Paleoclimate variations in Western Amazon based on Lago Airo (Brazil) diatoms from the last 13,300 years cal BP

LetíciaRizzettiPatrocínio¹1

JoãoCláudioCerqueiraViana¹1

DoriedsonFerreiraGomes¹1

DaianaFontes²1

LucianeSilvaMoreira²1

BrunoTurcq³1

PatriciaMoreira-Turcq³1

JuliaMaria

Aguiar1

DuarteContrera2

GabrielSouzaMartins²1

AbdelfettahSifeddine1

RenatoCampelloCordeiro²1✉Emailrccordeiro@id.uff.br

Instituto de BiologiaUniversidade Federal da Bahia

Departamento de GeoquímicaUniversidade Federal Fluminense

3Institut de Recherche Pour le Développement t, IRD- LOCEAN (UMR 7159 IRD, CNRS/UPMC/MNHN)BondyFrance

Letícia Rizzetti Patrocínio¹; João Cláudio Cerqueira Viana¹; Doriedson Ferreira Gomes¹; Daiana Fontes²; Luciane Silva Moreira², Bruno Turcq³; Patricia Moreira-Turcq³; Julia Maria de Aguiar Duarte Contrera ²; Gabriel Souza Martins², Abdelfettah Sifeddine³, Renato Campello Cordeiro²

¹ Instituto de Biologia, Universidade Federal da Bahia;

² Departamento de Geoquímica, Universidade Federal Fluminense,

³ Institut de Recherche Pour le Développement t, Bondy, France, IRD-LOCEAN (UMR 7159 IRD/CNRS/UPMC/MNHN)

Correspondence author: rccordeiro@id.uff.br

Keywords:

Pleistocene

Holocene

Paleolimnology

Eunotia spp

Aulacoseira spp

geochemistry

Rio Negro

Abstract

The study of paleolimnological and paleoclimatic variations in the Amazon from the last 13,300 cal yr BP is essential for interpreting the climatic and environmental history of this region, as it is one of the largest and most diverse biomes on the planet. Diatom assembly (density and relative abundance), sedimentological, bulk, and isotopic organic geochemistry analyses were obtained from a 146 cm-long core collected in a marginal lake nowadays disconnected from Rio Negro (Lago Airo; 00º19’37.225’’S, 66º08’33.266’’W). Seven organic matter samples were dated by using AMS radiocarbon determination. The 73 species found mostly comprised acidophilic species. We discerned four main phases: Phase IV, from ca. 13,400 cal yr BP to ca 11,800 cal yr BP: diatoms density was relatively low (i.e., ranging from 88 to 37.5 x 10³ valves/g) with mainly periphytic taxa. Sediments are comprised of coarser grains (sand > 99%). Geochemical data shows a TOC average of 2.2%, the C/N ratio is higher (average = 52.1), and chlorophyll derivatives are lower (average = 5.6 SPDU). 15N values show a source related to nitrogen atmospheric fixation metabolism, probably due to the extremely oligotrophic nature of the system. Our results suggest that Phase IV represented the Lago Airo as a lotic-like ecosystem with connections with the Rio Negro, characterized by high energy and thereby low diatom sedimentation. Phase III: 11800 to 8800 cal yr BP: The diatom community kept characteristics of low density with the domain of Eunotia spp (E. hirudo, E. bilunaris, and E. floweri). This phase exhibits a significant increase in clay content of around an order of magnitude, rising from 0.256% in phase IV to 2.05% in phase III. Although the absolute values are still low, they still constitute a sandy-textured sediment. TOC values increased compared to the previous phase, reaching 2.39%, with an increase in the C/N ratio denoting the transport of detrital organic matter from the basins. Phase II: 8800 to 4100 cal yr BP: The diatom community keep characteristics of phase III. This phase exhibits a decrease in clay content to 0.299% with a progressive increase in TOC, compared to the previous phase, reaching 3.47%. This increase in C/N ratio indicates the transport of detrital organic matter from the basins. On the onset of phase I, the Rio Negro became disconnected from Lago Airo, which is suggested by the predominance of silt, associated with low hydrodynamics, and an increase in diatom valve density. Phase I: 4100 to actual: This phase shows an increased diatom valves density (between 30.9 x 104 and 14.1 x 106), composed mainly by planktonic taxa. Silt became dominant (average = 59.1; clay = 10.2) simultaneously with. Higher silt contribution with a pronounced decrease in C/N ratios (average = 34.9) is indicative of declining influence of the Rio Negro over Lago Airo and an increase in autochthonous production, represented by the increase in TOC, changing from 3.47 in phase II to 25.0%, and chlorophyll derivatives (average = 12.7 SPDU). Our results suggest an increase in the mean water level in Lago Airo (characterized by higher valves/g, planktonic taxa dominance - Aulacoseira spp. - A. distans and A. granulata, mainly, and δ15N values increase), which was probably the result of higher precipitation regimes in the South American Monsoon System.

1. Introduction

The Amazon region has substantial participation on heat distribution and humidity release from the Tropics, influencing in climate and precipitation rates from global to local scales (Werth and Avissar 2002). The main climate influence of the Amazon is the presence of moisture from the Amazon Forest from eastern Brazil (e.g. South American Summer Monsoon - SASM) and the displacement of the Inter-Tropical Convergence Zone - ITCZ (Cruz et al. 2009). In respect of distinct climatic variabilities at different time scales, wet and dry conditions at the Amazon basin are related throughout the Holocene, and these periods are controversial (Silva et al. 2018).

The ITCZ is a band of clouds induced by low-level convergence of Tradewinds over equatorial oceans, that promotes a minimum in atmospheric pressure, a low mixed-layer depth, a maximum of Sea Surface Temperature (SST), and deep convective rainfall (Garreaud et al. 2008). The Amazon River basin and Amazon Forest contribute to climate dynamics on Tropical South America being affected by complex dynamics teleconnections (Potter et al. 2004), and astronomical forcing (e.g. insolation changes explained by Milankovitch cycles) (Silva Dias et al. 2009).

The Rio Negro (Amazon River basin) is the largest black water river in the world, with a basin area of around 700,000 km2 passing through Colombia, Venezuela, and Brazil (Latrubesse and Franzinelli, 2005). According to these authors its formation is dated back to the Late Pleistocene and Holocene, associated with activity during the last glacial in the Amazon Basin. However, paleoecological studies in remote areas like the upper part of Rio Negro are scarce (Rodríguez-Zorro et al. 2017). The analysis of three sediment records in three lakes of the Morro dos Seis Lagos stream in the upper Rio Negro showed no evidence of changes in vegetation physiognomy during the late Pleistocene and Holocene (Colinvaux et al. 1996, Bush et al. 2004), rather than significant evidence of changes in lake levels (Bush et al. 2004, Cordeiro et al. 2011). Changes in atmospheric patterns between the end of glaciation and the beginning of the warmer conditions of the Holocene were observed through changes in mercury fluxes in the Lagoa da Pata (Barbosa et al. 2004).

The aim of this study was to infer paleolimnological dynamics and paleoclimate conditions at western Amazon, based on Lago Airo (Rio Negro basin) sediment core and major studies in the area. This research was based on a multiproxy approach that included geochemistry, multiple sediment variables analyses and diatom assemblage. Our results assist in understanding paleolimnological and climate variations insights throughout the Holocene and Pleistocene/Holocene transition of the Amazon basin, from registers of a lake located at extreme northwestern Brazil.

2. Site Description

Lago Airo is a floodplain lake in Rio Negro basin (0°19’37.225” S, 66°8’33.266” W), Municipality of São Gabriel da Cachoeira city, Amazonas State (Fig. 1). Rio Negro is the largest tributary of Amazon River, considered the fifth largest river in the world (Franzinelli and Igreja 2002). The Negro River the most important black water river and the second largest tributary of the Amazon River drains an area of 696,810 km², with a discharge of about 28,400 m^− 3s^− 1 and provides an average of 8 Mt. yr^− 1 of suspended sediment flux.

Fig. 1

Principal Component Analysis (PCA) of diatom assemblages along the sedimentary profile of Lago Airo (Middle Rio Negro). The analysis displays the ordination of samples based on species composition, with Factor 1 (10.91%) and Factor 2 (5.50%) representing the main gradients. The different colors correspond to the four identified phases: Phase IV (13,400–11,800 cal yr BP, green): Characterized by low diatom density, dominance of periphytic taxa, and high-energy conditions. Sediments are predominantly sandy (> 99%), with low TOC (~ 2.2%) and high C/N ratios (~ 52.1), reflecting a lotic system connected to the Rio Negro; Phase III (11,800–8,800 cal yr BP, light green): Low-density diatom assemblages, mainly Eunotia spp., with a slight increase in clay content (~ 2.05%) and TOC (~ 2.39%). The system begins to transition towards reduced hydrodynamics and increasing organic matter deposition; Phase II (8,800–4,100 cal yr BP, light blue): Similar to Phase III, but with a gradual rise in TOC (~ 3.47%) and a shift towards increased detrital organic matter transport. The environment remains transitional, with decreasing influence from the Rio Negro; Phase I (4,100 cal yr BP–Present, dark blue): Marked by a significant increase in diatom density (30.9 × 10⁴ to 14.1 × 10⁶ valves/g) and dominance of planktonic taxa (Aulacoseira spp.). Sediment composition shifts to higher silt content (~ 59.1%), with rising TOC (~ 25.0%) and lower C/N ratios (~ 34.9), indicating a fully developed lentic system with increasing water levels, likely driven by intensified precipitation from the South American Monsoon System. The PCA plot reveals a clear separation between the earlier high-energy, lotic system (green and light green clusters) and the more recent low-energy, lentic conditions (light blue and dark blue clusters), reflecting long-term hydrological and climatic changes in Lago Airo.

Lago Airo has approximately three meters of depth in the collecting date, maximum width of 300 m, pH around 3.5, and nowadays has no connection with Rio Negro. Located on equatorial region, in the Amazon Forest biome, with a hot and humid climate. The annual rainfall average on São Gabriel da Cachoeira city area is 2,853 mm (Ribeiro et al. 1996).

The vegetation around the lake is Campinarana, which is composed by 12 m arboreum strata distributed continuously with scarce open fields. The sunlight reaches the soil. It is a vegetation which occurs on hydromorphic Podzolic soils, with hydromorphic quartz sands on the alluvial plain of the Negro River. The campinaranas occur on flat and flooded areas around the lake, and present quite varied physiognomy as observed in the region with campestral and forest formations. Vegetation is characterized by soil rich in organic matter, and slow decomposition due to low pH values and large aluminum concentration (Martins and Matthes1978).

Fig. 1

A. Satellite view showing the location of Lago Airo in relation to the Rio Negro, with the lake situated within the floodplain area. The white rectangle highlights the position of the lake. B. Close-up satellite image of Lago Airo and the surrounding floodplain landscape with campinarana vegetation in brown and ombrofilous vegetation ion green. The lake is shown in proximity to the Rio Negro, emphasizing its hydrological connection. C. Map of Brazil highlighting the state of Amazonas (shaded in gray), where Lago Airo is located. D. Hydrological map of the Rio Negro basin near Lago Airo, depicting the floodplain system (blue) and vegetative cover (green). The star symbol marks the location of Lago Airo within the basin.

3. Material and Methods

3.1. Core Sampling and Radiocarbon chronology.

Sediments from Lago Airo were collected in 2015 using a vibro-coring system equipped with aluminum tubes measuring 7.6 cm in diameter. The sampling platform consisted of two rubber boats, which facilitated the retrieval of the AIRO 12/01 core, measuring 146 cm in length (0°19'37.225" S, 66°8'33.266" W). The description of the sedimentary profile was carried out as soon as the core was opened, and its lithology was described with the help of a Munsell Colour Chart. The sediment core was sliced into 1.0 cm sections which were stored in plastic bags. Seven calibrated ¹⁴C dates of bulk organic matter based the age-depth model. Measurements of ¹⁴C activity were made by Acceleratory Mass Spectrometry (AMS) at the Radiocarbon Dating Laboratory Beta Analytic, located in Miami, Florida, Laboratoire de Mesure du Carbone 14 (Table 1). The program Calib 7.1 (Stuiver et. al. 2018) was used for calibration due to the natural variations of 14C (available at: http://calib.org/calib/), and dates were converted to cal BP. To model the age-depth relationship, the R package “Bacon” (Blaauw and Christen, 2011) was used, which adopts a Bayesian approach to establish a robust chronological sequence. This model integrates available radiometric dates and, through the IntCal 13 calibration curve, converts these dates into calibrated ages (cal yr BP), providing an accurate temporal reconstruction of the observed sedimentary variations along the profile.

3.2. Grain-size Distributions

A laser particle analyzer (CILAS Particle Analyzer® 1064) was used to measure the grain-size distribution of the mineral fraction (particles between 0.04 and 500 µm) after the organic matter and carbonate destruction with H₂O₂ and HCl, respectively, and it is dispersing in a Na₄P₂O₇ solution and sonifcation. The program GRADISTAT VERSION 4.0, was used to calculate the granulometric fraction (Blott 2000).

3.3. Bulk and isotopic organic geochemistry

The organic geochemical measurements made included Total Organic Carbon (TOC), Total Nitrogen (TN), δ¹³C (‰) and δ¹⁵N (‰). Samples were analyzed for TOC and stable isotopic (δ¹³C and δ¹⁵N) content to CHN analyzer (PDZ Europa ANCA-GSL) connected to an Isotope Ratio Mass Spectrometer (IRMS) PDZ Europa 20–20 (Sercon Ltd., Cheshire, UK) at UC Davis Stable Isotope Facility (SIF), Department of Plant Sciences, CA, USA. Samples were treated with 0.5 N HCl to remove carbonate prior to isotopic measurements. It was analyzed lithology, granulometry, total organic carbon (TOC), C:N ratio, δ¹⁵N and δ¹³C.

The analysis of chlorophyll derivatives in sediment samples was performed according to the method of Sanger and Gorham (1972), adapted by Cordeiro et al. (1997), (Chlorophyll derivatives were extracted with 90% acetone and measured at 667 nm in Perkin Elmer spectrometer). The results were expressed in arbitrary units as absorbance per gram organic matter, where one unit (SPDU) is equal to an absorbance of 1.0 in a 10 cm cell, when dissolved in 100 ml of solvent.

3.4. Diatom assemblages

3.4.1. Samples oxidation and slide preparation

Sediment samples oxidation followed the standard method of Battarbee et al. (2001). About 1 g of humid sediment and 30 ml of H₂O₂ at 37% were add to a beaker. This solution remained for 24h in hot plate (50°C ± 5) until all organic matter was removed. Samples were washed in distilled water until complete dilution of H₂O₂. Slides were prepared by dropping 1 ml of diatom suspension on a coverslip. The diatoms were allowed to settle and water to evaporate in a hot plate (50°C ± 5). When dry, slides were mounted using a resin Naphrax® (index = 1.74).

3.4.2. Identification and Counting

Diatoms were counted (500 valves per sample) in each 1 (one) cm depth intervals from 0 to 34 cm (except 5–6 cm and 29–30 cm due to lack of samples). After that, analyses were carried out each 2 cm intervals, totalizing 86 samples of sediment from Lago Airo. From 34 cm to 145 cm it was set a pattern of 4 transects per slide due to very few amount of diatoms in slides.

Diatoms in each sample were identified to the lowest possible taxonomic level along random transects using an Olympus CX31 light microscope at 1000x magnification. Valves were counted if more than one-third of their original structure was intact, including cases where the valve center was present or when specific morphological features characteristic of the species were observed, to minimize overestimation (Battarbee et al., 2001).

Identification of species was conducted using mainly the following references keys: Patrick and Reimer (1975), Lange-Bertalot (1966, 2007), Metzeltin and Lange-Bertalot (1998), DeNicola (2000) and Ferrari et al. (2007).

3.4.3. Diatom Analysis

The species’ relative abundance (%) and species’ density (valves/g) represented the diatom assemblage composition. Absolute abundance value of each species per slides represented the taxa relative abundances. The cut-off value for the species selected as the most representative within the community was 1% of relative abundance at least in 3 depths of the sediment samples. These species’ relative abundance was classified as benthic or planktonic according to autecological live characters. The diatom density per layer sampled was calculated from the total of valves counted per slide corrected by samples bulk density. The ratios of planktonic to benthic diatoms were calculated using the formula: P:B ratio = planktonic taxa / benthic taxa. A broken-stick model was used (Bennett 1996) to determine the number of diatom zones.

4. Results

Through cluster analyses that incorporate the total diatom assemblage distribution, four primary phases have been established. Accordingly, the subsequent parameter description will encompass these subdivisions (Fig. 2).

Fig. 2

Cluster analysis (paired group, Euclidean distance) using the distribution of diatom flora. The green arrows indicate the sections where lithological contacts occur, as described in Table 1. Some lithological units are associated with changes in the diamond assemblage, while others do not show significant ecological changes that could not incorporate changes in sedimentary facies.

Fig. 2

Distribution of diatoms along the sedimentary profile of Lago Airo (Middle Rio Negro). The scatter plot shows the ordination of diatom assemblages based on a factorial analysis, with Factor 1 (10.91%) and Factor 2 (5.50%) representing the main gradients in species composition. Blue-labeled points correspond to the present-day diatom community, while green-labeled points correspond to past assemblages. The transition from predominantly periphytic taxa in earlier phases (Phase IV and III) to planktonic taxa (e.g., Aulacoseira spp.) in Phase I reflects increased water levels and reduced influence of the Rio Negro over time. This shift is associated with changes in sediment composition, from predominantly sandy deposits (> 99%) in Phase IV to increasing silt content (59.1%) in Phase I, as well as geochemical shifts such as rising TOC (from 2.2% in Phase IV to 25.0% in Phase I) and decreasing C/N ratios (from ~ 52.1 to ~ 34.9). These changes suggest a transition from a high-energy, lotic system with low diatom sedimentation to a lower-energy, lentic system with increased autochthonous organic matter production, likely driven by increased precipitation associated with the South American Monsoon System.

4.1. Lithology, Chronology and Bulk Density

The Airo 12/01 sediment core has a length of 1.46 meters. According to the lithological description provided in Table 3, the presence of three main units and five sub-units can be observed. In lithological unit III, which spans from 146 to 91 cm, three sub-units can be identified. The lithological description indicates the occurrence of sandier sediment within this unit. Sub-units IIIc and IIIa exhibited a brown color pattern (referred to as "brown" according to the "Munsell Soil Color Chart").

Lithological unit II consists of two sub-units (IIa and IIb) and spans the interval from 90 to 30 cm. Like lithological unit III, the presence of sandy sediment was noted in unit II. The color pattern observed during the lithological description ranged from dark reddish-brown in sub-unit IIb to very dark gray in sub-unit IIa.

The first lithological unit represents the uppermost 29 centimeters of the sediment core. A noticeable change in grain size occurs between this unit and unit II, where finer sandy sediment is observed in unit I. The color pattern observed in this lithological unit was black, as per the Munsell Soil Color Chart.

Table 1
General description of the lithological units of Airo 12/01 core.
Lithological Units	Sub-units	Section (cm)	Sedimentological characterisation	Colour Standard	Colour Code
I		29 − 0 3000-Actual	Silt	Black	5YR 2.5/1
II	IIa	52 − 29 6910 − 3000	Sand with plant fragments	Very Dark Grey	5YR 3.1
II	IIb	90 − 52 11630 − 6910	Sand	Dark Reddish Brown	5YR 3.2
III	IIIa	101 − 90 12440 − 11630	Sand	Brown	10YR 4.2
	IIIb	112 − 101 12970 − 12440	Sand with plant fragments	Dark Grayish Brown	10YR 3.2
	IIIc	146 − 112 13360 − 12970	Sand	Brown	10YR 4.2

Three radiocarbon dates on total organic matter and four datations from leaf samples were obtained by Accelerator Mass Spectrometry (AMS) in the Airo 12/01 core.

Table 2
Description of radiocarbon ages on AIRO 12/01 core by AMS Method
Depth interval (cm)	Datation code	Measured sample	Conventional ¹⁴C age (years BP)	2-sigma intercept age (cal yr BP)	Age range 2-sigma
15 cm	Beta 388870¹	Bulk organic matter on Sediment	1170 ± 30	1065	986–1179
28 cm	SacA 50900²	Bulk organic matter on sediment	1575 ± 30	1410	1399–1537
50 cm	Beta 388871¹	Plant material	4930 ± 30	5650	5600–5719
70 cm	Beta 388872¹	Plant material	8550 ± 30	9520	9493–9547
78 cm	18OS/0403³	Bulk organic matter on sediment	8420 ± 30	9490	9334–9519
116 cm	Beta 388873¹	Plant material	11430 ± 30	13200	13255– 13300
143 cm	Beta 388874¹	Plant material	11270 ± 30	13130	13058–13212

¹Radiocarbon Dating Laboratory Beta Analytic; ²Laboratoire de Mesure du Carbone 14; and ³ International Chemical Analysis Inc.

The radiocarbon ages from the Airo 12/01 core indicate sedimentation for the last 13,300 years cal BP (Table 2, Fig. 3), corresponding to the Holocene, including the Pleistocene/Holocene transition.

Figure 3: Lithological profile, calibrated radiocarbon ages, sedimentation rate, and stratigraphic zones from Lago Airo. The lithology column shows sediment textures and colors with Munsell color notations. Radiocarbon ages (cal yr BP) are plotted against depth (cm) alongside the calibrated age model (solid red line) with confidence intervals (dashed gray lines). Sedimentation rates (cm/yr) are displayed on the right. Four stratigraphic zones (I–IV) are indicated based on changes in sedimentological characteristics and age-depth relationships.

4.2. Grain-size distribution

An average percentage of sand-size particles was 79.3 ± 30.4% cm with 17.4 ± 24.4% of silt and just 3.3 ± 12.9% of clay. Considering the coefficient of variation among particle size classes, it is evident that the clay class exhibits the highest variability among the distributions, With a variation coefficient of 393%. Following this, the silt class shows a variability of 140%, and lastly, the sand class displays the least variability at 38.3%.

Phase IV: 13,400 to 11,800 cal yr BP, 146 cm to 93 cm. In this phase, the sediment was predominantly composed of coarser grains, with sand constituting over 99% of the composition. The distribution of sand displayed low variability (variation coefficient less than 1%). The influx of sand intoLago Airo demonstrated uniformity throughout the designated period, suggesting a phase characterized by low variability and high energy, likely associated with a substantial intensification in the hydrological cycle of the Rio Negro. This period marks the transition between the Pleistocene and Holocene epochs and is globally recognized for its cold conditions, particularly the Younger Dryas, characterized by a rapid and drastic cooling of the climate, especially in the Northern Hemisphere (Rasmussen et al. 2006). Phase III: 11,800 to 8,800 cal yr BP, 92 to 68 cm. This phase shows a significant increase in clays and silt content in sediment and a great variability of sand. The transition from a period characterized by significant transport competence related to extreme climatic events shifts to a phase of substantial variability, as evidenced by a coefficient of variation that was previously less than 1% and has now increased to nearly 30% in terms of sand variability. This aspect indicates a period of intense changes in the lacustrine environment, alternating between extreme events and phases of substantial stability in the lacustrine system, accompanied by significant clay deposition, which was 0.285% in phase 4, increasing to a value of 2% in this phase. In Phase II: 8,800-4,100 cal yr BP, 67 to 38 cm. The sand sediments show an increase compared to the preceding phase and present a sediment rich in sand with average values of 97.9 ± 4.65%. Phase II is characterized by average sand values like Phase IV; however, it exhibits greater variability compared to phase IV, with a variation coefficient of 4.75% against 0.18% in Phase IV. Thus, during the mid-Holocene, a climate shift towards drier conditions in relation to precedent phase is observed, marked by extreme events of sandy sedimentation. Phase I: 4,100 to actual. Silt became dominant. At the onset of phase I, the Rio Negro became disconnected from Lago Airo, which is suggested by the predominance of silt, associated with low hydrodynamics, and an increase in diatom valve density.

4.3. Bulk and stable isotopic organic geochemistry

The total organic carbon (TOC) content, with an average of 7.89 ± 11.7% for the entire profile, varied from 0.4% at 13,130 cal yr BP (Phase IV) to 41.9% at 780 cal yr BP (Phase 1). The TOC values exhibit increasing mean values across the phases, with concentrations of 1.41% in Phase IV, 2.39% in Phase III, and 3.47% in Phase II. However, a substantial increase is evident in Phase I, with mean values reaching 28%, indicating the consolidation of a lentic environment.

The C/N ratio was greater than 20 throughout Airo 12/01 core with an average value of 47.8 ± 12.9 indicating a huge influence of terrestrial organic matter, with a strong degradation component. The C/N ratio varying between 4 and 10 (~ 8) commonly suggests an algae source, whereas C/N ratios greater than 20 are attributed to an origin of macrophytic margin vegetation or terrestrial plant source (Meyers 2003).

A trend towards an increase in the C/N ratio is observed from Phases IV to II, ranging between 49.3 ± 9.58 during Phase IV, 52.6 ± 7.08 during Phase III, and 58.1 ± 5.70 during Phase II, likely indicating a period when autochthonous influence was reduced because of a regional dry phase (Nascimento et al. 2021). An abrupt decrease in the C/N ratio values is observed during Phase I, with an average of 35.5 ± 7.67. This denotes a significant increase in autochthonous influence, as evidenced by high values of sedimentary chlorophyll.

In Phase IV (ca. 13,400 to 11,800 cal yr BP), geochemical data show a TOC average of 2.2%, high C/N ratios (average = 52.1), and low chlorophyll derivatives (average = 5.6 SPDU). δ15N values suggest a source related to atmospheric nitrogen fixation, likely due to the extremely oligotrophic nature of the system. These results suggest that Phase IV represented Lago Airo as a lotic-like ecosystem with strong connections to the Rio Negro. The δ15N values approaching 0‰ support the presence of predominantly terrestrial organic material. The loss of nitrogen induces that this organic matter is poor in Nitrogen favoring the biological process of nitrogen fixation (references) attested by δ15N values near 0‰. Negative values, particularly evident in Phase IV, suggest significant decomposition processes of the organic material, likely due to prolonged exposure in the podzolized soils prevalent in the Rio Negro basin. In Phase III (11,800 to 8,800 cal yr BP), exhibiting substantial variability. There was a significant increase in clay content, rising from 0.256% in Phase IV to 2.05%, though the sediments remained sandy-textured overall. TOC values increased slightly to 2.39%, and higher C/N ratios indicated the transport of detrital organic matter from the surrounding basin. Phase II (8,800 to 4,100 cal yr BP) there was a decrease in clay content (to 0.299%) alongside a progressive increase in TOC, reaching 3.47%. During this phase Chlorophyll derivatives values around 5,501 to 20.2 SPDU, showed lake-like conditions as well as the information given by C/N (Fig. 3). Higher C/N ratios continued to reflect detrital organic matter input. This phase marks the onset of the disconnection between Lago Airo and the Rio Negro and aligns with broader mid-Holocene dry conditions, with decreasing lake levels observed across Amazonia. Lago Airo’s ecosystem showed evidence of reduced fluvial connectivity and higher external organic input, consistent with high C/N ratios, and reflects progressive ecological isolation. Phase I (4,100 cal yr BP to present) Sediments transitioned to being predominantly silt (average = 59.1%) with an increase in clay (average = 10.2%). This phase also saw a pronounced increase in autochthonous production, as evidenced by rising TOC values (up to 25.0%) and chlorophyll derivatives (average = 12.7 SPDU). These changes suggest an increase in mean water level, likely driven by higher precipitation associated with the South American Monsoon System. The shift to finer sediments, coupled with increased diatom density and autochthonous production, highlights the establishment of a lentic system. This phase represents the wettest conditions in Amazonia over the past millennia, favoring ecosystem stabilization, higher productivity, and organic matter accumulation in Lago Airo.

Fig. 3

Description of the Airo 12/01 core, grain size distribution, Total Organic Carbon (TOC), C/N, δ¹³C and δ¹⁵N. Phases are represented with horizontal white and gray markings.

4.4. Diatoms Assemblage

It was identified 73 taxa. The 10 most frequent were Eunotia hirudo Metzeltin and Lange-Bertalot (mean 20.6 ± 21.4%), Aulacoseira distans (Ehrenberg) Simonsen (12.3 ± 14.3%), Eunotia didyma Grunow var. didyma (7.6 ± 12.6%), Aulacoseira sp. 3 (6.4 ± 12.1%), Aulacoseira cf. calypsi Tremarin, Torgan et Ludwig (3.4% ± 11.1%), Eunotia ventriosa Patrick (3.2%), Eunotia bilunaris (Ehenberg) Schaarschmidt (3.1 ± 4.7%), Eunotia gibbosa Grunow in Van Heurck (2.9% ± 4.4%), Aulacoseira granulata (Ehrenberg) Simonsen (2.8 ± 3.7%), Eunotia floweri Metzeltin and Lange-Bertalot (2.7 ± 6.2%). Figure 4 shows diatom diagram.

The diatoms density attained a peak (37,542 valves/g) around 10k cal. years BP (79 cm). Eunotia hirudo, Eunotia bilunaris, and Eunotia floweri were highly abundant during phases PIV, PIII, and PII. Phase II was marked with the disappearance of Aulacoseira cf. calypsi from records and a small decrease of Eunotia hirudo and Eunotia bilunaris. Phase I recorded the highest diatom densities, with average of 5.698,183 valves/g. Planktonic species dominated this phase with Aulacoseira distans and A. granulata, while A. cf. calypsi reappeared on record and Eunotia floweri abundance decreased.

Phase IV, from ca. 13,360 (145 cm) to 11,800 cal yr BP (92 cm). Diatoms valves density was relatively low (i.e., average = 1,684 valves/g ± 1,800; CV = 107%; minimum of 88 and maximum of 7,337) with mainly periphytic taxa. Sediments was dominated by coarser sediments (sand represents 99,7%). Geochemical data shows a TOC average of 1.4% ± 0,8; CV = 56,1%; minimum of 0.40% and maximum of 3.54%. C/N ratio presents high values (average = 50.3 ± 10.1; CV = 20.1%; minimum of 26.1 and maximum of 71.6., Chlorophyll derivatives are presents low values with average of 2.93 ± 1.6; CV = 54,3%; minimum of 0,76 SPDU and maximum of 6.12 SPDU. ¹⁵N values show a source related to a nitrogen atmospheric fixation metabolism, probably due to the extremely oligotrophic nature of the system. Our results suggest that Phase IV represented the Lago Airo as a lotic-like ecosystem with connections with the Rio Negro, characterized by high energy and thereby low diatom sedimentation.

Phase III from ca. 11,800 (145 cm) – 8,800 cal yr BP (68 cm). Diatoms valves density was relatively low (i.e., average = 5,589 valves/g ± 9,455; CV = 107%; minimum of 0 and maximum of 37,543 valves/g) with mainly periphytic taxa. This phase was dominated by coarse sediments (sand represents 93,3%) and showed a progressive increase in TOC compared to the previous phase, reaching an average value of 2.39%, with a variation coefficient around 52.4%; minimum of 0,7 and maximum of 5.3%. In this phase was observed an increase in the C/N ratio denoted the increase transport of detrital organic matter from the basins. The C/N ratio presents high values of 54.6 ± 7.1 and a variation coefficient with low variability, around 13,5%; a minimum of 33.1 and a maximum of 66.6. The increase in the C/N ratio denotes the transport of detrital organic matter from the basins. Chlorophyll derivatives present low values with an average of 5.86 ± 2.00 SPDU; CV = 26,6%; minimum of 2.14 SPDU and maximum of 11.9 SPDU.

Phase II: 8,800 (67 cm) to 4,100 (38 cm) cal yr BP. Diatoms valves density decrease in relation to the precedent phase (i.e., average = 1,454 valves/g ± 1,383; CV = 95%; minimum of 0 and maximum of 4,742 valves/g) with mainly periphytic taxa. Sediments was also dominated by coarse sediments (sand represent 97,9%). Geochemical data shows a TOC average of 3.47% ± 1.7%; CV = 49.7%; minimum of 1,36 and maximum of 5,3. The C/N ratio presents high values (average = 58,1 ± 5,7; CV = 9,81%; minimum of 48,6 and maximum of 70.8), and chlorophyll derivatives presents low values (average = 5,86 ± 3.74 SPDU; CV = 63,9%; minimum of 2,1 SPDU and maximum of 11,9 SPDU).

Phase I: From ca. 4,100 cal yr BP (37 cm) to actual. Here we saw a change in diatom structure, density, and composition, marking a transition from a lotic system to a lentic one. Diatom density increased (i.e., average = 4.69 x10⁶ valves/g ± 4.15; CV = 88%; minimum of 2.01 x10³ valves/g and maximum of 14.1 x10⁶ valves/g). These diatom communities were primarily composed of planktonic taxa. After approximately 4,100 cal yr BP there was a shift towards silt dominance in the sediment composition, with an average of 65.7 ± 30.4% (coefficient of variation = 46.3%), ranging from a minimum of 0.6% to a maximum of 93.0%, followed by clay = 11,1 ± 6,72%; CV = 60,1%; minimum of 0 and maximum of 20,3). The heightened contribution of silt and a noticeable decrease in C/N ratios, averaging 34.9 ± 7.7 (coefficient of variation = 21.9%), suggest a diminishing influence of the Rio Negro on Lago Airo and an increase in autochthonous production. This is evidenced by the rise in total organic carbon (TOC) content, which increased from 3.4% in Phase II to 25.0%. A substantial increase in chlorophyll derivative reaching was simultaneous reinforces the formation of a lacustrine environment with reduced influence from the Rio Negro, as we can see on diatom data.

Fig. 5

Plot of relative abundances (%) of main diatom species along the Airo 12/01 core and phases limits.

The planktonic-benthic ratio considered the mean relative abundance of the most abundant and significant species. All planktonic species were Aulacoseira spp. and the benthics were Eunotia spp. The P:B ratio also maintained the characteristic of two main phases formed by phases IV, III and II which is characterized by diatom low density and greater representation of Eunotia spp as opposed to phase I, which presents a high density of diatoms and the predominance of Aulacoseira spp.

Fig. 6

Plot of diatom density (valves/g), Planktonic (“P”) and Benthic (“B”) diatoms and P:B anomaly from the Airo 12/01 core (< 5.0) evidencing the following climate phenomena YD, DACP, MCA and LIA. Dates of the Pleistocene and Holocene are according the International Chronostratigraphy (Cohen et al., 2018).

A direct relationship can be observed among the alterations in the structure ofLago Airo diatom community, as represented by the first factor in the principal component analysis and their abundance, which is assessed based on cell concentration in the sediment. Factor 1 serves as a principal component considering all species, indicating a transition from a benthic to a planktonic community. These changes in the diatom community precede modifications in diatom biomass, primarily resulting from the establishment of a more productive lentic environment. The abrupt shift between basal phases PIV, PIII and PII and phase 1 appears to be primarily preceded by a transformation in community structure, particularly evident from 4100 cal yr BP. Substantial shifts in diatom biomass become noticeable from 3660 cal yr BP onward.

Fig. 7

Plot of diatom density (valves/g), and factor 1 from component principal analyses from the Airo 12/01 core. Factor 1 is controlled in order of importance by: Eunotia sp3,Aulacoseira ambigua, Aulocoseira distans, Aulocoseira granulate.

5. Discussion

5.1. Paleolimnological interpretations

Lago Airo records showed consequences of climate change on the Rio Negro basin and reveals a clear separation between the earlier high-energy, lotic system and the more recent low-energy, lentic conditions (Figs. 1 and 2 supplementary materials), reflecting high amplitude of hydrological and climatic changes in Lago Airo.

The paleoclimate evidence is discussed for the Younger Dryas, Late Holocene and Early-Mid Holocene. The Airo 12/01 core first phase coincides with the Bølling-Allerød (15 − 13 ka yrs) end, a warm period of the high latitude North Atlantic region. Its formation coincides with the significant increase in the level of Lagoa da Pata recorded on the LPT V core mainly after 14,500 cal yr BP (Cordeiro et al. 2011).

Since about 14,000 to 4,000 cal yr BP, the Rio Negro has behaved as a progradational system, infilling in downstream series a sequence of structurally controlled sedimentary basins or ‘compartments,’ creating alluvial floodplains and associated anabranching channel systems (Latrubesse and Franzinelli 2005). The Lago Airo seems to be part of this dynamic. About 750 km northeast of Lago Airo, in a region now dominated by cerrado vegetation (Brazilian classification), an integrative study of sediment cores from three lakes (Beltran et al. 2023) indicated savannah type ecosystems that have remained stable over the entire Holocene period, with minimal changes in vegetation. Lake sediment records indicate sensitivity to local water table fluctuations.

During Phase IV (13,360–11,800 cal yr BP), Lago Airo functioned as a lotic-like ecosystem connected to the Rio Negro. This phase was characterized by high-energy conditions and low diatom sedimentation. The low nutrient levels and high C/N ratio indicate an environment with limited organic matter input and reduced primary production, likely due to the high-energy, flowing water dynamics of the system. Around 13,400 cal yr BP the lake exhibited higher sediment deposition which may be explained by the contemporaneous high precipitation during and just after the last deglaciation. In the upper Rio Negro River at Lagoa da Pata, after 15,500 cal yr BP until 10,000 cal yr BP occurred a substantial increase in lake level attested by increase phytoplanktonic organic matter (Cordeiro et al. 2011).

During Phase III (11,800–8,800 cal yr BP), the sand percentage decreased, while higher proportions of silt and clay were observed (Fig. 4). These grain size variations reflect changes in the system's hydraulic energy, with reduced sand deposition indicating a decline in fluvial influence. This phase also shows a progressive increase in TOC and a high C/N ratio, suggesting enhanced transport of detrital organic matter from the surrounding basins. While the environment remained lotic, it exhibited signs of increased organic input and possibly higher productivity compared to Phase IV. In Acarabixi Lake (Rodríguez-Zorro et al. 2017) during the early Holocene, a highly dynamic floodplain lake in which forest vegetation occurred with direct influence by a channel of the Rio Negro from 10,840 to 8240 cal yr BP. According to Rodríguez-Zorro et al. (2017), Rio Negro C/N ratios are particularly high considering the record obtained in Acarabixi, due to the influence of detritic organic matter and humic acids, highly deficient in nitrogen. The high C/N ratios found on Lago Airo reveal also a great influence of Rio Negro over the lake at that specific time. Similarly, high C/N ratio values were found on Acarabixi lake records, situated nearby (Rodríguez-Zorro et al. 2017). In Roraima, North hemisphere Equatorial zone with Cerrado vegetation, Lake Caracaranã was formed around 11,500 cal yr BP (Simões Filho 2000, Cordeiro et al. 2014, Beltran et al. 2023) with evidence of dry periods around 10,000-7800 cal yr BP (Simões Filho 2000, Cordeiro et al. 2014, Beltran et al. 2023). This phase was characterized by relatively low diatom density (valves/g) with mainly epiphytic taxa specifically Eunotia spp. (E. hirudo, E. bilunaris, and E. flowerii). This genus occurs on acid, oligotrophic/distrophic and freshwater environments (Round et al. 1990, Spaulding et al. 2021), which explains its high occurrence in Lago Airo. The presence of Aulacoseira distans and A. granulata indicates that these species have ecological flexibility. While they are primarily limnophilic and prefer still water environments, their occurrence in flowing water suggests they can adapt to dynamic hydrological conditions (Moro and Fürstenberger, 1997)

Phase II (8,800–4,100 cal yr BP) is characterized by continued high-energy conditions, with coarse sediments predominating. The high C/N ratios and elevated TOC levels suggest a substantial contribution of detrital organic matter, indicating increased external organic input but limited primary production. This phase corresponds to the mid-Holocene dry period (Mayle and Power, 2008), during which many Amazonian sites experienced a decline in lake levels (Cordeiro et al. 1997, 2008) and a reduction in forest biomass (Bush et al. 2007, Behling et al. 2001, Absy et al. 1991, Fontes et al. 2017). In Lago Airo it is noted that at 8,000 cal yr BP there is an abrupt decrease in the density of diatom valves with subsequent variability and a substantial decrease until approximately 3,700 cal yr BP. A study carried out by Bush and Colinvaux (1988) on lake sediment in Ecuador (Lago Ayauch, Ecuador (500 m) showed horizons of laminated sediment and weathered gyttja (lake sediment rich in organic matter), showing a reduction in the water level in the period 4,200-3,150 years B.P., confirmed by the absence of open water diatoms and the abundance of sponge spicules. The resumption of gyttja deposition around 3,150 years B.P. indicated that the lake level, and presumably rainfall, had begun to increase

At phase I (ca 4,100 cal yr BP to present) the Rio Negro disconnects from Lago Airo, as suggested by the predominance of silt and clay, indicating a transition between a predominantly erosive environment (lotic-like condition) to a lower energy environment, producing a lentic ecosystem on Lago Airo. There is also a significant decrease on C/N ratio, which indicates a diminishing of the river influence over the lake. Analysis of Corg/Ntotal atomic ratio, δ¹³C and δ¹⁵N of organic matter can allow distinguishing between algae and vascular plant contributions to sediment organic matter (Meyers and Ishiwatari 1993, Meyers 2003). The higher δ¹⁵N rates during phase II additionally with the low values of C/N ratio, suggest the development of lacustrine productivity in agreement with the rising of silt and clay grain size on records. It was also recorded the highest diatom density during this phase, with planktonic species overcoming, such as Aulacoseira distans and A. granulata, whereas Aulacoseira cf. calypsi reappeared on record. Aulacoseira is the most representative genera in Lago Airo diatoms composition.

This phase marks a transition from a lotic to a lentic system. The increase in diatom density, shift to finer sediments, and higher TOC content reflect a more stable, low-energy environment with higher primary production. The lower C/N ratio indicates increased autochthonous production, suggesting diminished influence from the Rio Negro and the development of a lacustrine ecosystem.

Aulacoseira spp. are a good indicator of turbulent waters, because its heavy frustules require high water movement to remain on the photic zone (Wolin and Stone 2010). The dominance of Aulacoseira spp. indicated a lentic system on phase I after lateral displacement of the Rio Negro main channel or decrease of Rio Negro water level forming a shallow lake on Lago Airo region. This phase is considered the wettest in Amazonia principally after 3,000 cal yr BP, when the climate becomes progressively more humid and probably with less intranual hydric variability. This condition favoring the installation of a slower environment of greater productivity and potency of organic matter accumulation. Analogous to Lago Airo area, the Acarabixí lake (00º20' S; 64º29' O) located at a fluvial plain of the Rio Negro (AM) demonstrated lateral displacement of the main channels of the river throughout time (Rodriguez -Zorro et al. 2017).

During the Late Holocene, according to Helama et al. (2017) the ‘Dark Ages Cold Period’ (DACP) is probably the climate anomaly most frequently discussed of the first millennium. Also, the Medieval Climate Anomaly (MCA, ca. 950–1200 AD) and the Little Ice Age (LIA, ca. 1400–1700 AC) are important climate events that occurred during the Late Holocene (Mann et al. 2009). The MCA is considered the more recent warm interval from the pre-industrial period and had a great influence on US and Europe climate. Nevertheless, the causes of this climate anomaly are uncertain (Mann 2002). On the opposite, the most recent and longer episode of glacier advance and temperature drop throughout the Holocene is registered at the Little Ice Age with a global scale effect (Chambers et al. 2014). A climate simulation model confirmed the global scale effects of both climate events (Mann et al. 2009). Here we discuss these climatic events at the Rio Negro Basin (and Amazon Basin overall) and the possible forcings behind these paleoclimate periods.

5.1.1. The Younger Dryas

The Younger Dryas was an abrupt and global climate event that occurred in ca. 12.9–11.7 ka (Cohen et al. 2018). Meanwhile, during the same period, there is evidence of the Atlantic Meridional Overturning (AMOC) weakening (MacManus et al. 2004). AMOC reduction would have been responsible for cooling in the Northern Hemisphere and warming in the Southern Hemisphere by moving the ITCZ to South (Carlson 2010). In the Southern Hemisphere, the temperature in Antarctica showed warming at the Early Holocene period (between 11,000- and 9,000-years BP) (Masson et al. 2000).

Cariaco basin, on Venezuela, showed a shift from a warm and wet to a cold and dry condition at YD (Lea et al. 2003) and authors believe that there was a southward shift of the ITCZ during the period. Concurrent, Lake Titicaca indicated maximum water level (Baker et al. 2001). The results obtained in Lago Airo showed evidence that during YD climatic conditions in the region were more humid. The lowest average TOC (1.34%) as recorded. As seen previously, the low TOC concentration is related to high hydrodynamic energy, which may represent a strong fluvial influence at the lake.

5.1.2. Early (11,700- 8,200 Years BP) to Mid Holocene (8,200–4,200 Years BP)

The period of the early-mid-Holocene represented an intermediary climate condition in records all over the Amazon region (De Freitas et al. 2001, Moreira et al. 2013a, Cordeiro et al. 2008, 2014). Chemical analysis of Lake Valencia, Venezuela (Lewis and Webezahn 1981) showed that after the last glacial maximum the lake was ephemeral until 10,500 years B.P. From then on there was a decrease in salinity associated with a wet period in 8,000 years B.P. From 8,000 years B.P. to 7,500 years B.P. salinity increased suggesting a period of intensified dry conditions. Clayey clastic material was deposited between 7,400 and 6,000 years B.P. with low levels of sedimentary chlorophyll suggesting a return to dry conditions. Pollen analysis in the same lake showed changes in vegetation after the last glacial period, with forest-semideciduous vegetation appearing between 9,800 and 8,300 years B.P. (Leyden 1985). Between 8,300 and 5,200 years B.P. there was an expansion in savannah vegetation. In Lake Acarabixi, middle Rio Negro basin, a gap of sedimentation was observed between 8,200–1,500 cal yrs B.P. after a substantial decrease in carbon rates that coincides with the dry phase of the Holocene in Carajás and low water events in Roraima and Humaitá (Cordeiro et al. 2008).

Between 8,700 and 5,800 cal yr BP, it was observed dry events in Ecuador, western Amazon (Weng et al. 2002). Concurrently, the highest frequency of forest fires in the Amazon region were recorded for ca. 8,000–4,000 cal yr BP (Cordeiro et al. 2014). A middle Holocene dry event was also documented in the Peruvian Amazon that appears to have lasted from ca. 7,200 cal yr BP until ca. 3,300 cal yr BP (Bush et al. 2007). Climate simulations detected an intensification of seasonal winds at this period, which may have caused the diminishing of wind convergence upon the Amazonian region and this condition might explain the precipitation rates’ drop in this region during this time (Melo and Marengo 2008).

On early-mid-Holocene, the Lago Airo area showed low diatoms density, however at some periods presenting peaks of planktonic diatoms. These peaks on ca. 10,930 cal yr BP and 9,100 could represent a formation of a lentic system, intervals in which the Rio Negro disconnected of Lago Airo area, at a period when the Rio Negro connection with Lago Airo Area was strong. The records of the early-mid-Holocene on Lago Airo area represented a mainly lotic phase inferred by coarse grains results.

5.1.3. Late Holocene (4,200 Years BP to present).

During the late Holocene, the Amazon region was more humid than early mid-Holocene (Moreira et. al. 2013a). We interpreted the possible isolation of Lago Airo area from the Rio Negro as an indicative of a wetter condition, where highly dense vegetation might have isolated the connection with the river and formed a lake as is currently observed. This suggestion can be reinforced by the diminished C/N ratio and sand concentration. With lake isolation, the organic matter was less diluted, so the TOC presents high values during this period.

