Drought and water management in ancient Maya society

Abstract

Paleoclimate research in the Maya region of Mesoamerica provides compelling evidence of drought during key periods of cultural transition in Maya society. These include the transition from the Preclassic to the Classic, and from Classic to the Postclassic. Previous research emphasized a causal relationship between drought and cultural change, or so-called “collapse” in the Maya region. Recent advances in the range and precision of climate-sensitive proxies and the development of new archives have enabled quantitative reconstructions of past hydroclimate, as well as providing evidence of high impact, short-duration events, such as tropical cyclones. Simultaneously, archaeological research has unearthed widespread evidence of technologies used by the Maya to exert control over water resources in urban, rural, and agricultural settings. Evidence suggests that many of these water features were in use for multiple generations, possibly centuries, and many were constructed during the Terminal Preclassic and Terminal Classic periods. We suggest that, given the availability of new archaeological and paleoclimate records, these data can be combined to identify the full complexity of Maya adaptation to hydroclimate variability to emphasize adaptation and resilience to both water scarcity and over-abundance (e.g., flooding). Such syntheses, which can offer lessons for present-day efforts to grapple with regional climate change, will benefit from additional studies in data-poor zones of the Maya region, as well as public archiving of paleoclimate and archaeological data.

Keywords

Maya drought water management paleoclimatology archeology

Introduction

The Maya Lowlands of Central America, the region spanning the historic heart of the Maya civilization in southeastern Mexico, Belize, Guatemala, Honduras, and El Salvador, are projected to become more arid by the end of the 21st century. Current generation Earth system models (ESMs) suggest that this drying will extend across much of Central America and the Caribbean and will result from reductions in precipitation and increases in evaporative demand (Bhattacharya and Coats 2020; Fuentes-Franco et al., 2015; Herrera et al., 2018; Karmalkar et al., 2011; Magaña et al., 1999; Rauscher et al., 2011). Some analyses of observational data indicate that future changes in hydroclimate come on the heels of robust reductions in regional moisture availability over the 20th century (Herrera et al., 2018; Neelin et al., 2006). However, other work suggests that in many regions the magnitude of internal climate variability masked the signal of forced hydroclimate changes in the 20th century (Anderson et al., 2019). These analyses highlight the fact that variation in hydroclimate, with impacts for water resources, are a defining feature of both current, and most likely future, conditions in the region.

The societal consequences of hydroclimate variability and change are heavily influenced by fluctuations in environmental conditions as well as differences in forms of settlement, management practices, or governance. The overall climate of the Maya Lowlands is strongly seasonal, with a pronounced dry season in boreal winter. However, local climatologies vary spatially and are strongly influenced by orographic rainfall patterns. For instance, there is a steep north-south rainfall gradient from the northern tip of the Yucatán Peninsula into the interior of Guatemala: rainfall varies from 1.6 mm/day at the northern coast to over 13 mm/day at the base of the Maya Mountains and Sierra Madre (Douglas et al., 2015). In addition, while regions of Guatemala, Honduras, El Salvador, and Nicaragua experience a strong midsummer drought or canícula in the late summer, the magnitude of this caníıcula is quite weak in the Yucatán (Anderson et al., 2019; Magaña et al., 1999; Maldonado et al., 2016). Parts of the northeast Yucatán Peninsula receive approximately 10–15% of annual rainfall from tropical cyclones (Prat and Nelson, 2013). Variations in tropical cyclone activity may therefore have a stronger influence on water resources in this region of the Maya Lowlands. Some regions also receive occasional winter rainfall from storms known as “nortés.” Given this backdrop of spatial and temporal climate variability, the region features agricultural practices that are strongly attuned to changes in water availability. In fact, literature suggests that farmers across Central America more broadly are already altering agricultural practices in response to increasing drought and delayed onset of seasonal rains (Harvey et al., 2018; Rogé et al., 2014).

Past societies of Central America also grappled with significant climate variability, including extended periods of drought. Figure 1 maps key paleoclimate records spanning the Middle Preclassic through the Terminal Classic period (approximately 1000 BCE to 900 CE), as well as locations of Preclassic through Postclassic water management features that have been excavated and dated across the Maya Lowlands. Available paleoclimate records provide evidence of extended drought near or during the Terminal Classic and Late Preclassic period, coincident with societal upheaval in the Maya Lowlands (Aimers 2007; Brenner et al., 2002; Ebert et al., 2017). Many studies have posited a causal link between drought and societal collapse (Hodell et al., 1995; Kennett et al., 2012; Medina-Elizalde and Rohling 2012). However, an increasing volume of archaeological evidence points to widespread use of Maya technologies for adapting to hydroclimate variability and drought, reflecting a more nuanced and heterogeneous response to hydroclimate stress. Here, we review available paleoclimate evidence of hydroclimate changes from the Middle Preclassic through Terminal Classic period and review archaeological evidence for Maya water management (Figure 1). We suggest that this rich record of paleoclimate and archaeological data now allows for both qualitative and quantitative synthesis studies that analyze the full complexity of the relationship between climate and shifting human water management practices. We conclude by arguing that these efforts will be further bolstered by the creation of high-resolution, well-dated proxy records of novel climatic phenomena from understudied locations, application of spatiotemporal methods for analyzing regional patterns of change, and the intensive excavation and dating of human adaptation features at both the local and regional levels. Public archiving of both paleoclimate and archaeological data is also critical to these efforts.

Figure 1.

Map of the Maya region, showing locations of records of past climate discussed in text. Numbered diamonds correspond to paleoclimate records, whereas filled circles correspond to georeferenced, excavated water features with established chronologies. See text for more details. Background fill is TRMM satellite-derived rainfall rates (Bookhagen and Strecker 2008), and high-resolution topography. See Supplementary File S1 for details on water features, and Supplementary File S2 for links to all paleoclimate datasets. Site numbers: 1) Chichancanab; 2) Punta Laguna; 3) Chaltun Ha; 4) Itzamna; 5) Chaac; 6) Yok Balum; 7) Laguna Yaloch; 8) Macal Chasm; 9) Puerto Arturo.

Evidence for paleoclimate variability in the Maya region

The Maya region has a long history of paleoenvironmental research, stretching back to the 1960s. In the decades following a pioneering study of Lake Petenxil in the Petén region of Guatemala (Cowgill and Hutchinson 1966), improved spatial coverage of paleoclimate records and a wider variety of climate-sensitive proxies have facilitated a more detailed understanding of patterns of paleoclimate change. In much of the region, the paleoclimate signal in proxy records such as pollen or sediment chemistry may be difficult to disentangle from the signature of human modification of the landscape (Leyden 2002). For example, there has been long debate as to whether climatic drying or human activity was the primary driver of forest opening around 3000 BCE, as evidenced by many pollen records across the southern Lowlands (Anselmetti et al., 2007; Jones 1991; Mueller et al., 2009; Pohl et al., 1996; Wahl et al., 2006, 2016). The first appearance of Zea, or maize, pollen in the Maya Lowlands (Pohl et al., 1996; Wahl et al., 2006), coupled with records showing coeval large-scale burning (Anderson and Wahl 2016; Schüpbach et al., 2015) provide compelling evidence that vegetation change at that time was anthropogenic. Recent research shows that maize was a substantial and persistent component of human diet for some individuals in that region by 2750–2050 BCE, indicating agricultural intensification at that time period (Kennett et al., 2020).

Major advances in our understanding of late Holocene climate in the Maya region came with the application of stable oxygen isotope analysis (δ¹⁸O) on authigenic or biogenic carbonate minerals from closed-basin lakes. In these settings, increases in δ¹⁸O of lake water and the carbonate precipitated from it reflect enhanced evaporation relative to precipitation, or more arid conditions. Isotope-based studies from Lake Miragoane in Haiti and Lake Chichancanab and Punta Laguna in the northern Yucatán provided a climatic context for understanding the development of Maya societies (Curtis et al., 1996; Hodell et al., 1991, 1995, 2005). These studies provided some of the first evidence of dry conditions during the Maya Terminal Classic, a period of profound societal reorganization at approximately 900–1000 CE.

Since this initial work, additional hydroclimate proxy records have shed light on climate fluctuations over the last few millennia. Below, we review some broad trends in paleoclimate records available from the Maya Lowlands for the middle Preclassic through Terminal Classic period. We focus on geochemical data from closed-basin lakes, which broadly reflect the ratio of precipitation to evaporation (Curtis et al., 1996; Evans et al., 2018; Hodell et al., 1991, 1995, 2005; Wahl et al., 2013), δ¹⁸O of carbonate deposits in caves (e.g., speleothems), which reflect precipitation amount (Akers et al., 2016; Kennett et al., 2012; Medina-Elizalde et al., 2010; Medina-Elizalde and Rohling 2012), and records of tropical cyclone frequency (Frappier et al., 2014). We account for the age uncertainty inherent in each proxy record we review. In radiocarbon-dated lake sediment cores, calibrated ages for a given depth may have decadal- to centennial-scale uncertainty. This contrasts with much more precisely dated uranium-thorium (U-Th)-based chronologies for speleothem records. Therefore, accounting for age uncertainty by propagating error through the entire proxy record is a critical prerequisite for identifying robust patterns of variability across records, especially when integrating multiple proxy types. As a result, we focus on studies that make their raw age model and proxy data available, either via databases like the NOAA Paleoclimatology Database (https://www.ncdc.noaa.gov/data-access/paleoclimatology-data), or in supplemental files attached to publications, or provided directly by the authors. Details on how to access each record is provided in Supplementary File S2. However, we include a qualitative discussion of other studies that provide important context. For lake records dated via radiocarbon, we present each proxy record with age uncertainty, calculated using the BACON* age modeling package, propagated through the entire proxy timeseries (Blaauw and Christen, 2011). For speleothem records with precise U-Th dates (e.g., all sites excluding the Macal Chasm), we add an age error of 10 years to account roughly for the timescale of groundwater mixing in the epikarst. This approach is like the age modeling methods described in Bhattacharya and Coats (2020).

Whereas the mean climatology of the Maya lowlands is spatially heterogenous, recent syntheses of paleoclimate data, observations, and model simulations suggest that some component of the hydroclimate variability across the region can be explained by large-scale climate modes. These studies have identified Atlantic Multidecadal Variability, the El Niño Southern Oscillation, and the North Atlantic Oscillation as key drivers of interannual and decadal climate variability in this region (Bhattacharya et al., 2017; Bhattacharya and Coats, 2020). Here, we focus on describing qualitative patterns of similarity and difference across records, highlighting areas for future research. We separate our discussion into the northern and southern Maya Lowlands.

The paleoclimate records we reanalyzed from the northern Maya Lowlands, which include a speleothem record and two lake sediment records, indicate a climate shift to drier conditions near the end of the Preclassic, although the driest conditions in several records occur during the Terminal Classic (Figure 2). Details of these climate shifts, however, differ across records. At Punta Laguna, a δ¹⁸O record from biogenic calcite shows a centennial-scale drying trend starting at approximately 100 BCE, culminating in peak dry conditions during the Terminal Classic period (approximately 900 CE) (Curtis et al., 1996). This contrasts with Lake Chichancanab, which shows very little variability until the Terminal Classic (Hodell et al., 1991, 1995, 2005). The higher-resolution Itzamnah speleothem record, which only records the Preclassic and early Classic, suggests a shift to drier conditions at approximately 200 BCE (Medina-Elizalde et al., 2016). All records show evidence of dry conditions during the Terminal Classic. An open question remains as to whether past hydroclimate variability in the northern Yucatán was driven by changes in tropical storms. Medina Elizalde and Rohling (2012) suggest that reductions in tropical storms can explain drought during the Terminal Classic. One of the few available records of tropical cyclone frequency shows that the transition from the Preclassic to the Classic period was accompanied by a reduction in tropical cyclone frequency (Frappier et al., 2014). This record does not point to a significant reduction in storm activity during the Terminal Classic. A key area of disagreement concerns the magnitude of drying during the Terminal Classic, since quantitative reconstructions of hydroclimate from speleothems and lake sediment cores disagree about the amount of precipitation reduction needed to drive drought during this interval (Evans et al., 2018; Medina-Elizalde et al., 2010). Additional quantitative estimates of precipitation, evapotranspiration, and tropical storm frequency during the Terminal Classic are therefore critical for fully understanding climatic conditions during this interval.

Figure 2.

