(PDF) Questioning Postclassic Continuity at Baking Pot, Belize, Using Direct AMS 14C Dating of Human Burials

Radiocarbon, Vol 56, Nr 3, 2014, p 1057–1075 DOI: 10.2458/56.18100 © 2014 by the Arizona Board of Regents on behalf of the University of Arizona QUESTIONING POSTCLASSIC CONTINUITY AT BAKING POT, BELIZE, USING DIRECT AMS 14C DATING OF HUMAN BURIALS Julie A Hoggarth1,2 • Brendan J Culleton1 • Jaime J Awe3 • Douglas J Kennett4 ABSTRACT. Archaeologists working in the Belize Valley have argued for the persistence of Maya populations from the Classic (AD 300–900) through Postclassic (AD 900–1500) periods since Gordon Willey’s groundbreaking settlement survey and excavation work in the 1950s. This is contrary to the trajectory recorded in some parts of the Maya region where there is clear evidence for political disruption and population decline at the end of the Classic period. The argument for continuous Classic to Postclassic occupation in the Belize Valley remains ambiguous due to researchers’ reliance on relative ceramic chronologies. This article reports the results of direct accelerator mass spectrometry (AMS) radiocarbon dating of human skeletons (n = 12) from the important center of Baking Pot, Belize, which is thought to provide some of the best ceramic evidence for continuity in the valley. The AMS dates show a long span of mortuary activity between the Middle Preclassic and Late Classic periods (405 cal BC to cal AD 770), with a hiatus in activity during the Early Postclassic (cal AD 900–1200) and subsequent activity in the Late Postclassic (cal AD 1280–1420). These results are not consistent with the idea that Baking Pot was occupied continuously from the Classic through Postclassic periods. This work highlights the need for additional AMS 14C work at Baking Pot and elsewhere to establish absolute chronologies for evaluating the political and demographic collapse of Classic Maya regional centers. INTRODUCTION The political disintegration and depopulation of regional centers in the southern Maya lowlands at the end of the Classic period (AD 750–1000) has received a great deal of attention and remains one of the most debated topics in the literature (Demarest et al. 2004; Aimers 2007; Kennett et al. 2012; Webster 2012). Despite the large-scale abandonment of the central and southern lowlands by the end of the 10th century, some settlements appear to have persisted into the Postclassic period in coastal locations (Guderjan 1995; McKillop 1996), adjacent to lakes (Rice and Rice 2009), and along rivers and lagoons (Pendergast 1985; Graham and Bennett 1989; Graham 2004), although shallow stratigraphy, mixed deposits, and modern disturbance continue to plague age determina- tions for Postclassic contexts (Masson 1995; Pugh and Rice 2009:91). Radiocarbon dates have been used in a limited fashion to build regional and site chronologies in the Maya lowlands (but see Satterthwaite and Ralph 1960; Hammond et al. 1977; Awe 1992; Andrews et al. 2003; Peraza Lope et al. 2006; Prufer et al. 2011; Culleton et al. 2012; Aquino et al. 2013; Inomata et al. 2013), partic- ularly for understanding the timing of collapse (see Sidrys and Berger 1979 for a regional study). Long count calendar dates on carved monuments have also been relied on to date the breakdown of political institutions (Bove 1981; Premo 2004; Ebert et al., in press), and relative ceramic chronol- ogies have generally been used to understand the timing of the depopulation (or lack thereof) of sites in the region (e.g. Adams 1973; Sabloff 1973; Gifford 1976; Ball 1977; Rice 1979; Foias and Bishop 1997; Aimers 2004). Applying chronologies built solely on one of these methods may be problematic for several rea- sons. First, carved monuments with hieroglyphic texts are most commonly found at larger regional centers (e.g. Tikal, Calakmul, Caracol), while smaller centers often lack such written records. The end of sociopolitical activity at smaller centers is therefore more difficult to determine, and often overlooked using this approach. Second, ceramic chronologies were often based on excavations and 14 C dates from more than 50 yr ago when 14C dating was not as precise, required large and often mixed samples of charcoal or bone, and systematic errors occurred because sample preparation and purification techniques were less developed. In determining the chronology of demographic 1. Department of Anthropology, The Pennsylvania State University, University Park, PA 16802, USA. 2. Corresponding author. Email:

. 3. Northern Arizona University, Department of Anthropology, 5 E McConnell Drive, Flagstaff, AZ 86011, USA. 4. Department of Anthropology & Institutes of Energy and the Environment, The Pennsylvania State University, University Park, PA 16802, USA. 1058 J A Hoggarth et al. collapse, calibrated ranges on 14C dates from this time typically span hundreds of years (Sidrys and Berger 1979:Table 1). The challenges associated with 14C dating charcoal from cultural deposits are well known in Maya archaeology. This is best highlighted by the first published Preclassic chronology at Cuello (Ham- mond 1976; Hammond et al. 1979), where the amalgamated samples were contaminated with old charcoal. This issue was resolved through the dating of purified bone collagen from burials at the site (Housley et al. 1991; Law et al. 1991), a method that has received little attention in chronology building in the Maya lowlands since the early 1990s despite major methodological advances (e.g. ultrafiltration and XAD; Brown et al. 1988; Stafford et al. 1988). Because of these issues, problems in 14C dating continue to plague Maya archaeology, especially in Postclassic contexts where stratig- raphy is thin, close to the surface, and cultural remains are often mixed (Pugh and Rice 2009:91). This article assesses the occupational history of the site of Baking Pot, a medium-size Maya center in the Belize Valley (Figure 1) that has been subject to debate over whether occupation was episodic or continuous from the Classic to Postclassic period (Willey et al. 1965:303–4; Willey 1973:101; Aimers 2004:1–8; Audet and Awe 2004:50). We argue here that the direct AMS 14C dating of hu- man burials from the ceremonial center and settlement areas of Baking Pot provides an alternative and potentially complimentary way of assessing the occupational history of the site when combined with detailed analysis of ceramics and other diagnostic artifacts from burial lots. Figure 1 Map of the Maya lowlands, showing the location of Baking Pot and the Belize Valley based on excavations and radiocarbon dates from more than 50 years ago when radiocarbon dating was not as precise, required large and often mixed samples of charcoal or bone, and systematic errors occurred because sample preparation and purification techniques were less developed. In determining the chronology of demographic collapse, calibrated ranges on radiocarbon dates from this time typically span hundreds of years (Sidrys and Berger 1979:Table Questioning Postclassic Continuity at Baking Pot, Belize 1059 SITE BACKGROUND Baking Pot is one of several major centers that were occupied during the Classic period (AD 300– 900) in the Belize Valley, with evidence of monumental architecture, an emblem glyph, and elab- orate tombs that suggest it was the seat of power for a small polity (Audet 2006; Helmke and Awe 2008b; Helmke and Awe 2012). However, like many centers in this subregion, the site lacks carved monuments that are typical of regional capitals to the west and south of the Belize Valley (e.g. Naranjo and Caracol) (Helmke and Awe 2012; also see Andres et al. 2014 for discussions of carved monuments in the nearby Caves Branch region), thus making it impossible to date the end of elite activity in the ceremonial center based on dedication dates on monuments. Furthermore, the impre- cision of the ceramic chronology makes the timing of the abandonment of the settlement uncertain or at best dubious. Willey et al. (1965) were the first to propose Postclassic continuity at select sites in the Belize Val- ley. This hypothesis was based on their discovery of Postclassic New Town complex ceramics at 62 out of 65 mounds tested at Barton Ramie (Willey et al. 1965:291) and on their interpretation that these remains represented “the continued occupation of the Barton Ramie site zone from the end of the Classic period into the Postclassic period” (Willey et al. 1965:568). Gifford’s (1976) ceramic typology for Barton Ramie distinguished an early and late facet of the New Town complex for the Postclassic period (Figure 2), based on the seriation of ceramic groups including Augustine, Pax- caman, and Topoxte Red (among other types) in nearby regions (Bullard 1973; Rice 1987; but see Rice 2013:21–3) (Figure 2). More recently, Pugh and Rice (2009) suggest that these types overlap in time considerably in the Petén Lakes region. This research has revealed variation in the chrono- logical distribution (Pugh and Rice 2009:90) and sociopolitical organization (Cecil 2009) of popula- tions using these ceramic types, raising further questions about Postclassic occupation in the Belize Valley. However, the distinct aspects of the developmental and occupational histories of sites in the Petén Lakes region complicate the uniform application of these ceramic types as temporal markers in other