Introduction

Much archaeological attention on ancient Maya agriculture has tended to centre around one particular food crop, namely maize. Likewise, in recent decades, a great deal of interest has been paid to the role of intensive water and soil management in Maya cultivation. Initial analyses of a series of agricultural terraces in the Maya Mountains of southern Belize encourage us to expand our thinking on these fronts.

Discovered in 1996, the Maya Mountains terraces are beginning to provide key insights into alternative techniques of intensive agriculture employed by the Classic Maya (AD 250-900). Although terraces have been noted in the southern Maya Lowlands, in which the Maya Mountains chain is situated (Reference HealyHealy et al. 1983; Reference DunningDunning & Beach 1994), the complex way in which these particular terraces were used is unique. It suggests that the southern Maya Mountains were a region teeming with a large population employing a mixture of different intensive agricultural strategies in order to sustain itself.

Archaeological background

Terra incognita to archaeologists before the 1990s, the southern Maya Mountains (Figure 1) began to be thoroughly investigated by the Maya Mountains Archaeological Project (MMAP). Between 1992 and 2000, nineteen previously undocumented population centres were recorded by the MMAP nestled deep in the folds of the Maya Mountains' rugged topography (Reference DunhamDunham et al. 2000). The ancient inhabitants of this region exploited its unique montane mineral and biotic resources (Reference DunhamDunham 1996; Reference AbramiukAbramiuk & Meurer 2006). They sustained themselves through cultivating the few flatlands available to them.

Figure 1.

Map of Maya area with location of the Bladen Branch.

By the mid-1990s, the MMAP began to focus archaeological operations on the Classic Maya site of Muklebal Tzul located in the upper reaches of the Bladen Branch (Figure 2) of the southern Maya Mountains (Reference DunhamDunham et al. 1996). This site is modest in size compared to the great centres of the Maya heartland, but it is much larger and more complex than had previously been expected for such a rugged and remote setting with little arable flatland. Not long after the discovery of Muklebal Tzul, a means by which such a large settlement could be sustained revealed itself during reconnaissance of the surroundings: a significant complex of agricultural terraces located approximately 2km southwest of the Muklebal Tzul site core was identified (Reference DunhamDunham & Pesek 2000). The terraces originally dubbed Sahonak Tasar, were subsequently mapped and preliminary soil samples were taken.

Figure 2.

Map of the Upper Bladen Branch with Muklebal Tzul and location of terraces.

Over a decade later and under the auspices of the Maya Mountains Ethnobotany and Ecology Project (MMEEP) (Pesek et al. 2006), these terraces and their soils are being evaluated for new insights into ancient Maya agricultural strategies.

Terraces

The terraces in question (Figure 3) are located on a hillside alongside a nearby stream. They consist of a series of stone retaining walls supporting flat earthen shelves. The stone walls are made of neatly bedded limestone which provided the building material for most of the structures in the region. The terraces' proximity to the site of Muklebal Tzul suggests that food crops were grown for the expanding population of that community.

Figure 3.

Terrace wall.

Preliminary results from the analysis of the phytoliths in the soils of the terraces are compelling. They point to the fact that a myriad of different species were being cultivated, not just maize. Although it was likely that maize was grown, since phytoliths of the subfamily Panicoideae were present, the grass phytoliths are dominated by a morphotype that is diagnostic of bamboo species of the subfamily Bambusoideae. The most abundant morphotype identified (even more abundant than that which characterises the grasses) is the rugose sphere, a morphotype that in our case appears to be derived from the genus Canna (order Zingiberales). Other plant families can be identified but they will take some time to narrow to particular species. Pollen analysis indicates cultivation of avocado on one of the terraces.

Another surprising aspect of the terraces was evidence of a complex cultivation strategy that was previously undocumented and may have been of wider use to the Maya. Careful analysis of the stratigraphy, as well as the pollen analysis, has suggested a sequence of intentional burning and flooding. Nine strata representing alternating episodes of growing, burning and flooding have been identified. Evidence of the burning can be inferred from the presence of varying amounts of charcoal and vitrified plant tissue in the soil layers, and the presence of diatoms and fresh water sponge spicules in the sandy clay loam layers analysed suggests that the water from the adjoining stream flooded the terraces at different points in time (Figure 4). There appears to be a system of canals and possibly even water regulation gates for the purpose of dispersing water over the terraces and recharging the soils.

Figure 4.

East profile of terrace with colours and textures of soils.

To date, agriculture in the Maya area traditionally has been seen as consisting of four mutually exclusive strategies employed by the ancient Maya, namely: slash-and-burn, raised fields, irrigation and terracing (Reference FlanneryFlannery 1982). Of the four strategies, the latter three would have constituted the intensive approaches necessary to support the immense populations that flourished during the Late and Terminal Classic period (AD 600-900). Recent research conducted by Reference AtranAtran (1993), Reference FedickFedick et al. (2000) and others (e.g. Reference DemarestDemarest 1996) suggests, however, that ancient Maya agricultural techniques need not be limited to these four techniques. The evidence we have collected supports this assessment; it suggests that the Muklebal terraces were not simple run-off catchments, but constructions engineered in such a way as to allow for innovative soil building techniques. They were not simply soil retention devices but also hydroengineering facilities aimed at soil building from growing cycles interspersed with carbon and nitrogen deposition via cyclical burning and flooding (alluvial deposition).

More research needs to be conducted to reconstruct the precise details of the agricultural methods practiced on the Muklebal terraces, in particular documenting the water regulation system and deducing the time elapsed between the growing phases and the flooding and burning events, which are believed to be soil enrichment episodes. Provided this fine chronology can be reconstructed, an entirely 'forgotten' method by which the ancient Maya managed to harness the environment to sustain large populations may come to light, and perhaps be applied to benefit the contemporary Maya of the region in sustainable healthful cultivation scenarios.