Search for: All records

Tropical glaciers have retreated over recent decades, but whether the magnitude of this retreat exceeds the bounds of Holocene fluctuations is unclear. We measured cosmogenic beryllium-10 and carbon-14 concentrations in recently exposed bedrock at the margin of four glaciers spanning the tropical Andes to reconstruct their past extents relative to today. Nuclide concentrations are near zero in almost all samples, suggesting that these locations were never exposed during the Holocene. Our data imply that many glaciers in the tropics are probably now smaller than they have been in at least 11,700 years, making the tropics the first large region where this milestone has been documented. 

Paleoclimate records from the tropical Andes are scarce, and the variability of glacial-interglacial cycles is still not well characterized. Lake Junin, in the Peruvian Andes, offers a unique and continuous paleoclimate archive that spans the last 700,000 years. Here, we explore the potential of organic compounds in reconstructing Andean paleoclimate over the last 20,000 years. To address this, we first evaluated the preservation of organic matter in the lake’s sediments. The Carbon Preference Index (CPI) suggests that n-alkanes have not been altered, and their H isotope composition can be used as paleo precipitation proxies. Furthermore, biomarkers from Eustigmatophyte algae (long chain diols) and diatoms (loliolide/isololiolide) have been identified, and can be used to reconstruct the hydrogen isotopic composition of lake water. The contrast between rainfall and lake water will be a good tool for understanding lake water inputs through time as well as evaporation and aridity. Changes in n-alkane chain length will be used to identity the terrestrial plant (long chain n-alkanes) and aquatic macrophyte inputs (mid-chain n-alkanes), with potential implications for interpreting past lake level change as a function of climate. Finally, distributions of br-GDGTs (branched glycerol dialkyl glycerol tetraethers) will be used to reconstruct past temperature changes. With these proxies, we aim to characterize climate variability at the end of the Last Glacial Maximum (LGM) and the Holocene, with a focus on characterizing climate variability in the light of teleconnections between the South American Summer Monsoon and global climate patterns and their relationship with hydroclimate in the Amazon Basin. 

Atmospheric water vapor is predominately sourced from the tropics, such that characterizing the link between the tropical water cycle and global climate is of critical importance. Studies of central Andean climate from Lake Junín (11 °S, Peru) show that tropical glacial extent tracks global ice volume at a ~100 ka periodicity for the last 6 glacial cycles, indicating a tight coupling between tropical water balance and high latitude climate. However, it can be difficult to decouple temperature, precipitation, and water balance histories from records of glacial extent, especially for older intervals. In this work, we focus on one such interval, MIS 15 (621–563 ka), when the connections between tropical Andean water balance and global climate seem different than the last glacial cycle. Globally, MIS 15 was a weak interglacial, with cool temperatures and low GHG concentrations, however, the Lake Junín glacial record suggests an amplified hydroclimate response to this interglacial, stronger than any other over the last 700 ka. Causes for this apparent tropical amplification may be due to large, precession-paced changes in meridional insolation gradients that exceed other interglacials owning to enhanced orbital eccentricity. Given that the role of precession on South American monsoon strength over the last glacial cycle is well established, we hypothesize that monsoon strength may have been highly variable during MIS 15 and forced changes in central Andean water balance and glacial extent. To test this, we reconstructed temperature and evaporation histories using carbonate clumped and triple oxygen isotopes of Lake Junín sediments. Preliminary results suggest temperatures were relatively stable, but possibly lower than both the present and Holocene, consistent with cool global climate at that time. Triple oxygen isotope values vary substantially, indicating massive swings in lake hydrology, between open and (nearly?) closed basin hydrology on a ~12 ka cycle that exactly match insolation variations. From this work, we conclude that hydrologic change in the central Andes was rapid and extreme during MIS 15, owning to profound changes in monsoon strength. Given that monsoons in other sectors are also sensitive to insolation changes, our work could suggest pervasive hydrologic variability throughout the tropics at this time. 

The El Niño Southern Oscillation (ENSO) is a major source of global interannual climate variability, however our ability to predict the response of ENSO to changes in the mean state of climate is limited in part by a paucity of long-term records of ENSO. The sediment record from Laguna Pallcacocha in El Cajas National Park, southern Ecuador (4060 masl; 2°46’S; 79°14’W) records El Niño floods spanning the Holocene (Rodbell et al., 1999; Moy et al., 2002; Mark et al., 2022). The sediment record is unusual for the nearly continuous dark- and light-colored laminations (0.1-2.0 cm thick) that comprise the Holocene section. Light laminae represent deposition during periods of increased precipitation, mobilization of unvegetated sediment above the lake, and increased stream discharge, all of which generate density-driven undercurrents. Conversely, dark laminae are deposited relatively slowly by sedimentation of organic matter. To date, no other lake has yielded such a high-resolution record of rainfall events, and here we review sediment cores from the western-most lakes in the region that are most likely to also be influenced by convective driven precipitation during coastal (Pacific) El Niño events. All lakes in this region contain multiple distal tephra 0.1-1.0 cm thick that enable precise correlation among records. Cores from Laguna Pampiada, Pampiada Bog, and Laguna Narigüiña are all located between 3500 and 4000 masl, all are located within steep catchments, and cores from these lakes reveal high-resolution records of clastic sediment delivery as recorded in bulk magnetic susceptibility. The stratigraphy of flood plain alluvium in the catchment of lakes studied provides an independent record of the geomorphic response to intense rainfall events. Buried soil A-horizons are clear indicators of major flooding events, and radiocarbon dates from charcoal in the uppermost sections of these A horizons can provide age estimates for large magnitude rain events and resultant floods that are comparable to the record from lake sediment cores. A 3-year record of atmospheric pressure and temperature from a data logger located at 4143 masl in El Cajas National Park provides the basis for comparing atmospheric conditions in the region of the studied lakes with those at sea level on the Ecuador coast, ~50 km to the west. 

Lake Junín, located in the uppermost Amazon Basin in central Peru, was drilled as part of the International Continental Drilling Program in 2015. A piston core with a composite length of ~95 m provides a continuous archive of upstream glacial activity spanning ~700,000 years. The age-depth model was established with 80 AMS 14C dates, 12 U-Th dated intervals of authigenic calcite, and 17 geomagnetic relative paleointensity tie points, and yields an age of 677±20 ka at 88 m. Four samples from near the base of the core reveal normal polarity paleomagnetic directions, consistent with an age younger than ~773 ka. The composite section comprises intervals of siliciclastic sediment intercalated with intervals dominated by authigenic calcite. The siliciclastic-rich intervals have a consistent signature, with relatively low concentrations of carbonate and organic carbon, and high values of bulk density, magnetic susceptibility and concentrations of elements derived from glacial erosion of the non-carbonate fraction of the regional bedrock. We find that tropical glaciers tracked changes in global ice volume and followed a clear ~100,000-year periodicity. Two caves, Huagapo and Pacupahuain, are located within 25 km of Lake Junín and provide a basis for testing and refining the age model of the Lake Junín drill core based on the high precision and accuracy of Uranium series dates for speleothems from these caves. The assumption here is that significant changes in regional ice volume will also be recorded in the 18O of cave drip water and thus in speleothems. Our initial target interval is the 9-8 marine isotope stage (MIS) boundary (~300 ka), which is recorded in the Junín drill core as an abrupt increase in the influx of glacigenic sediment, and in stalagmite 22-22 from Huagapo Cave as an abrupt 4.5‰ decrease in 18Ocalcite. The age of the onset of this transition in the Junín drill core is about 25 kyr older than that in Stal 22-22, and this difference is within the age model error envelope for the Junín drill core. Similar MIS boundaries provide the basis for adjustments in the Junín age model, which will improve the precision of correlation of this continuous record of tropical glaciation with paleoclimate archives in extra tropical regions. 

Glacial-interglacial transitions and abrupt millennial-scale events are the most prominent features in many paleoclimate records. Understanding these oscillations requires high-resolution time series from multiple locations to constrain the latitudinal response to forcings. Few high-resolution records exist from the Southern Hemisphere tropics that predate the last two glaciations. We present a high-resolution speleothem oxygen and carbon isotope record from Huagapo Cave in the Central Peruvian Andes covering Marine Isotope Stage (MIS) 8 glacial and MIS 9 interglacial (339 to 249 ka). Uranium-series dates on three stalagmites (n=18) with small age uncertainty ±1% allows us to resolve abrupt climate events similar in structure and duration to Dansgaard-Oescchger and Heinrich events. The South American Summer Monsoon (SASM) controls modern hydroclimate variability in the Andes, and previous records from Huagapo Cave have provided records of past SASM variability. Termination three (T-III) in our record has a steep increase in δ18O values of 5‰, punctuated by two stadial event decreases of ~3‰ (S8.1 and S8.2). This pattern is mirrored in the δ13C record, indicating that these millennial-scale events record hydroclimate and vegetation productivity changes. The same structure as our T-III record is found in other records globally, where they are noted to be Heinrich-like events. Frequency analysis indicates that the occurrence of these abrupt events changes between glacial cycles. Precession is weakly expressed in the δ18O record during MIS 8; similar to speleothem records from the region dating to the Last Glacial Maximum (LGM). Global ice cover and sea levels were similar in the LGM and MIS 8, but the Milankovitch insolation forcing differed. This change in SASM behavior is not observed in the East Asian monsoon, where the precession signal is dominant throughout. Interglacial precessional control is apparent during the latter half of MIS 9 and during Huagapo Cave intervals dating to MIS 6 and 7. These data indicate that the response to high-latitude forcing in the Southern Hemisphere tropics fluctuates through time, and potential explanations for low-latitude sensitivity to forcing factors are further explored. 

Recent studies have improved our understanding of how karst hydrology impacts variability in modern cave drip water δ18O values and the resultant calcite δ18O values of speleothem paleoclimate records. Monitoring of cave drip water isotope values reveals that flow path controls the differences in drip site values in many caves worldwide. We present a case study of three caves from the central Peruvian Andes where isotopic differences between sites are informed by monitoring data. Relative humidity at Huagapo and Pacupahuain caves is 100% year-round with no fluctuations, so any isotopic fractionation of waters must occur in the vadose zone or epikarst. Precipitation isotope data from the 2022-2023 year show differences with elevation, where annual mean precipitation at 3600 masl (meters above sea level) is, on average, 2‰ greater than precipitation at 4100 masl. Cave drip water was sampled four times (April, June, and November 2022, and June 2023). Average drip water δ18O values were lowest at the high elevation (4004 masl) cave of Antipayargunan -14.7 ± 2.5‰; similar values were found at the lowest elevation (3600 masl) cave of Huagapo -14.5 ± 1.2‰. Pacupahuain cave had the highest values with an average of -13.9 ± 1.7‰. The higher values at Pacupahuain Cave (3800 masl) may be attributed to higher evaporation due to vadose zone residence time, a lower average recharge elevation for this catchment and/or potential contribution from a sinkhole lake (Lago Gallerina) above the cave. Huagapo Cave is large, and sampling sites over 1 km in distance show that the δ18O value of drip water increases by 0.5‰ with increasing distance from the cave entrance. Drip counting sensor data and a continuous SYP autosampler at Pacupahuain Cave provide a time series showing that drip rate peaks during the monsoon season. More specifically, the data show a maximum of 2 ‰ difference in drip water at the autosampler site between the end of the wet season in May and the middle of the dry season in August – at which point drips cease for six months. Seasonal recharge dominates most drip water sites, while drip counters show evidence for fracture and diffuse flow-dominated drip sites. These data suggest that, similar to other cave sites, flow path is important for intra-cave differences in drip water isotope values. However, we find that karst hydrology plays a more dominant role between caves.