regions (Pugh and Rice 2009:92). As such, the timing of the transition between these facets, as well as the end of Postclassic occupation, remains undefined in the Barton Ramie sequence (Gif- ford 1976:288–9). While scholars have speculated that the New Town complex at Barton Ramie ex- tended no later than the Middle Postclassic (Rice 1979:80–1; Chase and Chase 1985:14), temporal distinctions of the ceramics remain unclear (Willey et al. 1965:570–1; Chase and Garber 2004:8–9). Figure 2 Ceramic chronology and Postclassic types of the Belize Valley and Petén Lakes region (modified from Chase and Chase 1985:Figure 4). (Willey et al. 1965:303-304; Willey 1973:101; Aimers 2004:1-8; Audet and Awe 2004:50). We argue here that the direct AMS 14C dating of human burials from the ceremonial center and settlement areas of Baking Pot provides an alternative and potentially complimentary way of assessing the occupational history of the site when combined with detailed analysis of ceramics and other diagnostic artifacts from burial lots. SITE BACKGROUND Baking Pot is one of several major centers that were occupied during the Classic period (AD 1060 J A Hoggarth et al. Recent research at Baking Pot (Figure 3) has yielded comparable Postclassic materials to nearby Barton Ramie. These materials are most evident in Plaza 1 and 2 of Group A of the site core (Ricket- son 1931:4–5; Willey et al. 1965:306; Aimers 2003:156) and at 68% of house groups, concentrated in the central and eastern settlement (Willey et al. 1965:309; Ehret and Conlon 2000, 2001; Hog- garth 2012). Surface collections and horizontal excavations of domestic groups have also uncovered deposits containing Postclassic ceramics and materials (Ehret and Conlon 2000, 2001; Audet and Awe 2004; Hoggarth 2009), with evidence for shifts in material culture and mortuary patterns from the Classic to Postclassic periods (Hoggarth 2012). Figure 3 Map of Baking Pot site core and settlement, showing the locations of structures with raising further dated questions burials. about Postclassic occupation in the Belize Valley. However, the distinct aspects of the developmental and occupational histories of sites in the Petén Lakes region Early facet New complicate Town ceramics, the uniform applicationparticularly Augustine of these ceramic typesRedasvessels, temporaloccur in higher markers proportions in other regions than those from the late facet of the complex (particularly Paxcaman and Topoxte (Pugh and Rice 2009:92). As such, the timing of the transition between these facets, as Red) in well Groupas A of the the end of Postclassic site core occupation,4.18) (Aimers 2004:Table remains undefined and in the easternin settlement the Barton (Ehret Ramieand sequence Conlon(Gifford 2000:Ta- ble1976:288-289). 1; Ehret and ConlonWhile 2001:Table scholars have 1).speculated that has This evidence the been New used Towntocomplex infer thatatPostclassic Barton Ramieoccu- extended pation peakedno inlater the than thepart earlier Middle of thePostclassic period and(Rice slowly1979:80-81; Chaseseveral declined within and Chase 1985:14), hundred years, a temporalwhich, situation, distinctions of thewould if accurate, ceramics lendremain supportunclear (Willey to Willey’s et al. 1965:570-571; hypothesis of PostclassicChase and continuity at Garber the site.2004:8-9). Aimers (2003:160) has suggested a collapse of elite activity at the end of the Classic period coupled with gradual depopulation thereafter, with small populations remaining throughout Recent research at Baking Pot (Figure 3) has yielded comparable Postclassic materials to nearby the Postclassic and Historic periods. As indicated above, however, the temporal uncertainties of Barton Ramie. These materials are most evident in Plaza 2 of Group A of the site core (Bullard these ceramic-based relative chronologies, combined with the shallow and mixed nature of deposits, and Bullard 1965:21; Willey et al. 1965:308; Aimers 1997) and at 68% of house groups, make these interpretations concentrated in the central uncertain and may and eastern provide(Willey settlement us withet an al.imprecise 1965:309;timing Ehret of andPostclassic Conlon occupation at the site. 2000; Ehret and Conlon 2001; Hoggarth 2012). Surface collections and horizontal excavations of domestic groups have also uncovered deposits containing Postclassic ceramics and materials Excavations in the monumental site core of Baking Pot (Ricketson 1931; Bullard and Bullard 1965; (Ehret and Conlon 2000; Ehret and Conlon 2001; Audet and Awe 2004; Hoggarth 2009), with Willey et al. evidence for 1965; Audet shifts 2006; Helmke in material culture2008) uncoveredpatterns and mortuary several from elite interments. the Classic Investigations to Postclassicin theperiods periphery (Willey (Hoggarth 2012). et al. 1965; Conlon 1994; Conlon and Ehret 2000) are encompassed within a 9-km2 area around the ceremonial center and 20% of house groups were tested to identify areas of Early Postclassic occupation facet New (Hoggarth Town ceramics, 2012). Horizontal particularly Augustineexcavations Red vessels,ofoccur houseingroups higher in Settlement proportions than those from the late facet of the complex (particularly Paxcaman and Topoxte Red) in Group A of the site core (Aimers 2004:Table 4.18) and in the eastern settlement (Ehret and Conlon 2000:Table 1; Ehret and Conlon 2001:Table 1). This evidence has been used to infer that Postclassic occupation peaked in the earlier part of the period and slowly declined within several hundred years, a situation, which if accurate, would lend support to Willey’s hypothesis of 5   Questioning Postclassic Continuity at Baking Pot, Belize 1061 Cluster C, where Postclassic occupation was concentrated, identified two primary burials (99E-1 and 101-1) in association with New Town ceramics. Previous excavations by Audet (2002) also recorded a burial (M-198, Burial 2) in association with mixed Classic and Postclassic materials in the Yaxtun (M-198) house group. These samples were supplemented by burials recovered from Preclassic and Classic contexts to ascertain the span of occupation in the settlement of Baking Pot. METHODS Bone collagen for 14C and stable isotope analyses was extracted and purified using the modified Longin method with ultrafiltration (Brown et al. 1988). Bone samples were initially cleaned of adhering sediment and the exposed surfaces were removed with an X-ACTO® blade. Samples (200– 400 mg) were demineralized for 24–36 hr in 0.5N HCl at 5°C followed by a brief (<1 hr) alkali bath in 0.1N NaOH at room temperature to remove humates. The pseudomorph was rinsed to neutrality in multiple changes of Nanopure H2O, and then gelatinized for 12 hr at 60°C in 0.01N HCl. Gel- atin solution was pipetted into precleaned Centriprep® 30 ultrafilters (retaining >30kDa molecular weight gelatin) and centrifuged three times for 30 min, diluted with Nanopure H2O and centrifuged three more times for 30 min to desalt the solution (ultrafilter cleaning methods are described in McClure et al. 2010:28–9). Ultrafiltered collagen was lyophilized and weighed to determine per- cent yield as a first evaluation of the degree of bone collagen preservation. Carbon and nitrogen concentrations and stable isotope ratios were measured at the Penn State University Light Isotope Laboratory with a Costech EA (ECS 4010), Thermo Finnigan Conflo IV gas handling device, and a Thermo Finnigan Delta V analyzer. Sample quality was evaluated by % crude gelatin yield, %C, %N, and C:N ratios before AMS 14C dating. C:N ratios for all samples fell between 3.19 and 3.42, indicating good collagen preservation (DeNiro 1985; van Klinken 1999). C samples (~2.5 mg) were combusted for 3 hr at 900°C in vacuum-sealed quartz tubes with CuO 14 wire and Ag wire. Sample CO2 was sent to KCCAMS (University of California, Irvine) where it was reduced to graphite at 550°C using H2 and a Fe catalyst, with reaction water drawn off with Mg(ClO4)2 (Santos et al. 2004). Graphite samples were pressed into targets in Al boats and loaded on the target wheel for AMS analysis. 14C ages were corrected for mass-dependent fractionation with measured δ13C values (Stuiver and Polach 1977), and compared with samples of Pleistocene whale bone (background, >48 14C kyr BP), late Holocene bison bone (~1850 14C BP), late AD 1800s cow bone, and OX-1 oxalic acid standards for calibration. Dates were calibrated with OxCal v 4.2.3 (Bronk Ramsey 2009) using the IntCal13 Northern Hemi- sphere curve (Reimer et al. 2013). It is possible that the atmosphere over Central America could have been a mixture of Northern Hemisphere (NH) and Southern Hemisphere (SH) air, as the boundary between the two, the Intertropical Convergence Zone (ITCZ), has shifted over the last several mil- lennia (Haug et al. 2001; Kennett et al. 2012). Because SH air is on average about 45 ± 15 14C yr older than contemporary NH air between 500 cal BC and cal AD 1500, a mixture of the two would tend to make the dates reported here slightly older than their true age (Hogg et al. 2013; Reimer et al. 2013). Recent work to confirm the correlation between the Maya and Gregorian calendars proposed a mixture of 67% NH and 33% SH based on wiggle-matching between cellulose isotope data and the YOK-I speleothem record (Kennett et al. 2013). For the majority of our samples, calibrating with that mixture tends to have little effect on the overall 2σ calibrated range, though it does tend to pull the probability densities later within that range. Cases where the alternative calibration would affect interpretations are noted in the text. Given Baking Pot’s proximity to the Belize River and the likelihood of some amount of riverine food in the Maya diet, an unquantified freshwater reservoir effect (Rf) may be at work with some 1062 J A Hoggarth et al. or all of the skeletons. The Macal and Mopan rivers, which join to form the Belize River upstream from Baking Pot, both flow over Cretaceous and Early Tertiary limestones at the edge of the Vaca Plateau, which could contribute dead carbon to aquatic organisms such as river snails (Pachychi- lus spp., Nephronaias spp., and Pomacea flagellata), turtles (Dermatemys mawii and Staurotypus triporcatus), and fish, all of which have been recovered in small numbers in the excavations (see Hoggarth 2012:157–9). Unpublished data from paired charcoal and Pachychilus shells in two karst- ic watersheds in southern Belize indicated Rf on the order of 1000–1500 14C yr. If the Belize River reservoir was of similar magnitude, even a small proportion of aquatic protein in the diet could lead to older apparent ages. This is a matter we intend to pursue going forward, but evidence suggests that the offsets are not very large and/or not of variable magnitude, as might be expected for indi- viduals with differing diets and indicated by the agreement between a hieroglyphic date and 14C age in sample BKP01, and the identical ages of the M-096 and Late Postclassic burials. The OxCal models Sequence and Phase were used to constrain the calibrated ages for the case of a sequence of three stratigraphically separated burials from M-096 assuming serial interment (Bronk Ramsey 2009). An alternative model of a single coeval interment was considered by combining the three dates following the procedure of Ward and Wilson (1978). For a single burial with chrono- logical information on an associated ceramic vessel, the calibrated distribution was constrained and renormalized using the reigns of possible rulers referred to in the hieroglyphic text, based on the timeline for Naranjo rulers listed in Martin and Grube (2008). The chronology of ceramic complexes (and subcomplexes) is not typically viewed as absolute, as it is commonly recognized that individual ceramic types were initiated, flourished, and declined at different times (and in different locations). The Barton Ramie ceramic typology was primarily based on seriation from stratigraphic levels, supplemented by cross-dating of at least one major type in each ceramic complex (Gifford 1976:37) with the ceramics from Uaxactun (Smith 1955), where carved monuments with calendar dates are present. In addition, New Town complex ceramics were compared by Gifford to those at Tikal and the Petén Lakes region. Additional research in the Belize Valley has focused on refining Gifford’s ceramic typology with an emphasis on identifying diag- nostics that distinguish the Terminal Classic period (LeCount 1996:123–71; Aimers 2004) largely lacking at Barton Ramie. These works have slightly shifted the timing of some phases designated at Barton Ramie (Willey et al. 1965:Figure 3), with LeCount et al. (2002) defining the Terminal Clas- sic between AD 780–890 based on 14C dates. This article builds on efforts towards chronological refinement for individual ceramic types and complexes in the Belize Valley. In addition, we argue that high-precision AMS 14C dating of human skeletons is more sensitive to multiple waves of aban- donment and reoccupation that may go undetected using ceramics alone. In cases of dated burials with identifiable ceramic types, temporal differences between the accepted ceramic chronology and 14C dates are explored using the Difference model in OxCal. In optimal cir- cumstances, a larger sample of directly dated burials would be used to assess the temporal discrep- ancies between burials and the established time periods associated with ceramic phases. However, even this small sample reveals discrepancies between the timing for ceramic types that have been used as temporal markers in the Belize Valley, and our goal is to continue directly dating human burials at the site as additional materials become available. RESULTS Baking Pot burials yielded calibrated dates that span from the Middle Preclassic to the Late Post- classic periods (Figure 4). Table 1 shows the 2σ calibrated dates, with the majority of dates falling into the Classic period (AD 300–900). Burials from Group A (formerly designated as Group I, see Questioning Postclassic Continuity at Baking Pot, Belize 1063 Helmke and Awe 2008a) in the ceremonial center date to the Late Preclassic (sample BKP03) and Late Classic periods (sample BKP01). Although only two burials from the monumental center have been dated so far, we intend to include more in the future. Burials from the settlements span the Mid- dle Preclassic (sample BKP21) to the Late Postclassic periods (samples 99E-1, BKP18, BKP22), with a large temporal gap separating the Classic and Postclassic burials. Figure 4 Calibrated 2σ dates for Baking Pot burials, with dates from burials at M-96 that were placed in vertical su- one majorperposition typeininthe each same location, modeled ceramic in OxCal. (Gifford 1976:37) with the ceramics from Uaxactun complex (Smith 1955), where carved monuments with calendar dates are present. In addition, New Town complex ceramics Site Core Burialswere compared by Gifford to those at Tikal and the Petén Lakes region. Additional research The earliest in the date from theBelize site coreValley has focused comes from on refining Burial 1 (sample BKP03)Gifford’s in Groupceramic A, Plaza typology 2 (pre- with an emphasis on identifying viously designated Group I, diagnostics Plaza II). Thethat adultdistinguish individual ofthe Terminalsex indeterminate Classic period was buried along(LeCount 1996:123-171; Aimers the central axis of this2004) plaza. largely lacking Jade inlays in the at Barton upper teethRamie. These suggest that the works individualhave wasslightly from a shifted high-status lineage (Swain 2005). Vertical excavations (directly below the timing of some phases designated at Barton Ramie (Willey et al. 1965:Fig.3), with the location of Burial 1) toLeCount sterile sediment in this plaza recovered Late Preclassic ceramic types in the and colleagues (2002) defining the Terminal Classic between AD 780-890 based on radiocarbon earliest levels, and a 14 C date from charcoal places deposits in this location in the Middle Preclassic period (700–400 BC) dates. This paper builds on efforts towards chronological refinement for individual ceramic types (Audet 2006:217–8). The Late Preclassic/Early Classic date of sample BKP03 at cal AD 130–325 and complexes in the Belize Valley. In addition, we argue that high-precision AMS 14C dating of (95.4%) is within expectations, as it is stratigraphically well above this Middle Preclassic level. human skeletons is more sensitive to multiple waves of abandonment and reoccupation that may go undetected Sample BKP01 using ceramics comes from analone. adult individual of indeterminate sex buried in a plastered cist in the western shrine (Str. A1) of Group A. Elaborate grave goods included a carved K’awiil jade pendant In casespolychrome and of dated ceramic burialsvessels with sourced geochemically identifiable ceramicandtypes, stylistically to the central temporal Petén region differences between the (Reents-Budet et al. 2005; Audet 2006:212). The wealth and location of accepted ceramic chronology and radiocarbon dates are explored using the Differencethis burial suggest that thismodel in individual may have been one of the rulers of Baking Pot. Audet assigned a Terminal Classic date OxCal. In optimal circumstances, a larger sample of directly dated burials would be used to to this burial based on the presence of a Daylight Orange: Darknight variety vessel in the tomb (Fig- assessure 5a), the which temporal has beendiscrepancies used as one of thebetween diagnosticburials and forthe ceramic types established the late time periods facet of the Spanish associated Lookout with ceramic complex (AD phases. 800–900)However, in the regioneven this2004:89,113). (Aimers small sample reveals discrepancies between the timing for ceramic types that have been used as temporal markers in the Belize Valley and our goal is to continue directly dating human burials at the site as additional materials become available. RESULTS Baking Pot burials yielded calibrated dates that span from the Middle Preclassic to the Late Postclassic periods (Figure 4). Table 1 shows the 2σ calibrated dates, with the majority of dates 1064 J A Hoggarth et al. Table 1 Calibrated AMS 14C dates from human burials at Baking Pot. UCI- Sample 14 C age cal age1 p Gelatin C:N AMS # ID Context (BP) ± 2σ 2σ yield %C %N ratio 127393 99E-1 M-099b, Burial 99E-1 565 20 1315–1360 (51.8%) 9.3% 40.6 14.9 3.19 1385–1420 (43.6%) 132226 BKP01 Grp A, Str. A1 Burial 1 1315 20 655–715 (75.1%) 4.3% 40.8 14.5 3.29 740–770 (20.3%) Modeled date: 660–680 (54.9%) 695–715 (15.7%) 745–765 (24.8%) 132227 BKP03 Grp A, Plaza 2 Burial 1 1805 25 130–260 (87.6%) 0.9% 47.4 16.2 3.42 280–290 (0.4%) 295–325 (7.4%) 132228 BKP12 M-096, Burial 96-1 1325 20 650–715 (82.6%) 11.5% 44.8 15.9 3.29 745–765 (12.8%) Modeled date: 660–715 (82.6%) 745–765 (12.8%) 132229 BKP13 M-096, Burial 96-2 1335 25 645–715 (85.5%) 7.6% 43.8 15.5 3.31 745–765 (9.9%) Modeled date: 650–695 (95.4%) 132230 BKP14 M-096, Burial 96-3 1305 25 660–725 (66.5%) 6.6% 44.4 15.7 3.30 735–770 (28.9%) Modeled date: 650–700 (95.4%) Combined (BKP12, 13, 14) Modeled: 655–695 (95.4%) 132231 BKP16 M-099a, Burial 99N-1 1945 20 5–90 (89.2%) 3.7% 47.4 16.6 3.34 100–125 (6.2%) 132232 BKP17 M-099d, Burial 99W-1 1330 20 650–710 (87.0%) 6.1% 46.8 16.5 3.31 745–765 (8.4%) 132233 BKP18 M-101, Burial 101-1 650 20 1280–1320 (41.9%) 3.6% 43.1 15.3 3.29 1350–1390 (53.5%) 132234 BKP19 M-112, Burial 112-1-1 1510 20 430–455 (2.7%) 2.0% 44.5 15.6 3.33 470–490 (3.6%) 530–610 (89.1%) 132235 BKP21 M-198, Burial 1 2290 20 405–355 BC (88.5%) 1.2% 44.4 15.4 3.36 285–255 BC (6.9%) 132236 BKP22 M-198, Burial 2 650 20 1280–1320 (41.9%) 5.5% 44.8 15.7 3.32 1350–1390 (53.5%) Poorly preserved burials: BKP02 Grp A, Str A5 Burial 1 0.3% BKP04 Grp B, Plaza 1 Burial 1 0.1% BKP05 Grp B, Plaza 1 Burial 2 0.0% BKP06 Grp B, Str. B1, Burial B1-8 0.5% BKP07 Grp B, Str B1 Burial B1-9 0.0% BKP08 M-209, Burial 1 0.0% BKP09 M-209, Burial 2 0.0% BKP10 M-209, Burial 3 0.0% BKP11 M-209, Burial 4 0.0% BKP15 M-096, Burial 96-4 0.0%   BKP20 M-112, Burial 112-1-2       0.1%       1. All ages are cal AD, except for the two indicated for Sample ID BKP21, which are given in cal BC. Questioning Postclassic Continuity at Baking Pot, Belize 1065 The date of cal AD 655–715 (75.1%) and cal AD 740–770 (20.3%) from sample BKP01 calls into question the temporal assignment of the Daylight Orange: Darknight variety, and possibly the late facet Spanish Lookout ceramic complex to the traditional temporal span of the Terminal Classic period in the Belize Valley. Additional chronological data in the form of a glyphic text from the burial assemblage can be brought to bear on the chronology as well. A miniature Cabrito Cream polychrome vase (Figure 5b) with a Primary Standard Sequence (PSS) text (Figure 5c) names the patron of the vessel as one of a line of lords at Naranjo who ruled between AD 644 and 780 with a variation of the K’ahk’ [name variant] Chan Chaak nominal/titular hieroglyphic sequence (Figure 5d) (Audet and Awe 2005:362; Helmke and Awe 2008b:79–80). Helmke and Awe (2008b:80) argued that the glyphic elements in the PSS text most closely matches K’ahk’ Ukalaw Chan Chaak, who reigned at Naranjo from AD 755–780 (Martin and Grube 2008:80–1). We modeled the date for sample BKP01 in OxCal v 4.2.3, using the initiation of the Naranjo titular tradition in AD 644 as a terminus post quem with the Before command to constrain the calibrated age, as we assume that the vessel could not have been made before the initiation of the K’ahk’ [name variant] Chan Chaak tradition. Assuming that the vessel was interred with the individual, this modeling shifts the 14C date to cal AD 660–765 (95.4%) (Figure 5d). Although it is possible that this vessel may have been deposited at a later time through tomb re-entry, there was no evidence for the disturbance of the skeletal remains or other materials in the burial, and several intact floors are noted above the burial. The possibility exists that the polychrome vase may have been curated as an heirloom; however, the agreement between the 2σ date for the burial with the Naranjo dynastic titular tradition, along with the Late Classic temporal assignments for nearly all of the other ceramics in the burial (including Belize Red, Puhui-Zibal Composite, and three other Cabrito Cream polychrome vessels), renders this prospect unlikely, as it is equally unlikely that nearly all of the vessels in the burial would have been curated. However, a Balanza Black vessel in the burial, associated with the Hermitage complex, may have been an heirloom vessel. Together, the correspondence between the 14C date Figure 5 Ceramic vessels from Burial 1, Structure B at Baking Pot; (a) Daylight Orange: Darknight Variety vessel (modified from Audet 2006:Figure 5.27); (b) Miniature Cabrito Cream vase (modified from Helmke and Awe 2008:Figure 13a); and (c) Group Roll-out I, view seeof Helmke vessel, showing the PSS and Awe text (modified (2008a)) from in the Helmke and center ceremonial Awe 2008:Figure date to 13b); and (d)Preclassic the Late Calibrated 14C date(sample for BKP01, modeled using the Naranjo titular tradition, breaks in this tradition, BKP03) and Late Classic periods (sample BKP01). Burials from the settlements and noting the reign of the span ruler that maythebe Middle mentionedPreclassic in the text. (sample BKP21) to the Late Postclassic periods (samples 99E-1, BKP18, BKP22), with a large temporal gap separating the Classic and Postclassic burials. Site Core Burials The earliest date from the site core comes from Burial 1 (sample BKP03) in Group A, Plaza 2, (previously designated Group I, Plaza II). The adult individual of indeterminate sex was buried along the central axis of this plaza. Jade inlays in the upper teeth suggest that the individual was from a high status lineage (Swain 2005). Vertical excavations (directly below the location of Burial 1) to sterile sediment in this plaza recovered Late Preclassic ceramic types in the earliest levels, and a radiocarbon date from charcoal places deposits in this location in the Middle 1066 J A Hoggarth et al. and ceramics in this burial point to a date prior to AD 800, suggesting that the initiation of some Terminal Classic ceramic diagnostics (especially Daylight Orange) may be earlier than previously interpreted. Additional dating of burials with this ceramic type is necessary to more definitively assess the chronological discrepancy. Overall, burials from the site core of Baking Pot provide evidence supporting a history of elite activity throughout the Classic period. Initial construction in the plaza began during the Middle Preclassic period (Audet 2006), with the later mortuary activity along the central axis of the plaza during the Early Classic. The presence of Cabrito Cream polychrome vessels in the burial of one of the rulers of Baking Pot at the end of the Classic period suggests connections to larger regional centers in the Petén such as Naranjo and Holmul (Reents-Budet et al. 2000), which may have served to foster alliances between these large regional centers and smaller Belize Valley centers through the gifting of elaborate ceramics. Settlement Burials Dates from burials in the settlement at Baking Pot show a long span of occupation from as early as the end of the Middle Preclassic to the Late Postclassic period. The earliest date, at 405–255 cal BC (95.4%), comes from sample BKP21, a child (3–5 yr) recovered from the M-198 house group. Au- det (2002:98) assigned this individual a Late Preclassic designation due to the presence of Sierra Red ceramics in associated contexts, so the Middle to Late Preclassic date is not unexpected. One additional burial (sample BKP16) falls within the Late Preclassic period at cal AD 5–125 (95.4%), a secondary burial recovered on top of the terminal floor of Structure M-99a in a context with Spanish Lookout and New Town ceramics (Hoggarth 2012). The Preclassic date of this sample was not expected for this context; however, Maya practices of ancestor veneration could explain this discrepancy due to the secondary nature of the deposit. The majority of burials in the settlements surrounding the ceremonial core date to the Classic period, with one burial (sample BKP19) dating to the latter part of the Early Classic period at cal AD 430–610 (95.4%). This adult female was interred in an elaborate crypt at M-112, with Bar- ton Creek, Floral Park, and Hermitage ceramic vessels, carved limestone spindle whorls, obsidian, and over 300 carved greenstone beads (Lamb 2009). In addition, the remains of a secondary individ- ual (which were not dated) were placed at the feet of the primary individual. Together, the 14C date is in agreement with the Early Classic date that was assigned based on the ceramics. Burial 99W-1 (sample BKP17), the remains of a young child (2–3 yr) interred in Mound 99b, dates to the Late Classic period at cal AD 650–765 (95.4%). This date falls within expectations for the context of the burial, as Spanish Lookout ceramics were found above the burial. Five burials associated with domestic groups date to the Late Classic period, and the dates of four of these (samples BKP12, BKP13, BKP14, and BKP17) are statistically indistinguishable (cal AD 660–770). Three of these were recovered at M-96, in vertical superposition at the same location in prone positions with heads oriented to the south, which is the dominant burial pattern of the Belize Valley from the Late Preclassic to the end of the Classic period (Willey et al. 1965:534–5; Awe 2013:34). Burial 96-3 (sample BKP14) was the deepest of the three burials, the remains of an adult female buried in a simple pit, with Belize Red sherds and several carved shell ornaments (Pe- layo 2010; Hoggarth 2012:Appendix B). Burial 96-2 (sample BKP 13) was recovered 4 cm directly above Burial 96-3, and the adult individual (indeterminate sex) was buried with Belize Red ceram- ics, limestone spindle whorls, and an undiagnostic miniature ceramic vessel. Burial 96-1 (sample BKP12) was located 10 cm above Burial 96-2, with the adult male interred with five obsidian blades near the pelvis, although no ceramics were present in this burial. Questioning Postclassic Continuity at Baking Pot, Belize 1067 There are (at least) two interpretations of the burial practices at M-96 that lead to slightly different chronological outcomes, both of which we consider here. On the one hand, the three interments may have been distinct events separated in time (i.e. a sequence). On the other, all three deaths and burials may have occurred at the same time. Neither can be ruled out on the basis of available stratigraphic or chronometric information. The lack of comingled human remains, disarticulation, or disturbance of primary positions, and the presence of matrix between the burials could be consistent with either burial scenario, assuming later interments never disturbed earlier ones. Each scenario invokes dif- ferent a priori assumptions about deposition of the human remains and requires different chrono- logical models to resolve the age of the deposits. In the first case, i.e. a sequence of burials, the close stratigraphic positions of Burial 96-2 and 96-3 suggest that these interments may have occurred in close succession. We modeled this scenario in OxCal by including Burials 96-2 and 96-3 in a Phase, making no assumptions about the order of interment within a modeled Sequence that placed Burial 96-1 later. The modeled dates fall between the 7th to 8th centuries (sample BKP13: cal AD 650–695 (95.4%), sample BKP14: cal AD 650–700 (95.4%), and sample BKP 12: cal AD 660–765 (95.4%). Agreement indices (A) well above A′ for each modeled posterior and Amodel = 114%; outlier analysis detects no outliers in the sequence (General Outlier Model; Bronk Ramsey 2009). Despite the pres- ence of Belize Red ceramics in Burial 96-2 (sample BKP13) and 96-3 (sample BKP14), the dates for these burials slightly precede the accepted start of the Spanish Lookout ceramic complex (at AD 700, Willey et al. 165:371). In contrast, Burial 96-1 falls into the accepted temporal assignment of the Spanish Lookout complex, although it lacked any ceramics with which to compare the date. No other identifiable ceramics were included in these burials. The conventional 14C ages of the three burials are statistically identical (χ2 = 0.76, df = 2, Tcrit = 5.99; Ward and Wilson 1978) and when combined give a weighted mean age of 1322 ± 9 BP. Under the assumption that the burials are contemporaneous and interred in a single event, this yields a cali- brated age of cal AD 655–695 (95.4%). This places the burials in the timeframe of the Tiger Run phase (AD 600–700), which is unexpected due to the presence of Belize Red ceramics. However, this overlaps with LeCount et al.’s (2002) Samal (AD 600–670) and Hats’ Chaak (AD 670–780) Late Classic ceramic phases. Both the successive interment and single event interpretations for the burials at M-96 reveal unexpected dates that precede the traditional initiation of the Spanish Lookout complex. As a caveat, however, if the combined age is calibrated with a 67% to 33% Northern-Southern Hemisphere mix, the result is a discontinuous distribution that falls partly in the Spanish Lookout complex at cal AD 665–695 (56.1%), 700–710 (1.8%), and 740–765 (37.5%). In the future, additional dating of burials in the settlement and site core will test whether the dates of the Tiger Run and Spanish Lookout ceramic complexes (or individual ceramic types within these groups) need to be pushed earlier in time. Three burials (samples 99E-1, BKP18, BKP22) from Baking Pot’s settlement date to the Postclassic period, spanning the interval between cal AD 1280 and 1420. Sample BKP22 comes from M-198, Burial 2, a young child (20 months ± 8 months) buried near a mixed midden deposit with Spanish Lookout and New Town ceramic types. Audet (2002) surmised that the burial dated to the Classic period (no ceramics were present in the burial), although she suggested that the Postclassic materials in the midden might have dated to the Middle to Late Postclassic. Burial 99E-1 (sample 99E-1) was an adult male in a fully flexed position interred adjacent to the wall of the eastern structure of the M-99 group, with a greenstone adze and Paxcaman Red ceramics associated with the late facet New Town complex (Hoggarth 2012:Appendix B). A similar burial was identified at M-101, an adult in a semiflexed position with Augustine Red sherds. Hoggarth (2012) suggested that the atypical locations (off-platform), orientations, and po- 1068 J A Hoggarth et al. sitions of these burials may represent a distinct shift in burial patterns from extended burials with head oriented to the south during the Classic period to flexed burials oriented to the north during the Postclassic period. The presence of scroll-feet from Augustine Red dishes suggested an Early Postclassic date for Burial 101-1 and Paxcaman Red ceramics indicated a Late Postclassic date for Burial 99E-1. Similarly, the presence of both early and late facet New Town ceramics in a midden above the level of Burial 2 at M-198 were interpreted as Middle to Late Postclassic in age. Howev- er, the 14C dates for all three Postclassic burials fall distinctly in the Late Postclassic period. The temporal clustering of these burials from the end of the 13th century to the beginning of the 15th century, coupled with the temporal gap between Classic and Postclassic interments at the site, estimated at 532–559 yr (7.6%) and 585–725 yr (88.3%) (calculated using the Difference command in OxCal to take the probability distributions of the dates into consideration along with the timespan between dates), provides support for an abandonment of Baking Pot during the Late to Terminal Classic period followed by reoccupation during the Late Postclassic. This evidence does not support hypotheses of Postclassic continuity, raising questions about previous interpretations of the demo- graphic collapse of the Belize Valley and occupation during the Postclassic period. DISCUSSION 14 C dates from 12 human burials in the site core and settlement of Baking Pot span the interval between the Preclassic and Postclassic periods, with the majority of burials falling in the Late Clas- sic period. This evidence, coupled with epigraphic and archaeological evidence, indicates that the apogee of the center occurred between the 7th and 8th centuries. In the case of the burials at M-96, all three burials likely date to the latter half of the 7th century, a period of time that is typically associated with the Tiger Run ceramic complex (Figure 6). These results are not in alignment with long-held assumptions of site occupation, as these dates suggest that the traditional timing of the Spanish Lookout ceramic complex (or perhaps simply Belize Red, the dominant type of Late Classic ceramics at Baking Pot) may be too late. Calibrating with 100% Northern Hemisphere atmosphere, Spanish Lookout ceramics in the dated burials are between 5 and 45 yr (95.4%) earlier than expect- ed, calculated using the Difference command in OxCal measured from the beginning of the Spanish Lookout complex in AD 700 to the combined dates from burials at M-96. These results suggest that the burials at M-96, as well as the associated ceramics, are statistically earlier than expected, although this difference could potentially be on the lower end of the distribution. This pattern is re- peated in the royal burial in Group A, with the date for the burial corresponding well with the written text naming the ruling lineage of Naranjo, but in disagreement with the Terminal Classic (AD 800– 900) designation of the Daylight Orange ceramic group. Calculating the Difference between the beginning of the late facet of the Spanish Lookout complex in AD 800 and the royal burial (BKP01) gives a range of 30–60 yr (19.7%) and 80–145 yr (75.7%) earlier than expected. This is a large difference in the expected dates for Daylight Orange ceramics and prompts a major re-evaluation of the timing for some diagnostics associated with the late facet of the Spanish Lookout phase as well. Although this sample consists of only four dated burials, the close temporal correspondence of dates considered too early suggests that both the Tiger Run and Spanish Lookout ceramic complexes may require re-evaluation, shifting the timeframes for these phases earlier in time. However, the small burial sample included here does not yet provide enough evidence for such a shift at this point. This proposition has implications for understanding sociopolitical change at Baking Pot, as well as for interpreting the interactions between sites in the Belize Valley. Baking Pot has often been heralded as having a strong Late to Terminal Classic occupation with evidence of construction evident through the presence of late facet Spanish Lookout ceramics in construction fill of plazas and monumental architecture (Aimers 1997, 2003; Audet 2006; Helmke 2008) and domestic groups Questioning Postclassic Continuity at Baking Pot, Belize 1069 Figure 6 14C dates from Late Classic and Postclassic burials at Baking Pot, with included ceramics and their traditional ceramic phases. Ceramic phases were based on the Barton Ramie ceramic chronology (Willey et al. 1965; Gifford 1976), with Terminal Classic designations by Aimers (2004). Ceramic types from dated burials include Belize Red (from Gif- ford 1976:Figure 161b); Cabrito Cream (figures modified from Audet and Awe 2005:Figures 3 and 4; Helmke and Awe 2008:Figure 13a); Puhui-Zibal Composite (from Audet 2006:Figure 5.26); Daylight Orange: Darknight variety (from Audet 2006:Figure 5.27); Augustine Red (from Gifford 1976:Figure 189z); Unknown scroll-foot polychrome; Paxcaman Red (from Gifford 1976:Figure 194h); Picu Incised: Paxcaman variety (from Audet 2002:Figure 3); and Ixpop Polychrome (from Gif- ford 1976:Figure 196a). 15   1070 J A Hoggarth et al. (Audet 2002; Hoggarth 2012). The 14C dates do not support this proposition; although the sample of dated burials is relatively small, not one date falls within the timeframe associated with the Terminal Classic period in the Belize Valley. The temporal correspondence between the Late Classic burials, with the 2σ ranges of all of the burials not extending past AD 800, suggests that the political and demographic collapse at Baking Pot may have been earlier than previously interpreted. 14C dates for some other sites in the Belize Valley (e.g. Xunantunich, San Lorenzo, and Chan) date as late as AD 1000 (Figure 7). However, the large error ranges associated with many of the charcoal 14C dates in the region currently lack the precision to gauge the timing of the collapse that we have proposed for Baking Pot. Figure 7 Calibrated 14C dates from various sites in the published literature of the Belize Valley (Actuncan: LeCount et evident through the presence of late facet Spanish Lookout ceramics in construction fill of plazas al. 2002:Table 3; Baking Pot: this article; Blackman Eddy: Garber et al. 2004:Table 2, Brown 2008:Table 1; Buenavis- taand del monumental architecture Cayo: Helmke et al. 2008:45, (Aimers 1997; Audet Peuramaki-Brown 2006; 2012:683; Cahal Helmke 2008) Pech: Awe and domestic 1992:Table groups 1 and Appendix 2, (Audet 2002; Hoggarth 2012). The radiocarbon dates do not support this proposition; although Awe and Healy 1994:196, Healy and Awe 1995:Table 1, Ebert et al. 2014; Chaa Creek: LeCount et al. 2002:Table 3; Chan: the sample of dated Kosakowsk burials3.2; y 2012:Table Las Ruinassmall, is relatively de Arenal: not one date falls Tascheck within and Ball Pacbitun: the time 1999:228; frameHealy associated 1990:Ta- ble 1, withHohmann and Powis the Terminal 1999:Table Classic period1;inSan theLorenzo: Belize LeCount Valley. et al. 2002:Table The 3) with the bone collagen temporal correspondence betweenAMSthe 14 C dates LateatClassic Baking Pot showing burials, withthe the gap 2σ between theof ranges Classic all ofand thePostclassic burials notperiods. extending past AD 800, suggests that the political and demographic collapse at Baking Pot may have been earlier than previously interpreted. Dates Radiocarbon from burials recovered dates for some other in Postclassic sites inatthe contexts BelizePot Baking Valley (e.g. Xunantunich, revealed Sanover a temporal gap 500Lorenzo, andlatest yr after the Chan) dateClassic dated as lateperiod as AD 1000at(Figure burials the site,7). However, falling the cal between large ADerror 1280ranges and 1420 associated with many of the charcoal radiocarbon dates in the region currently lack the precision during the Late Postclassic. Previous interpretations of the Postclassic populations at the site have to gauge the timing of the collapse that we have proposed for Baking Pot. Dates from burials recovered in Postclassic contexts at Baking Pot revealed a temporal gap over 500 years after the latest dated Classic period burials at the site, falling between cal AD 1280 and 1420 during the Late Postclassic. Previous interpretations of the Postclassic populations at the 16   Questioning Postclassic Continuity at Baking Pot, Belize 1071 argued for a strong Early Postclassic occupation (Willey et al. 1965) with waves of abandonment and small-scale repopulation during the Late Postclassic period (Aimers 2003:160). The presence of Augustine Red ceramics, which have been widely accepted as being Early Postclassic in age, in Burial 101-1 contrasts with the 14C date, suggesting that the transition between Augustine Red and Paxcaman Red, if a temporal sequence in types exists at Baking Pot (and if these ceramics were not heirloom pieces), may have occurred between cal AD 1330 to 1385 (95.4%). The clustering of all three dates during the Late Postclassic suggests that early facet New Town ceramics at Baking Pot may be much later than has been presented. No other burials with New Town ceramics, or 14C dates that are distinctly within the Postclassic (Figure 7), have been reported in the literature of the area. Willey et al. (1965:545–8) lists 16 burials that they assigned to the broad temporal span of the Spanish Lookout/New Town complexes, although none include Postclassic ceramics or show the same burial pattern as the Baking Pot Postclassic burials. As with other Postclassic sites in the Maya lowlands, the mixed contexts of deposits with New Town ceramics at Baking Pot and Barton Ramie have made chronological associations difficult. Directly dating human burials from Baking Pot provides one approach for testing the hypothesis of Postclassic continuity in the Belize Valley that avoids many of the difficulties that have long plagued the archaeology of the Postclassic Maya. While these results appear to support a re-occu- pation scenario in the Late Postclassic rather than continuity, the sample remains small and we are designing a research program to evaluate this hypothesis more completely at Baking Pot. These investigations will focus on dating a larger number of burials, in addition to faunal remains from Classic and Postclassic deposits in the settlement. This will help to determine whether the large temporal gap between Classic and Postclassic burials at Baking Pot is the result of reoccupation or instead represents differential burial activity. CONCLUSIONS Direct AMS 14C dating of 12 burials at Baking Pot has provided new information about the nature and timing of occupation at the site. While most dates were in agreement with the archaeological record, several unexpected dates suggest that the Terminal Classic and Postclassic ceramic chronol- ogy in the Belize Valley may require re-evaluation. Dates from Postclassic burials at Baking Pot do not conform to Willey’s hypothesis of Postclassic continuity, and these results may have impli- cations for understanding the Postclassic occupation of Barton Ramie and the population history of the Belize Valley in general. This research demonstrates the utility of using 14C dating human or faunal remains for understand- ing the problems associated with dating shallow and often mixed contexts that have long plagued archaeological research on the Postclassic Maya. Analysis of bone from tropical contexts is now more feasible because new filtration methods allow purification of samples from challenging pres- ervational contexts like the lowland tropics. Quality control measures using the % yield of collagen and stable nitrogen and carbon isotopic analysis provide greater confidence in the results. AMS 14C dating of human remains also has the advantage of providing more precise temporal control that may aid in further refinement of the ceramic chronology from primary deposits in the Belize Valley and elsewhere in the Maya lowlands. Although the utility of 14C dating for refining ceramic chronol- ogies continues to face resistance in Maya archaeology, we argue that when high-precision AMS 14 C dating is used in conjunction with a priori contextual information, the comparison of direct dates from burials with associated ceramic types may provide better chronological control than ceramic seriation alone. The results of AMS 14C dating of the burials from Baking Pot have provided new in- sights into the occupation of the Belize Valley, showing a complex history of occupation over nearly 1072 J A Hoggarth et al. 2 millennia, and posing new questions about the relationships between the Classic and Postclassic inhabitants at the site with nearby regions. ACKNOWLEDGMENTS Funding for this research was granted by Penn State, the National Science Foundation (Kennett: BSC-0940744) and the Center for Latin American Studies at the University of Pittsburgh. Exca- vation and survey research was conducted under the auspices of the Belize Valley Archaeological Reconnaissance (BVAR) project. We thank the staff and students of the 2007 to 2010 field seasons of the BVAR field school for their contributions to the fieldwork, as well as the Belize Institute of Archaeology for permission to conduct research at Baking Pot. We extend our gratitude to several individuals who have had a part in the Baking Pot research, especially Carolyn Audet, Jim Con- lon, and Christophe Helmke. Osteological analysis of the burials at Baking Pot was conducted by Jennifer Piehl, Carolyn Freiwald, and Anna Novotny. We also thank Karen and Jenny Trippett for their work processing the bone samples. Mark Robinson and Carolyn Freiwald, as well as the three anonymous reviewers, offered helpful comments on the manuscript, which has served to clarify several critical points. REFERENCES Adams REW. 1973. Maya collapse: transformation and lize Valley Archaeological Reconnaissance Project: termination in the ceramic sequence at Altar De A Report of the 2001 Field Season. Belmopan: Be- Sacrificios. In: Culbert TP, editor. The Classic Maya lize Institute of Archaeology. p 91–110. Collapse. Albuquerque: School of American Re- Audet CM. 2006. Political organization in the Belize search, University of New Mexico Press. p 133–63. Valley: excavations at Baking Pot, Cahal Pech and Aimers JJ. 1997. Preliminary investigations of architec- Xunantunich [unpublished PhD dissertation]. Nash- ture in Plaza 2 of Group 1 at Baking Pot, Belize. ville: Vanderbilt University. In: Awe JJ, Conlon JM, editors. The Belize Valley Audet CM, Awe JJ. 2004. What’s cooking at Baking Pot: Archaeological Reconnaissance Project: A Report a report of the 2001 to 2003 seasons. Research Re- of the 1996 Field Season. Peterborough: Trent Uni- ports in Belizean Archaeology 1:49–59. versity. p 7–20. Audet CM, Awe JJ. 2005. The political organization Aimers JJ. 2003. Abandonment and nonabandonment at of the Belize Valley: evidence from Baking Pot, Baking Pot, Belize. In: Inomata T, Webb RW, edi- Belize. Research Reports in Belizean Archaeology tors. The Archaeology of Settlement Abandonment in 2:357–64. Middle America. Salt Lake City: University of Utah Awe JJ. 1992. Dawn in the land between the rivers: Press. p 149–62. formative occupation at Cahal Pech, Belize, and Aimers JJ. 2004. Cultural Change on a Temporal and its implications for Preclassic development in the Spatial Frontier: Ceramics of the Terminal Classic central Maya lowlands [unpublished PhD disserta- to Postclassic Transition in the Upper Belize River tion]. London: Institute of Archaeology, University Valley. BAR International Series 1325. Oxford: of London. Archaeopress. Awe JJ. 2013. Journey on the Cahal Pech time machine: Aimers JJ. 2007. What Maya collapse? Terminal Classic an archaeological reconstruction of the dynastic se- variation in the Maya lowlands. Journal of Archaeo- quence at a Belize Valley polity. Research Reports in logical Research 15(4):329–77. Belizean Archaeology 10:33–50. Andres CR, Helmke C, Morton SG, Wrobel GD, Awe JJ, Healy PF. 1994. Flakes to blades? Middle For- González JJ. 2014. Contextualizing the glyphic mative development of obsidian artifacts in the up- texts of Tipan Chen Uitz, Cayo District, Belize. Lat- per Belize River valley. Latin American Antiquity in American Antiquity 25(1):46–64. 5(3):193–205. Andrews AP, Andrews EW, Castellanos FR. 2003. The Ball JW. 1977. The Archaeological Ceramics of Becan, Northern Maya collapse and its aftermath. Ancient Campeche, Mexico. New Orleans: Tulane Universi- Mesoamerica 14(1):151–6. ty, Middle American Research Institute. Aquino VV, Prufer KM, Meredith C, Culleton BJ, Ken- Bove FJ. 1981. Trend surface analysis and the low- nett DJ. 2013. Constraining the age of abandonment land Classic Maya collapse. American Antiquity of Uxbenk Á site core using archaeological stratig- 46(1):93–112. raphy and AMS 14C dates. Research Reports in Be- Bronk Ramsey C. 2009. Bayesian analysis of radiocar- lizean Archaeology 10:269–80. bon dates. Radiocarbon 51(1):337–60. Audet CM. 2002. Excavations of Structure 198, Baking Brown MK. 2008. Establishing hierarchies in the middle Pot, Belize. In: Griffith CS, Awe JJ, editors. The Be- Preclassic Belize River valley. Research Reports in Questioning Postclassic Continuity at Baking Pot, Belize 1073 Belizean Archaeology 5:175–84. tion program at Baking Pot: the Northeast Baking Brown TA, Nelson DE, Vogel JS, Southon JR. 1988. Pot and North Caracol farm settlement clusters. In: Improved collagen extraction by modified Longin Griffith CS, Ishihara R, Awe JJ, editors. The Western method. Radiocarbon 30(2):171–7. Belize Regional Cave Project a Report of the 1999 Bullard WR. 1973. Postclassic culture in central Peten Field Season. Durham: Department of Anthropolo- and adjacent British Honduras. In: Culbert TP, edi- gy University of New Hampshire. p 55–72. tor. The Classic Maya Collapse. Albuquerque: Uni- Ehret JJ, Conlon JM. 2001. Results of the second season versity of New Mexico Press. p 245–308. (2000) surface collection program at Baking Pot: the Bullard WR, Bullard MR. 1965. Late Classic Finds at Northeast Baking Pot and North Caracol farm set- Baking Pot, British Honduras. Toronto: University tlement clusters. In: Ishihara R, Griffith CS, Awe JJ, of Toronto Press. editors. The Western Belize Regional Cave Project a Cecil LG. 2009. Technological styles of Slipped Pottery Report of the 2000 Field Season. Durham: Depart- and Kowoj identity. In: Rice PM, Rice DS, editors. ment of Anthropology University of New Hamp- The Kowoj: Identity, Migration, and Geopolitics in shire. p 309–24. the Late Postclassic Petén, Guatemala. Boulder: Foias AE, Bishop RL. 1997. Changing ceramic produc- University Press of Colorado. p 238–75. tion and exchange in the Petexbatún region, Gua- Chase AF, Chase DZ. 1985. Postclassic temporal and temala: reconsidering the Classic Maya collapse. spatial frames for the Lowland Maya: a background. Ancient Mesoamerica 8:275–91. In: Chase AF, Rice PM, editors. The Lowland Maya Garber JF, Brown MK, Awe JJ, Hartman CJ. 2004. Mid- Postclassic. Austin: University of Texas Press. dle Formative prehistory of the central Belize Val- p 9–22. ley: an examination of architecture, material culture, Chase AF, Garber JF. 2004. The archaeology of the Be- and sociopolitical change at Blackman Eddy. In: lize Valley in historical perspective. In: Garber JF, Garber JF, editor. The Ancient Maya of the Belize editor. The Ancient Maya of the Belize Valley: Half Valley: Half a Century of Archaeological Research. a Century of Archaeological Research. Gainesville: Gainesville: University Press of Florida. p 25–47. University Press of Florida. p 1–14. Gifford JC. 1976. Prehistoric Pottery Analysis and the Conlon JM. 1994. The final frontier: settlement survey Ceramics of Barton Ramie in the Belize Valley. at the Maya site of Baking Pot. In: Conlon JM, Awe Memoirs of the Peabody Museum of Archaeology JJ, editors. The Belize Valley Archaeological Re- and Ethnology. Cambridge: Harvard University. connaissance Project: A Report of the 1993 Field Graham E. 2004. Lamanai reloaded: alive and well in Season. Peterborough: Trent University. p 81–102. the Early Postclassic. Research Reports in Belizean Conlon JM, Ehret J. 2000. Ancient Maya settlement at Archaeology 1:223–42. Baking Pot, Belize: result of the continually expand- Graham E, Bennett S. 1989. The 1986–1987 excavations ing survey program in the search for the end of the at Negroman-Tipu, Belize. Mexicon 11(6):114–7. final frontier. In: Griffith CS, Ishihara R, Awe J, ed- Guderjan TH. 1995. Maya settlement and trade of Am- itors. The Belize Valley Archaeological Reconnais- bergris Caye, Belize. Ancient Mesoamerica 6:147– sance Project: A Report of the 1999 Field Season. 59. Durham: University of New Hampshire. p 43–55. Hammond N. 1976. Radiocarbon chronology for ear- Culleton BJ, Prufer KM, Kennett DJ. 2012. A Bayesian ly Maya occupation at Cuello, Belize. Nature AMS 14C chronology of the Classic Maya center of 260(5552):579–81. Uxbenká, Belize. Journal of Archaeological Science Hammond N, Donaghey S, Berger R, De Atley S, Swit- 39(5):1572–86. sur V, Ward A. 1977. Maya Formative phase radio- Demarest AA, Rice PM, Rice DS, editors. 2004. The carbon dates from Belize. Nature 267(5612):608–10. Terminal Classic in the Maya Lowlands: Collapse, Hammond N, Pring D, Wilk R, Donaghey S, Saul FP, Transition, and Transformation. Boulder: Universi- Wing ES, Miller AV, Feldman LH. 1979. The earli- ty Press of Colorado. est lowland Maya? Definition of the Swasey phase. DeNiro MJ. 1985. Postmortem preservation and al- American Antiquity 44:92–110. teration of in vivo bone collagen isotope ratios in Haug GH, Hughen KA, Sigman DM, Peterson LC, Röhl relation to palaeodietary reconstruction. Nature U. 2001. Southward migration of the Intertropical 317(6040):806–9. Convergence Zone through the Holocene. Science Ebert CE, Kennett DJ, Awe JJ. 2014. Excavations at 293(5533):1304–8. Tzutziiy K’in: Formative period settlement and Healy PF. 1990. Excavations at Pacbitun, Belize: pre- Classic development at an ancient Maya house liminary report on the 1986 and 1987 investigations. group. Poster presented at the 79th Annual Meeting, Journal of Field Archaeology 17(3):247–62. Society for American Archaeology, Austin. Healy PF, Awe JJ. 1994. Radiocarbon dates from Cahal Ebert CE, Prufer KM, Macri MJ, Winterhalder B, Ken- Pech, Belize: results from the 1994 field season. In: nett DJ. In press. Terminal long count dates and the Healy PF, Awe JJ, editors. Belize Valley Preclassic disintegration of Classic Period Maya polities. An- Maya Project: Report on the 1994 Field Season. cient Mesoamerica. Peterborough: Department of Anthropology Trent Ehret JJ, Conlon JM. 2000. Results of the surface collec- University. p 198–215. 