A wet condition in the Late Holocene was also registered at several regions at the Amazon rainforest (Fig. 8): Mayle et al. (2000) reported that humid evergreen rainforests on the southern margin of Amazonia expanded due to increased precipitation after 2790 cal yr BP. Similarly, increased lacustrine productivity was observed in the Carajás region (~ 2800–1300 cal yr BP) at 800 m elevation outside the Amazon Basin's floodplain (Cordeiro et al., 2008), indicated by higher chlorophyll derivatives and TOC accumulation rates. Comparable trends were reported in the eastern Colombian Amazon (Behling and Hooghiemstra, 2000), and central Amazon (Behling et al., 2001). Some authors indicate that this condition was linked with the northward shift of ITCZ position (Behling and Hooghimiestra 2000, Silva Dias et al. 2009).

In Calado Lake, 80 km from the confluence of the Amazonas River and the Rio Negro, Behling et al. (2001) found longer periods of high water in 4,070 cal yr BP. In the Llanos Orientales, Behling and Hooghiemstra (2000) found wetter climatic conditions after ca. 3,500-year BP, when according to them, grassland savanna decreased. In Lago Airo area a true-lake formation was inferred since the last ca. 2000 years due to more frequent isolation events of Lago Airo from the Rio Negro. In Lake Maracá in eastern Amazonia Sedimentation resumed (~ 3600–2700 cal yr BP) with reduced Amazon River inflow. Full hydrological connection with the river was established by 1880 cal yr BP, increasing sediment inputs from the Andes, altering mineral composition, organic matter sources, and soil pH, reflecting rising water levels (Moreira et al., 2013b). Geochemical analyses of Preto Lake sediments revealed Solimões River hydrological changes during the late Holocene. From 3600–400 cal yr BP, high river inflow maintained lake stage, contributing smectite-rich sediments and phytoplankton-derived organic matter. Over the last 400 cal yr, reduced river inflow and increased C3-plant runoff marked the lake's semi-isolation, likely caused by sediment accretion forming a natural levee (Moreira et al. 2020).

The ‘Dark Ages Cold Period’ (DACP) is the most frequently discussed climate anomaly of the first millennium, apparently, it occurred around 1,540 cal BP until 1,175 cal BP. (410 AD and 775 AD). This event has been considered as cold, reported by studies around the Northern Hemisphere (Helama et al. 2017). The same authors indicate that DACP might have proceeded because of a combination of factors, particularly volcanic and reduced solar activity.

Over the DACP beginning, the diatoms density records (valves/g) increased like never seen before on Lago Airo and the P:B diatoms ratio resulted on planktonic diatoms dominancy that indicates open water on a lentic system. These results showed that a true lentic system was in formation under this climatic event, that started with a lotic-like system at the beginning (ca. 4,200–3,600 years BP) of the Late Holocene, being influenced by the Rio Negro floodplain.

After 1,000 cal yr BP in an isolate lake from the river influence at Carajás region (Cordeiro et al. 2008a) as well as in the floodplain areas of the Amazon River (Moreira et al. 2004), the last 1000 years cal BP were the wettest period on record. This pattern is already identified at low latitudes in the Northern Hemisphere in Roraima Savannas (Beltran et al. 2023) that also experienced wetter conditions and established the current lake level and climate after 1000 years.

In the Amazonas River lowlands, Lake Grande do Curuai and Maracá Lake showed the higher rate of sedimentation recorded since 2,700 cal yr BP due to the discharge of the Amazon River, caused by changes in ITCZ frequency southwards (Moreira-Turcq et al. 2014). According to the planktonic diatoms’ dominancy interpretation, at DACP Lago Airo water-level started to increase to nowadays’ levels, which corroborates with the argument of the shift of ITCZ frequency southwards.

It seems that during the Little Ice Age (ca. 300 years BP) a strong SASM probably reinforced the South Atlantic Subtropical Anticyclone, limiting the southward shift of the ITCZ and bringing drier conditions to northeast Brazil, a mechanism already suggested for the early and middle Holocene (Cruz et al. 2009) and for the late Holocene (Novello et al. 2012). P:B diatoms ratios on Lago Airo indicated that LIA had less wet conditions than DACP and MCA (See Fig. 8). In Fig. 8 we presented a plot of humid and dry periods from the end of Pleistocene to Holocene on South America latitudinal geographic coordinates 4ºN to 8ºS according to several authors.

Fig. 8

Plot of humid and dry periods from end of Pleistocene to Holocene (International Chronostratigraphic - Cohen et al., 2018) on South America latitudinal geographic coordinates 4ºN to 8ºS according to: 1- Llanos Orientales - Colombia (Behling and Hooghiemstra 1999); 2- Caracanã lake (Cordeiro et al. 2008b, 2014); 3- Roraima state (Desjardins et al. 1996); 4- Airo Lake (current paper); 5 - Acarabixi lake (Cordeiro et al. 2008a); 6- Yasuni National Park (Weng et al. 2002); 7- Santa Maria Lake and Sacuri Lake (Bush et al. 2007); 8- Santa Ninha Lake (Cordeiro et al. 2008b); 9a- Curuai floodplain (Moreira-Turqc et al. 2014); 9b- Maracá Lake (Moreira-Turqc et al. 2014) 10- Comprido Lake (Moreira et al. 2013a); 11 - Santarem (Soubies 1980); 12 - Tapajós Lake (Irion et al. 2006); 13 - Calado Lake (Behling et al. 2001); 14 - Quistococha Lake - Peru (Aniceto et al. 2014); 15 - Santa Ninha Lake (Moreira et al. 2012); 16 - Serra Sul Carajás (Alizadeh et al. 2017); 17 - Serra Sul Carajás (Sifeddine et al. 2001); 18 - Serra Sul Carajás (Hermanowski et al. 2012); 19 – Serra Norte de Carajás (Period seco entre 8000 − 4000. 1200, 700, 400 cal BP_ Cordeiro et al. 2008a). 19 - Humaitá (De Freitas et al. 2001). YD = Younger Dryas and P = Pleistocene.

6. Conclusion

The integration of diatom assemblages, sedimentological data, and geochemical indicators reveals a clear trajectory of environmental change in Lago Airo over the past 13,300 years, shedding light on paleohydrological variability in the Amazonian Rio Negro basin. This provides valuable insights into the climatic variability from the end of the Pleistocene to the Holocene in the Amazon basin. The integration of the data highlights the lake's ecological dynamics.

Over the past 13,300 years, Lago Airo has undergone significant environmental changes driven by hydrological and sedimentary dynamics. During Phase IV (ca. 13,400–11,800 cal yr BP), the lake exhibited high-energy lotic conditions, dominated by oligotrophic and periphytic species such as Eunotia hirudo and E. bilunaris, with coarse sediments and low diatom density reflecting strong fluvial influence and low nutrient availability. In Phase III (11,800–8,800 cal yr BP), the system transitioned toward more stable conditions, marked by increased clay content, slightly higher TOC values, and continued dominance of periphytic diatoms. These changes indicate reduced fluvial influence, moderate detrital input, and a gradual shift toward greater sedimentation stability and slightly higher productivity. Phase II (8,800–4,100 cal yr BP) saw the emergence of planktonic species like Aulacoseira spp., signaling a transition to more lentic conditions with diminished Rio Negro influence. This phase aligns with mid-Holocene dry conditions, lower lake levels, and increased organic input with terrestrial source. By Phase I, 4,100 cal yr BP to present, the lake had fully transitioned to a lentic environment, characterized by finer sediments, higher diatom density, and dominance of planktonic species. This phase reflects eutrophic conditions, increased productivity, and the wettest climatic period in Amazonia over the past millennia, promoting ecosystem stabilization and organic matter accumulation.

This hydrological variability indicates a dry early-mid-Holocene climate with extreme events at the Medium Rio negro Basin compared to a wetter and more stable late Holocene, which would be attributed to shifts in the position of the ITCZ and South American Monsoon System.

Around 4,000 years ago, under a wetter climate and more regular rainfall, Lago Airo transitioned to a lentic environment as a result of this new hydroclimatic regime, establishing more eutrophic conditions characterized by finer sediments, higher diatom density, and increased productivity. This contrasts with the Middle Holocene, when the Rio Negro likely transported coarser sediments during extreme climatic events.

Our study showed lower levels of water along with less influence of Rio Negro water’s on Lago Airo from 13,400 years to 4,100 years BP (Phases IV, III, and II, which could be inferred by fewer diatom valves/g, increasing percentages of silt, high C/N ratio and low values of δ15N. Whereas from 4,000 years BP to recently (Phase 1, the Rio Negro disconnected from Lago Airo and an increase in water level occurred, characterized by higher valves/g, planktonic taxa dominance, and δ¹⁵N values increase.

Acknowledgement

We are grateful for the support of the Bilateral Scientific Cooperation Project IRD-França/CNPq, Laboratório Mixto Internacional PALEOTRACES and the "INSU-EC2CO Dynamique du carbone et les changements climatiques dans le Bassin Amazonien. The authors acknowledge research support from the “Conselho Nacional de Pesquisa” (CNPq) process: 3479873/2013-5, 308222/2015-6 and “Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro” (FAPERJ) process: E-26/111.625/2011. Also, for the financial aid of the Project "RED0026/2014 (Fapesb) - Reconstrução Paleoambiental da Baía de Camamu: Implicações para a Gestão Ecossistema" and for the support of the EcoPaleo lab (Biology Institute, UFBA, BR). We are thankful for the assistance of Dr. Abdelfettah Sifeddine (IRD, France), and Fundação Oswaldo Cruz - Fiocruz (Salvador/BA, BR). This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) - Finance Code 001. We also acknowledge the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, Brazil) for a Research Productivity Fellowship (PQ)

Electronic Supplementary Material

Below is the link to the electronic supplementary material

Supplementary Material 1

Author Contribution

LRP, JCCV, DFG, DF, and RCC: wrote the main manuscript textPM-T, BT, LSM, and RCC: conceptualization JDADC, GSM, LRP, JCCV, DFG, DF, AB: Data curation , Formal analysis

References

Absy, ML, Cleff, A, Fournier, M, Martin, L, Servant, M, Sifeddine, A, Ferreira DA Silva,M, Soubies, F, Suguio, K, Turcq, B, Van Der Hammen, TH, (1991) Mise en évidence dequatre phases d'ouverture de la forêt dense dans le sud-est de l'Amazonie au coursdes 60000 dernières années. P Première comparaison avec d'autres régions tropicales.C.R. Acad. Sci. 312, 673–678.

Alizadeh K, Matthias I, Rodríguez-Zorro PA, et al (2017) Forest-savanna boundary shift on the plateau of Serra Sul dos Carajás (southeastern Amazonia) since the mid-Holocene; driving forces and limiting factors. Quat Int 449:12–21.

Aniceto K, Moreira-Turcq P, Cordeiro RC, Fraizy P, Quintana I, Turcq B (2014) Holocene paleohydrology of Quistococha Lake (Peru) in the upper Amazon Basin: Influence on carbon accumulation. Palaeogeogr. Palaeoclimatol. Palaeoecol 415: 165–174.

Baker PA, Seltzer GO, Fritz SC, Dunbar RB, Grove MJ, Tapia PM, Cross SL, Rowe HD, Broda JP (2001) The History of South American Tropical Precipitation for the Past 25,000 Years. Science 291: 640–643.

Barbosa, J.A., Cordeiro, R.C., Silva Filho, E.V., Turcq, B., Gomes, P.R.S., Santos, G., Sifeddine, A., Albuquerque, A.L.S., Lacerda, L.D., Hausladen, P.A., Tims, S.G., Levchenko, V.A., Fifield, L.K. (2004) 14C-AMS as a tool for the investigation of mercury deposition at remote Amazon location. Nuclear Instruments & Methods in Physics Research. Section B. Beam Interactions with Materials and Atoms 223 – 22, 528–534.

Battarbee RW (2000) Palaeolimnological approaches to climate change, with special regard to the biological record. Quat Sci Rev 19:107–124.

Battarbee RW, Jones VJ, Flower RJ et al (2001) Diatoms. In: Smol JP, Birks HJB, Last, WM (eds) Tracking Environmental Change Using Lake Sediments, Vol 3, Terrestrial, Algal, and Siliceous Indicators. Kluwer Academic Publishers, Dordrecht, pp.155–202.

Behling, H and Hooghiemstra, H (1999) Environmental history of the Colombian savannas of the Llanos Orientales since the Last Glacial Maximum from lake records El Piñal and Carimagua. Journal of Paleolimnology 21, 461–476.

Behling H, Hooghiemstra H (2000) Holocene Amazon rainforest-savanna dynamics and climatic implications: High-resolution pollen record from Laguna Loma Linda in eastern Colombia. Journal of Quaternary Science: 15, 687–695.

Behling H, Keim G, Irion G, et al (2001) Holocene environmental changes in the central amazon basin inferred from Lago Calado (Brazil). Palaeogeogr Palaeoclimatol Palaeoecol 173:87–101.

Bennett, K. D. (1996) Determination of the number of zones in a biostratigraphical sequence. New Phytologist, 132: 155–170.

Beltran, J., Palace, M., Dibb, J., & Bush, M. B. (2023). Holocene histories of biome stability in northern Amazonian savannas. The Holocene. https://doi.org/10.1177/09596836231211878

Blott S (2000) GRADISTAT Version 4.0: A Grain Size Distribution and Statistics Package for the Analysis of Unconsolidated Sediments by Sieving or Laser Granulometer. Surface Processes and Modern Environments Research Group, Department of Geology, Royal Holloway University of London (Available at http://www.kpal.co.uk/gradistat.html).

Bush, MB and Colinvaux, PA (1988). 7000-year record from the Amazon low lands, Ecuador. Vegetation. 76. 141–154.

Bush MB, Silman MR, Listopad CMCS (2007) A regional study of Holocene climate change and human occupation in Peruvian Amazonia. Journal of Biogeography: 34, 1342–1356.

Bush, MB., De Oliveira, PE, Colinvaux, PA, Miller, M.C., Moreno, JE (2004) Amazonian paleoecological histories: one Hill, three watersheds. Palaeogeography, Paleoclimatology, Paleoecology 214, 359–393.

Carlson AE (2010) What caused the Younger Dryas cold event? Geology: 38, 383–384

Chambers FM, Brain SA, Mauquoy D, et al (2014) The “Little Ice Age” in the Southern Hemisphere in the context of the last 3000 years: Peat-based proxy-climate data from Tierra del Fuego. The Holocene 12:1649–1656.

Cohen KM, Harper DAT, Gibbard PL (2018) ICS International Chronostratigraphic Chart 2018/08. International Commission on Stratigraphy, IUGS. www.stratigraphy.org (visited: 2018/08/20).

Colinvaux, P.A., De Oliveira, P.E., Moreno, J.E., Miller, M.C., Bush, M.B., 1996. A longpollen record from lowland Amazonia Forest and cooling in glacial times. Science 274, 85–88.

Cordeiro, R.C., Turcq, B., Suguio, K., Ribeiro, C.V., Silva, A.O., Martin, L., Sifeddine, A., 1997. Holocene environmental changes in Carajás Region (Pará, Brasil) record by lacustrine deposits. Verhandlungen des Internationalen Verein Limnologie 26, 814–817).

Cordeiro, R.C., Turcq, B., Suguio, K., Silva, A.O., SIfeddine, A., Volkmer-Ribeiro, C., (2008a). Holocene fires in east Amazonia (Carajás), new evidences, chronology and relation with paleoclimate. Glob. Planet. Chang. 61, 49–62.

Cordeiro RC, Moreira Turcq PF, Turcq BJ, Moreira LS, Rodrigues RdeC, da Costa RL, Sifeddine A, Simões Filho FFL, (2008b) Acumulação de carbono em lagos Amazônicos como indicador de eventos paleoclimáticos e antrópicos. Oecol Bras: 12, 130–154 (b)

Cordeiro RC, Turcq B, Moreira LS, et al (2014) Palaeofires in Amazon: Interplay between land use change and palaeoclimatic events. Palaeogeogr Palaeoclimatol Palaeoecol 415:137–151.

Cordeiro, R.C., Turcq, B.J., Sifeddine, A., Lacerda, L.D., Silva Filho, E.V., Gueiros, B.B., Cunha, Y.P.P., Santelli, R.E., Pádua, E.O., Pachinelam, S.R., 2011. Biogeochemical indicators of environmental changes from 50 ka to 10 ka. Paleogeogr. Paleocalimatol. Paleoecol. 299, 426–436.

Cruz FW, Vuille M, Burns SJ, Wang X, Cheng H, Werner M, Edwards RL, Karmann I, Auler AS, Nguyen H (2009) Orbitally driven east–west antiphasing of South American precipitation. Nat Geosci: 2, 210–214.

De Freitas H.A.; Pessenda, L.C.R.; Aravena, R.; Gouveia, S.E.M.; Ribeiro, A.S.; Boulet, R. 2001. Late Quaternary vegetation dynamics in the southern Amazon Basin inferred from carbon isotopes in soil organic matter. Quaternary Research: 55, 39–46.

Denicola DM (2000) A review of diatoms found in highly acidic environments. Hydrobiologia 433: 111–122.

Desjardins, T, Carneiro Filho A, Mariotti A, Chauvel A, Girardin C (1996) Changes of the forest-savanna boundary in Brazilian Amazonia during the Holocene revealed by stable isotope ratios of soil organic carbon. Oecologia 108:749–756.

Ferrari F, Procopiak LK, Alencar YB, Ludwig TAV (2007) Eunotiaceae (Bacillariophyceae) em igarapés da Amazônia Central, Manaus e Presidente Figueiredo, Brasil. Acta Amaz 37: 1–16.

Franzinelli E, Igreja H (2002) Modern sedimentation in the Lower Negro River, Amazonas State, Brazil. Geomorphology 44: 259–271.

Fontes, D., Cordeiro, R. C., Martins, G. S., Behling, H., Turcq, B., Sifeddine,A., Rodrigues, R. A. (2017). Paleoenvironmental dynamics in South Amazonia, Brazil, during the last 35,000 years inferred from pollen and geochemical records of Lago do Saci. Quaternary Science Reviews, 173, 161–180. https://doi.org/10.1016/j.quasc irev.2017.08.021

Garreaud RD, et al (2008) Present-day South American climate. Paleogeogr Palaeoclimatol Palaeoecol 281: 180–195.

Helama S, Jones PD, Briffa KR (2017) Dark Ages Cold Period: A literature review and directions for future research. Holocene 27: 1600–1606.

Hermanowski B, da Costa ML, Behling H (2012) Environmental changes in southeastern amazonia during the last 25,000 year revealed from a paleoecological record. Quaternary Research 77: 138–148.

Irion G, Bush MB, Nunes De Mello JA et al. (2006) A multiproxy palaeoecological record of Holocene lake sediments from the Rio Tapajos, eastern Amazonia. Palaeogeography, Palaeoclimatology, Palaeoecology 240: 523–536.

