 mm – except during El Niño years when precipitation can be 25 times higher and usually cause massive flooding. Normally, when rain falls - there are also years without any rain – it is over a short period between December and March. Consequently, most rivers carry water for only a couple of months. Farmers have a saying: if it has not rained by Saint Joseph Day (March 19), it won’t at all. Either way, agriculture on the coast is irrigation-depended. It has been for 3000 years.

Sketch 1: Family farm with pozas in Motupe (Credit: Carolina Dominguez)

Early chroniclers documented the immense indigenous irrigation systems on the coast and two key infrastructural features: large earthen (inter-valley) canals and permanently-diked plots, which today are referred to as pozas (see sketch 1). Decay followed shortly after the Spanish conquest, and it would take until the second half of the 20th century to surpass the pre-Columbian irrigated area (Vos, 2002). Yet, these trunk canals and pozas continued to be used until today. Over time, they amazed and intrigued many who traveled this desert land.

One traveler to the Ica Valley observed that “all fields are inside boxes made of the same earth… one vara (83 cm) in depth. The seasonal rains from the mountains fill these boxes that contain up to 2000 plants. They [farmers] leave water bounded. Until harvest, they irrigate once…without fear, for the yield will undoubtedly be abundant” (Vicuña Mackenna, 1860).

In 1894, German anthropologist Ernst Middendorf wrote about “earthen walls, two feet (61 cm) high, circumventing a plot…which they call poza” and which receive water only twice a year. He continues: “After 10 days, when the water has been absorbed by the ground, it is irrigated again and then the farms no longer receive water throughout the year” (Middendorf, 1973).

These accounts can make you think: how is it possible to harvest in a desert with only one irrigation turn; or why would a farmer irrigate abundantly twice in ten days? While engineers and water managers in Peru have often referred to poza-irrigation as archaic, backward and hugely inefficient in comparison to, for example, drip irrigation, we were curious about those who used these systems and what their reasons were for doing so. After all, tens of thousands of smallholders continue to rely on their pozas up to today.

It was David who took two of the authors, Carolina and Andres, to see pozas for the first time. They went to Occucaje (see photo 1). Here pozas were big, up to several soccer fields; and deep, well over one meter. This was the tail-end of the Ica Valley. It received water after everyone else, sporadically, and only in years of abundance. Back at David’s, we discussed its size and depth - and figured it was to take in as much water as possible when it was available to replenish local wells.

Photo 1: Four filled pozas in Ica/Occucaje in 2008 (Credit: David Bayer)

Maria Teresa’s first encounter with pozas dates back to the mid-nineties. She travelled to the Ica Valley to investigate the legend about the construction of a large pre-Hispanic canal, called La Achirana. Its origins are attributed to the Inca Pachacútec, and it is still in use today (see also Ore 2005). One smallholder, Alejandro Tipismana, who received water from the canal, told her back then:

“We filled our pozas to have humidity last us four to five months, until we harvest. When the year was good, we could fill twice, when the year was bad there was barely enough for one irrigation”. In general, “people harvested good field beans, good grapes, good chickpeas. Everything came out good. We had enough to live”.

Another small farmer, Roque Velázques, explained that irrigation rotation was from head to tail-end: “Once the [higher] poza was filled … water was passed to one’s neighbour. That is how we irrigated, no water was wasted and there was union”.

David’s house was a kind of home-base for visiting researchers to share these kinds of stories and exchange information about Ica’s water problems. He and his wife, Rosa, were hospitable and helped with connections and the latest information and insights. Rosa’s family had pozas of their own, and David kept himself involved in local water policy debates and smallholder complaints.

One of the water problems in Ica involved the booming agro-export sector, which grew rapidly since the end of the 1990s. The Peruvian government had adopted an economic model that encouraged large foreign companies to colonize the desert areas that bordered the valley, extending into the plains of Villacuri (see map 1). Through the use of borewells and sophisticated irrigation technologies, the companies, with vast estates, started exploiting the aquifer underneath, often for high water-consuming cash crops, like asparagus and avocado (Ore, Bayer et al 2014). By 2007 groundwater levels in the Ica Valley had dropped dramatically, leading to a looming water crisis. In relation, the large companies created a groundwater user association, called JUASVI, and demanded corresponding recognition and privileges from the government, which eventually gave them the ability to claim state funds and hydraulic improvement projects to recharge the aquifer (Damonte and Gonzales 2014).

