Shell Games: A Middle Preclassic Shell Deposit at the Minor Center of Tutu Uitz Na in the Upper Belize River Valley

Abstract

Recent excavations at the ancient Maya minor center of Tutu Uitz Na in the Belize River Valley revealed an especially large—about 20 million shells—Middle Preclassic (900–300 BC) shell deposit underlying the plaza. Although marine shell species make up a small percentage of the assemblage, most shells are Pachychilus spp., a common freshwater snail known in the southern Maya Lowlands as jute. This report describes the architectural context and assemblage of the deposit and compares it to similar examples in the region. We propose that the Tutu Uitz Na deposit provides one of the earliest examples of depictions of the Maya primordial sea in an architectural context.

Resumen

Excavaciones recientes en el antiguo centro menor Maya de Tutu Uitz Na en el valle del río Belice revelaron un depósito de conchas del Preclásico Medio (900–300 aC) especialmente grande (~20 millones de conchas) que subyace a la plaza. Mientras que las especies de conchas marinas constituyen un pequeño porcentaje del conjunto, la mayoría de las conchas son Pachychilus spp., un caracol común de agua dulce conocido en las tierras bajas mayas del sur como jute. Este informe describe el contexto arquitectónico y el ensamblaje del depósito, y lo compara con ejemplos similares en la región. Proponemos que el depósito Tutu Uitz Na proporciona uno de los primeros ejemplos de representaciones del mar primordio maya en un contexto arquitectónico.

The species of freshwater snail commonly known in the southern Maya Lowlands as jute (Pachychilus spp.) is a familiar sight in archaeological contexts throughout the Belize River Valley. Sizable deposits of jute shells, often intermixed with their marine counterparts and figurines, are a common feature of Middle Preclassic (900–300 BC) contexts, particularly in caches and deposits in caves (Boileau and Stanchly 2020; Healy et al. 1990; Keller 2012:254): their presence reflects purposeful efforts by the Maya to imbue spaces with connections to fertility and characteristics of the watery underworld (e.g., Halperin et al. 2003). One locale also known to be associated with aquatic contexts is the Mesoamerican plaza (Freidel et al. 1993). We draw on new data from the minor center of Tutu Uitz Na in the Upper Belize River Valley of west-central Belize (Figure 1) to propose that the presence of jute and marine shell in plaza deposits reflects Middle Preclassic Maya intentions to imbue these spaces with symbolic properties of the primordial sea.

Figure 1. Regional map.

Maya ceremonial spaces were built to reflect the Maya cosmos, a quadripartite world bordered in the four cardinal directions by sacred mountains and divided vertically into the heavenly upperworld, the middleworld or living world, and a watery underworld. The Popol Vuh, the Maya creation story, describes all of creation as arising from a vast primordial sea, with the middleworld portrayed as resting on the back of a great turtle or crocodile floating in this sea. The sea was a place of both death and of rebirth (Freidel et al. 1993:92). The underworld existed both beneath the ground surface and within bodies of water (Chase and Chase 2009), and aquatic creatures often appear in underworld iconography. Archaeologically documented caches include representations and remains of both freshwater and marine animals and may be physical representations of this underworld. In the Maya worldview all waters are interconnected: rainwater, freshwater, and saltwater all originate in the same primordial sea. This connection means that the attributes and inhabitants particular to one body of water could be shared by all. Water lilies (Nymphaea apla), although they are exclusively freshwater plants, occur in multiple representations of the primordial sea (Houston 2010:70). The Maya constructed plaza spaces to represent the surface of this primordial sea, with the surrounding temples acting as witz, or sacred mountains (Freidel et al. 1993:139).

The Tutu Uitz Na Jute Deposit

Tutu Uitz Na is a tertiary minor center located on a hilltop 600 m southeast of the Late Classic (AD 600–900) center of Lower Dover. The Belize Valley Archaeological Reconnaissance (BVAR) project placed six excavation units in the plaza at Tutu Uitz Na (Figure 2). Ceramic and radiocarbon dating indicate that the center was occupied from the Middle (765–535 cal BC) through the Terminal Classic period (around 890 cal AD; Walden 2021:335). Late Classic Tutu Uitz Na consisted of four structures situated around a sizable plaza (700 m²). The Middle Preclassic component was smaller, consisting of several low-cut stone platforms around the plaza (about 600 m³ of architecture). Excavation of Structure N1 revealed a 50 cm high Middle Preclassic residential platform that sat above the jute deposit, which dated to 540–400 cal BC (reported in Walden 2021:Table 6.11). This earliest structure was twice the size of contemporaneous commoner households and included cut limestone blocks that were similar to architecture at larger nearby contemporaneous centers (Figure 3; see also Walden 2021:329).

Figure 2. Map of Tutu Uitz Na showing locations of excavations.

Figure 3. East profile of Structure N1. Jute deposit shown in green; Middle Preclassic platform shown in orange. (Color online)

A jute deposit was documented in all six plaza excavation units. It consisted of a dense, approximately 30 cm layer of jute shells, pebbles, and a dark loamy matrix (7.5YR 6/3) placed directly on top of bedrock and filling the whole of the unit. The placement of these shells on an area that had been scraped to bedrock potentially indicates a founding event. The diagnostic ceramic assemblage within the deposit is entirely Middle Preclassic (n = 37); it includes Jocote Orange Brown and Savana Orange sherds, as well as fragments of Savana Orange paste figurines (n = 2; Awe 1992; Gifford 1976).