Lange-Bertalot H (1966) Iconographia Diatomologica, Annoted Diatom Micrographs Edited by Horst Lange‐Bertalot. Koeltz Scientific Books, 2: 226

Lange-Bertalot H (2007) Iconographia Diatomologica. Tropical Diatoms of South America II, Stuttgart: Koeltz Scientific Books, 18

Latrubesse EM, Franzinelli E (2005) The late Quaternary evolution of the Negro River, Amazon, Brazil: Implications for island and floodplain formation in large anabranching tropical systems. Geomorphology 70: 372–397

Lea, DW, Pak DKP, Larry CH, Konrad A (2003) Synchroneity of Tropical and High-Latitude Atlantic Temperatures over the Last Glacial Termination. Science 301: 1361–1364

Leyden, BW (1985) Late Quaternary Aridity and Holocene Moisture Fluctuation in the Lake Valencia Basin, Venezuela. Ecology. 66.1279–1295.

Lewis, M. & Webezahn, F.H. (1981) Chemistry of a 7,5-m sediment core from Lake Valencia, Venezuela. Limnol. Oceanogr.

Mann ME, Zhang Z, Rutherford S, Bradley RS, Hughes MK, Shindell D, Ammann C, Faluvegi G, Ni F (2009) Global Signatures and Dynamical Origins of the Little Ice Age and Medieval Climate Anomaly. Science 78: 1256–1260

Mann, ME (2002) Little Ice Age. Volume 1: The Earth system: physical and chemical dimensions of global environmental change (Eds Michael CM, Dr John SP) In: Encyclopedia of Global Environmental Change pp 504–509

Martins FR, Matthes LAF (1978) Respiração edáfica e nutrientes na Amazônia (Região de Manaus): floresta arenícola, campinarana e campina. Acta Amaz 8:233–244

Masson V, Vimeux F, Jouzel J, Morgan V, Delmotte M, Ciais P, Hammer C, Johnsen S, Lipenkov VY, Mosley-Thompson E, Petit JR, Steig EJ, Stievenard M, Vaikmae R (2000) Holocene Climate Variability in Antarctica Based on 11 Ice-Core Isotopic Records. Quaternary Research 54: 348–358

Mayle FE, Burbridge R, Killeen TJ. Millennial-scale dynamics of southern Amazonian rain forests. Science. 2000;290(5500):2291-4. doi: 10.1126/science.290.5500.2291. PMID: 11125139.

Mayle, FE., Power, MJ, 2008. Impact of a drier early mid-Holocene climate upon Amazonian forests. Philos. Trans. Royal Soc. B 363 (1498), 1829–838.

McCormac F, Hogg A, Blackwell P, Buck C, Higham T, Reimer P (2004). Shcal04 Southern Hemisphere Calibration, 0–11.0 Cal Kyr BP. Radiocarbon, 46(3), 1087–1092

Melo MLD, Marengo JA (2008) The influence of changes in orbital parameters over South American climate using the CPTEC AGCM: simulation of climate during the mid-Holocene. The Holocene 18: 501–516

Metzeltin D, Lange-Bertalot H (1998) Iconographia diatomologica: tropical diatoms of South America I, 5

Meyers P, (2003) Applications of organic geochemistry to paleolimnological reconstructions: a summary of examples from the Laurentian Great Lakes. Org Geochem 34: 261–289

Meyers P, Ishiwatari R (1993) Lacustrine organic geochemistry—an overview of indicators of organic matter sources and diagenesis in lake sediments. Org Geochem 20: 867–900

Moreira, L.S., Moreira-Turcq, P, Cordeiro, RC., Turcq, B., Aniceto, KC, Moreira-Ramirez, M, Cruz, APS, Caquineau, S, Silva, VC (2020). Climate and hydrologic controls on late Holocene sediment supply to an Amazon floodplain lake. J Paleolimnol 64, 389–403.

Moreira LS, Moreira-Turcq P, Cordeiro RC, et al (2013a) Holocene paleoenvironmental reconstruction in the Eastern Amazonian Basin: Comprido Lake. J South Am Earth Sci 44:55–62.

Moreira LS, Moreira-Turcq, P, Turcq, B, Cordeiro, RC, Kim, J-H, Caquineau, S., Mandengo Yogo, M., Sinninghe Damsté, J.S., (2013b) Palaeohydrological controls on sedimentary organic matter in an Amazon floodplain lake, Lake Maracá (Brazil) during the late Holocene. The Holocene 0:1–12

Moreira, L. S. Moreira-Turcq, P., Turcq, B., Caquineau, S., Cordeiro, R.C., (2012) Paleohydrological changes in an Amazonian floodplain lake: Santa Ninha Lake. Journal of Paleolimnology: 48, 339–350

Moreira-turcq P, Turcq B, Moreira LS, Amorim M, Cordeiro RC, Guyot J (2014) A 2700 cal yr BP extreme flood event revealed by sediment accumulation in Amazon floodplains. Palaeogeography, Palaeoclimatology, Palaeoecology 415: 175–182

Moro RS and Fürstenberger CB, (1997) Catálogo dos Principais Parámetros Ecológicos de Diatomáceas Naomarinhas, Universidade Estadual de Ponta Grossa, Parana.

Nascimento, M. N, Peters-Schulze, G., Martins, G. S, Cordeiro, R. C, Turcq, B., Moreira, L. S, & Bush, M. B. (2021). Limnological response to climatic changes in western Amazonia over the last millennium. Frontiers of Biogeography, 13(2). http://dx.doi.org/10.21425/F5FBG50860

Novello VF, Cruz FW, Karmann I, Burns SJ, Stríkis NM, Vuille M, Cheng H, Edwards RL, Santos RV, Frigo E, Barreto EAS (2012) Multidecadal climate variability in Brazil's Nordeste during the last 3000 years based on speleothem isotope records. Geophys Res Lett: 39

Patrick R, Reimer CW (1975) The diatoms of United States. Philadelphia: Academy of Natural Sciences, 23

Potter C, Klooster S, Steinbach M, Tan PN, Kumar V, Shekhar S, de Carvalho CR (2004) Understanding global teleconnections of climate to regional model estimates of Amazon ecosystem carbon fluxes. Global Change Biology 10: 693–703

Rasmussen, S O, Andersen, KK, Svensson, AM, Steffensen, JP, Vinther, BM., Clausen, H. B., Siggaard-Andersen, M-L., Johnsen, S. J., Larsen, LB, Dahl-Jensen, D, Bigler, M. (2006). "A new Greenland ice core chronology for the last glacial termination" (PDF). Journal of Geophysical Research. 111 (D6).

Ribeiro A, Villa NA, Lu N, et al (1996) Análise do regime pluviométrico da região Amazônica a partir de dados de onze localidades. Rev Bras Meteorol 1:25–35

Rodríguez-Zorro PA, Turcq B, Cordeiro RC, Moreira LS, Costa RL (2017) Forest stability during the early and late Holocene in the igapó floodplains of the Rio Negro, northwestern Brazil. Quat Res 1–15

Round FE, Crawford RM, Mann DG (1990) The Diatoms: biology and morphology of the genera. Cambridge University Press, New York. pp 747

Sanger JE, Gorham E (1972) Stratigraphy of fossil pigments as guide to the postglacial history of Kirchner marsh, Minesota. Limnology and Oceanography 17: 840–854

Sifeddine, A.; Martin, L.; Turcq, B.; Volkmer-Ribeiro, C.; Soubiès, F.; Cordeiro, R. C.; Suguiu, K. (2001). Variations of the Amazonian rainforest environment: a sedimentological record covering 30,000 years. Palaeogeography, Palaeoclimatology, Palaeoecology 168: 221–235.

Silva Dias PL, Turcq B, Silva Dias MAF, Braconnot P, Jorgetti T (2009) Mid-Holocene Climate of Tropical South America: A Model-Data Approach. In: Vimeux F., Sylvestre F., Khodri M. (eds) Past Climate Variability in South America and Surrounding Regions. Developments in Paleoenvironmental Research, vol 14. Springer, Dordrecht.

Silva M, Cohen M, Rossetti D, Pessenda L (2018). Did Sea-Level Changes Affect the Brazilian Amazon Forest during the Holocene? Radiocarbon 60: 91–112

Simões Filho, F.F.L., 2000. Sedimentação lacustre e implicações paleoambientais na região de contato floresta-savana de Roraima durante o Holoceno. (Ph.D. thesis) Universidade Federal Fluminense, Niterói, Brazil.

Soubies F (1980) Existence d'une phase sèche en Amazonic Brésilien ne datée para la presenze de charbons dans les sols (6000 – 3000 anos B.P.). Cah. O. R. S.T. O. Pi., Sér Géol: 11:133–146.

Spaulding, S.A.; Potapova, M.G.; Bishop, I.W.; Lee, S.S.; Gasperak, T.S.; Jovanoska, E.; Furey, P.C.; Edlund, M.B. 2021. Diatoms.org: supporting taxonomists, connecting communities. Diatom Research, 36: 4, 291–304, DOI: 10.1080/0269249X.2021.2006790.

Stuiver M, Reimer PJ, Reimer RW (2018) CALIB 7.1 [WWW program] at http://calib.org, accessed 2018-6-20.

Weng C, Bush MB, Athens JS (2002) Two histories of climate change and hydrarch succession in Ecuadorian Amazonia. Review of Palynology and Paleobotany 120: 73–90

Werth D, Avissar R (2002) The local and global effects of Amazon deforestation, J Geophys Res Atmospheres, 107(D20), LBA-55

Wolin JA, Stone JR (2010) Diatoms as indicators of water level change in freshwater lakes In:Smol JP, Eugene FS (eds) The diatoms: applications for the environmental and earth science. Cambridge University Press, United Kingdom

Supplementary Material

Yes