Map 1: The Ica Watershed (Credit: Project Estate and Scarcity. Alejandra Cuentas (PUCP) and (GIZ) 2017)

The stories farmers told Maria Teresa were about a time before the arrival of agro-export companies in the Ica Valley; A time when, according to farmer Alejandro Tipismana, “with pozas humidity was kept and with a single irrigation turn [farmers] were able to harvest”. However, a declining water table and increased competition complicated life of smallholder farmers. While they did not irrigate with groundwater directly, they complained that the water that was bounded by their pozas infiltrated much quicker (due to lower groundwater levels). Therefore, one or two irrigation turns no longer assured abundant harvest. Many looked for livelihood opportunities outside of agriculture because they did not want to risk crop failure. Many pozas remained empty.

David was the first to promote the use of these pozas to recharge the aquifer (see Llosa, n.d.). Even when not cultivating anything, smallholders should be motivated and compensated for filling their pozas in times of abundance (Bayer 2015). He shared his concerns and ideas for action in his typical David-way by e-mailing them to a very long list of professionals, scientists and policymakers; and update this every time he got responses or feedback. He was the first, to our knowledge, to question the efficiency of drip irrigation over that of pozas, stating (in his e-mails) that water that infiltrates in Ica is not lost. In contrast, drip technology, which threatens Ica valley ecology, should not be considered more productive. It was his voice against those of agribusinesses who insisted their knowledge and technology was the best way to save water: “not a single drop is lost”. And others should follow their example. David’s initiative never really took off, but it inspired us and influenced further developments.

Carolina and Andres went on to study agro-export practices on the North coast of Peru. In the Motupe valley, to be more precise, which is one of the most important sites of mango production in the country. This ethnographical research allowed us to explore further the questions about pozas and their use (Dominguez et al. 2017). There are distinct differences with Ica. Both the valley and agro-companies are smaller in terms of size, aquifer exploitation and political pull. Another significant distinction with Ica is that Motupe smallholder farmers are engaged in agro-export.

In the early 1990s, new mango cultivars were introduced in Motupe by large farmers, which smallholders could easily copy to their own farms. Mango trees resist drought and waterlogging better than most other cash crops and proved advantageous to small-scale farming in this valley. Local leaders and NGOs have worked with smallholders on how to commercialize their produce and set up cooperatives. Today, their pozas with mango trees make up the landscape, interspersed with several large company-owned estates (see map 2). We talked to one of the large company owners, Elias Baca. He took us on a tour of the estate where borewells and drip technology were used to irrigate thousands of small trees that were two, three meters in height. Refuting pozas, Elias Baca said he prefers to keep his estate clean:

“We don’t dirty the farm, there is no need to weed, it reduces labor costs, and it [drip irrigation] is more precise … when you flood your field [using pozas], the trees do not benefit from all the water that you put in, the rest goes to the ‘camellón’ (the space underneath the roots) and is lost” (Dominguez, forthcoming).

Map 2: Motupe area (Credit: Dominguez et al 2017)

Bordering the estate is the small farm of the Manayay family. They don’t have a borewell but used an open well for domestic purposes. To irrigate their tall mango trees, about 8-10 meters in height, they relied exclusively on pozas (see sketch 1; photo 2). Motupe pozas are not as deep as in Ica: about 30 centimeters (see also Hatch 1974). When asked about the possibility of irrigating his mango trees with a drip system, Octavio, who ran the Manayay farm, reacted with another question: “What would happen to the roots of the trees?” He is convinced that pozas are needed to give water in abundance and is aware of connections between the deep mango taproot, the groundwater level, which he sees in his open well, and his irrigation method. He irrigates only three times a year and his harvests are plentiful. When talking about pozas and groundwater, Octavio mentioned a place downstream from the village where water appears in the river while upstream the river runs dry. Other hamlets use this water because pozas stored it during floods, intensive rains or irrigation and slowly released it downstream. When we visited Octavio again a few years later, he mentioned a prolonged drought and spoke proudly of his trees. They had given him fruit while not receiving irrigation water for over 12 months.

Photo 2: Pozas in Motupe (Credit: C. Dominguez)

In Motupe, pozas allow farmers like Octavio to engage in the production mangos for agro-export and compete with large farms like those of Elias. But only when done collectively and jointly are groundwater levels maintained. Farmers debate this as professionals of drip technology companies are traveling the countryside trying to persuade smallholders to go for a more ‘efficient’ and modern technology where “no drop is lost”.

In Ica, there is a renewed valorization of these pre-Hispanic pozas and their recharge potential, both in public and private sectors. Due to today’s water crisis, even large companies create them on their lands. They buy lands of smallholders to utilize their pozas. The idea that David promoted first in 2008 is seemingly being picked up by the government officials and agribusinessmen that criticized him earlier. Even more so, through a program of the groundwater user association, JUASVI, smallholders can rent out their land (and pozas) and receive a compensation. In 2017 over 800 pozas, scattered across the valley, were estimated to contribute 17 MMC in recharge (Navarro and Fernandez, 2017). The downside, among others, is the permeation of large companies and JUASVI interests in matters of surface water user associations and distribution. They utilize canal infrastructure, obtain irrigation turns and shape decisions on rotation schedules which were, until recently, the domain of smallholder organization.