Paleoclimate records from northern Yucatán spanning the Preclassic and Classic Maya periods, numbers correspond to locations in Figure 1. Age uncertainty has been propagated through all records to provide error bounds for each record, with light colors reflecting 2-σ (e.g. 95%) confidence intervals, and darker colors reflecting 1-σ confidence bounds. Y axes are oriented where needed so that down indicates drier conditions (explicitly labeled on the Y axes), except in the case of the Chaltun Ha speleothem, where up on the Y axis reflects the frequency of tropical cyclones. X axis reflects BCE/CE, with BCE on the left side of the plot (see axis label). Vertical dashed lines indicate the boundaries between the Preclassic and Classic, and Classic and Postclassic.

In contrast to the northern region, fewer records are available from the southern Maya Lowlands. Those that exist include two speleothem records from the Macal Chasm and Yok Balum Cave, Belize (Akers et al., 2016; Kennett et al., 2012). Lacustrine records from the Petén region include a sediment geochemistry (% carbonate) record from Laguna Yaloch, and an oxygen isotope record from Puerto Arturo (Wahl et al., 2013, 2014) (Figure 3). Both Yok Balum and Laguna Yaloch show drying following 750 CE, and the Yok Balum stable isotope record also reflects significant drought during the Preclassic to Classic transition (Figure 3). This period of drying between the Preclassic and Classic is reflected in the δ¹⁸O record of biogenic carbonate at Lago Puerto Arturo, and in a compound-specific hydrogen isotope record from Lake Salpetén (Douglas et al., 2015; Wahl et al., 2014). Results from Lago Puerto Arturo also suggest that the Middle to Late Classic was characterized by a drier climate. We note, however, that while Laguna Yaloch, Puerto Arturo, and the Macal Chasm suggest relatively stable conditions during the Classic, Yok Balum reflects a transition to the wettest conditions within the record (Figure 3). The record from the Macal Chasm features an age model with radiocarbon dates and U-Th dates with higher error as a consequence of the high detrital thorium present in this speleothem (Akers et al., 2016). However, this record also shows a drying trend during the Classic and Terminal Classic. More high-resolution, well-dated paleoclimate records from the southern Maya Lowlands are needed to better understand these apparently conflicting results. The scarcity of records from the southern sector of the Maya Lowlands highlights the need to develop additional high-resolution records in this region.

Figure 3.

Paleoclimate records from southern Yucatán spanning the Preclassic and Classic Maya periods, numbers correspond to locations in Figure 1. Age uncertainty has been propagated through all records to provide error bounds for each record, with light colors reflecting 2-σ (e.g. 95%) confidence intervals, and darker colors reflecting 1-σ confidence bounds. Axes are reoriented where needed so that down indicates drier conditions. Vertical bars indicate the boundaries between the Preclassic and Classic, and Classic and Postclassic.

Despite differences between records, some consistent patterns emerge across paleoclimate reconstructions from the Maya Lowlands. First, many records suggest a drying trend or shift in hydroclimate towards the end of the Preclassic. Second, records stretching from the northern to southern Lowlands indicate that the Terminal Classic was an interval of drought and hydroclimatic instability, even accounting for age uncertainty in each record (Bhattacharya et al., 2017). Taken together, available paleoclimate evidence suggests that hydroclimate variability was a persistent feature of regional climate for both Preclassic and Classic Maya society.

Evidence for Maya adaptation to hydroclimate variability and change

Heavy reliance on physical systems for water management is a defining characteristic of historic urban centers in the seasonally dry tropics globally (Scarborough and Isendahl 2020; Scarborough et al., 2012). In the Maya world, water management infrastructure included aguadas (both natural and constructed reservoirs), dams and channels, and wetland raised field/canal systems. Natural or anthropogenic features including wells, springs, chultunes, and natural cenotes were also exploited. The Maya harvested rainwater and managed groundwater in a variety of ways, and much research has focused on the cultural meaning of water within Maya society (Lucero, 2006; Lucero et al., 2014; Luzzadder-Beach et al., 2016; Scarborough, 1998). Another line of research focuses on the ways in which water quantity and quality placed constraints on agricultural practices (Dunning et al., 2002; Luzzadder-Beach and Beach, 2010). Archaeological and landscape survey efforts demonstrate extensive human control over water resources in urban, rural, and agricultural settings. In addition, many water management features were in use for decades or possibly centuries. Many of these features coincide with the transitions from the Preclassic to the Classic, and from the Classic to Postclassic (Lucero et al., 2011; Luzzadder-Beach et al., 2012).

Although certain regions have been extensively studied, there are significant spatial gaps in our understanding that derive, in part, from a lack of survey and excavation outside of major urban centers, often because of the lack of access and difficult field conditions. Figure 1 includes only those sites that have been dated and excavated, though many others have been mapped (see Supplementary File) (Beach and Dunning 1995, 1997; Beach 1998; Beach et al., 2009, 2011, 2015c,b,d, 2018; Brewer et al., 2017; Brewer 2018; Dahlin and Chase 2014; Davis-Salazar 2003; Dunning et al., 1997; Fedick et al., 2000; Ferrand et al., 2012; Garrison and Dunning 2009; Hammond et al., 1987; Hansen et al., 2002; Jacob 1995; Krause et al., 2019a, 2021; Lohse and Findlay 2000; Luzzadder-Beach and Beach 2009; Luzzadder-Beach et al., 2017; Pohl et al., 1990; Pope and Dahlin 1989; Scarborough et al., 1994; Smyth et al. 2017; Scarborough et al., 1995; Turner and Harrison 1981; Wahl et al., 2007; Weiss-Krejci and Sabbas 2002; Wyatt 2014). The majority of features date to the Classic. In many settings, survey and mapping efforts have provided evidence of water features in urban spaces and nearby settlement zones, but additional excavation and dating of these features is often required to provide information on how they were constructed, and the chronologies of their use and abandonment. Analogous work on the 9–15th century CE cities of the Khmer, in Cambodia, provide an example of this analytical focus on the history and operation of water management infrastructure (e.g., Evans 2016; Penny et al. 2019). In the Maya world, extant studies have already yielded critical data on water management features (Brewer et al., 2017; Ferrand et al., 2012; Lentz et al., 2020), but more excavation and reconstruction is needed at a variety of spatial scales, including efforts to reconstruct hinterland and non-elite water features and link these to more urban structures. Recent acquisition of LiDAR data for broad swaths of the Maya Lowlands is revealing the near universal presence of water management features across the region (Krause et al., 2021); these findings provide ample opportunity for focused fieldwork to better understand their age and function.

Aguadas are characterized as natural ponds, or karst sinks that were later modified, or as excavated and engineered depressions. Often these features were clay or plaster lined, and today are visible across the landscape and Maya built environments as water-filled depressions of various sizes. Within both urban and rural settings, such aguadas have been studied and modeled to better understand when they were built and how much water they could have stored during use (Beach et al., 2015b; Ferrand et al., 2012). Dams and canals, usually associated with aguadas, are often mapped as part of a larger archaeological landscape, although few studies have incorporated excavation or radiocarbon dating of these features. The character of aguadas changed over the course of Maya history: Middle to Late Preclassic aguadas were “concave” in structure and located in depressions adjacent to settlements, whereas later Classic aguadas exhibited a “convex” form and were elevated within urban areas to control the release of rainwater through residences (Scarborough and Burnside 2010).

A particularly high density of Maya water features has been excavated in the southern Maya Lowlands, including locations in Belize and the Peten. This likely reflects fewer archaeological surveys and excavations in the northern Maya Lowlands, rather than a true difference in the density of water features. This suggests that synthesis studies on water management in the northern Maya Lowlands will only be possible after more extensive excavation of these features in this region. Simultaneously, while the majority of water features have been excavated in the southern Lowlands, fewer paleoclimate records are available from this region. Additional well-dated, high-resolution records from the southern region would be invaluable for facilitating detailed studies of human adaptation to climate variability.

Despite these limitations, significant progress has been made towards understanding water management in the face of hydroclimate change in Preclassic and Classic Maya society. We highlight two key sites in the southern Lowlands that illustrate the relationship between water quality and quantity and human management practices. At the site of Tikal, in the southern Maya Lowlands, the onset of the Late Preclassic drying coincides with the construction of interconnected canals, reservoirs, and other management structures to collect and distribute spring and rainwater throughout the site center (Scarborough and Grazioso, 2015). Comparisons of population estimates at this urban center with regional climate records have suggested that urban planning responses occurred during periods of drying (Akers et al., 2016; Lentz et al., 2018). Geoarchaeological excavations have also shown that at Tikal, the reservoir system was continuously modified and used into the Classic period, serving to irrigate a variety of crops, or even serve as a place to grow cacao (Lentz et al., 2018). Despite this, the Late Classic drought recorded in the Tzabnah cave speleothem does coincide with the final abandonment at Tikal’s site core (Lentz et al., 2018; Medina-Elizalde et al., 2016). New approaches to understanding water at Tikal involve detailed analyses of the relationship between climate, management, and water quality. Ongoing work involves multiproxy analyses that quantify the role of pollutants like mercury and phosphate, as well as bacterial loads, in altering water quality in the Late Classic (Lentz et al., 2021). Research at Tikal therefore provides a template for future work combining climate reconstructions and archeology to reconstruct human water management in urban centers during periods of drought. Other studies focus on smaller Classic drainage/reservoir systems, such as one recorded and excavated at the site of Dos Hombres. This system was likely used by individual households (Lohse and Findlay 2000). The Dos Hombres site provides a look at water management systems and drought adaptation at the residential (albeit elite) scale. This system consisted of a channel dug into bedrock that linked two small depressions. Ceramic and architecture style date this system to the Late Classic, an interval of drying (Lohse and Findlay 2000). Other reservoirs similar to the Dos Hombres system have been excavated, such as the canal and reservoir system at nearby Chan Cahal, a much smaller non-elite settlement (Beach et al., 2015c; Krause et al., 2019a). The site of Chan Cahal was continuously occupied from the Middle Preclassic to possibly the Postclassic, with palynological and carbon isotopic evidence of intensive cultivation for much of the occupation, despite evidence for regional hydroclimate change.

Maya wetland agricultural strategies also represent a large-scale response to hydroclimate variability. These farming systems occurred in a variety of watersheds, and geoarchaeological reconstructions illustrate the different functions of each system. Past farmers considered factors such as height of the water table, groundwater quality, and soil chemistry of these wetland agroecosystems (Beach et al., 2015a, 2019; Krause et al., 2021). Ditched wetland agriculture was used at a wide variety of spatial scales throughout the Maya world as early as the Preclassic (Turner and Harrison 1981) and as late as the Postclassic (Krause et al., 2019b) and was an important component to the agricultural sector of Maya society, displaying some resilience to drought. Current research within wetland agroecosystem studies demonstrates an increased reliance on these kinds of systems during cultural transitions, especially during the Terminal Classic and Postclassic (Krause et al., 2021; Turner and Sabloff, 2012). The Central Maya Lowlands (including the aforementioned Tikal, as well as Calakmul and many other large urban centers) all depopulated at the end of the Terminal Classic due to complex stressors that were intensified by regional drying, and archaeological evidence suggests that this resulted in a shifting of trade and political power to the coastal regions (Turner & Sabloff, 2012). In the Birds of Paradise wetlands along the Rio Bravo, inferences based on multiple lines of evidence demonstrated the continued use of canal systems even into the Postclassic, when major urban centers had already been abandoned as a result of droughts during the Terminal Classic (Krause et al., 2021). While some field complexes were abandoned during the sociopolitical upheavals of the Terminal Classic, many of these field systems highlight the persistence of Maya society, and water management technologies during an episode of drought thought to have contributed to total societal “collapse.”

In addition to the complex adaptive responses demonstrated by Maya communities, the variable landscapes inhabited by the Maya imposed constraints and opportunities that were heterogenous over space. Large-scale landscape characteristics—geographical location, elevation, topography, geology and soil type, for example, acted to mediate between climate forcing and the adaptive decisions made by local human communities. The site of Lamanai, for example, survived the Terminal Classic with no discernible loss of population (Graham 2012) and with apparently limited evidence of hydroclimate stress (Metcalfe et al., 2009; Rushton et al., 2020). This remarkable resilience was due, presumably, to the buffering effect of the enormous New River Lagoon, on the western shore of which Lamanai was established. How far into the hinterland this hydroclimate buffering extended is not known in detail and is a matter of ongoing research, but the ceremonial core of Ka’kabish—a site just 10 km northwest of Lamanai—was abandoned in the Terminal Classic (McLellan 2020; McLellan and Haines 2013), as was Altun Ha, 40 km to the east (Pendergast 1981). This kind of large-scale variation in landscape resilience influences both the impact of climatic variability and the range of adaptive choices open to human populations and underscores the complexity of social transformation across large geographic areas. We note that, despite the fact that data archiving in public databases is becoming a more common practice in the field of paleoclimatology, it is not always a common practice in archaeological research. With that in mind, since properly quantifying sources of uncertainty is a key aspect of rigorous studies of human-environment relations (e.g., Degroot et al., 2021), making raw data (chronological and proxy) from archaeological features publicly available would facilitate such efforts.

Conclusions

Existing paleoclimate data from the Maya region illustrate that periods of key societal change in the Maya Lowlands, including the transitions from the Preclassic to the Classic and Classic to the Postclassic, were also intervals of climate change. Despite regional differences, many records indicate a shift towards drier conditions near the end of the Preclassic. In addition, even accounting for age uncertainty in many records, paleoclimate data from the southern Maya Lowlands and through the northern region suggest that the terminal Classic period was unusually dry. Hydroclimate variability and change were likely a source of stress and stimulus to both Preclassic and Classic Maya society. Additional insights on the nature of past climatic variability will come from new records that clarify the role of higher-resolution transients (e.g., tropical cyclones) in the water budget of local regions as well as the application of existing methods to understudied regions like the southern Maya Lowlands. Spatiotemporal statistical methods will also be useful for analyzing local and region-scale climate patterns, as well as synthesizing these data with available archaeological data. Recent work has also begun to analyze the relationships between past climate and water quality, adding a new dimension to our understanding of water resource fluctuations in ancient Maya society.

Archaeological remote sensing, ground surveys, and excavations have revealed a plethora of water management features across the Maya region. These include adaptations to both water scarcity and water excess, suggesting a broad range of technologies employed by Maya society to smooth the sharp seasonality of the tropical savanna climate. Most of these features remain unexcavated, but some (e.g., those mapped in Figure 1) provide chronologies and secondary environmental proxies by which we can understand site-specific responses to shifting climate conditions. New directions of research could target hinterland or non-elite water features or structures and reconstruct their use alongside more urban structures, but often such undertakings are resource and time intensive. Ongoing geoarchaeological excavation, coupled with LiDAR and other remote sensing techniques, will likely continue to fill these gaps. Coupling a better understanding of the construction and use of water features with regional climate trends to develop secondary and tertiary proxies that focus on site-specific response to regional environmental change could shed light on the dynamic human-environment relationship in the Maya Lowlands.

Previous efforts to synthesize archaeological and paleoenvironmental data have yielded insights into Maya warfare and human/environment relations (Ebert et al., 2017; Wahl et al., 2019). Here, we suggest similar efforts can identify changes in Maya water management in response to hydroclimate variability; a critical component in broader social adaptation to climate stress. Qualitative and quantitative syntheses such as these will become more achievable as researchers, particularly in archeology, take advantage of publicly accessible data archives that will permit re-analysis as additional data become available. Such a synthesis would, we argue, reveal greater heterogeneity and complexity in social adaptation and highlight examples of social resilience and cultural continuity as an alternative to deterministic conceptions of a homogenous climate-driven catastrophe. These studies will, in turn, offer critical lessons as societies in the present grapple with hydroclimate change and variability.

Supporting data

Supporting data for this paper is included in two supplementary excel files. File S1 contains all details of water features included in Figure 1 and discussed in the text, and File S2 contains links to the paleoclimate datasets included in Figures 1–3 and discussed in the text. *Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

Supplemental Material

Supplemental material - Progress report: Drought and water management in ancient Maya society

Supplemental material for Progress report: Drought and water management in ancient Maya society by Tripti Bhattacharya, Samantha Krause, Dan Penny, and David Wahl in Progress in Physical Geography: Earth and Environment.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: USGS Climate and Land Use Research and Development Program (n/a), Australian Nuclear Science and Technology Organisation (AP12578 and AP12409), Australian Research Council (DP180101986).

ORCID iD

Tripti Bhattacharya

Supplemental Material

Supplemental material for this article is available online.

References

Aimers

(2007) What Maya collapse? Terminal classic variation in the Maya Lowlands. Journal of Archaeological Research 15(4): 329–377.

Akers

Brook

Railsback

, et al. (2016) An extended and higher-resolution record of climate and land use from stalagmite MC01 from Macal Chasm, Belize, revealing connections between major dry events, overall climate variability, and Maya sociopolitical changes. Palaeogeography, Palaeoclimatology, Palaeoecology 459: 268–288.

Anderson

Wahl

(2016) Two Holocene paleofire records from Peten, Guatemala: Implications for natural fire regime and prehispanic Maya land use. Global and Planetary Change 138: 82–92.

Anderson

Anchukaitis

Pons

, et al. (2019) Multiscale trends and precipitation extremes in the Central American midsummer drought. Environmental Research Letters 14(12): 124016.

Anselmetti

Hodell

Ariztegui

, et al. (2007) Quantification of soil erosion rates related to ancient maya deforestation. Geology 35(10): 915–918.

Beach

(1998) Soil catenas, tropical deforestation, and ancient and contemporary soil erosion in the Peten, Guatemala. Physical Geography 19(5): 378–405.

Beach

Dunning

(1997) An ancient Maya reservoir and dam at Tamarindito, el Peten, Guatemala. Latin American Antiquity 8(1): 20–29.

Beach

Dunning

(1995) Ancient Maya terracing and modern conservation in the Peten rain forest of Guatemala. Journal of Soil and Water Conservation 50(2): 138–145.

Beach

Luzzadder-Beach

Cook

, et al. (2015a) Ancient Maya impacts on the Earth’s surface: An Early Anthropocene analog? Quaternary Science Reviews 124: 1–30.

10.

Beach

Luzzadder-Beach

Cook

, et al. (2018) Stability and instability on Maya Lowlands tropical hillslope soils. Geomorphology 305: 185–208.

11.

Beach

Luzzadder-Beach

Dunning

, et al. (2009) A review of human and natural changes in Maya Lowland wetlands over the Holocene. Quaternary Science Reviews 28(17–18): 1710–1724.

12.

Beach

Luzzadder-Beach

Flood

, et al. (2015b) A neighborly view: water and environmental history of the El Zotz region. In: Lentz

Dunning

Scarborough

(eds) Tikal: Paleoecology of an Ancient Maya City. Cambridge, UK: Cambridge University Press, pp. 258–279.

13.

Beach

Luzzadder-Beach

Guderjan

, et al. (2015c) The floating gardens of Chan Cahal: Soils, water, and human interactions. Catena 132: 151–164.

14.

Beach

Luzzadder-Beach

Krause

, et al. (2019) Ancient Maya wetland fields revealed under tropical forest canopy from laser scanning and multiproxy evidence. Proceedings of the National Academy of Sciences 116(43): 21469–21477.

15.

Beach

Luzzadder-Beach

Krause

, et al. (2015d) ‘Mayacene’ floodplain and wetland formation in the Rio Bravo watershed of northwestern Belize. The Holocene 25(10): 1612–1626.

16.

Beach

Luzzadder-Beach

Terry

, et al. (2011) Carbon isotopic ratios of wetland and terrace soil sequences in the Maya Lowlands of Belize and Guatemala. Catena 85(2): 109–118.

17.

Bhattacharya

Chiang

Cheng

(2017) Ocean-atmosphere dynamics linked to 800–1050 CE drying in Mesoamerica. Quaternary Science Reviews 169: 263–277.

18.

Bhattacharya

Coats

(2020) Atlantic-Pacific gradients drive last millennium hydroclimate variability in Mesoamerica. Geophysical Research Letters 47(13): e2020GL088061.

19.

Blaauw

Christen

(2011) Flexible paleoclimate age-depth models using an autoregressive gamma process. Bayesian Analysis 6(3): 457–474.

20.

Bookhagen

Strecker

(2008) Orographic barriers, high- resolution TRMM rainfall, and relief variations along the eastern Andes. Geophysical Research Letters 35(6): L06403. DOI: 10.1029/2007GL032011.

21.

Brenner

Rosenmeier

Hodell

, et al. (2002) Paleolimnology of the Maya Lowlands: Long–term perspectives on interactions among climate, environment, and humans. Ancient Mesoamerica 13(1): 141–157.

22.

Brewer

(2018) Householders as water managers: A comparison of domestic-scale water management practices from two central Maya lowland sites. Ancient Mesoamerica 29(1): 197–217.

23.

Brewer

Carr

Dunning

, et al. (2017) Employing airborne lidar and archaeological testing to determine the role of small depressions in water management at the ancient Maya site of Yaxnohcah, Campeche, Mexico. Journal of Archaeological Science: Reports 13: 291–302.

24.

Cowgill

Hutchinson

(1966) La Aguada de Santa Ana Vieja: the history of a pond in Guatemala. Arch. Hydrobiol 62: 335–372.

25.

Curtis

Hodell

Brenner

(1996) Climate variability on the Yucatan Peninsula (Mexico) during the past 3500 years, and implications for Maya cultural evolution. Quaternary Research 46(1): 37–47.

26.

Dahlin

Chase

(2014) A tale of three cities: effects of the AD 536 event in the lowland Maya heartland. In: Iannone

(ed) The Great Maya Droughts in Cultural Context: Case Studies in Resilience and Vulnerability. Boulder, CO: University Press of Colorado, pp. 127–156.

27.

Davis-Salazar

(2003) Late Classic Maya water management and community organization at Copan, Honduras. Latin American Antiquity 14(3): 275–299.

28.

Degroot

Anchukaitis

Bauch

, et al. (2021) Towards a rigorous understanding of societal responses to climate change. Nature 591(7851): 539–550.

29.

Douglas

Pagani

Canuto

, et al. (2015) Drought, agricultural adaptation, and sociopolitical collapse in the Maya Lowlands. Proceedings of the National Academy of Sciences 112(18): 5607–5612.

30.

Dunning

Beach

Rue

(1997) The paleoecology and ancient settlement of the Petexbatun region, Guatemala. Ancient Mesoamerica 8: 255–266.

31.

Dunning

Luzzadder-Beach

Beach

, et al. (2002) Arising from the bajos: the evolution of a neotropical landscape and the rise of Maya civilization. Annals of the Association of American Geographers 92(2): 267–283.

32.

Ebert

May

Culleton

, et al. (2017) Regional response to drought during the formation and decline of Preclassic Maya societies. Quaternary Science Reviews 173: 211–235.

33.

Evans

(2016) Airborne laser scanning as a method for exploring long-term socio-ecological dynamics in Cambodia. Journal of Archaeological Science 74: 164–175.

34.

Evans

Bauska

Gázquez-Sánchez

, et al. (2018) Quantification of drought during the collapse of the classic Maya civilization. Science 361(6401): 498–501.

35.

Fedick

Morrison

Andersen

, et al. (2000) Wetland manipulation in the Yalahau region of the northern Maya Lowlands. Journal of Field Archaeology 27(2): 131–152.

36.

Ferrand

Dunning

Lentz

, et al. (2012) Use of aguadas as water management sources in two southern Maya lowland sites. Ancient Mesoamerica 23(1): 85–101.

37.

Frappier

Pyburn

Pinkey-Drobnis

, et al. (2014) Two millennia of tropical cyclone- induced mud layers in a northern Yucatán stalagmite: multiple overlapping climatic hazards during the Maya terminal classic “megadroughts”. Geophysical Research Letters 41(14): 5148–5157.

38.

Fuentes-Franco

Coppola

Giorgi

, et al. (2015) Inter-annual variability of precipitation over Southern Mexico and Central America and its relationship to sea surface temperature from a set of future projections from CMIP5 GCMs and RegCM4 CORDEX simulations. Climate Dynamics 45(1–2): 425–440.

39.

Garrison

Dunning

(2009) Settlement, environment, and politics in the San Bartolo-Xultun territory, el Peten, Guatemala. Latin American Antiquity 20(4): 525–552.

40.

Graham

(2012) Collapse, Conquest and Maya survival at Lamanai, Belize. Archaeology International 4(1): 52–56.

41.

Hammond

Donaghey

Gleason

, et al. (1987) Excavations at Nohmul, Belize, 1985. Journal of Field Archaeology 14(3): 257–281.

42.

Hansen

Bozarth

Jacob

, et al. (2002) Climatic and environmental variability in the rise of Maya civilization: A preliminary perspective from northern Peten. Ancient Mesoamerica 13(2): 273–295.

43.

Harvey

Saborio-Rodrıguez

Martinez-Rodrıguez

, et al. (2018) Climate change impacts and adaptation among smallholder farmers in central america. Agriculture & Food Security 7(1): 57.

44.

Herrera

Ault

Fasullo

, et al. (2018) Exacerbation of the 2013–2016 pan- caribbean drought by anthropogenic warming. Geophysical Research Letters 45(19): 10–619.

45.

Hodell

Brenner

Curtis

(2005) Terminal Classic drought in the northern Maya Lowlands inferred from multiple sediment cores in Lake Chichancanab (Mexico). Quaternary Science Reviews 24(12–13): 1413–1427.

46.

Hodell

Curtis

Brenner

(1995) Possible role of climate in the collapse of Classic Maya civilization. Nature 375(6530): 391–394.

47.

Hodell

Curtis

Jones

, et al. (1991) Reconstruction of Caribbean climate change over the past 10, 500 years. Nature 352(6338): 790–793.

48.

Jacob

(1995) Ancient Maya wetland agricultural fields in Cobweb Swamp, Belize: construction, chronology, and function. Journal of Field Archaeology 22(2): 175–190.

49.

Jones

(1991) Pollen evidence of prehistoric forest modification and Maya cultivation in Belize. PhD Thesis. Texas A&M University.

50.

Karmalkar

Bradley

Diaz

(2011) Climate change in Central America and Mexico: regional climate model validation and climate change projections. Climate Dynamics 37(3–4): 605–629.

51.

Kennett

Breitenbach

Aquino

, et al. (2012) Development and disintegration of Maya political systems in response to climate change. Science 338(6108): 788–791.

52.

Kennett

Prufer

Culleton

, et al. (2020) Early isotopic evidence for maize as a staple grain in the Americas. Science Advances 6(23): eaba3245.

53.

Krause

Beach

Luzzadder-Beach

, et al. (2019a) Ancient Maya wetland management in two watersheds in Belize: Soils, water, and paleoenvironmental change. Quaternary International 502: 280–295.

54.

Krause

Beach

Luzzadder-Beach

, et al. (2019) Wetland geomorphology and paleoecology near Akab Muclil, Rio Bravo floodplain of the Belize coastal plain. Geomorphology 331: 146–159.

55.

Krause

Beach

Luzzadder-Beach

, et al. (2021) Tropical wetland persistence through the Anthropocene: Multiproxy reconstruction of environmental change in a Maya agroecosystem. Anthropocene 34: 100284.

56.

Lentz

Dunning

Scarborough

, et al. (2018) Imperial resource management at the ancient Maya city of Tikal: A resilience model of sustainability and collapse. Journal of Anthropological Archaeology 52: 113–122.

57.

Lentz

Hamilton

Dunning

, et al. (2020) Molecular genetic and geochemical assays reveal severe contamination of drinking water reservoirs at the ancient Maya City of Tikal. Scientific Reports 10(1): 1–9.

58.

Lentz

Hamilton

Dunning

, et al. (2021) Environmental DNA reveals arboreal cityscapes at the ancient Maya center of Tikal. Scientific Reports 11: 12725. DOI: 10.1038/s41598-021-91620-6.

59.

Leyden

(2002) Pollen evidence for climatic variability and cultural disturbance in the Maya Lowlands. Ancient Mesoamerica 13(1): 85–101.

60.

Lohse

Findlay

(2000) A Classic Maya house-lot drainage system in northwestern Belize. Latin American Antiquity 13(1): 175–185.

61.

Lucero

(2006) Water and Ritual: The Rise and Fall of Classic Maya Rulers. Austin, TX: University of Texas Press.

62.

Lucero

Fedick

Dunning

, et al. (2014) 3 water and landscape: ancient maya settlement deci- sions. Archeological Papers of the American Anthropological Association 24(1): 30–42.

63.

Lucero

Gunn

Scarborough

(2011) Climate change and classic Maya water management. Water 3(2): 479–494.

64.

Luzzadder-Beach

Beach

(2009) Arising from the wetlands: Mechanisms and chronology of landscape aggradation in the northern coastal plain of Belize. Annals of the Association of American Geographers 99(1): 1–26.

65.

Luzzadder-Beach

Beach

Garrison

, et al. (2017) Paleoecology and geoarchaeology at El Palmar and the El Zotz region, Guatemala. Geoarchaeology 32(1): 90–106.

66.

Luzzadder-Beach

Beach

Hutson

, et al. (2016) Sky-earth, lake-sea: climate and water in Maya history and landscape. Antiquity 90(350): 426–442.

67.

Luzzadder-Beach

Beach

(2010) Wetlands as the intersection of soils, water and indigenous human society in the Americas. In: McNeill

Winiwarter

(eds) Soils and societies: Perspectives from Environmental History. Winwick, UK: White Horse Press, pp. 91–117.

68.

Luzzadder-Beach

Beach

Dunning

(2012) Wetland fields as mirrors of drought and the Maya abandonment. Proceedings of the National Academy of Sciences 109(10): 3646–3651.

69.

Magaña

Amador

Medina

(1999) The midsummer drought over Mexico and Central America. Journal of Climate 12(6): 1577–1588.

70.

Maldonado

Rutgersson

Alfaro

, et al. (2016) Interannual variability of the midsummer drought in Central America and the connection with sea surface temperatures. Advances in Geosciences 42: 35–50.

71.

McLellan

(2020) From Lamanai to Ka’kabish: Human and Environment Interaction, Settlement Change, and Urbanism in Northern Belize. PhD Thesis. London, UK: UCL (University College London).

72.

McLellan

Haines

(2013) Casting a light in the wilderness: the ancient maya site of Ka’Kabish, Northern Belize. Research Reports in Belizean Archaeology 10: 187–198.

73.

Medina-Elizalde

Burns

Lea

, et al. (2010) High resolution stalagmite climate record from the Yucatán peninsula spanning the maya terminal classic period. Earth and Planetary Science Letters 298(1–2): 255–262.

74.

Medina-Elizalde

Burns

Polanco-Martınez

, et al. (2016) High- resolution speleothem record of precipitation from the Yucatan Peninsula spanning the Maya Preclassic Period. Global and Planetary Change 138: 93–102.

75.

Medina-Elizalde

Rohling

(2012) Collapse of Classic Maya civilization related to modest reduction in precipitation. Science 335(6071): 956–959.

76.

Metcalfe

Breen

Murray

, et al. (2009) Environmental change in northern Belize since the latest Pleistocene. Journal of Quaternary Science 24(6): 627–641.

77.

Mueller

Islebe

Hillesheim

, et al. (2009) Climate drying and associated forest decline in the Lowlands of northern guatemala during the late holocene. Quaternary Research 71(2): 133–141.

78.

Neelin

Münnich

, et al. (2006) Tropical drying trends in global warming models and observations. Proceedings of the National Academy of Sciences 103(16): 6110–6115.

79.

Pendergast

(1981) Lamanai, Belize: Summary of excavation results, 1974–1980. Journal of Field Archaeology 8(1): 29–53.

80.

Penny

Hall

Evans

Polkinghorne

(2019) Geoarchaeological evidence from angkor, cambodia, reveals a gradual decline rather than a catastrophic 15th-century collapse. Proceedings of the National Academy of Sciences 116(11): 4871–4876.

81.

Pohl

Bloom

Pope

(1990) Interpretation of wetland farming in northern Belize: excavations at San Antonio Rio Hondo. In: Ancient Maya Wetland Agriculture: Excavations on Albion Island, Northern Belize, pp. 187–237.

82.

Pohl

Pope

Jones

, et al. (1996) Early agriculture in the maya Lowlands. Latin American Antiquity 7(4): 355–372.

83.

Pope

Dahlin

(1989) Ancient Maya wetland agriculture: new insights from ecological and remote sensing research. Journal of Field Archaeology 16(1): 87–106.

84.

Prat

Nelson

(2013) Mapping the world's tropical cyclone rainfall contribution over land using the TRMM Multi-satellite Precipitation Analysis. Water Resources Research 49(11): 7236–7254.

85.

Rauscher

Kucharski

Enfield

(2011) The role of regional SST warming variations in the drying of Meso- America in future climate projections. Journal of Climate 24(7): 2003–2016.

86.

Rogé

Friedman

Astier

, et al. (2014) Farmer strategies for dealing with climatic variability: a case study from the Mixteca Alta region of Oaxaca, Mexico. Agroecology and Sustainable Food Systems 38(7): 786–811.

87.

Rushton

Whitney

Metcalfe

(2020) A tale of maize, palm, and pine: changing socio-ecological interactions from pre-classic maya to the present day in Belize. Quaternary 3(4): 30–51.

88.

Scarborough

(1998) Ecology and ritual: water management and the Maya. Latin American Antiquity 9(2): 135–159.

89.

Scarborough

Becher

Baker

, et al (1995) Water and land at the ancient Maya community of La Milpa. Latin American Antiquity 6(2): 98–119.

90.

Scarborough

Burnside

(2010) Complexity and sustainability: perspectives from the ancient maya and the modern Balinese. American Antiquity 75(2): 327–363.

91.

Scarborough

Chase

(2012) Low density urbanism, sustainability, and IHOPE-maya: can the past provide more than history? UGEC Viewpoints 8: 20.

92.

Scarborough

Connolly

Ross

(1994) The pre-Hispanic Maya reservoir system at Kinal, Peten, Guatemala. Ancient Mesoamerica 5(1): 97–106.

93.

Scarborough

Isendahl

(2020) Distributed urban network systems in the tropical archaeological record: Toward a model for urban sustainability in the era of climate change. The Anthropocene Review 7(3): 208–230.

94.

Scarborough

Grazioso

(2015) The evolution of an ancient waterworks system at Tikal. In: Lentz

Dunning

Scarborough

(eds)Tikal: Paleoecology of an Ancient Maya City, p. 16–45.

95.

Schüpbach

Kirchgeorg

Colombaroli

, et al. (2015) Combining charcoal sediment and molecular markers to infer a Holocene fire history in the Maya Lowlands of Petén, Guatemala. Quaternary Science Reviews 115: 123–131.

96.

Smyth

Dunning

Weaver

, et al. (2017) The perfect storm: climate change and ancient Maya response in the Puuc Hills region of Yucatán. Antiquity 91(356): 490–509.

97.

Turner

Harrison

(1981) Prehistoric raised-field agriculture in the Maya Lowlands. Science 213(4506): 399–405.

98.

Turner

Sabloff

(2012) Classic Period collapse of the Central Maya Lowlands: Insights about human–environment relationships for sustainability. Proceedings of the National Academy of Sciences 109(35): 13908–13914.

99.

Wahl

Anderson

Estrada-Belli

, et al. (2019) Palaeoenvironmental, epigraphic and archaeological evidence of total warfare among the Classic Maya. Nature Human Behaviour 3(10): 1049–1054.

100.

Wahl

Byrne

Anderson

(2014) An 8700 year paleoclimate reconstruction from the southern Maya Lowlands. Quaternary Science Reviews 103: 19–25.

101.

Wahl

Byrne

Schreiner

, et al. (2006) Holocene vegetation change in the northern Peten and its implications for Maya prehistory. Quaternary Research 65(3): 380–389.

102.

Wahl

Estrada-Belli

Anderson

(2013) A 3400 year paleolimnological record of prehispanic human–environment interactions in the Holmul region of the southern Maya lowlands. Palaeogeography, Palaeoclimatology, Palaeoecology 379: 17–31.

103.

Wahl

Hansen

Byrne

, et al. (2016) Holocene climate variability and anthropogenic impacts from Lago Paixban, a perennial wetland in Petén, Guatemala. Global and Planetary Change 138: 70–81.

104.

Wahl

Schreiner

Byrne

, et al. (2007) A paleoecological record from a Late Classic Maya reservoir in the north Peten. Latin American Antiquity 18(2): 212–222.

105.

Weiss-Krejci

Sabbas

(2002) The potential role of small depressions as water storage features in the central Maya Lowlands. Latin American Antiquity 13(3): 343–357.

106.

Wyatt

(2014) The scale and organization of ancient Maya water management. 449–467. Wiley Interdisciplinary Reviews: Water 1(5): 449–467.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB