Speleothems from Warm Climates –
Holocene Records from the
Caribbean and Mediterranean Regions
Meighan Boyd
© Meighan Boyd ISSN 1653-7211
ISBN 978-91-7649-246-8 Paper I © Meighan Boyd
Paper II © Open Access. First published in the USA in International Journal of Speleology. Paper III © Oxbow Books
Paper IV © Meighan Boyd Paper V © Meighan Boyd Paper VI © Meighan Boyd
Cover illustration: Lake in Alepotrypa Cave, Photo by Giorgos Maneas Published papers typeset by respective publishers, reprinted with permission Printed by Holmbergs, Malmö 2015
Doctoral dissertation 2015 Meighan Boyd
Department of Physical Geography Stockholm University
Abstract
This thesis contributes to increased knowledge on Holocene climate and environmen-tal variability from two complex and sparsely studied areas. Using a speleothem from Gasparee Cave, Trinidad, as a paleoclimate archive, the local expression of the 8.2 ka (thousand years before 1950) climate event and associated patterns of the inter-tropical convergence zone (ITCZ) and rainfall is provided. Subsequent speleothem studies using multi-proxy analysis of stalagmites from Kapsia Cave and Alepotrypa Cave, Greece, pro-vide records of climate, vegetation and human induced changes in the cave environment during parts of the Holocene.
The speleothems from the well-studied Neolithic habitation site, Alepotrypa Cave, have produced a climate and habitation record which covers the period of 6.3-1.0 ka. The cave was inhabited between 8.0-5.2 ka and was closed by a tectonic event, which has preserved the settlement. The stable oxygen record shows the first well-dated and robust expression of the 4.2 ka dry event in the Peloponnese, places the timing of the 3.2 ka dry event within an ongoing dry period, and shows a final dry event at 1.6 ka. The North Atlantic as well as more regional drivers, such as the North Sea Caspian Pattern Index is proposed to, in a complex interplay, govern many of the climate trends and events observed.
Trace element variation after the site is abandoned indicate what is interpreted as two volcanic eruptions, the Minoan eruption of Thera (Santorini) around 3.6 ka and the 2.7 ka eruption of Somma (Vesuvius). Variations in trace elements during the habitation pe-riod show clear human influence, indicating an association with specific cave activities. One of the most interesting prospects for continued work on Alepotrypa Cave is this successful marriage of speleothem studies and archeology. A framework of dates which constrain some behavior of people living in the cave is only the beginning, and there is great potential to continue finding new clues in the speleothem data.
Keywords: Stable isotopes; U-Th dating; trace elements; stalagmite; speleothem;
Mid-Holocene; Caribbean; ITCZ; rapid climate change; climate; Eastern Mediterranean; Peloponnese; Santorini; Neolithic; Alepotrypa Cave
Sammanfattning
Denna avhandling bidrar till ökad kunskap om klimatets variationer och miljön i två geografiskt skilda områden på låga breddgrader och under tidsperioder inom den Holo-cena epoken. Genom att använda en droppsten (stalagmit) från Gasparee-grottan, Trini-dad, som ett paleoklimatarkiv, har det bland annat varit möjligt att visa att Trinidad upplevde torrare förhållanden under den snabba klimatförändring som observerats ske för 8200 år sedan på många platser i världen. Denna torrare klimatsituation i Trinidad föreslås vara ett resultat av en sydlig förflyttning av den intertropiska konvergenszonen. Övriga stalagmiter som studerats för denna avhandling kommer från Kapsia-grottan och Alepotrypa-grottan som finns på Peloponnesos-halvön i Grekland. Resultaten däri-från speglar dels klimat- och vegetationsvariatoner och dels graden av mänsklig aktiv-itet, under tiden för ca 8000 år sedan till för 1000 år sedan. Alepotrypa-grottan är känd för att vara en av de större Neolitiska boplatserna i Grekland. Isotop- och spårämne-sanalyser av stalagmiterna har bidragit med ny kunskap om tidpunkten för mänsklig aktivitet, hur människorna påverkade grott-miljön samt hur klimatet varierat efter det att grottan, genom en tektonisk händelse, stängdes för människans inverkan. Snabba klimatförändringar, för 4200 och 3200 år sedan, observerade i andra regioner, rekonstru-eras här för första gången på Peloponnesos. En snabb förändring mot torrare förhål-landen observeras även för 1600 år sedan. De klimatstyrande processerna föreslås vara en kombination av storskaliga processer som den nordatlantiska oscillationen och mer regionala processer som det så kallade North Sea Caspian Pattern Index. Variationer i spårämnen i stalagmiterna efter att Alepotrypa-grottan stängdes kan kopplas till två vulkaniska utbrott, nämligen det Minoiska utbrottet av Thera på ön Santorini kring 3600 år sedan och utbrottet av Somma (Vesuvius) kring 2700 år sedan. Spårämnesvariationer under bo-perioden ger tydliga indikationer på människans påverkan på grottmiljön och som delvis kan länkas till specifika aktiviteter, som eldning av dynga i grottan. Avhan-dlingen är ett resultat av en framgångsrik kombination av klimatstudier och arkeologisk kunskap och utgör ett viktigt underlag för fördjupat interdisciplinärt forskningssamar-bete i Alepotrypa-grottan.
Contents
1. Introduction 1
1.1. Aims and objectives of the thesis . . . 4
2. Site selection, speleothems, and analytical techniques 5 2.1. Caribbean . . . 5
2.2. Mediterranean . . . 6
3. Speleothems as climate archives 11 3.1. Caves and speleothem formation . . . .11
3.2. Proxy types . . . .12 4. The samples 17 4.1. Gasparee Cave . . . .17 4.2. Kapsia Cave . . . .17 4.3. Alepotrypa Cave . . . .17 5. Analytical Methods 27 5.1. Stable isotopes . . . .27 5.2. U-Th Dating . . . .27
5.3. Petrographic and SEM images . . . .30
5.4. Modern-day cave environment . . . .30
5.5. Trace elements . . . .31
6. Results 33 6.1. Constraining the timing of a rapid climate change . . . .33
6.2. Moving on from the Caribbean – Novel techniques applied to challenging material . . . .35
6.3. Alepotrypa Cave – Combining methods to broaden the ap-proach to speleothem studies . . . .36
6.4. Mid-Holocene Climate in the Peloponnese . . . .43
7. Discussion and Supporting Data 45 7.1. The Caribbean . . . .45
7.2. Speleothems in the Peloponnese . . . .46
8. Final remarks –
Perspectives on future research 57
9. Conclusions 59
10. Acknowledgements 61
11. References 65
Speleothems from Warm Climates –
Holocene Records from the
Caribbean and Mediterranean Regions
Meighan Boyd
Department of Physical Geography, Stockholm University, Sweden
List of papers
This doctoral thesis consists of this summary and the following papers, which are re-ferred to by their Roman numerals in the text.
Paper I
Boyd, M., Holmgren, K., Shaw, P., Hoffmann, D., Mangini, A., Mudelsee, M., Spötl, C.
manuscript. Early Holocene patterns of rainfall, vegetation and soil conditions, inferred
from a southern Caribbean stalagmite.
Paper II
Finné, M., Kylander, M., Boyd, M., Sundkvist, H.S., Löwemark, L. 2015. Can XRF scanning of speleothems be used as a non-destructive method to identify paleoflood events in caves? International Journal of Speleology, 44, 17-23.
Paper III
Boyd, M. and Holmgren, K. in press. Speleothems from Alepotrypa Cave: Towards cli-mate reconstruction. In: Alepotrypa Cave in the Mani, Greece: A festschrift to honor Dr.
G. Papathanasopoulos on the occasion of his 90th birthday. 2015. eds. Α. Papathanasiou, M.
Galaty, P. Karkanas, W. Parkinson, D. Pullen, Oxbow Books
Paper IV
Boyd, M., Karkanas, P., Papathanasiou, A., Hoffmann, D., Holmgren, K. manuscript. U-Th dating of calcite on human bones from Alepotrypa Cave, Greece.
Paper V
Boyd, M., Holmgren, K., Finné, M., Hoffman, D., Jochum, K.P., Karkanas, P., Papathanasiou, A., Scholz, D., Stoll, B., Spötl, C. manuscript. Stable isotopes and phosphorus patterns in calcite stalagmites from Alepotrypa Cave, Peloponnese, Greece as indicators of Holocene changes in rainfall and vegetation.
Paper VI
Boyd, M., Hoffman, D., Jochum, K.P., Karkanas, P., Krusic, P.J., Papathanasiou, A., Scholz, D., Stoll, B., and Holmgren, K. manuscript. Trace elements as recorders of human activity and environmental indicators at Alepotrypa Cave, Greece.
Author contributions
Paper I
Conceived and designed by KH, PS, and MB. MB wrote the paper in close collaboration with KH and PS. Stable isotope lab work carried out by MB under the guidance of CS. Dating of material by AM and DH, statistical analysis by MM. All authors contributed with commenting on the manuscript.
Paper II
Conceived and designed by MF, MB, HSS, MK and LL. MF wrote the paper and de-signed figures following discussions with MK. MB, HSS and LL contributed to discussion around interpretations and commented on the manuscript. Lab work conducted by MF, MB with supervision and help from MK and LL.
Paper III
Written by request by MB with input and editing from KH.
Paper IV
Conceived and designed, and written by MB with input from PK and AP on archeological issues. Lab work carried out by DH.
Paper V
Conceived and written by MB through discussion and comments from KH. Lab work performed by MB and made possible by the generous donation of lab time by DH, CS, KPJ and DS. Improvements in structure and discussion on interpretation from MF. Work made possible through collaboration with PK and AP, and support from their archeologi-cal data.
Paper VI
Conceived and written by MB with support and comments from KH, PJK, and KPJ. Lab work performed by MB and made possible by the generous donation of lab time by DH, KPJ, BS and DS. Statistics and MATLAB work contributed by PJK. Work made possible through collaboration with PK and AP and support from their archeological data.
1. Introduction
Speleothems (secondary cave carbonates, e.g. stalagmites and flowstones) provide a number of proxy records of past climate and cave environment variability. As they be-have as a closed system, speleothems can be dated using very precise radiometric decay methods, which makes them ideal for use in paleoclimate reconstruction. They occur in carbonate bedrock all over the world in locations where temperatures are above freezing for at least part of the year. They fill a particularly important gap in the global climate re-cord map by providing a terrestrial climate data in tropical regions where other archives of high-resolution, such as ice cores, are extremely rare. Low latitudes are traditionally underrepresented in climate research and are therefore subject to greater uncertainties in climate scenario modeling, and speleothems from these regions are particularly valu-able as climate proxies.
Models predicting future humidity patterns, rainfall seasonality, intensity of precipi-tation levels and atmospheric moisture require paleoclimate data as test cases to see how well the model produces known conditions. The lack of long-term observational data is a source of uncertainty when trying to constrain models (Knutti and Sedláček, 2012), and without accurate modelling results it will be difficult to secure water supplies and to mitigate potential inundation, flooding, and salinization which lead to severe impacts on ecosystems, infrastructure, agricultural production and human health.
To help improve the spatial and temporal availability of paleoclimate data in these sparsely studied regions, speleothems were collected from Gasparee Cave, Trinidad and Tobago, and from Alepotrypa Cave, Greece (Fig. 1). The former lies directly on the pres-ent northernmost position of the Intertropical convergence zone (ITCZ), experiencing a dry warm climate from January-June, and wetter conditions from July-December, pro-viding a seasonal distribution of rainfall. The latter experiences the dry hot summers and cool wet winters common in the coastal eastern Mediterranean, also with a distinct seasonal change in precipitation regime. Both of these regions are densely populated, and also face large challenges in the face of predicted global climate change. For this reason, and to address the uncertainty in the models stated by the Intergovernmental Panel on Climate Change (IPCC) in its 5th assessment report, more paleoclimate data from the tropics and Mediterranean regions could help to minimize uncertainty caused by large amounts of natural variability in the regions (Christensen et al., 2013). Through additional paleoclimate data sets to test climate models, it will be possible to improve the accuracy of modelled future climate scenarios, which is vital for planning and adaptation strategies to be effective.
An unusual and exciting aspect of the project is that Alepotrypa Cave is a well-studied and unique site of a Neolithic human settlement. In order to understand how past popula-tions coped with and adapted to climate change, it is ideal to combine paleoclimate stud-ies with archeological excavations, in particular if those climate proxstud-ies studied provide
a high chronological resolution and are of a length which is capable of capturing climate events of both long and short duration and frequency. By capturing the amplitude and frequency of climate variation and climate change and providing a secure chronology of the extent of these events, it is possible to untangle the complexity of how humans in the past have moved, adapted, or simply carried on in the face of droughts, cold periods and other challenges.
The climate of the Earth has varied and changed from a completely ice-covered snow-ball Earth to hothouse conditions. Massive changes on geological time scales concerning the arrangement of continents have often been tied to such changes. Within the qua-ternary period the climate of the earth more closely resembles that of today, owing to continental and oceanic extents which closely resemble those of the present. The state of
! ! ! B A 20° E -30° W -80° W 60° N 35° N 10° N -15° S 0 5,000 10,000 km
±
Figure 1. Map showing the two study sites. A) Gasparee Cave, Trinidad and Tobago. B) Pelopon-nese Peninsula, Greece.
understanding about past climate is constantly improving, and through the use of paleo-climate proxies, researchers have been able to produce long records of past paleo-climate from a multitude of sources. Marine sediment cores and ice cores have provided significant advances in the understanding of large-scale circulation patterns and climate variability, allowing for the recognition of glacial/interglacial cycles and their connection of insola-tion changes resulting from variainsola-tion in the orbit of the Earth (Berger and Loutre, 1991). Orbital forcing is not the only factor influencing climate, with sunspot activity, ocean and atmospheric circulation, and more recently, anthropogenic input of aerosols and greenhouse gases into the atmosphere also being involved (Kaufmann et al., 2011).
In order to test and disentangle the complexities and influence of these factors, and to test the reliability of climate models, paleoclimate proxy data are used. Climate archives such as tree rings, corals, ice cores, lake, bog, marine sediments, and speleothems are storage vehicles for past climate information. Each archive type can provide multiple proxies, including stable isotopes (e.g. δ13C, δ18O), pollen, tephra particles, macro and
microfossils such as seeds and diatoms, variation of trace elements, and growth rings or layer thickness. Each proxy may be related to a different climate variable (or variables) and can preserve information about these variables at different resolutions, represent single events, seasonal variation, and annual or decadal cycles and so on. It is by disen-tangling the complex processes which control the chemical and physical properties of climate archives that we can identify which climate signals are found therein. We can then use paleoproxies, which provide climate data back beyond the time for which em-pirical observations are available, to contribute to a deeper understanding of the climate picture.
By providing climate modelers with high-resolution climate reconstructions of the past, it is possible to test performance of models and also to see how well they reflect known conditions as a given point. Accurate future climate scenarios require testing against known values, and this is vital to meet the requests of policymakers and citizens who are facing an increasingly uncertain climate situation in the future.
The IPCC has identified the tropics and Mediterranean as an area where people are expected to experience major challenges in the face of climate change (Christensen et al., 2013). Sea-level rise leading to inundation of coastal areas, saltwater intrusion, increased severe storms, extreme temperatures, and a concentration of rainfall into short intense events with more arid conditions prevailing are all potential consequences of climate change. Overall, the coverage of paleoclimate research in these regions leaves much to be resolved. The prevalence of marine studies, which do not capture local to regional variation, and the relative lack of long, high-resolution terrestrial datasets presents an excellent opportunity for work with speleothems to greatly contribute to understanding of climate in these regions. For speleothems, advances in dating techniques have opened the door for better chronologies based on accurate and highly resolved dating (Hoffmann et al., 2009). As speleothems are a well-protected archive type due to the properties of the cave environment, they can survive for many hundreds of thousands of years (van Breukelen et al., 2008; Bajo et al., 2012), and provide near-continuous climate records.
1.1. Aims and objectives of the thesis
This project is the result of the combination of studies of speleothems from Gasparee Cave, Trinidad, and Kapsia Cave and Alepotrypa Cave, Greece. The unique human his-tory at Alepotrypa Cave presents both challenges and new opportunities when working with speleothems from this site, and these have both come to play a large role in the thesis work.
The aim of this thesis is to increase the spatial coverage of Holocene paleoclimate data from terrestrial archives in the Caribbean and Mediterranean, and to apply speleothem studies to the unique questions which are presented at archeological sites. The dual re-gion nature of the project has resulted in two main foci.
Climate and climate variability
• Are speleothems from each site suitable for paleoclimate reconstruction?
• If so, what do speleothems from each site contribute to increase our knowledge and understanding of past climate conditions in the regions?
These questions are addressed by Papers I and V.
Changing cave environments
• Which methods can be used to see how changes in the cave environment are expressed in speleothems?
• Which techniques can be applied to speleothems to reduce uncertainties in the archeological timeline of the Neolithic habitation at Alepotrypa Cave?
2. Site selection, speleothems, and
analytical techniques
In order to study changes in the climate system, and to understand the governing pro-cesses behind observed changes, a wide spatial coverage is required. Sites which lie on borders between climate zones, e.g. the ITCZ or monsoon regions, can provide insights into the extent and placement of these strong influences over time. Research from Oman (Fleitmann et al., 2007), China (Dykoski et al., 2005; Cosford et al., 2008; Zhang et al., 2013) and South and Central America (Leduc et al., 2009; Schmidt and Spero, 2011) have contributed to a more complete understanding of the dynamics of the ITCZ. The place-ment and extent of the ITCZ controls the rainfall of many tropical regions, and so un-derstanding ITCZ response to other changes in the climate system is vital if we are to be able to predict future climate change scenarios with less uncertainty.
Both the Caribbean and Mediterranean (Fig. 1) experience moderate temperatures and receive highly seasonal rainfall, and are situated in complex regions with many topographical factors affecting rainfall distribution. The geographical and climatological setting of these sites indicates they could provide important insight into the occurrence of what are considered global climate events.
2.1. Caribbean
The Caribbean region is dominated by ocean surface, with the climate being dominated by the interaction of oceanic and atmospheric systems. The southern islands are strongly influenced by the annual passage of the ITCZ (Mangini et al., 2007). Easterly trade winds combine with the ITCZ and provide the greatest influences on climate on the island of Trinidad. From May to December these winds reach their greatest strength, and pre-cipitation is at a maximum. Relatively high topographical relief influences wind expo-sure and causes orographic effects leading to significant local variation in precipitation. The Caribbean low level jet, and North Atlantic high-pressure combine with subsidence caused by convection in Central America to influence weather patterns of the region (Gamble and Curtis, 2008; Wang et al., 2008).
The island of Trinidad, 5128 km2, is located between 10°N and 11°N latitude, and
be-tween 61°W and 62°W longitude. It extends around 80 km from north to south, and 60 km from east to west and at its closest point is only 15 km offshore from Venezuela. Average annual precipitation near the cave site is 1500 mm (Trinidad Meteorological Service). Rainfall variation within the Caribbean is strongly controlled by topographical effects, with many islands having strong rainfall gradients from east to west.
2.1.1. Gasparee Cave
The entrance to Gasparee Cave is situated within Gaspar Grande Island (Fig. 2) with an entrance at 30 m above sea level (a.s.l). The bedrock of the area is Cretaceous Laventille Formation limestone, and Gasparee Cave likely formed during the Pleistocene. Today, it is the most visited show cave in Trinidad. Vegetation on Gaspar Grande Island is domi-nated by deciduous seasonal forest of the Bursera-Lonchocarpus association (Day and Chenoweth, 2004), and the island has thin soil coverage with many bedrock exposures. The lower parts of the cave lie at sea level, and the saltwater lagoon in the cave is in-fluenced by tidal movement through connections to the Gulf of Paria. It is decorated mainly with flowstones and columns, though some modern stalagmites are seen around the rim of the saltwater lagoon, which is a popular swimming spot for cave visitors. Preliminary investigations indicate that the cave extends well below the current sea level. Influenced by karst processes in more modern times, the cave has several large roof openings formed by collapses, and the cave is today open to the outer environment. Visitors enter the cave by descending 30 m of stairs in one of these collapse openings to the main chamber. The stalagmite, called GC2, formed in a space created by roof collapse toward the back of the cave, and was collected with its tip above the current tidal range.
2.2. Mediterranean
The Mediterranean region hosts a high density of archeological studies in comparison to many other populated regions, but boasts comparatively few robust paleoclimate re-constructions (Finné, 2014). It is also a region where the human impact on the environ-ment is evident reaching back for many thousands of years (Wick et al., 2003; Jalut et al., 2009). It can be seen in the form of vegetation type change (e.g. introduction of the olive
!Gasparee Cave -55° W -60° W -65° W -70° W -75° W 20° N 15° N 10° N 0 750 1,500 km
±
Figure 2. Map showing the location of Gasparee Cave on Gaspar Grande Island, off the northern west coast of Trinidad.
trees, Olea), agricultural animal husbandry, irrigation, and aqueducts. In order for hu-man societies to develop and thrive in an arid climate prone to extreme changes in water availability, adaptation strategies must have been employed (Weiss and Bradley, 2001).
The present-day climate of Greece is generally characterized as Mediterranean, with hot dry summers and mild wet winters. Temperatures are highest in the costal southern regions, with a gradient in temperature showing cooler annual averages on a northerly path, and also with increasing altitude and distance from the coast (Harding et al., 2009; Dotsika et al., 2010). Precipitation also changes spatially across the Mediterranean ba-sin, and is highly influenced by the orographic effects of the mountain ranges, with the western Mediterranean receiving more rain. Within the eastern Mediterranean region, a precipitation gradient ranges from over 2000 mm per year to less than 120 mm per year (Rohling et al., 2009). Precipitation is strongly associated with cyclogenesis in the Atlan-tic and western Mediterranean, and is strongly seasonal, with 70-80% of rainfall occur-ring between October and April. The main precipitation driver is the winter southward shift in the subtropical high-pressure systems. During the summer, the more northerly position of these highs causes a subsidence of air over the eastern Mediterranean, pre-venting cloud formation and severely limiting precipitation. By contrast, the winter con-ditions see an increase in the intrusion of North Atlantic synoptic low-pressure systems, bringing moist air into centers of cyclogenesis in the Gulf of Genoa, the Ionian Sea, and Cyprus (Kutiel and Benaroch, 2002; Argiriou and Lykoudis, 2006; Harding et al., 2009). Most cyclones passing over the Peloponnese have tracked southeastward from the Gulf of Genoa, though the same system of North Atlantic origin can experience consecutive cyclogenesis at all three centers (Trigo et al., 2002). Together, these systems bring pre-cipitation to the eastern parts of the basin.