The JUASVI program also does not seem to stimulate smallholders to continue cultivating. When Maria Teresa visited the Ica Valley, early in 2020, numerous of smallholder pozas were filled, but without crops, while large companies (only) inundated part of their land. We see how agro-export companies are transforming the (concept of) pozas from the need of maintaining collective humidity to the need for more individual borewell pumping. Because of the scatteredness of inundated plots, the joined poza-system cannot be maintained. There is no union anymore. We are left with the question to what extent the idea of David is transformed or coopted by agro-exporters who keep smallholders from practicing irrigated agriculture.

The rehabilitation of poza knowledge and practices indeed comes at a crucial time for agro-export companies that, belatedly, realized the necessity of aquifer recharge. But it also raises questions about water sharing and the common good of a groundwater table and soil humidity amidst unequal power relations in the Ica Valley. Given the situation, we ask ourselves whether the potential benefit for both smallholder cultivation and an agro-export sector could work in the Ica Valley.

As we remember David, we imagine a long conversation at his house: a spirited talk with dried olives and pisco, where we share and exchange our experience and views about large companies, overexploitation, high-technology agriculture; and also, smallholders, cooperatives, connected pozas and favorable crops. We envision that later that night – they were usually sent at night – David would leave us to write an e-mail to all kinds of stakeholders with a plan that celebrates meaningful participation of smallholders. He would point out examples on the north coast and emphasize that a livable valley ecology, historically, started with them and it is from them that we should learn. But these e-mails are no longer going around. At present, we continue to look out for new advocates of the smallholder cause in the Ica Valley. To those persons also, we offer these musings.


Bayer, D. (2015). Las pozas de Ica como sistema de riego y recarga del acuífero. Leisa, 31(3), 21–23.

Damonte, G. and Gonzales, I. (2018). Politicas públicas, relaciones de poder e intereses en conflictos en la gestión de las aguas subterráneas en Ica. In Ore, M.T.and Muñoz, I. (eds.), Aguas en Disputa: Ica y Huancavelica entre el enfrentamiento y el diálogo (pp. 75-110). Lima: Fondo editorial de la PUCP y GIZ.

Dominguez Guzman, C. (forthcoming). Techno-diveristy: On Purification and Pozas of mango irrigation practices in Northern Peru.

Dominguez Guzmán, C., Verzijl, A., and Zwarteveen, M. (2017). Water Footprints and ‘ Pozas ’: Conversations about Practices and Knowledges of Water Efficiency. Water, 9(16).

Hatch, J.K. (1974). The Corn Farmers of Motupe: A Study of Traditional Farming Practices in Northern Coastal Peru. University of Wisconsin, Madison.

Middendorf, E.W. (1973) Peru,Observaciones generales y estudios del país y sus habitantes durante una permanencia de 25 años . Tomo II La Costa. Universidad Nacional Mayor de San Marcos, Lima

Navarro, R., and Fernandez, E. (2017). Implementacion de mas de 500 pozas de recarga artifical para la sostenibilidad del regadio en el acuifero de Ica (Peru). In O. Escolero, C. Gutierrez, & E. Mendoza (Eds.), Manejo de la recarga de acuiferos: un enfoque hacieLatinomaerica (pp. 565–626). Mexico: IMTA.

Ore, M.T. (2005). Agua Bien Común y Usos Privados . Riego, Estado y conflictos en La Achirana del Inca. Lima: Fondo Editorial de la PUCP, Universidad de Wageningen UR Netherlands, Soluciones Practicas. ITDG.

Ore M.T., Bayer, D., Chiong, J. and Rendón, E. (2014). Emergencia hídrica y conflictos por el agua  en una cuenca peruana : La cuenca del rio Ica. In  Guevara, A. and Verona, A. (eds.), El derecho frente a la crisis del agua en el Perú . Primeras jornadas de Derechos de Aguas (pp. 269-294). Lima: CICAJ, INTE -PUCP.

Vicuña Mackenna, B. (1860). La revolución de la independencia del Perú. Lima.Perú 

Vos, J. (2002). Metric Matters. The performance and organization of volumetric water control in large-scale irrigation in the North Coast of Peru. Wageningen University, Wageningen.

[1] PhD Candidate, Anthropology Department, University of Amsterdam (UvA), the Netherlands

[2] Professor and researcher, Department of Social Sciences of the Pontifical Catholic University of Perú

[3] Governance and inclusive development group, University of Amsterdam (UvA) & International development studies group, Utrecht University (UU).


Funding by: