UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Hologene evolution of the changuinola peat deposit, Panama: sedimentology of a marine-influenced tropical… Phillips, Stephen 1995

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1995-98340x.pdf [ 12.2MB ]
JSON: 831-1.0052918.json
JSON-LD: 831-1.0052918-ld.json
RDF/XML (Pretty): 831-1.0052918-rdf.xml
RDF/JSON: 831-1.0052918-rdf.json
Turtle: 831-1.0052918-turtle.txt
N-Triples: 831-1.0052918-rdf-ntriples.txt
Original Record: 831-1.0052918-source.json
Full Text

Full Text

HOLOCENE EVOLUTION OF THE CHANGUINOLA PEAT DEPOSIT, PANAMA:SEDIMENTOLOGY OF A MARINE-INFLUENCED TROPICAL PEAT DEPOSIT ON ATECTONICALLY ACTIVE COASTbyStephen PhillipsB .A., University of Waterloo, 1971B.Sc., University of Victoria, 1990A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FORTHE DEGREE OF DOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDEPARTMENT OF GEOLOGICAL SCIENCESWe accept this thesis as conforming to the required standardTHE UNIVERSITY OF BRITISH COLUMBIAMarch, 1995© Stephen Phillips, 1995In presenting this thesis in partial fulfillment of therequirements for an advanced degree at the University of BritishColumbia, I agree that the Library shall make it freely availablefor reference and study. I further agree that permission forextensive copying of this thesis for scholarly purposes may begranted by the head of my department or by his or herrepresentatives. It is understood that copying or publication ofthis thesis for financial gain shall not be allowed without mywritten permission.(Signature)_____________________Department of (E&-çfc4-LThe University of British ColumbiaVancouver, CanadaDate /3///9s_:ABSTRACTThe evolution and structure of a large peat deposit on the Caribbean coast of western Panama,Central America is evaluated as a possible analogue for the deposition of low-ash, low sulphur coals.Effects of earthquake-driven subsidence events on the peat and the peat-forming vegetation areinvestigated, and implications of tectomc subsidence on the evolution of this deposit and on thecurrently developing model of coastal tropical coal deposition are described.The deposit is approximately 80 km2 in extent, averages 6.5 m in thickness, and occupies thewidth of the narrow coastal plain between the Talamanca Cordillera and barrier beach on a seismicallyactive part of the Caribbean coast. Based on vegetation zonation , topography and hydrology, themodern Changuinola mire complex can be divided into a raised, concentrically zoned, ombrotrophicwestern section, and a dissected and partially rheotrophic eastern section. Differing hydrologicalregimes of these two sections are reflected in the physical and chemical stratigraphy of the peat. In thewestern section, a vertical succession of peat types, highly humified at the base and margins, and morefibric in the upper central part, is the result of internal hydrological boundaries, created by density andpermeability variations in the peat. The mire is insulated from marine and fluvial influences bytopography and hydrology, and displays no evidence of fluctuating sea level. Coal formed in such anenvironment would be low in sulphur and ash, dull and massive at the base and margins, and finelybanded in the upper central part. The eastern section of the mire is in part rheotrophic, with a complexmosaic of vegetation types, and is segmented into distinct drainage areas by tidal blackwater creekchannels. Effects of this marine influence are localized to the bay and channel margins. Coals formed11in this environment would have large variations in sulphur over distances of a few metres laterally, anda few centimetres vertically.Earthquake-driven coastal subsidence is greatest in the southeast, and has lead to drowning ofthe deposit. Subsidence events raise the level in the blackwater creeks, moving the front of marineinfluence to the northwest (inland), and leading to the replacement of freshwater vegetation withmangroves. The degree ofpenetration of marine waters remains restricted, however, to marginal peats.An increase in the scale of subsidence events may overcome the response capability of the mangrovesand lead to disruption of internal hydrological boundaries and ultimate deflation and drowning of themire.111TABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTS ivLIST OF FIGURESLIST OF PLATES xiLIST OF TABLES xiiDEDICATION xiiiACKNOWLEDGEMENTS xivFOREWORD xvCHAPTER 1: iNTRODUCTION 11.1 OBJECTiVES 21.2 PRESENTATION 61.3 REFERENCES CITED 8CHAPTER 2: SEDIMENTOLOGY OF THE CHANGUINOLA PEAT DEPOSIT: 9ORGANIC AND CLASTIC SEDIMENTARY RESPONSE TO PUNCTUATEDCOASTAL SUBSIDENCE2.1 ABSTRACT 102.2 INTRODUCTION 122.3 METHODS 14a) Measures ofEarthquake-induced Subsidence 14b) Levelling Surveys 17c) Remote Sensing 17d) Hydrology 19e) Clastic Sediments 19J) Vegetation Survey 19g) Peat Sampling 21h) Peat Characterization 212.4 GEOLOGICAL SETTING 262.5 GEOMORPHOLOGY 282.6 VEGETATION 302.7 AGE AND GEOMETRY OF THE DEPOSIT 31iv2.8 SEA LEVEL CHANGE AND PUNCTUATED SUBSIDENCE 34a) Subsidence in 1991 35b) Long Term Sea Level Change 382.9 ORGANIC M”D) CLASTIC SEDIMENTOLOGY 402.9.1 CLASTIC SEDIMENTS 40a) Changuinola River Floodplain 40b) Barrier System 43i) Sedimentary Structures 43ii) Grain Size andMineralogy 43c) Peat Basal Sediments 45d) Almirante Bay 452.9.2 ORGANIC SEDIMENTATION 48a) Peat Characterization 48b) Geochemical Parameters 50c) Degree ofHum/Ication, Ash and Moisture Content 51i) Western Section 54ii) Eastern Section 562.10 DISCUSSION 592.10.1 CLASTIC SEDIMENTARY RESPONSE TO RECENT TECTONISM 59a) Changuinola River Floodplain 59b) Barrier Beach 612.10.2 HYDROLOGICAL CONTROLS ON PEAT ACCUMULATION 62a) Western Section 63b) Eastern Section 672.10.3 IMPLICATIONS FOR COAL GEOLOGY 71a) Transgressive and Regressive Signatures and the Relation ofPeat Deposition to 71Active Clastic Sedimentation2.11 CONCLUSIONS 742.12 REFERENCES CITED 77CHAPTER 3: VEGETATION ZONES AND DIAGNOSTIC POLLEN PROFILES OF A 82COASTAL PEAT SWAMP, BOCAS DEL TORO, PANAMA3.1 ABSTRACT 833.2 iNTRODUCTION 853.3 GEOGRAPHY AND CLIMATE 883.4 METHODS 90a) Problems 90b) Satellite Imagery 92c) Levelling Surveys 92d) Coring ofPeat and Collection ofSurface Samples 92e) Collection and Identification ofPlants 95fi Peat Classfication, Pollen Preparation and Identification 95g) Pollen Counts and Pollen Diagrams 963.5 RESULTS AND INTERPRETATION 97a) Vegetation Distribution 971) Mappable Vegetation Zones 98ii) Description ofthe Phasic Communities 993.6 POLLEN DISTRIBUTION 110Va) Surface Pollen Diagrams 110b) Pollen From Cores 1163.7 DISCUSSION 1213.8 SUMMARY AND CONCLUSIONS 1263.9 REFERENCES CITED 129CHAPTER 4: SULPHUR IN THE CHANGUINOLA PEAT DEPOSIT AS AN 133INDICATOR OF THE ENVIRONMENTS OF DEPOSITION OF PEAT ANDCOAL4.1 ABSTRACT 1344.2 iNTRODUCTION 1354.3 REGIONAL SETTING 137a) Geography, Climate and Biology 137b) Geomorphology 139c) Geological Setting 1404.4 SAMPLiNG AND EXPERIMENTAL 1404.5 RESULTS 142a) Geochemical Parameters 1421) Sulphur and Salinuy 142ii) Sulphur andp1-I 145b) Total Sulphur Distribution 148I) Sulphur andpH in the Western Section of the Deposit 148ii) Sulphur, Salinity andpH in the Eastern Section of the Deposit 152c) Forms ofSulphur 1541) B13: An Example ofa Group III Peat 154ii) BDD 23: An Example ofGroup land Group IlPeats 1594.6 DISCUSSION 162a) Sulphur, pH and Marine Influence 162b) Sulphur and Vegetation 163c) Sulphur and Climatic Influence 166d) Sulphur and Tectonic Influence 1674.7 IMPLICATIONS FOR ENVIRONMENTAL STUDIES OF COAL 1684.8 CONCLUSIONS 1704.9 REFERENCES CITED 172CHAPTER 5: EARTHQUAKE-INDUCED FLOODING OF A TROPICAL COASTAL 176PEAT SWAMP: A MODERN ANALOGUE FOR HIGH SULPHUR COALS5.1 ABSTRACT 1775.2 INTRODUCTION 1785.3 SAMPLING AND EXPERIMENTAL PROCEDURES 1805.4 RESULTS 180a) Total Sulphur, Salinity andp1-f 180b) Forms ofSulphur 183z) Organic Forms 183ii) Inorganic Forms 1835.5 DISCUSSION 1855.6 CONCLUSIONS 188vi5.7 REFERENCES CITED 190CHAPTER 6: CONCLUSIONS 1916.1 CONCLUSIONS 1916.2 SUGGESTIONS FOR FURTHER RESEARCH 193APPENDICES 195APPENDIX A SAMPLE SITE LOCATIONS AND ANALYTICAL KEY 195APPENDIX B SAMPLE DESCRIPTIONS 198APPENDIX C RESULTS OF WET SIEViNG 216APPENDIX D POLLEN COUNTS FROM CORES 222APPENDIX E POLLEN DESCRIPTIONS AND PLATES 225APPENDIX F RESULTS OF VEGETATION SURVEY 233APPENDIX G ANALYTICAL PROCEDURES AND DEFiNITIONS 235APPENDIX H REMOTE SENSING 243APPENDIX I LEVELLING SURVEYS 244viiLIST OF FIGURESFigure 1.1 Regional map of southern Central America showing major tectonic elements. 3Figure 1.2 Landsat satellite image of western Panama and eastern Costa Rica. 4Figure 2.1 Large map shows general geology of part of the Limon - Bocas del Toro Basin, 15and bathymetry of the continental shelf off the study area in northwestern Panama.Insert shows the location of the study area.Figure 2.2 Areas of observed and modelled uplift and subsidence along the Caribbean coast 16of Panama and Costa Rica as a result of the April 22, 1991 Ms=7.5 earthquake.Figure 2.3 Colour infrared air photo of part of the subsided margin of the peat deposit along 18the shore of Almirañte Bay.Figure 2.4. Map of the study area showing sample sites referred to in the text. 20Figure 2.5. Cross sections and high and low discharge rates in the three largest brackish 22blackwater creeks.Figure 2.6. The western half of the barrier beach, and cross sections of the barrier at the 23beach end of the surveyed transect.Figure 2.7. False colour SPOT satellite image of the area of the detail map marked in Figure 242.1, including the Changuinola peat deposit.Figure 2.8. False colour SPOT satellite image of the narrow coastal plain and barrier 24shoreline.Figure 2.9. Upper panel. SW-NE cross section of the peat deposit along survey line. Lower 33panel: Results of particle size analysis (wet sieving) of 8 cores across the westernpart of the deposit.Figure 2.10. Photograph of drowned tree on the shore side of the beach berm at site Beach 2. 36Figure 2.11. Photograph of laminated silts, sands, peats and cm-thick leaf beds of the alluvial 41plain.Figure 2.12. NW-SE cross section from the Changuinola River to Almirante Bay. 42Figure 2.13. Photograph of bedding structures in a trench across the beach berm at Beach 2. 44Figure 2.14. Comparison of the results of wet sieving (particle size distribution) with 55palynological analysis of core Ed 3.viiiFigure 2.15. A generalized view of wet sieving results across the marine margin of Almirante 57Bay.Figure 2.16. A comparison of the wet sieving results for core BDD 23 with palynological 58analysis.Figure 2.17. Proposed model of peat development on the barrier coastline as a result of 69periodic punctuated subsidence.Figure 3.1 Location map of the Changuinola peat deposit showing sample sites used in 89vegetation and pollen analyses.Figure 3.2 NE-SW Cross-section of the deposit along levelling transect, showing surface 91vegetation zoned into phasic communities (PC’s).Figure 3.3 False-colour SPOT multispectral image of the mire showing concentrically zoned 93vegetation.Figure 3.4 Interpretive map of the area shown in the satellite image (Fig. 3.3). 94Figure 3.5 Lateral sequence of phasic communities from mangrove fringe to bog plain. 100Figure 3.6-a Photograph of vegetation in Phasic Community 2: Back-mangrove forest. 101Figure 3.6-b Photograph of vegetation in Phasic Community 5: Campnosperma forest 101Figure 3.6-c Photograph of vegetation in Phasic Community 3: Raphia palm swamp 104Figure 3 .6-d Photograph of vegetation in Phasic Community 6: sawgrass-stunted forest 106Figure 3.6-e Photograph of vegetation in Phasic Community 7: Myrica-Cyrilla bog plain. 109Figure 3.7-a Pollen fingerprint, from surface peat, of Phasic Community 1. 111Figure 3.7-b Pollen fingerprint, from surface peat, of Phasic Community 2. 111Figure 3.7-c Pollen fingerprint, from surface peat, of Phasic Community 3. 113Figure 3.7-d Pollen fingerprint, from surface peat, of Phasic Community 4. 113Figure 3.7-e Pollen fingerprint, from surface peat, of Phasic Community 5. 115.Figure 3.7-f Pollen fingerprint, from surface peat, of Phasic Community 7. 115Figure 3.8 Pollen diagram of core ED-3. 118Figure 3.9 Pollen diagram of core BDD-23. 120Figure 3.10 NW-SE Cross-section of the Changuinola peat deposit as interpreted from pollen 122in cores and surface vegetation zones.ixFigure 4.1 Site map: Site name is followed by the mean total sulphur content in brackets, 136with the standard deviation in italics.Figure 4.2 Scatter plot of sulphur (logarithmic scale) vs pH. n = 216. 143Figure 4.3 Scatter plot of sulphur (logarithmic scale) vs depth, western section of deposit.n = 144244.Figure 4.4 Scatter plot of sulphur (logarithmic scale) vs depth, eastern section of deposit.n =144.Figure 4.5 Scatter plot of total sulphur (logarithmic scale) vs salinity; n = 213. Groups I, II 146and II are sulphur - salinity groups defined in the text.Figure 4.6 Areal and depth distribution of sulphur Groups I, II and III in the Changuinola 147peat deposit.Figure 4.7-1 pH vs depth in the western section of the deposit; n=32 1. 149Figure 4.7-2 pH vs depth in the eastern section of the deposit; n=3 87. 149Figure 4.8 Variation in particle size (humification), pH and sulphur concentration in an 870 151cm core from the central part of the deposit (MILE 5).Figure 4.9 Variation in particle size (humification), pH and sulphur concentration in a forest- 151swamp core from the margin (LAKE 2).Figure 4.10 Total sulphur and pH vs Depth for a Group II channel-margin peat, core CS 3. 153Figure 4.11 Geochemical characteristics of a coastal mangrove peat core (B13). 156Figure 4.12 Wt % salinity vs Depth for 5 mangrove cores. 158Figure 4.13 Geochemical characteristics of core BDD23. 160Figure 5.1 Site of study on Caribbean coast of Panama. Block diagram illustrates extent of 179marine transgression following 1991 earthquake-induced subsidence.Figure 5.2 Spatial variation in total sulphur and salinity of peat across marine margin, Insert 184columns depict variation in sulphur fractions for eight samples, based on data fromTable 5.2.xLIST OF PLATESPLATE 1 Fern spores: 1) Tnchomanus crinitum; 2) Cyclopeltis semicordata Sw.; 3) 227Salpichlaena sp.; 4) Cyathea multzflora(?); 5) Lindsaea sp.; 6) Monolete type 1;7) Monolete type 2 8) Monolete type 3 9) Monolete type 8 10) cf. Metaxya sp.Palmae: 11) Raphia taedigera; 12) Raphia phytolith; 13) Euterpe precatoria.PLATE 2 Angiosperm Pollen: 1) Campnosperma panamensis; 2) Campnosperma-type; 3) 230Tricolporate type 1; 4) Tricolporate type 2; 5) Tricolporate type 26; 6)Tricolporate type 54; 7) Myrica mexicana; 8) Myrica type A; 9) Triporate type33; 10) Periporate type 39; 11) Triporate type 40; 12) Triporate type 48.PLATE 3 (10 pm scale bar): 1) Campnosperma panamensis low and high focus in both 232equatorial and polar views; 2) Detail of exine decoration of Campnosperma (x1780); 3) Periporate type 39, low and high focus;; 20 jim scale bar: 4)Fusiformisporites sp. A 5) Fusformisporites sp. CxiLIST OF TABLESTable 2.1 Radiocarbon ages of 4 peat samples. 32Table 2.2 Results of thin section analysis of barrier and basal sands. 46Table 2.3 Grain size distribution of sands. 47Table 2.4 Summary of some physical and chemical characteristics of the Changuinola peats. 49Table 2.5-A Core descriptions, western section. 52Table 2.5-B Core descriptions, eastern section. 53Table 4.1 Forms of sulphur at four levels in core BI 3. 155Table 4.2 Forms of sulphur at four levels in core BDD 23. 155Table 4.3 Total sulphur content of some plant parts from this and other studies. 165Table 5.1 Total sulphur, salinity and pH of 46 peat samples. 181Table 5.2 Sulphur fractions in 8 peat samples from BI 3. 182xiiDEDICATIONtoMarlene Ricewithout whom this would not have begun,Sammy Sanchezwithout whom it would undoubtedly have ended abruptly,andDoris Phillipswithout whom it would never have been completed.Gracias a todos.xiiiACKNOWLEDGEMENTSI must first thank Marc Bustin for the inspiration and the instigation of this project,which began as a Master’s thesis and became a way of life. He was a superb supervisor,unflinching in his support of a challenging project. Thanks also to the members of mycommittee, who never hesitated to help when asked, and whose expertise and good sensemade the thesis possible. I admire and am grateful to you all.The research was supported by graduate fellowships from The University ofBritishColumbia, the Natural Sciences and Engineering Research Council of Canada and theInternational Development Research Council of Canada. Logistical support was provided byThe Smithsonian Tropical Research Institute, Panama, el Instituto de Recursos Hidraulicos yElectrificacion of the Government of the Republic ofPanama, the Chiriqui Land Company,Changuinola, Panama and Transworld Explorations, Panama.A great many individuals contributed to the completion of this project. I amparticularly indebted to the people of Boca del Drago, Panama, without whose help thisproject would have been impossible. Thanks to the entire Serracin family, particularly to Sr.Jose Wilfredo Serracin de Leon, and to the Sanchez family, in particular Sr. Gilberto Sanchez.Also in Panama, I express my deepest gratitude to Eduardo Reyes, Lulu and Bibi Ferranbach,Isabel Barra.za, Enrique Moreno, Paul Colinvaux, Tony Coates, Gonzalo Cordoba, ArturoRamirez, Andres Hernandez, Mark and Connie Smith, Cameron Forsythe, Alberto Cheng,Julio Benedetti, Clyde Stevens, Ruben Chaw, Jo-Anne Ng, Paul and Anna Shaffer, Roman,Julio, Temistocle, Nicodemo, and to the memories of Captain Pete Burch, and Sra. LigiaPaget. Finally, to Sr. Samuel Sanchez, who knows as much about the Changuinola peatdeposit as I do, and to whom I refer all further inquiries.In Canada, I am indebted to all those in the Department of Geological Sciences at TheUniversity of British Columbia who eased this project along. In particular, thanks to TimEngland, to Michelle Lamberson, who was always ready with advice and suggestions, toGlenn Rouse, John Ross, and Bill Barnes, each of whom contributed much to both the workand the necessary attitude adjustment. Thanks to Lawrence Lowe and Carol Dyck in the soilslab, and to Carter Kagume and Sophie Weldon, who did the bulk of the lab work. Finally,thanks to my dear friend Judith Knight, who encouraged and supported me when I needed itmost, and my mother Doris Phillips, who had faith in me.xivFOREWORDScience is a collaborative discipline, and this study has attempted to take a multidisciplinary approach to a large subject. As a result, the expertise of many individuals hascontributed to the end result. In addition, in accordance with University guidelines, and withthe agreement of the Committee, the thesis has taken the form of a series of four researchpapers, each with a distinct focus, and each forming a chapter of the thesis. All four of thesepapers have been submitted for publication in professional journals, each naming as co-authorsthe individuals who have contributed.The papers on which Chapters Two, Three, Four and Five are based are co-authoredby Dr. R.M. Bustin, thesis supervisor, who has provided advice, support and editorialguidance throughout the research. The paper which forms the basis of Chapter Three, onvegetation and palynology of the deposit, is co-authored by Dr. G.E. Rouse of theDepartments of Botany and Geological Sciences, The University of British Columbia. Dr.Rouse provided guidance in preparation techniques, photography and palynologicalterminology. He also provided full access to his laboratory and darkroom, and shared his bestmicroscope, as well as his editorial expertise. The paper which forms the basis of ChapterFive is co-authored by Dr. L.E. Lowe of the Department of Soil Science, The University ofBritish Columbia. Dr. Lowe’s laboratory performed analysis of sulphur forms on 12 peatsamples, and he patiently and critically guided the manuscript through its many revisions. Iam grateful to each of my co-authors for their invaluable assistance.xvAll of the research, analysis and interpretation not specifically mentioned above wasperformed by Stephen Phillips, in accordance with the guidelines of the University. Thepapers which relate to the aforementioned Chapters are as follows:Publications:1. Phillips, S., Bustin, R. M. and Lowe, L. E., 1994, Earthquake-induced flooding of a tropical coastalpeat swamp: a modem analogue for high-sulphur coals. Geology, v. 22, No. 10, p. 929-932.2. Phillips, S. and Bustin, R. M., (accepted with minor revisions) Sulphur distribution in theChanguinola peat deposit, Panama as an indicator of the environments of deposition of peatand coal. Journal of Sedimentary Research.Publications submitted:1. Phillips, S., Rouse, G.E, and Bustin, R.M, (in review). Vegetation zones and diagnostic pollenprofiles of a coastal peat swamp. Bocas del Toro, Panama.Palaeogeography.Palaeoclimatology.Palaeoecology.2. Phillips, S. and Bustin, R. M., (submitted). Sedimentology of the Changuinola peat deposit: organicand elastic sedimentary response to punctuated coastal subsidence. Geological Society ofAmerica Bulletin.xviCHAPTER 1: INTRODUCTIONThe last half of the twentieth century has seen an increasing awareness world-wide of the fragilityof ecosystems and the implications of enviromnental