The fishermen of Sechura: excavations at a specialised site from the Early Intermediate Period, extreme northern Peru
The pre-Hispanic societies of the Central Andes developed in a geographical setting which forms a unique environmental mosaic. Despite its inhospitable nature, the desert of Sechura, located in the extreme north of Peru, has been occupied for at least 5000 years (Milla 1989; Cárdenas et al. 1991, 1993). It constitutes a special area in which to investigate human subsistence strategies and adaptive responses to environmental stresses, such as aridity and very limited terrestrial resources.
Under the ‘Programme Archéologique Désert de Sechura’, a French-Peruvian team has resumed work in this area with excavations in 2012 and 2013 at the Bayovar-01 site (Figure 1), dating from the transition between the Early Intermediate Period (200 BC–AD 600) and the Middle Horizon (AD 600–1000). This small settlement is on a Pleistocene marine terrace at 9–18m above sea level, located 6km from the shore and separated from the ocean by a very flat sandy area, lower than sea level.
The extensive excavations of several areas have allowed us to characterise the occupation of the site, which includes two structures and a large midden. The structures are formed by several rows of marine formation blocks (conchuelas) up to 1.40m high and 1m wide. South of these rows of blocks, posts that were probably used as wind screens were found (Figure 2). Some posts, which collapsed after the site was abandoned, measured up to 4m in length. The wood species could not be determined, but the sinuous shape of some trunks is similar to the shape of the American carob (Prosopis pallida) which is still present in the area. Pits excavated in the north, south and in between these block alignments revealed no evidence of domestic occupation. The position and organisation of the blocks, the presence of screens and the existence of an access corridor to a central space suggest that these structures were more likely some kind of small public place than domestic structures. Among several unusual finds from the excavation were a deposit of 12 bivalves (Trachycardium procerum) in the same layout, and a ceramic vessel, deposited upside down (on its neck rather than its base), probably representing an offering made at the time of the abandonment of the area or site (Figure 3).
The midden excavation provided data on the use of natural resources in this desert environment. It quickly became apparent that this area was not a simple midden but that it fulfilled a more complex function. Thousands of fish remains were discovered during the excavations intermixed with a very large amount of charcoal. Extensive excavation of a 50m2 area uncovered very large burnt zones, nearly 0.5m thick (Figure 4). We observed the superposition of hearths indicating that these work areas were reused. The form of the hearths also evolved; the older examples had a circular shape and measured approximately 3m2 (Figure 5) while the later hearths measured 0.8m in diameter or represented larger combustion areas (>10m2). Burnt material was mainly charcoal, but also included charred seeds and, more surprisingly for the desert area, camelid faeces.
The very high densities of fish bones and otoliths, particularly of a Sciaenidae (Micropogonias altipinnis), raises many palaeoenvironmental questions. The presence of this species, which today inhabits only the warm and coastal or brackish waters bordering Ecuador, could indicate different climatic and environmental conditions, such as warmer waters, higher sea levels, or the presence of a lagoon or embayment.
The analysis of ichthyological remains (Figure 6) is underway, but the midden excavation alone has uncovered more than 35 000 otoliths and many more bones. Among the species, we note the abundance of Micropogonias altipinnis, Albula sp. and Mugil cephalus. The first two taxa live preferentially in warm and shallow coastal waters over sandy/muddy bottoms, a very different environment compared to the one that exists around the site today. Flooding of the sandy plain by fluvial and/or marine inputs would have allowed the development of lagoons. These inputs are an expression of specific and local hydromorphological dynamics (coastal and continental), presumably related to the intensity of the ENSO (El Niño-Southern Oscillation) phenomenon. The occupation of the site, dated to cal AD 547–766 (2σ; ULB23689, ULB23690, ULB23691, ULB23693, ULB23694 and ULB23695), would be contemporary with high intensity El Niño events, enabling the perennial flooding of the sandy plain that separates the site from the current shoreline and therefore the entry of marine fish.
Fishermen who occupied this site benefited from favourable environmental conditions for settlement along the lagoon (now called Las Salinas) or the shore. More than a domestic occupation site, Bayovar-01 would have been a specialised site for fishing and the preparation of fish for transport to other areas by llama caravans. This possible specialisation would be unprecedented for the region and the period and perhaps indicates organisation at a regional level. Thus, this settlement integrates a network of interactions among sites of the region allowing the circulation of products, which form the basis for the development of pre-Hispanic complex societies in the Central Andes.
The Programme Archéologique Désert de Sechura is funded by the French Ministry of Foreign Affairs with additional contributions from the University of Paris 1 (Paleosech Grant) and the CNRS. The fieldwork would not have been possible without the support of the Peruvian Ministry of Culture, the French Embassy in Peru, and the French Institute of Andean Studies (IFEA).