Preclassic environmental degradation of Lake Petén Itzá, Guatemala, by the early Maya of Nixtun-Ch'ich'

Paleolimnological studies in the lowland Maya regions of eastern Mexico, Guatemala, and Belize have documented extensive human impacts on terrestrial environments through widespread deforestation for urbanization and agriculture during the Preclassic and Classic periods. Land clearance led to rapid soil erosion and deposition of inorganic "Maya clay" in lakes, and increased fire frequency likely linked to swidden agriculture. These pronounced land-use changes have motivated proposals of an early Anthropocene ("Mayacene"). Yet, most prior work emphasized terrestrial transformations and largely overlooked lacustrine ecological responses to Maya disturbances. Despite dense ancient populations around central Petén lakes, prior lake studies have not revealed large-scale aquatic ecosystem alterations attributable to Maya activities, with the prevailing hypothesis that severe siltation suppressed primary production by limiting light and adsorbing nutrients. In contrast, modern observations from Guatemalan lakes show eutrophication and macrophyte proliferation associated with recent population growth and nutrient loading, suggesting that ancient urban centers might also have altered lake ecological status. However, multi-proxy sediment core datasets explicitly addressing this issue in the lowland Maya region have been lacking. To address this gap, the authors conducted a multi-proxy paleolimnological investigation of a radiocarbon-dated sediment core from the western arm of Lake Petén Itzá adjacent to Nixtun-Ch'ich'. They aimed to determine the timing, magnitude, and drivers of aquatic ecosystem change associated with Maya occupation and urban development, and to evaluate lake recovery following population decline in the Terminal Preclassic.

Prior studies documented: (1) widespread deforestation and soil erosion across the Maya Lowlands, depositing thick Maya clay units during Preclassic and Classic times; (2) increased fire activity associated with swidden agriculture; and (3) proposals of an early Anthropocene driven by extensive Maya land modification. However, lacustrine responses were often interpreted as depressed primary production due to heavy siltation reducing light and adsorbing nutrients. Modern records from Lake Petén Itzá and Lake Izabal, in contrast, show recent eutrophication and abundant aquatic macrophytes linked to human and livestock disturbances and wastewater inputs. Highland Maya lakes have shown clear anthropogenic ecological effects, underscoring a discrepancy with lowland interpretations and motivating a multi-proxy approach near Nixtun-Ch'ich'.

Study site and core collection: Lake Petén Itzá (northern Guatemala) consists of a deep northern basin and a shallow southern basin. Core PI-NC-1 (515 cm) was collected in July 2018 from 8.4 m water depth in the narrow western arm of the southern basin, ~200 m south of Nixtun-Ch'ich' (16°56′36″ N, 89°55′56″ W), in a position downgradient from the city’s paved Avenues G and H. A 72-cm mud-water interface (MWI) core was collected using a modified piston corer to preserve undisturbed surface sediments; deeper sections (50–515 cm) were retrieved with a modified Livingstone-type piston corer. MWI sediments were extruded and sectioned at 2-cm intervals; deeper cores were stored in polycarbonate tubes. Age-depth modeling: Twelve radiocarbon measurements were attempted; six additional terrestrial wood charcoal dates were added to a prior model, three of which were discarded for stratigraphic reversals. A Bayesian age-depth model (rbacon in R) calibrated with IntCal20 and the 2018 collection date was produced, avoiding hard-water effects by dating terrestrial charcoal. The 515-cm record spans ~7000 years; the interval analyzed here (313–58 cm) covers ca. 2050 BCE to 1490 CE, divided into three zones by lithology. Sedimentation during the Maya clay interval was ~0.15 cm/yr. Geochemical analyses: Samples (313–53 cm) were taken every 5 cm (and every 4–6 cm in the upper 50 cm). After drying (60 °C, 15 h) and homogenization, total carbon (TC) and total nitrogen (TN) were measured by elemental analyzer; total inorganic carbon (TIC) by coulometry; total organic carbon (TOC) as TC–TIC; C/N as TOC/TN. Bulk δ13Corg (acidified to remove carbonates) and δ15Norg (unaltered) were measured by EA-IRMS (values reported relative to VPDB for carbon and AIR for nitrogen). Zonal means and 2σ were calculated. n-Alkane quantification and isotopes: Thirty-eight samples (13–310 cm) were solvent-extracted (ASE; 9:1 DCM:MeOH), with apolar fractions isolated via alumina and Ag+-silica columns. Quantification used GC-FID with internal standard (5α-androstane), reported as sum of C16–C35 per dry weight (µg/g). Compound-specific δ13C (C21–C35) used GC-C-IRMS (VPDB-referenced), with primary reference materials and internal lab standards; precision ~0.3–0.4‰; sample standard errors ~0.22–0.26‰. Fecal stanols/sterols: Fifty-nine samples were microwave-extracted (9:1 DCM:MeOH), saponified (KOH), neutral fraction isolated, derivatized (BSTFA+pyridine), and analyzed by GC-FID; select samples confirmed by GC–MS. Concentrations (cholestanol, cholesterol, stigmastanol, coprostanol+epicoprostanol) were expressed per dry weight and normalized to TOC (µg/g of OC); coprostanol+epicoprostanol was used as the primary human fecal indicator. Charcoal: Macroscopic charcoal (>125 µm) was isolated by H2O2 oxidation and wet sieving, and counted microscopically at 25× from 1 cm3 samples at defined depth intervals, providing local fire history. Inorganic proxies: XRF-derived lithogenic elements (e.g., Ti; reported as cps) and magnetic susceptibility were used to track detrital inputs and erosion. Lithostratigraphy defined Zones 1–3: Zone 1 (gray laminated carbonate-rich silty clays), Zone 2 (dark brown-gray massive carbonate-rich clay; Maya clay), Zone 3 (light brown thinly bedded organic- and carbonate-rich silty clays).

• Early occupation and land-use signals: In Zone 1 (ca. 2050–710 BCE), TOC and TN were stable (TOC 2.8 ± 0.6 wt%; TN 0.2 ± 0.03 wt%); C/N averaged 13.5 ± 1.1; δ15Norg 2.7 ± 0.4‰; δ13Corg −25.0 ± 0.7‰. Charcoal increased sharply after ~1070 BCE (mean 573 ± 308 particles/cm³ at top of Zone 1; peaks 1,476 at ~1070 BCE and 1,062 particles/cm³ at ~890 BCE), indicating intensified local burning consistent with swidden agriculture. Coprostanol+epicoprostanol (human fecal indicator) remained low but variable (avg. 25.8 ± 17.5 µg/g OC before ~1900 BCE; 28.5 ± 20.2 µg/g OC after ~1900 BCE), implying sparse early human presence beginning by ca. 1400–1300 BCE.
• Maya clay and lake ecosystem alteration (Zone 2, ca. 710 BCE–280 CE): Lithology shifts to massive carbonate-rich clay coeval with the regional Maya clay. TOC and TN halved (TOC 1.4 ± 0.4 wt%; TN 0.1 ± 0.03 wt%); total n-alkanes decreased (0.6 ± 0.3 µg/g), consistent with dilution by clastics. C/N declined to ~9.5 ± 0.1 (ca. 150 BCE–200 CE). δ15Norg and δ13Corg rose progressively to 6.0 ± 0.1‰ and −23.1 ± 0.3‰ (ca. 100 BCE–200 CE). Mid- and some long-chain n-alkanes became 13C-enriched (e.g., C25 to −26.8‰; C27 to −26.6‰; C33 to −27.6‰; C35 to −27.0‰ by ~200 CE), while C29 and C31 remained ~−30.5‰ and −29.7‰, implying enhanced aquatic macrophyte/algal contributions and altered productivity rather than increased C4 plant inputs. Charcoal dropped (62 ± 49 particles/cm³), indicating low local fire activity after urbanization. Coprostanol+epicoprostanol increased (avg. 41.6 ± 17.9 µg/g OC), evidencing sustained human fecal inputs. Ti and magnetic susceptibility increased from ~710 BCE, indicating localized erosion tied to construction and maintenance of the urban grid.
• Peak disturbance: Maximum eutrophication/macrophyte expansion inferred from low C/N and high δ13Corg/δ15Norg and enriched n-alkane δ13C lasted ~350 years (~150 BCE–200 CE), coinciding with dense Late Preclassic occupation and monument expansion.
• Drivers and site sensitivity: Likely drivers include nutrient-rich runoff (phosphorus, human waste) from paved streets/avenues/canals, reduced riparian vegetation, shallow water depth (~8–10 m), and poor hydrologic exchange in the western arm, facilitating nutrient accumulation and macrophyte proliferation.
• Recovery and persistence (Zone 3, 280–1490 CE): Geochemical variables shifted back toward pre-disturbance values (TOC 3.6 ± 1.2 wt%; TN 0.3 ± 0.1 wt%; total n-alkanes 2.4 ± 0.7 µg/g; C/N 12.4 ± 0.6; δ15Norg 2.8 ± 0.7‰; δ13Corg −24.9 ± 0.7‰). Coprostanol+epicoprostanol declined (28.0 ± 11.1 µg/g OC), consistent with reduced population following Terminal Preclassic depopulation. Charcoal remained low (63 ± 24 particles/cm³). Statistical comparisons show Zone 3 δ13Corg and C/N are similar to Zone 1 but C/N is intermediate, indicating swift but incomplete recovery by ca. 400 CE. Continued low-level occupation during Classic/Postclassic likely limited full ecological return to pre-disturbance conditions.

The multi-proxy record demonstrates that early Maya activities at Nixtun-Ch'ich' altered aquatic ecosystems in the western arm of Lake Petén Itzá. Contrary to prior interpretations that Maya-induced siltation suppressed lake productivity, this study documents cultural eutrophication and macrophyte expansion during the Middle to Late Preclassic concurrent with urban grid construction and dense occupation. Elevated δ13Corg and δ15Norg with declining C/N and enriched mid-chain n-alkane δ13C reflect increased contributions from aquatic producers and altered nitrogen cycling consistent with higher productivity and aquatic weed proliferation. Elevated Ti and magnetic susceptibility indicate localized erosion due to construction/land clearance. The site’s sensitivity—shallow depths, restricted hydrologic exchange, proximity to the urban shoreline, and paved drainage corridors—likely amplified nutrient delivery and retention, producing ecological degradation not captured in cores from less isolated or deeper basins. Following Terminal Preclassic depopulation (by ~280 CE), the lake exhibited relatively rapid geochemical recovery toward pre-disturbance conditions by ~400 CE, although incomplete, likely due to continued sparse occupation. These findings reconcile modern observations of eutrophication in Petén lakes with ancient anthropogenic impacts, highlighting spatial heterogeneity in lacustrine responses to Maya land use.

A high-resolution, multi-proxy sediment core adjacent to Nixtun-Ch'ich' reveals that early Maya urbanization and land-use changes degraded aquatic ecosystems in the western arm of Lake Petén Itzá during the Middle and Late Preclassic. The record shows increased local fires and early occupation in the late Early Preclassic, followed by a Maya clay interval with elevated nutrient inputs, cultural eutrophication, and macrophyte expansion peaking ~150 BCE–200 CE. After Terminal Preclassic depopulation, lake conditions recovered swiftly but incompletely, stabilizing by ~400 CE, with persistent low-level human presence limiting full return to pre-disturbance states. This study provides the first clear lowland Petén evidence of ancient Maya-mediated lacustrine deterioration, emphasizing the importance of basin morphology, hydrologic isolation, and urban drainage design in mediating ecological outcomes. Future research should: (1) expand multi-core transects across different Petén basins to assess spatial heterogeneity; (2) integrate diatom, pigment, and ancient DNA proxies to refine ecological reconstructions; (3) quantify nutrient fluxes and model hydrologic residence times to link urban infrastructure to lake responses; and (4) obtain higher-density dating within the Maya clay to reduce chronological uncertainty.

• Chronology: The middle of Zone 2 has relatively large age uncertainty (mean ± ~300 years) due to paucity of datable material; three radiocarbon samples were out of order and discarded. Age control relies on terrestrial charcoal and Bayesian modeling; fine-scale event timing remains uncertain.
• Spatial representativeness: Results derive from a single core in a hydrologically isolated, shallow sub-basin; findings may not generalize to deeper or better-flushed parts of Lake Petén Itzá or other lakes.
• Proxy constraints: Low abundances/preservation of short-chain n-alkanes may limit direct algal productivity inferences; coprostanol and epicoprostanol co-elution necessitated summing the isomers; fecal stanols can be influenced by preservation and mineral dilution (mitigated by OC normalization but not eliminated).
• Alternative explanations: While multiple lines of evidence favor eutrophication/macrophyte expansion, contributions from 15N-enriched soil organic matter, vegetation shifts, or lake-level changes cannot be entirely ruled out.
• Heterogeneous Maya clay deposition: Thickness and expression vary regionally, complicating direct inter-lake comparisons.