Study Sites in the BEAST Permit Area

The BEAST permit area encompasses 560 km² within the southeastern Three Rivers adaptive region (TRR) in northwestern Belize. Reconnaissance, survey, and visual inspection of 650 km² of lidar data have identified 43 major or minor centers in and adjacent to the permit area (see Houk et al. Reference Houk, Thompson and Novotny2026). The preliminary analysis of the lidar data reveals nuanced details about localized topography and features at each site, allowing us to further refine previous settlement survey data from the BEAST area such as those from Chan Chich, as well as settlement data for new sites such as Ayiin Winik.

Chan Chich

Chan Chich is the largest site in the BEAST permit area. It is near the headwaters of the Rio Bravo, with the dense settlement system focused on low hills and ridges. Archaeologists first recorded Chan Chich in 1989 (Guderjan Reference Guderjan and Guderjan1991), and the CCAP surveyed more than 1.5 km² of the site in 1996 (Houk Reference Houk2015:190). The site’s monumental precinct includes a public plaza, a ballcourt, an acropolis, a marketplace, and causeways. Most of the settlement comprises courtyards, plazuelas, and individual structures (Figure 2). The surveyed area includes 130 residential groups; although more are visible on the lidar data, these groups are yet to be field verified. Excavation data demonstrate occupation in the site core beginning as early as 800 BC and extending to approximately AD 850, with a peak during the Late Classic (Houk Reference Houk2015).

Figure 2.

Settlement map of Chan Chich (top) and Ayiin Winik (bottom) illustrating the location of major plazas and residential groups (gray polygons) near the site core used in the analysis. Lidar-derived basemaps are a 0.5 m DEM over a simple local relief model (SLRM). Lidar data courtesy of BEAST. (Color online)

Figure 2 Long description

The first map shows the Chan Chich site with labeled features such as North Plaza, Main Plaza, Upper Plaza, Back Plaza and Norman′s Temple. Structures and residential groups are marked, with the Rio Bravo river visible. The second map depicts the Ayiin Winik site, highlighting the Main Plaza, Acropolis, Palace and Double Ball Court. Both maps use symbols to differentiate structures and residential groups, with a scale bar indicating distances in meters.

Ayiin Winik

Ayiin Winik, another major site in the BEAST area, is approximately 14 km north of Chan Chich (see Figure 1). Archaeologists first identified Ayiin Winik in the BEAST lidar data in December 2022 and field verified and mapped the monumental core in spring 2023 (Houk et al. Reference Houk, Thompson and Novotny2026). Ayiin Winik’s monumental core includes a large plaza, flanked by an acropolis on the west and an elevated palace group on the east, a rare double ballcourt, and a wide parapet-lined causeway. Nearby settlement includes primarily courtyards and plazuelas, many on hilltops (Figure 3). A cluster of residences of well-defined house lots with delineated entryways and prominent linear stone features or terraces is located 1.5 km southeast of the Main Plaza. Preliminary excavation data from the civic ceremonial core suggest a Classic period construction, although Postclassic revisitations occurred as well (Ingalls and Houk Reference Ingalls and Houk2026).

Figure 3.

Residential groups of varying size and complexity at Chan Chich (left) and Ayiin Winik (right). Top images show the acropolis, and bottom images display residential groups of varying size (yellow arrows). All maps are the same scale. The basemap is a lidar-derived visualization for archaeological topography (VAT) image. Lidar data courtesy of BEAST. (Color online)

Figure 3 Long description

The top left image shows the Chan Chich Acropolis, featuring rectangular structures. The top right image displays the Ayiin Winik Acropolis, with similar rectangular formations. The bottom left image illustrates Chan Chich Households, marked by yellow arrows indicating various residential structures. The bottom right image presents Ayiin Winik Households, also marked with yellow arrows highlighting residential features. A scale bar indicates distances in meters, ranging from 0 to 50 and a north arrow is present for orientation.

Methods: Measuring Inequality

To evaluate inequality, I calculated a Gini coefficient using house size metrics. First, I measured the areas and volumes of residential groupsFootnote ¹ at Chan Chich and Ayiin Winik (see Supplementary Table 1), combining data from traditional pedestrian survey and digital data in GIS. Then, I calculated Gini coefficients for each site using R. The Gini coefficient assesses the unevenness of a distribution of data (Peterson and Drennan Reference Peterson, Drennan, Kohler and Smith2018)—in this case, house sizes—with scores ranging from 0 (perfect equality) to 1 (perfect inequality); it is used among archaeologists in the Maya region (Chase et al. Reference Chase, Thompson, Walden and Feinman2023) and beyond (Kohler et al. Reference Kohler, Bogaard, Ortman, Crema, Chirikure, Cruz and Green2025), thereby enabling broader regional comparisons. Given the varied occupational longevity of Chan Chich and Ayiin Winik, I applied a “snapshot” approach, which assumes that most of the settlement visible on the lidar dates to the final period of occupation during the Late Classic (Canuto et al. Reference Canuto, Estrada-Belli, Garrison, Houston, Acuña, Kováč and Marken2018), around AD 800, before the widespread abandonment of the Maya Lowlands (see Supplementary Text 1 for additional details).

Results: Preliminary Observations of Inequality in the Southeastern TRR

In this section I report preliminary results of inequality at Chan Chich and Ayiin Winik. For the sake of brevity and comparison to other studies, I discuss residential group area and volume. The differences between area and volume help reveal some households’ increased ability to invest in labor for more complex architectural constructions compared to others; it likely reflects social signaling in which elites reinforced their claims of lineage and political importance through domestic spaces. I acknowledge that house size metrics are but one proxy for evaluating inequality. Other scholars have evaluated abundance and suites of artifacts, distance to urban services, and grave goods and architecture to evaluate inequality, resulting in a deeper understanding of the nuances of ancient Maya behaviors and displays of wealth (Hutson Reference Hutson and Scott R.2023; Munson and Scholnick Reference Munson and Scholnick2022; Thompson, Feinman, and Prufer Reference Thompson, Feinman and Prufer2021). Future BEAST household excavations will target residential groups of varying size and complexity to corroborate these findings.

The Gini coefficients for both Chan Chich and Ayiin Winik allude to high degrees of wealth inequality. Chan Chich has an area Gini score of 0.65 and a volume Gini score of 0.89 (Table 1; Figure 4). Ayiin Winik shows similar trends, with Gini coefficients of 0.56 and 0.77 for area and volume, respectively. These values are corroborated by the distribution in the size of residential groups evidenced by the mean and median values. Both sites have higher mean house sizes than median house sizes, indicating that a few residential groups at the site are substantially larger than most residential groups. Furthermore, Lorenz curves for area and volume indicate that the top 10% of the population held 60%–85% of the wealth at Chan Chich and 50%–70% of the wealth at Ayiin Winik (Figure 4).

Figure 4.

Box-and-whisker plots of Gini coefficients with 95% confidence intervals (top). Lorenz curves show the proportion of the population and proportion of wealth (bottom). (Color online)

Figure 4 Long description

A box and whisker plot and four Lorenz curves. The box and whisker plot displays Gini coefficients for Chan Chich and Ayiin Winik. Chan Chich shows a Gini coefficient of 0.65 for Area and 0.89 for Volume, indicating higher inequality in Volume. Ayiin Winik has a Gini coefficient of 0.56 for Area and 0.77 for Volume, also showing greater inequality in Volume. Overall, Chan Chich exhibits more inequality than Ayiin Winik. The Lorenz curves illustrate the proportion of population versus wealth. For Chan Chich, the Area curve bows less from the equality line than the Volume curve, indicating less inequality. Similarly, for Ayiin Winik, the Area curve bows less than the Volume curve. The horizontal axis for the Lorenz curves is labeled Proportion of Population and the vertical axis is labeled Proportion of Wealth, both ranging from 0.0 to 1.0. The Gini values are consistent across the box plot and Lorenz curves, with the exception of Ayiin Winik Area, which is 0.56 in the box plot and 0.57 in the Lorenz curve, likely due to rounding. The visual differentiation uses distinct styles to represent Area and Volume for each group.

Table 1.

House Size Metrics and Gini Coefficients for Study Sites (see Supplementary Table 2).

Table 1 Long description

Metrics compare Chan Chich and Ayiin Winik for documented temporal occupation, surveyed area, residential mapped structures and groups, and group area and volume statistics with Gini values. Documented temporal occupation is 800 BC to AD 850 for Chan Chich and AD 400 to 850 for Ayiin Winik. Surveyed area is 1.57 square kilometers at Chan Chich and 1.44 square kilometers at Ayiin Winik. Residential mapped structures are 247 at Chan Chich and 200 at Ayiin Winik, and residential mapped groups are 130 and 99. Mean area in square meters is 521.68 at Chan Chich and 490.15 at Ayiin Winik, while median area is 207.75 and 229.97. Mean volume in cubic meters is 836.96 at Chan Chich and 653.8 at Ayiin Winik, while median volume is 85.15 and 146.26.

High degrees of wealth inequality were expected for Chan Chich and Ayiin Winik, so these findings are not surprising. Chan Chich is the largest site in the area, encompassing five minor centers (see Figure 1) with an estimated 25,000 occupants during the Late Classic (Houk et al. Reference Houk, Thompson and Belli2025). It hosts several large plazas and a regional marketplace, with obsidian sourcing data alluding to its interconnectedness in the regional political economy by the Late Classic (Degnan Reference Degnan2026). Similarly, for Ayiin Winik the presence of a rare double ballcourt, massive central plaza, and two palace areas suggest it was a social reactor. Archaeologists have documented elaborate tombs and stone monuments at both sites.

Chan Chich and Ayiin Winik have Gini scores that are similar to those of other sites located in the broader Petén elevated interior region (EIR), such as El Perú-Waka’ (area: 0.64; volume: 0.80), El Pilar (area: 0.56; volume: 0.80), Tikal (area: 0.56), and La Corona (area: 0.54; volume: 0.66; see Supplementary Table 3).Footnote ² In the Maya region, area Gini scores tend to be lower than volume Gini scores by approximately 0.20 ± 0.05 (Thompson et al. Reference Thompson, Chase and Feinman2023), and the BEAST and EIR sites adhere to this trend. Compared to other subregions, such as southern Belize and the Vaca Plateau, the EIR sites have higher Gini coefficients (Figure 5), suggesting higher degrees of wealth inequality in this more centralized region (see Figure 1). The EIR is the Maya heartland, hosting many of the major Maya centers; these sites were occupied for centuries and were well connected, which may have bolstered rulers’ ability to harness power and accumulate wealth.

Figure 5.

Comparison of area and volume Gini scores for Maya sites in different regions (see Figure 1 and Supplementary Table 3). (Color online)

Figure 5 Long description

The scatter plot titled ′Area vs Volume Gini Coefficient and Region′ displays data representing different sites (points) regions (colors). The horizontal axis is labeled ′Area Gini Coefficient′ and ranges from 0 to 1. The vertical axis is labeled ′Volume Gini Coefficient′ and also ranges from 0 to 1. Each point on the plot corresponds to a specific site , with labels such as Chan Chich, Copan, El Peru-Waka, Ayiin Winik and Tikal. The data points are scattered across the plot, generally with a positive trend. Tikal does not have volume Gini data and therefore ispositioned at 0 on the y-axis. The plot provides a visual comparison of the Gini coefficients for different sites and regions.

However, differences in Gini scores may be the outcome of several variables, such as the impact of environmental variability on the construction of residential spaces, differences in what communities value in terms of wealth, or shifts in governance and fiscal financing of sites based on their political and economic ties (Thompson, Feinman, et al. Reference Thompson, Feinman, Lemly and Prufer2021). For example, Caracol was among the most powerful Classic Maya cities, with its political reach extending into the EIR, yet its Gini scores are comparatively low. Chase and Chase (Reference Chase and Chase2017) argue for a form of governance that resulted in “symbolic egalitarianism,” rather than strictly autocratic rule from AD 700 to 800. This “symbolic egalitarianism” is manifested in the distribution of wealth items to households of varying size; small houses may contain prestige and nonlocal goods, and large houses may appear “poor” in regard to their material culture. Similar trends have been noted at Chunchucmil (Hutson Reference Hutson and Scott R.2023). However, in southern Belize, house size often correlates with an abundance of prestige and nonlocal goods (Thompson, Feinman, and Prufer Reference Thompson, Feinman and Prufer2021), highlighting the diversity in human behaviors as they relate to wealth.

Concluding Thoughts

Wealth inequality was undoubtedly present in Classic Maya cities. How wealth manifested across the Maya lowlands varied, but new lidar and settlement surveys provide robust datasets for evaluating inequality in the absence of excavation data, thereby enabling broader regional comparisons. In this article, I evaluated wealth inequality at two sites in the BEAST permit area—Chan Chich and Ayiin Wink—using house size metrics for area and volume to calculate Gini coefficients and Lorenz curves. Chan Chich and Ayiin Winik both display remarkably high Gini scores based on the differences in house size, alluding to high degrees of wealth inequality at these sites, bolstered by their civic ceremonial architecture. Their Gini scores largely parallel those from other sites in the EIR and are higher than those from more peripheral Maya subregions. My findings provide the first quantitative evaluation of residential wealth inequality in the TRR, creating a foundation for future studies that will take a multiproxy approach, expanding on these findings through the integration of other datasets such as artifact suites and burial patterns.