The North Atlantic low-pressure systems are tied to the mode of the North Atlantic Oscillation (NAO), with NAO+ phases bringing cooler and drier conditions to the Eastern
Mediterranean (Xoplaki et al., 2004; Harding et al., 2009). NAO- conditions bring warmer
wetter conditions, as this mode allows for a greater penetration of the low-pressure systems. Another factor affecting the regional climate is the North Sea Caspian Pattern Index (NCP) (Kutiel et al., 2002; Kutiel and Benaroch, 2002) which has a marked effect on temperature in the western Peloponnese throughout the year. NCP mode is also respon-sible for precipitation changes over other parts of Greece, Turkey, and the Eastern Medi-terranean. A NCP- mode brings wetter conditions to most of Greece, along with warmer
temperatures, as warmer moist air flows inland from the Ionian Sea. With NCP+ cooler
drier continental air with a northeasterly wind brings cooler drier continental air into the Peloponnese but shows comparatively little influence on precipitation in the western Peloponnese (Kutiel et al., 2002; Kutiel and Benaroch, 2002). The Mediterranean Oscilla-tion (MO), which is defined by the pressure differences between staOscilla-tions at Algiers and Cairo, produces a see-saw of conditions between the western and eastern Mediterranean (Harding et al., 2009). For Greece, this results in more frequent southerly flow of cool air during a positive MO phase. Winter surface air temperatures are strongly related to the MO, when low-pressure systems are passing through the area, but no significant correla-tion is seen with summer surface air temperatures (Nastos et al., 2011).
2.2.1. Kapsia Cave
The Peloponnese peninsula (Fig. 3) makes up the southern part of mainland Greece, and is rich in archeological sites, such as Franchthi Cave which includes Upper Paleolithic through Mesolithic deposits (Vitelli, 1999; Stiner and Munro, 2011; Colonese et al., 2013), and evidence of Neanderthals from Lakonis (Panagopoulou et al., 2004). The
proxim-ity to the Hellenic arc contributes to high influences of tectonics on the Peloponnesian landscape, with high mountains at the central areas, to lowland river valleys which are found along the coast.
Kapsia Cave is located in the central Peloponnese (N37.623°, E22.354°) close to the city of Tripoli. The cave entrance is located around 700 m a.s.l. at the base of the Mainalo Mountains. The cave is located on the Mantinea Plain, which is drained by five sinkholes, one of which lies directly outside the natural cave entrance. In 2004 an artificial entrance was made in the cave and it was developed as a tourist attraction, and from 2010 onwards the cave has been open for visitors. The proximity to the sinkhole makes the cave vulner-able to flooding events when the sinkhole becomes plugged, and the most recent of these occurred in 2001. Over a meter of clay has been deposited in the cave by these floods, and there is a high-water mark seen on the cave walls in clay deposits related to massive flooding in the past. A more detailed description of the cave and associated speleothem studies there can be found in Finné (2014).
2.2.2. Alepotrypa Cave
Alepotrypa Cave is located on Diros Bay on the western side of the Mani Peninsula (N36.638°, E22.380°), on a promontory which is around 2 km wide, and it has maximum thickness of around 200 m. The modern-day cave entrance sits around 30 m a.s.l. and is set back about 50 m from the current seafront.
Orographic effects from the nearby 1000 m a.s.l. peaks of the Taygetos Mountains induce rainfall on the western side of the peninsula. Average rainfall and temperatures 64 km from the site at the nearby Methoni station for the period 1951-2008 are around 700 mm ± 150 mm per year, with over 90% of precipitation occurring between October and April, giving a negative water balance between May and September. Temperature is
! ! ! Athens Kapsia Cave Alepotrypa Cave 26° E 24° E 22° E 20° E 40° N 39° N 38° N 37° N 36° N 0 250 500 km
±
moderated by the sea, and the average annual temperature is 18.0 ± 0.4°C, with summer temperatures averaging 23.2 ± 2.5°C for May-September and winter average tempera-tures of 11.9 ± 1.4°C for December-February.
Vegetation above the cave is of C3 type, with old non-producing olive groves, and much thorny scrub dominating the area. On the steeper slopes vegetation is more sparse, but directly above the cave, the dominance of thorny scrub makes passage impossible.
The area immediately around Alepotrypa Cave includes Glyfada Cave and the Neo-lithic archeological site of Ksagounaki. At the time of its discovery by Petrocheilou of the Greek Speleological Society in 1958, the cave only had a very narrow opening. The pres-ence of Neolithic artifacts and a settlement site within the cave led to the site becoming the object of development as a promising tourist attraction. While not as highly deco-rated as its neighbor, Glyfada Cave (which is very popular with tourists), it was altered and much of the entrance chamber was modified to permit access. However, the site was recognized as having a great importance and so development was halted when manage-ment was taken over in AD 1970 by the Greek Ministry of Culture. Since AD 1970 the cave has been the site of ongoing excavation.
The cave seems to have been an important center for Neolithic culture in the Pelo-ponnese. It has been determined that occupation of the site began around 8.0 ka, and continued until abandonment of the site around 5.2 ka (Papathanasiou et al., 2000; Pa-pathanasiou, 2009; Papathanasopoulos, 2011). A tectonic event closed the cave either in conjunction with or shortly after this time as settlement deposits are, with the exception of work conducted after AD 1958, undisturbed (Paper IV, this volume). The history of the occupation of the cave is still being studied, but occupants explored the entirety of the mapped and much of the unmapped passageways. Throughout the cave, burials and ossuaries were found, and these have provided much evidence of human health and cul-tural habits relating to internment (Papathanasiou et al., 2000; Papathanasopoulos, 2011). Evidence of extensive trade networks supplying goods from across the Aegean is seen at the site, and finds of silver and copper jewelry, imported pottery, and obsidian give some insight into the wealth and importance of the Alepotrypa settlement (Papatha-nasopoulos, 2011). One possible explanation for the choice of the cave for settlement is the availability of fresh water through the very dry summer months. Within the cave a large brackish lagoon and several groundwater pools are found. Despite a negative sum-mer water balance there is active dripping at many sites within the cave at the end of the dry season, and this combined with moderate temperatures and good defensive position would contribute to its appeal.
The preservation of pottery from the site is outstanding, and many of the finds may be viewed in the Neolithic Museum of Diros adjacent to the cave entrance. Work by The Diros Project, a collaboration of the Ephorate of Paleoanthropology and Speleology with the American Field Museum, has produced more data on the extent and complexity of the Ksagounaki settlement just north of the cave entrance, and results from this project will be detailed in the forthcoming volume1 which includes Paper III from this thesis. The
Diros Project continues the work of G. Papathanasopoulos who has headed work at the cave for many years.
1 Papathanasiou, A., Galaty, M., Karkanas, P., Parkinson, W., Pullen, W. 2015.
3. Speleothems as climate archives
3.1. Caves and speleothem formation
Caves, most often found in limestone and dolomite bedrock, are a common feature in karstic landscapes. Named after the type locality in Slovenia, karstic landscapes are char-acterized by high porosity, fracturing, and dissolution pathways leaving very little surface water flow and drainage. Precipitation (either rain or snow, and then melt water upon thawing) and surface waters move through the bedrock. The vadose zone (including soil, epikarst and transmission zones), where waters can move through either fracture, fis-sure or seepage flows, sits above the water-saturated phreatic zone (Williams, 2008). The upper epikarst, along with the soil zone, functions as the primary source for CO2 in cave drip waters. Within the epikarst, high partial pressure of carbon dioxide (pCO2) is pres-ent as a result of biological respiration, and decomposition of organic materials (Fairchild and Baker, 2012) water in this zone is made slightly acidic. This slightly acidic water then moves through the limestone bedrock where it dissolves carbonate minerals, causing the water to become supersaturated in calcium while a high pCO2 is present. As the waters percolate deeper into the bedrock, they enter into areas of lower partial pressure, like caves. When the water seeps out into the cave pCO2 is lower, causing carbon dioxide to degas from the water and resulting in the precipitation of calcium carbonate (CaCO3) as speleothems (Fairchild et al., 2006) (Fig 5). While many other minerals can form speleo-thems, the most common ones are composed of calcite.
The term speleothem (Greek: spelaion, cave; thema, deposit) is used to describe sec-ondarily precipitated deposits within caves, but in the context of climate studies this term generally refers to either flowstone or stalagmite deposits. The most common for-mations which appear in the literature regarding climate proxy studies include:
• stalactites - which form on the roof of the cave chambers and extend towards the cave floor
• flowstones - which form on the walls and floors of the cave, deposited by the flow of water
• stalagmites - those formations deposited on the cave floor by water dripping (usually) from feeder soda straw stalactites
Of these three, it is stalagmites which are most frequently used for paleoclimate proxy studies as the layering of stalagmites is generally less complex than that of either stalac-tites or flowstones.
3.2. Proxy types
The waters entering the cave environment are the carriers of climate information. As caves generally present a stable environment (eg. reflecting average mean outside face air temperature), changes in the water entering the cave reflect changes at the sur-face. Seasonal variation can be reflected in the changing chemistry of the drip water, the rate of dripping, and in the rate of speleothem growth (Borsato et al., 2007; Baldini et al., 2008; Sherwin and Baldini, 2011; Frisia et al., 2012; Hartland et al., 2012). These in turn can be enhanced by changes in cave pCO2 which can be driven by increasing tempera-ture differences between the cave air and the outside air driving ventilation of the cave (Spötl et al., 2005).
3.2.1. Stable oxygen and carbon isotopes in speleothems
Stable isotopes (variations of chemical elements with different neutron numbers but the same number of protons which do not undergo radioactive decay) in climate systems provide a valuable variety of climate proxies. In many archives they are used to trace the source of dust and pollutants (Belli et al., 2013), the source of precipitation (Cruz Jr et al., 2006; Breitenbach et al., 2010), and amounts of precipitation (Fairchild and McMillan, 2007). The use of stable isotopes in speleothem studies has become more prevalent over the last 30 years as today only very small sample sizes are required for analysis, thus providing the potential for high-resolution climate studies.
In speleothems, the most commonly used stable isotopes are carbon (C) and oxygen (O).
Ratios of isotopes in carbonates are given as a δ parts per mille (‰) relative to the Vienna Pee Dee Belemnite (δ18O or δ13C ‰ V-PDB) standard, while water samples are
relative to the Vienna Standard Mean Ocean Water (δ18O ‰ V-SMOW).
The ratios are determined using the following formulas: δ18O ‰ V-PDB = (R
sample / Rstandard -1) × 1000 where R = 18O/16O
δ13C ‰ V-PDB = (R
sample / Rstandard -1) × 1000 where R = 13C/12C
The water from which speleothems precipitate ultimately reflects the δ18O of meteoric
waters and therefore can reflect the δ18O of precipitation (Rozanski et al., 1993; Lachniet,
2009).
Carbon isotopes are more ambiguous than oxygen isotopes and are less frequently in-terpreted in speleothems. Contributing sources of δ13C in speleothems are the sources of
CO2 dissolving in the ground water, which include soil zone and vegetation, atmosphere, and bedrock. Of these, the soil zone and vegetation account for around 80-90% of the δ13C
observed (Cosford et al., 2009).
In order to preserve these signals, speleothem growth (precipitation of calcite) should occur under conditions which do not introduce additional fractionations of these sta-ble isotopes (Hendy, 1971; McDermott, 2004; Fairchild et al., 2006; Lachniet, 2009). Such conditions are referred to as equilibrium or near-equilibrium conditions. Despite this being a textbook requirement for good proxy data, and possibly occurring in the cave environment, it has become apparent that such conditions are the exception rather than the rule (Kim and O’Neil, 1997; Mickler et al., 2004; Lachniet, 2009; Day and Henderson, 2011). Evaporative effects, open/closed bedrock dissolution, cave pCO2 levels which en-hance rapid degassing of CO2 from drip water, changing drip rate, prior calcite precipi-tation (PCP), and processes within the karst, epikarst and soil layers can all contribute
to fractionation before calcite precipitation on the speleothem surface, and can poten-tially complicate climate proxy interpretations (Bar-Matthews et al., 1996). To address this, monitoring studies at cave sites aim to improve the understanding of the processes which interplay during speleothem growth and to shed light on the relationship between the modern climate and cave records (Spötl et al., 2005; Mattey et al., 2008; Mattey et al., 2010; Baker et al., 2014).
Both δ18O and δ13C will vary in what they represent at each individual cave site, and
this needs to be tested in order to produce a good interpretation and climate proxy re-cord. For example, the δ18O signal of precipitation can reflect precipitation amount, the
so-called amount effect, where a more negative δ18O indicates increased precipitation
amount, or temperature, where a more negative δ18O indicates colder temperatures. In
very broad terms, tropical and subtropical coastal speleothems often record an amount affect signature (e.g. Bar-Matthews et al., 1997; Lachniet et al., 2004; Griffiths et al., 2010).
δ13C as an indicator of transitions between dominance of C
3 or C4 is made possible due
to the differences in the photosynthetic pathway used by the plants resulting in more negative (C3) or less negative (C4) δ13C values (Cosford et al., 2009). This mechanism
provided the potential to observe changes in vegetation type between C3 and C4 plants above caves (Lee-Thorp et al., 2001; Holmgren et al., 2003). In the case of sparse vegeta-tion cover, or low biological activity, speleothem δ13C values will be less negative. Less
negative δ13C values can also result from kinetic effects such as evaporation, PCP, or
rap-id degassing of CO2, and so during drier conditions the δ13C signal can be enhanced by a
combination of factors all working to pull the signal in the same direction (Fairchild and McMillan, 2007). δ13C has also been used as a proxy for relative humidity in regions with
high rainfall and steady temperature and vegetation assemblages (Göktürk et al., 2011).
3.2.2. Trace elements
In other climate archives, such as ice cores and trees, trace elements are used to provide measurements of atmospheric fluxes, movement and source of air masses, and pollut-ants. They also provide information on dust abundance (Gabrielli et al., 2005). Tree ring trace element records can provide highly resolved archives of volcanic eruptions (Pear-son et al., 2009) and pollutants (Balouet et al., 2007).
In addition to the well-established use of stable oxygen and carbon isotopes for envi-ronmental and climate proxies, speleothems can provide additional information in the form of the variations of trace elements (Hellstrom and McCulloch, 2000; Borsato et al., 2007; Fairchild and Treble, 2009). A number of methods are available for looking at trace element compositions of speleothems (Fairchild and Treble, 2009). In the work done for this thesis, both micro X-ray fluorescence spectrometry (µXRF)(Finné et al., 2015) and Laser Ablation Inductively Coupled Mass Spectrometry (LA-ICP-MS) were used (Jochum et al., 2012). While trace elements have not been as extensively studied as stable oxygen and carbon, they present an opportunity to study hydrological conditions, dust events, pollution, and volcanic activity at very high resolutions, but good interpretations are dependent on an increase in the number of modern-day cave monitoring studies to pro-vide a better understanding of the exact controls on their behavior. Trace elements are present in the speleothem in several different forms. Magnesium (M), strontium (Sr), and barium (Ba) are often transported in solution as divalent cations and substitute out Ca in the crystal lattice of calcite. These are the most studied trace elements, with their behavior in solution and mineral form dependent on temperature, precipitation rate and crystal morphology (Fairchild and Treble, 2009). Other elements are transported with the drip water and occur as very fine colloidal particles (1 nm to 1 µm) (Zhou et al., 2008).
Caves present a complex environment with many factors influencing the trace ele-ment signal in the speleothem with bedrock and soil as primary sources of both calcium and trace elements. The various elements may have more than one source, as ground cover, vegetation (Cosford et al., 2009), sea spray (Baldini et al., 2015), and atmospheric particles are all potential sources (Ayalon et al., 1999; Fairchild and Treble, 2009). Water interactions with the bedrock can vary spatially throughout the cave system, and as in-teraction time of groundwater with the bedrock, and maximal dissolution occurs when
pCO2 is highest, differences between fracture, seepage, and fissure flows can be revealed
in the speleothem (Fairchild et al., 2000). In arid conditions, the possibility for prior cal-cite precipitation (PCP) is increased, and this can be seen in the covariation of Mg and Sr (Verheyden et al., 2000). Instances of high flow and flushing of colloidal transported elements, such as those carried by small detrital particles, (Hartland et al., 2012) is an-other aspect to consider. In cases of high flow and flushing, it would be expected that elements which are associated with these small particles would occur in higher concen-trations directly in conjunction with a high infiltration occurrence (Fairchild and Treble, 2009). Phosphorus (P) (and phosphate) in speleothems has been related to environmental changes (Mason et al., 2007; Jones, 2009; Frisia et al., 2012). Mason et al. (2007) showed that P can be present in several forms within speleothems, incorporated into the calcite crystal in defects and also as phosphate inclusions.
3.2.3. Petrographic analysis and Scanning Electron Microscopy (SEM)
Speleothem petrography, i.e. the identification and classification of crystal morphologi-cal changes within a speleothem, provides a valuable tool for the understanding of geo-chemical proxy data (Frisia et al., 2000). The development of the microstratigraphic log (Frisia, 2015) presents researchers with the opportunity to characterize fabrics with a standard framework.
Most speleothems are composed of CaCO3, in the form of calcite or aragonite. Within these two mineral types, variations in crystal fabric indicate a number of important envi-ronmental controlling factors (Frisia et al., 2000). It is this information which can provide a valuable basis for interpretations of stable isotope and trace element data.
Calcite speleothems generally exhibit formation of crystals which are elongated with respect to the growth axis. The degree of crystal growth surface irregularities is gener-ally no more than 10 µm, and these tend to occur between crystals. Irregularities allow for the formation of spaces termed inclusions, which can contain air and/or water. Fluid-filled inclusions can indicate changes in drip water chemistry or cave microclimate, as they can occur in discretely laminated formations. They are also a source of additional paleoproxy information through the use of δ18O thermometry performed on the tiny
amounts of water trapped in these inclusions (McGarry et al., 2004). Laminations or banding can occur on seasonal, annual, or supraannual scales, particularly in regions with strong seasonal variations. Lamina can result from an influx of humic and fluvic ac-ids from the soil zone, generally occurring in the early winter months in boreal climates, from alterations between calcite and aragonite (termed couplets), from changes in cave
pCO2 on a seasonal scale, and from changes in trace element composition of the drip
waters (McMillan et al., 2005; Baker et al., 2008). However, not all lamina are seasonal or annual, and so great care must be used in making such interpretations (Shen et al., 2013).
Calcite fabrics in speleothems have been studied and parallels between fabric and formation environments have been established (Frisia et al., 2000; Frisia, 2015). Initially, petrographic analysis may be used to ensure no post-depositional alteration has oc-curred in the speleothem, as this can result in element movement within the sample, and
erroneous and misleading data. Changes in fabric can indicate the presence of lamina, changes in cave microclimate, periods of increased ventilation, and influxes of detrital particles, stops in growth or hiatuses, corrosion events caused by acidic or undersatu-rated drip waters, and so on.
Beyond the standard petrographic microscope, scanning electron microscopy (SEM) imaging is used to examine changes in surface morphology, elemental variation as well as microbe and soot particle presence (Cañveras et al., 2001; Jones, 2001; Jeong et al., 2003; Jones, 2009) within speleothems.
3.2.4. U-Th dating method
To provide a useful paleoclimate record, a good chronology is paramount. The most com-mon method for dating speleothems is uranium-thorium disequilibrium (U-Th) dating. This method provides an absolute date based on radioactive decay of uranium which has been incorporated into the crystal lattice of the speleothem calcite at the time the crystal formed.
By determining the relative abundance of U and Th isotopes it is possible to use the ratio of the evolved nuclides to determine the age since the calcite was deposited. The uranium which occurs in speleothems is dissolved by percolating waters from within the bedrock, and so the amount of uranium in the host rock (and also the residence of the water within the bedrock) will be reflected in the uranium concentration in the speleo-them. This method of dating is made possible by the different behavior of U and Th. Th is insoluble, and is primarily transported while adsorbed onto particles, such as dirt or clay (this is often called “detrital thorium” as it is not a daughter product of the radioactive decay of uranium within the calcite crystal). By contrast, U is highly soluble, allowing it to be transported in solution in the drip water and then incorporated into the calcite crystal lattice. Once the crystal has formed, the U within the lattice begins decaying into Th and the ratio of U to Th tells us how long ago the calcite formed. In the absence of detrital thorium, extremely precise dating is possible, and using U-Th it is possible to date samples up to around 600 ka with precision of between 0.1-1% for samples with low levels of detrital contamination (Fairchild and Baker, 2012).
When selecting the area of the speleothem to take samples from for dating, it is im-portant to consider that material from the smaller area will produce a date which has less uncertainty, and that sampling should occur along visible discreet lamina (layers). Fur-ther, the cleaner the sample (i.e. less detrital thorium) the smaller the uncertainty will be. When selecting places to sample from, it is standard to initially take a top, middle, and bottom date to provide an idea of the time over which the speleothem was growing. These samples should, if at all possible, be placed in layers or sections of the stalagmite which are without visible detrital (dirt) contamination, and which are dense and without voids. Once initial results have been obtained, samples can ideally be placed above and below any visible or suspected hiatuses (growth stops, often indicated by a change in color, texture, or change in crystal shapes). Ideally these initial dates will allow for the fine tuning of sample sizes to optimize both the preservation of the intact speleothem as much as possible, and also to improve efficiency in the lab portion of the analysis.
With solid samples, taken using a wire or band saw, the sample may be first cleaned using an ultrasound bath to remove any surface detritus which can either have come from the cave environment or from handling the sample after collection. This type of pre-cleaning is only possible on solid samples, and so in cases where controlled sampling environments, such as laminar flow hoods, are not available, solid samples are preferable.
4. The samples
4.1. Gasparee Cave
Stalagmite GC2 (Fig. 5, page 19) from Gasparee Cave, Gaspar Grande Island, Trinidad and Tobago was collected in two sections from a tidal saltwater lagoon in 2008 and 2010 (Fig. 6). The speleothem was not actively growing at the time of collection, though spe-leothem precipitation was active within the cave at other locations. GC2 measures 293 mm and is composed of white to brownish calcite, with evidence of corrosion on the outer surface of the speleothem. The sample was cut in half down the growth axis and two central slabs were made. One half of each was used for stable isotope, U-Th dating, and petrographic thin sections.
4.2. Kapsia Cave
For an initial trace elements study, speleothems from Kapsia Cave were collected (Finné, 2014). Samples from Kapsia are visibly laminated, with porous whitish calcite. The sam-ple from the lowest position in the cave, GK0901 contains a large number of clayey ho-rizons between sections of white porous laminated calcite (Fig. 16, page 37). GK0901 was sawn in half down the growth axis, and two central slabs around 1 cm thick were made. This was used for µXRF analysis, while the facing central slab was used to make petrographic thin sections.
4.3. Alepotrypa Cave
Five speleothems covering the period of the Mid-Holocene were analyzed from Alepo-trypa Cave (Fig. 4). At the time of collection none of these speleothems were situated under active drip sites. A1 (Fig. 7) and A2 (Fig. 8) were collected in May 2013. EH1 (Fig. 9) (Collected by P. Karkanas), A6 (Fig. 10), and A7 (Fig. 11) were collected in July 2014. Sample collection sites (Fig. 12) are situated at differing proximities to intense activity in the cave during the Neolithic habitation period. Color images of Fig. 7-11 are found on pages 20-24.
A1 is composed of compact whitish to honey-colored calcite, with a black band around 121 mm distance from top (dft) which is interpreted as a hiatus. A2 is composed of simi-larly colored calcite but contains no visible hiatus features. EH1 has a striking variation in color and texture, from compact brown, to white and clear, and to very porous black. It contains two black layers, the lower of which we interpret as strong indications of human activity, and the upper of which is interpreted as relating to a hiatus occurring around the time of the collapse of the cave. A6 and A7 each contain two compact black
colored layers of calcite with clear to honey-colored calcite between, though in A6 this clear middle layer is much thicker. Above the upper black layer, which is interpreted as occurring simultaneously in all three specimens, compact clear/whitish to honey colored calcite is seen. EH1 is topped by a thin layer of white
calcite, which dates to modern times.
All specimens were sawn in half and two central thick sections made, around 1 cm thick.
Modern speleothems, one from a staircase (MA1) and one next to A2 (MA2) were also collected, as was a large Pleistocene age stalagmite (T1). Detailed de-scriptions and results from these specimens are not presented.
Figure 4. Time periods covered by stalagmites from Alepotrypa cave.
A1 A2 EH1 A6 A7 MA1 MA2
0 10 00 20 00 30 00 40 00 50 00 60 00 70 00 80 00 90 00 10 00 0 11 00 0 12 00 0 13 00 0 14 00 0 15 00 0 16 00 0 17 00 0 18 00 0 19 00 0 20 00 0 21 00 0 Years before 2015
Figure 5. Scan of thick section central slab from Gasparee Cave GC2 stalagmite.
2 cm
Figure 6. Photograph of Gasparee Cave from entry staircase. Photograph shows light coming in the collapsed hole in the roof above the seawater lagoon within the cave. Brown color above the water surface shows the range of the tide within the cave. Photo: Fredrik Ljung
2 cm
Figure 7. Scan of stalagmite A1 indicating dating sample sites (yellow ovals), sample track for mi-cromilling and LA-ICP-MS sampling. Dotted line indicates habitation period, rectangles indicate areas shown in petrographic thin section in Paper V. Black arrows indicate the layer of darker material which marks a growth hiatus.
Figure 8. Scan of stalagmite A2 indicating dating sample sites (yellow ovals) and isotope sampling track.
2 cm
Figure 9. Scan of stalagmite EH1 indicating dating sample sites (yellow ovals), micromilling track, and LA-IPC-MS analysis track overlap. Black arrows indicate the black layers mentioned in the text. Note the very porous black middle layer, indicated by the yellow line. This section is topped by clean white calcite, and thin black layer.
Figure 10. Scan of A6 stalagmite, indicating dating sample sites. Micomilling and LA-ICP-MS track overlap and are indicated by the black line. The yellow rectangle indicates the sampling site for SEM material, and the black arrows indicate the black layers discussed in the text.
Figure 11. Scan of stalagmite A7 with yellow oval indicating dating sample site. Black line indi-cates stable isotope sampling track, and yellow rectangle indiindi-cates area used for SEM mapping. Black arrows indicate black layers mentioned in the text.
Figure 12. Map of the front chambers of Alepotrypa Cave showing collection sites. Chamber G (Γ) and D (Δ) marked in Greek alphabet. Samples T1, A3, MA1 and MA2 have not been fully analyzed and are not presented here. (Map by R. Seifried).
5. Analytical Methods
In order to produce a good paleoclimate proxy record from speleothems a combination of techniques will produce the best results. Most commonly used in speleothem studies are U-Th dating and stable isotope studies, but trace elements and petrographic analysis are today seen as important signposts for robust interpretations. By combining absolute dating and high-resolution paleoproxy studies it is possible to achieve a high-quality climate archive from speleothems spanning well past the Holocene period, if growth conditions are favorable.
5.1. Stable isotopes
Stalagmites from Gasparee Cave and Alepotrypa Cave (A1, A2, A6, A7, EH1) were sam-pled for 1 mm low-resolution stable isotope analysis using a Dremel hand drill with a 0.5 mm diamond-coated dental drill bit. Based on these initial results, micromill sampling at resolutions between 0.01 mm and 0.25 mm was conducted at University of Innsbruck, Austria.
Samples of modern calcite from Alepotrypa cave were collected by placing glass slides and a granite pebble in the cave and removing them at the next cave visit. From each collection medium a small amount of calcite was taken using a scalpel.
Stable isotope concentrations were measured on a Thermo Fisher Finnigan Delta-plusXL isotope ratio mass spectrometer equipped with an automated carbonate prepara-tion system (Gasbench II) at the University of Innsbruck, Austria. Values are reported in δ notation relative to the V-PDB standard, with a precision better than 0.08‰ and 0.06‰ for δ18O and δ13C, respectively (Spötl and Vennemann, 2003).
5.2. U-Th Dating
5.2.1. Gasparee Cave
For the Gasparee Cave stalagmite, initial dates at the top, middle, and bottom of the sam-ple were produced by the thermo-ionization mass spectrometry method (TIMS). These provided a baseline for the size and placement of subsequent TIMS dating samples, and later for the placement of more precise MC-ICP-MS (multi-collector inductively coupled mass spectrometry) dating samples.
400 mg and 250 mg samples taken from solid 8 mm cores were analyzed at the Uni-versity of Heidelberg, with preparation chemistry following the procedures in Frank et al. (2000) and analyzed using a Finnigan MAT 262 RPQ mass spectrometer.
Additional dates were made on 200 mg samples drilled using a diamond-tipped dental drill, and taken along discreet layers. Preparation chemistry was done following
Hoff-mann et al. (2007). Measurements of the purified fractions, also following procedures outlined in Hoffmann et al. (2007), were carried out at the National Research Centre for Human Evolution (CENIAH) in Burgos, Spain on a Neptune Thermo Finnigan multi-collector inductively coupled mass spectrometer (MC-ICP-MS). Detrital correction was done assuming a typical upper crustal composition with a 238U/232Th activity ratio for the
detritus of 0.8 ± 0.4 (Wedepohl, 1995) and the 238U decay chain in secular equilibrium.
Ages were calculated using the half-lives reported in Cheng et al. (2000).
5.2.2. Alepotrypa Cave
For the Alepotrypa Cave speleothems, sample sizes ranged between 50 mg and 200 mg. Samples were taken using a Dremel hand drill fitted with a 0.9 mm diamond tipped den-tal drill bit.
Initial dates from speleothem A1 (Fig. 7) were taken in 2013. These were placed in the top, middle (above hiatus) and bottom of the sample. In 2015, 12 additional dates were taken.
Speleothem A2 (Fig. 8) was dated at the top, middle, and bottom of the stalagmite. Ad-ditional dating samples were not prioritized at this stage, as high-resolution dating and analysis of all samples was not feasible.
Speleothem A6 (Fig. 10) was sampled at 6 sites, including one sample in the dark layer associated with human influence on the cave environment. Sample sizes were taken based on results from A1, with 150 mg and 100 mg samples.
Speleothem EH1 (Fig. 9) was sampled in three sites, and speleothem A7 (Fig. 11) was sampled at its top to provide a youngest age.
Table 1. List of materials used in measuring conditions in Alepotrypa Cave and their accuracy, as stated by the manufacturers.
Measured
parameter Equipment Accuracy Remark
Drip rate Digital stopwatch ± 1s
Relative humidity
of cave air Vaisala HM70 ± 1.7% HMP75 probe
Cave air temperature –
discrete Vaisala HM70 ± 0.2°C HMP75 probe
Cave air CO2 Vaisala GM70 ± 1.5% of range and ± 2% of
reading GMP222 probe
a)
b)
Figure 13. Photos of large human bone (a) D968 and animal horn (b) Z595 samples within the calcite crust (Photos: P. Karkanas). Inset are images of the smaller crust samples taken from these larger sections (Photos: D. Hoffmann). Note placement and size of cuts where final samples were extracted.
Three samples were taken from the widespread calcite crust which topped the materi-als of the Neolithic habitation sediment (Fig. 13). These samples were taken from calcite which grew on human bone (Fig. 13a) and a piece of animal horn (Fig. 13b). Samples were cut using a small band saw under a laminar flow hood. As these samples were solid slices and not powder, they were cleaned using an ultrasound bath prior to being dis-solved for preparation chemistry.
Preparation chemistry for all samples was done following Hoffmann et al. (2007). In brief, samples were dissolved in 7M HNO3, spiked, and organics were removed through the addition of H2O2 and (where necessary) centrifuging. Fractions of U and Th were separated using resins and purified. Measurements of the purified fractions following procedures outlined in Hoffmann et al. (2007) were carried out at the National Research Centre for Human Evolution (CENIAH) in Burgos, Spain and at the Max Planck Institute for Human Evolution in Leipzig, Germany, both on a Thermo Finnigan Neptune MC-ICP-MS. Detrital correction was done assuming a typical upper crustal composition with a 238U/232Th activity ratio for the detritus of 0.8 ± 0.4 (Wedepohl, 1995) and the 238U decay
chain in secular equilibrium. Ages were calculated using the half-lives reported in Cheng et al. (2000).
5.3. Petrographic and SEM images
Stalagmite A1 was analyzed in 16 20 µm thick petrographic slides on a standard Leica petrographic microscope at the University of Newcastle, Australia. Images were taken to identify hiatuses which would have implications for the age models.
Additional analysis was done on material drilled from the black layers of stalagmites A6 and EH1, as well as thick section slices of A1 and A7, and a solid chip representing the bottom black layer of A6.
Material drilled from the black layers of A6 and A7 was dissolved in 6% HCl and pre-pared following procedures in Dredge (2014), and placed onto aluminum stubs for analy-sis. Additional material was scraped from the black layers of A6 and EH1 and placed, untreated, onto carbon tape pins for analysis.
SEM images showing contents of the black layers contained in stalagmites EH1, A6, and A7 were taken using a FY Quanta FEG650 Field Emission Gun and EDS Detector X-Maxs at 80 nm2 in a 0.9 millibar chamber using a large field detector at the Department
of Geological Sciences, Stockholm University.
5.4. Modern-day cave environment
During field visits measurements were taken of air pressure, cave air CO2, temperature, and relative humidity. Drip rates at three sites in Alepotrypa Cave were measured using a stopwatch.
5.4.1. Water δ18O and δD
Drip water samples were collected during field visits to Alepotrypa Cave and were analyzed for δ13C and δ13C using a Laser Water Isotope Analyzer at the Department of
Geological Sciences, Stockholm University. Reproducibility was calculated to be better than 0.6‰ for δD and 0.15‰ for δ18O.
5.5. Trace elements
5.5.1. XRF
The novel method of applying XRF core scanning was applied to a speleothem from Kap-sia Cave (Paper II). A polished thick section of speleothem GK0901 was scanned along the growth axis using the ITRAX µ-XRF core scanner at the Department of Geological Sciences at Stockholm University. Scans were conducted using a Molybdenum tube set to 30 kV and 30 mA with a step size of 200 µm and an exposure time of 40 s.
5.5.2. LA-ICP-MS
This microanalysis technique was applied to samples A6 and A7, and EH1, using a high-resolution sector field ICP-MS Thermo Element2 combined with a UP213 (213 nm, Nd:YAG) laser ablation system at the Max Plank Institute for Chemistry (MPI), Mainz. Synthetic silicate glasses NIST SRM 610, NIST SRM 612, and carbonate reference mate-rial MACS3 were used as reference matemate-rials. Measurements of 40 elements were made at low and medium mass resolution modes. Five passes of each reference material were made. A pre-ablation pass at 80 µm/second was made along the stalagmite sampling axis at 80% laser output with a spot size of 110 µm prior to ablation for data collection. Abla-tion was conducted in a helium (He) atmosphere, with an argon (Ar) carrier gas flow up line of the plasma torch. Data reduction from counts per second (cps) to parts per million (ppm) was done by calculating the cps intensity relative to the internal standard isotope of 43Ca.
6. Results
The major results from this thesis are presented as an overview of the sites combined with a summary of papers, followed by a discussion and summation of the most im-portant findings of the project. Dating results for all stalagmites and calcite crust from Alepotrypa Cave are presented in supplementary tables 1, 2, 3 and 4, and age models constructed using StalAge (Scholz and Hoffmann, 2011) for stalagmites GC2, A1, and A6, and supplementary figures 1-3. Modern-day cave environment and drip water data are found in supplementary table 5.
6.1. Constraining the timing of a rapid climate change
6.1.1. Paper I: Early Holocene patterns of rainfall, vegetation and soil conditions, inferred from a southern Caribbean stalagmite
Paper I contains the results of the analysis carried out on the GC2 stalagmite, collected from Gasparee Cave, Trinidad. Stable O and C isotope results indicate that Trinidad, with its position in relation to ITCZ movement, is a good site for future studies to constrain the timing and extent of ITCZ related to changes in precipitation amount at the site. δ18O
in precipitationis dominated by the amount effect, and wet/dry periods in the GC2 re-cord agree to some extent with those in the Cariaco Basin rere-cord (Haug et al., 2001) and in Cuban speleothems (Fensterer et al., 2013) (Fig. 14). By applying spectral analysis to the speleothem stable isotope series it was seen that the site shows a strong periodicity at 1000-1800 years, similar to that seen in many other Holocene climate archives (e.g. Bond et al., 2001; Wanner et al., 2011). Using rampfit analysis (Mudelsee, 2000; Mudelsee, 2010) the onset, duration, and recovery time from the dry conditions was constrained at 8.2 ka to a period of 170 years, occurring between 8.44 – 8.27 ka (±39 a) thousand years before 2000 (Fig. 15) and reflecting the same double peak structure observed in other archives (Alley and Ágústsdóttir, 2005).
-41 -39 -37 -35 7500 8000 8500 9000 9500 10000 10500 11000 11500 12000 -33 a) GRIP δ 18 O ‰ VSMOW -4 -2 0 c) Gasparee Cave δ18 O ‰ VPDB -8 -4 0 4 d) Gasparee Cave δ 13C ‰ VPDB d) e) f) 160 170 180 190 200 e) Cariaco Grey scale
0.1 0.2 0.3 0.4 7500 8000 8500 9000 9500 10000 10500 11000 11500 12000 f) Cariaco Ti % Years before 2000 -2.5 -1.5 -0.5
b) Dos Anas Cave δ
18O ‰ VPDB
-3.5
a)
b)
c)
Figure 14. Comparison of Caribbean and Greenland archives. (a) GRIP δ18O record (Rasmussen
et al., 2006; Vinther et al., 2006), (b) Dos Anas Cave, Cuba, stalagmite δ18O rainfall proxy record
(Fensterer et al., 2013), (c) Gasparee Cave, Trinidad stalagmite δ18O rainfall proxy record, (d)
Gas-paree Cave, Trinidad stalagmite δ13C rainfall proxy record (e) Cariaco Basin grey scale upwelling
(wind strength) record (Hughen et al., 2000) (f) Cariaco Basin runoff record (Haug et al., 2001). Note inverse isotope axes for Gasparee Cave and Dos Anas Cave records. In these records more negative values indicate wetter conditions. Grey boxes indicate timing of the 8.2 ka event, with dating uncertainties (1σ) around 8.0 ka after Rohling and Pälike, (2005) for (a) and (f) and after the original authors for (b) and (c).