1074 J A Hoggarth et al. Helmke CGB. 2008. Excavations of Structures B1 and S, Neff H, Awe J, Graham HV, Freeman KH, New- B7 at Baking Pot, Belize. In: Helmke CGB, Awe som L, Lentz DL, Anselmetti FS, Robinson M, Mar- JJ, editors. The Belize Valley Archaeological Re- wan N, Southon JR, Hodell DA, Haug GH. 2013. connaissance Project: A Report of the 2007 Field Correlating the ancient Maya and modern European Season. Belmopan: Belize Institute of Archaeology. calendars with high-precision AMS 14C dating. Na- p 109–44. ture Scientific Reports 3:1–5. Helmke C, Awe J. 2008a. New site description and struc- Kosakowsky LJ. 2012. Ceramics and chronology at ture designations of Baking Pot, Belize. In: Helmke Chan. In: Robin C, editor. Chan: An Ancient Maya C, Awe J, editors. The Belize Valley Archaeological Farming Community. Gainesville: University Press Reconnaissance Project: A Report of the 2007 Field of Florida. p 42–70. Season. Belmopan: Belize Institute of Archaeology. Lamb C. 2009. Test pit excavations at M-112, Baking p 81–102. Pot, Belize. In: Hoggarth JA, Awe JJ, editors. The Helmke CGB, Awe JJ. 2008b. Organización territorial de Belize Valley Archaeological Reconnaissance Proj- los antiguos Mayas de Belice central: confluencia de ect: A Report of the 2008 Field Season. Belmopan: datos arqueológicos y epigráficos. Mayab 20:65–91. Institute of Archaeology. p 60–70. Helmke CGB, Awe JJ. 2012. Ancient Maya territorial Law IA, Housley RA, Hammond N, Hedges REM. organisation of central Belize: confluence of archae- 1991. Cuello: resolving the chronology through di- ological and epigraphic data. Contributions in New rect dating of conserved and low-collagen bone by World Archaeology 4:57–88. AMS. Radiocarbon 33(3):303–15. Helmke C, Ball JW, Mitchell PT, Taschek JT. 2008. LeCount L. 1996. Pottery and power: feasting, gifting, Burial BVC88-1/2 at Buenavista del Cayo, Belize: and displaying wealth among the Late and Terminal resting place of the last king of Puluul? Mexicon Classic Lowland Maya [unpublished PhD disserta- XXX(2):43–9. tion]. University of California, Los Angeles. Hogg AG, Hua Q, Blackwell PG, Niu M, Buck CE, Guil- LeCount LJ, Yaeger J, Leventhal RM, Ashmore W. 2002. derson TP, Heaton TJ, Palmer JG, Reimer PJ, Reimer Dating the rise and fall of Xunantunich, Belize: a RW, Turney CSM, Zimmerman SRH. 2013. SHCal13 Late and Terminal Classic lowland Maya regional Southern Hemisphere calibration, 0–50,000 years cal center. Ancient Mesoamerica 13(1):41–63. BP. Radiocarbon 55(4):1889–903. Martin S, Grube N. 2008. Chronicle of the Maya Kings Hoggarth JA. 2009. Settlement and community orga- and Queens. London: Thames & Hudson. nization in the Classic to Postclassic transition: re- Masson MA. 1995. Context of the Maya Postclassic in search methodology and aims of the 2007 to 2010 Belize. Geoarchaeology 10(5):389–404. settlement research at Baking Pot, Belize. In: Hog- McClure SB, Puchol OG, Culleton BJ. 2010. AMS dat- garth JA, Awe JJ, editors. The Belize Valley Archae- ing of human bone from Cova de la Pastora: new ological Reconnaissance Project: A Report of the evidence of ritual continuity in the prehistory of 2008 Field Season. Belmopan: Belize Institute of eastern Spain. Radiocarbon 52(1):25–32. Archaeology. McKillop H. 1996. Ancient Maya trading ports and the Hoggarth JA. 2012. Social reorganization and household integration of long-distance and regional econo- adaptation in the aftermath of collapse at Baking mies: Wild Cane Cay in south-coastal Belize. An- Pot, Belize [unpublished PhD dissertation]. Univer- cient Mesoamerica 7(1):49-62. sity of Pittsburgh. Pelayo PA. 2010. 2008–2009 excavations of Mound 96 Hohmann B, Powis T. 1999. Excavations of Plaza B at at Baking Pot, Belize. In: Hoggarth JA, Awe JJ, ed- Pacbitun, Belize. In: Healy PF, editor. Belize Valley itors. The Belize Valley Archaeological Reconnais- Maya Project: Report on the 1996 and 1997 Field sance Project: A Report of the 2009 Field Season. Seasons. Peterborough: Department of Anthropolo- Belmopan: Institute of Archaeology. p 69–100. gy Trent University. p 1–18. Pendergast DM. 1985. Lamanai, Belize: an updated Housley RA, Hammond N, Law IA. 1991. AMS radio- view. In: Chase AF, Rice PM, editors. The Low- carbon dating of Preclassic Maya burials at Cuello, land Maya Postclassic. Austin: University of Texas Belize. American Antiquity 56(3):514–9. Press. p 85–90. Inomata T, Triadan D, Aoyama K, Castillo V, Yonenobu Peraza Lope C, Masson MA, Hare TS, Delgado Kú PC. H. 2013. Early ceremonial constructions at Ceibal, 2006. The chronology of Mayapan: new radiocar- Guatemala, and the origins of lowland Maya civili- bon evidence. Ancient Mesoamerica 17(2):153–75. zation. Science 340(6131):467–71. Peuramaki-Brown MM. 2012. The integration and dis- Kennett DJ, Breitenbach SF, Aquino VV, Asmerom Y, integration of ancient Maya urban centres: charting Awe J, Baldini JU, Bartlein P, Culleton BJ, Ebert C, households and community at Buenavista del Cayo, Jazwa C, Macri MJ, Marwan N, Polyak V, Prufer Belize [unpublished PhD dissertation]. University of K, Ridley HE, Sodemann H, Winterhalder B, Haug Calgary. GH. 2012. Development and disintegration of Maya Premo LS. 2004. Local spatial autocorrelation statistics political systems in response to climate change. Sci- quantify multi-scale patterns in distributional data: ence 338(6108):788–91. an example from the Maya lowlands. Journal of Ar- Kennett DJ, Hajdas I, Culleton BJ, Belmecheri S, Martin chaeological Science 31(7):855–66. Questioning Postclassic Continuity at Baking Pot, Belize 1075 Prufer KM, Moyes H, Culleton BJ, Kindon A, Kennett search, University of New Mexico Press. p 107–31. DJ. 2011. Formation of a complex polity on the east- Santos G, Southon J, Druffel-Rodriguez K, Griffin S, ern periphery of the Maya Lowlands. Latin Ameri- Mazon M. 2004. Magnesium perchlorate as an al- can Antiquity 22(2):199–223. ternative water trap in AMS graphite sample prepa- Pugh TW, Rice PM. 2009. Zacpetén and the Kowoj: field ration: a report on sample preparation at KCCAMS methods and chronologies. In: Rice PM, Rice DS, at the University of California, Irvine. Radiocarbon editors. The Kowoj: Identity, Migration, and Geo- 46(1):165–74. politics in Late Postclassic Petén, Guatemala. Boul- Satterthwaite L, Ralph EK. 1960. New radiocarbon dates der: University Press of Colorado. p 85–122. and the Maya correlation problem. American Antiq- Reents-Budet D, Bishop RL, Taschek JT, Ball JW. 2000. uity 26(2):165–84. Out of the palace dumps: ceramic production and Sidrys R, Berger R. 1979. Lowland Maya radiocar- use at Buenavista Del Cayo. Ancient Mesoamerica bon dates and the Classic Maya collapse. Nature 11(1):99–121. 281(5730):269–74. Reents-Budet D, Bishop RL, Audet C, Awe J, James M. Smith RE. 1955. Ceramic Sequence at Uaxactun, Gua- 2005. Act locally, think internationally: the pottery temala. New Orleans: Middle American Research of Baking Pot, Belize. Research Reports in Belizean Institute, Tulane University. Archaeology 2:365–86. Stafford TW, Brendel K, Duhamel RC. 1988. Radiocar- Reimer PJ, Bard E, Bayliss A, Beck JW, Blackwell PG, bon, 13C and 15N analysis of fossil bone: removal of Bronk Ramsey C, Buck CE, Cheng H, Edwards RL, humates with XAD-2 resin. Geochimica et Cosmo- Friedrich M, Grootes PM, Guilderson TP, Haflida- chimica Acta 52(9):2257–67. son H, Hajdas I, Hatté C, Heaton TJ, Hoffmann DL, Stuiver M, Polach HA. 1977. Discussion: reporting of Hogg AG, Hughen KA, Kaiser KF, Kromer B, Man- 14 C data. Radiocarbon 19(3):355–63. ning SW, Niu M, Reimer RW, Richards DA, Scott Swain L. 2005. Test excavations of Plaza 2, Group 1, EM, Southon JR, Staff RA, Turney CSM, van der Baking Pot. In: Audet CM, Awe JJ, editors. The Be- Plicht J. 2013. IntCal13 and Marine13 radiocarbon lize Valley Archaeological Reconnaissance Project: age calibration curves 0–50,000 years cal BP. Radio- A Report of the 2004 Field Season. Belmopan: Be- carbon 55(4):1869–87. lize Institute of Archaeology. p 13–20. Rice PM. 1979. Ceramic and nonceramic artifacts of Taschek JT, Ball JW. 1999. Las Ruinas de Arenal: pre- Lakes Yaxha-Sacnab, El Peten, Guatemala, Part I, liminary report on a subregional major center in the the ceramics, Section B, Postclassic pottery from western Belize Valley (1991–1992 excavations). An- Topoxte. Ceramica de Cultura Maya 11:1–86. cient Mesoamerica 10(2):215–35. Rice PM. 1987. Mancanche Island, El Peten, Guatema- van Klinken GJ. 1999. Bone collagen quality indicators la: Excavations, Pottery, and Artifacts. Gainesville: for palaeodietary and radiocarbon measurements. University of Florida Press. Journal of Archaeological Science 26(6):687–95. Rice PM. 2013. Type variety: what works and what Ward G, Wilson S. 1978. Procedures for comparing and doesn’t. In: Aimers JJ, editor. Ancient Maya Pottery: combining radiocarbon age determinations: a cri- Classification, Analysis, and Interpretation. Gaines- tique. Archaeometry 20(1):19–31. ville: University Press of Florida. p 11–28. Webster DL. 2012. The Classic Maya Collapse. The Ox- Rice PM, Rice DS, editors. 2009. The Kowoj: Identi- ford Handbook of Mesoamerican Archaeology. Ox- ty, Migration, and Geopolitics in Late Postclassic ford: University of Oxford Press. p 324–34. Petén, Guatemala. Boulder: University Press of Willey GR. 1973. Certain aspects of the Late Classic to Colorado. Postclassic periods in the Belize Valley. In: Culbert Ricketson OG. 1931. Excavations at Baking Pot, British TP, editor. The Classic Maya Collapse. Albuquer- Honduras. Washington, DC: Carnegie Institution. que: School of American Research, University of Sabloff JA. 1973. Continuity and disruption during Ter- New Mexico Press. p 93–106. minal Late Classic times at Seibal: ceramic and other Willey GR, Bullard WR, Glass JB, Gifford JC, editors. evidence. In: Culbert TP, editor. The Classic Maya 1965. Prehistoric Maya Settlements in the Belize Collapse. Albuquerque: School of American Re- Valley. Cambridge: Cambridge University Press.