Although no charcoal was recovered from the deposit itself, samples from surrounding architecture corroborate the ceramic/stratigraphic dating of the deposit. Charcoal from the fill of the second construction phase at Structure N1 provides a reliable terminus ante quem of 540–400 cal BC for the jute deposit. Structure E2 is built on a natural spur in the limestone bedrock that rises above the surrounding deposit (Figure 4). A small Middle Preclassic cist cut into this rise contained 20 jute and charcoal dating to 765–540 cal BC (reported in Walden 2021:Table 6.12). Although this does not provide a terminus ante quem for the jute deposit, the placement of the deposit adjacent to the cist and the inclusion of jute within the cist suggest they were roughly coeval.

Figure 4. North profile of Structure E2; see sample A.

The six units yielded about 91,000 jute, with a median density of 6,735 per m². The distribution of excavation units suggests that the jute deposit underlays the entire eastern half of the plaza. Similar densities are evident in the western half of the plaza, where the deposit and its vast numbers of jute are visible in the eroded hillside. Extrapolating from the test units, the extensively excavated eastern plaza could contain about 10 million jute. If shell densities remain consistent throughout the plaza, the entire deposit would contain more than 20 million jute. These jute were likely consumed prior to deposition, because 88% show evidence of extraction techniques such as puncturing and “spire-lopping” (Figure 5; Keller 2012:258–259; see also Solis 2010:26–27, 92–94; Supplemental Table 1).

Figure 5: Illustration showing (a) Pachychilus glaphyrus, (b) Pachychilus indiorum, (c) spire-lopped jute, and (d) punctured jute.

In addition to jute, the deposit contains freshwater species such as apple snail (Pomacea flagellata) and river clam (Nephronaias sp.). High proportions of marine shell beads and production debris are also present in the jute deposit (Figure 6), including 74 finished shell beads, 32 partially finished shell beads, and 188 pieces of shell debitage (Supplemental Table 1). The larger marine shell beads may represent earlier stages in a production sequence (Keller 2012). Alternatively, larger, rougher beads may represent finished products dating to the Early Middle Preclassic (900–600 BC), whereas the smaller refined ones were Late Middle Preclassic (Boileau and Stanchly 2020:51). Unlike other examples of Preclassic bead production in the region, there is a paucity of chert drills at Tutu Uitz Na (Hohmann et al. 2018), indicating that the shell was removed from its primary production locale (see Keller 2012:266). Although it is possible that local residents were importing beads and debitage to place in the jute deposit, it is more likely that bead manufacture was occurring somewhere near Tutu Uitz Na.

Figure 6. Marine shell beads. The arrows represent either the hypothesized production sequence based on Keller (2012) or diachronic stylistic changes based on Boileau and Stanchly (2020).

Construction and Cosmology: Preclassic Jute Deposits in the Belize River Valley

Middle Preclassic jute deposits have been uncovered in the Belize River Valley at early centers, including Blackman Eddy (Brown et al. 1999), Cahal Pech (Ferguson et al. 1996:44), Las Ruinas de Arenal (Horowitz et al. 2020:292), Pacbitun (Boileau and Stanchly 2020), and Xunantunich (Jamison 1993:25). Other examples of deposits at smaller peripheral settlements associated with nascent elite households (like at Tutu Uitz Na) include Callar Creek (Kurnick 2013:130) and Chan (where Middle Preclassic jute had been redeposited in the Late Classic; see Keller 2012:257). Although these examples vary slightly, jute deposits with marine shell primarily occur beneath plaza spaces (Brown et al. 1999:43). The Middle Preclassic in the Belize River Valley saw increasingly sedentary populations and the introduction of monumental architecture. Public ceremonial architecture served to imbue spaces with cosmological significance and to tie a formerly disparate people to a common location, thereby engendering a sense of communal identity (Awe 1992). In many centers, elaborate subfloor dedicatory deposits and caches were laid to consecrate plaza space (Reilly 1994). We argue that a similar logic underlies the placement of the jute deposit at Tutu Uitz Na.

There are alternate explanations, however, that could explain the presence of this deposit. The jute at Tutu Uitz Na could reflect feasting episodes, given that the plaza deposit contained sufficient jute for some 800,000 meals (Healy et al. 1990). If this deposit reflected generic feasting remains, however, we would expect high proportions of nonaquatic fauna, high proportions of serving vessels in relation to storage vessels, and an absence of such unconsumed aquatic items as marine shell beads and debitage (see Supplemental Table 2 and Supplemental Figure 1). If the deposit reflected generic architectural fill, we would expect higher densities of cobbles, limestone, and utilitarian trash such as ceramics and chert (see Supplemental Tables 3 and 4). If the jute were brought up in leveling fill dredged from nearby rivers, we would expect to see river sediment, fish bones, a lower overall proportion of jute (with little to no puncturing or spire-lopping), and a paucity of items such as marine shell, ceramics, and chert. Artifact proportions support neither the feasting remains nor the architectural fill hypotheses. What we see in this deposit is the placement of shells and shell debitage in an area known by the Maya to represent the primordial sea; these aquatic items, although associated with activities such as feasting and craft production, appear to have been removed from their original sites of consumption and production. This removal, along with a relative dearth of other quotidian trash items commonly encountered in architectural fill, suggests that these items were placed in the deposit with specific intent. We hypothesize that the Maya's intent with this deposit was to fashion the plaza at Tutu Uitz Na as a representation of the primordial sea.

The Middle Preclassic jute deposit at Tutu Uitz Na was likely constructed to symbolically imbue the hilltop with cosmological power. We propose that the other Middle Preclassic jute deposits in the region may likewise reflect attempts to imbue plazas spaces with qualities of the primordial sea. The construction of the plaza and the surrounding structures mirrors Maya cosmology in miniature and legitimizes the rituals and ceremonies occurring within this space. This example represents, we believe, one of the first architectural depictions of the primordial sea in the Maya Lowlands.