change. Along with this interest in presentenvironmental change has grown a realization of the importance of the geological record as an indicator ofpast changes in local as well as global enviromnents. Coal deposits have long been utilized as sources ofpalaeoenvironmental information, comprising as they do a detailed and sometimes continuous record ofvegetation and climate in palaeo-wetlands, over large areas of the earth’s surface, for tens to hundreds ofthousands of years. In turn, in recent years, extensive modem wetlands, particularly in tropical latitudes,are being increasingly studied as analogues for ancient coal depositional systems. By detailed, multidisciplinary examinations of modern peat-forming systems, scientists are coming to an understanding ofthe processes by which thick deposits of peat can accumulate, in what manner they record changingenvironmental conditions, and by what mechanisms they might be preserved.In the normal measure of things, it is expected that when an organism dies, it will decompose andits components return to the ecosystem as raw material for the continuation of the system. Were all deadorganic matter preserved, the earths surface would be obscured by it. Thus a peat deposit is an indicationof an imbalance, of the inability of normal decompositional processes to keep up with the rate ofaccumulating organic debris. This can be due to rapid accumulation, or to suppression of the mechanismsof decomposition. In peat deposition the latter predominates. Peat is a sediment, an accumulation ofpredominantly organic material derived from the preserved remains of plants along with associatedinorganic detrital material, and other inorganics generated biochemically and geochemically during theprocess of accumulation. In order for this process, which may be termed ‘peatification’, to proceed, certainphysical and chemical conditions must be present. Thus peat sedimentation is associated with particularenvironments which meet these conditions. In turn, variations Within these environments are associatedwith variations within the resulting peat deposits, and given the right conditions, within coal measures thatmay eventually form from the peat. Most of the material that makes up a peat deposit is added at the top,within a few centimeters of the growing surface. Changes occurring at the surface, such as in the type ofvegetation or the proportions of organic to inorganic deposition, will be reflected in a horizontalstratification of the deposit which is evident both macroscopically in cores or hand specimens, andmicroscopically in the assemblage of preserved palynomorphs and in microtomed thin sections. Thus peatdeposits can appropriately be studied in stratigraphic columns as are other sediments, variations in thestratigraphy and sedimentology allowing for inferences as to the environment of deposition.The peat deposit which is the focus of this study is located near the town of Changuinola on theCaribbean coast of Panama, in the humid tropics of Central America (Fig. 1.1). The site is a wetlandlying within a few centimetres of sea level in a region which is tectonically active and subject to periodicearthquakes. It is one of many coastal mires to be found in this region of Central America - several aremarked in Figure 1.2, a Landsat satellite image of western Panama and eastern Costa Rica - and is ofparticular interest because its lateral extent (about 80 km2) and thickness are comparable to palaeo-mireswhich have given rise to economic coal deposits. The deposit is thus felt to provide an appropriateanalogue for the deposition of tropical, coastal coals in tectonically active settings. It is hoped that thepresent work may contribute to the continuing development and refinement of a process model for theaccumulation of coal deposits, particularly in tropical coastal environments.1.1 OBJECTIVES AND METhODSThis investigation into the evolution and structure of the Changuinola peat deposit has two mainobjectives. The first is to evaluate the deposit as a possible analogue for the deposition of low-ash, low2PLATh—-Limon-l3ocas:1Toro.Bain--00’ WFigure1.1.SouthernCentral America.Major tectonicfeaturesonboththeCaribbeanandPacificsidesoftheCentral America Arcareshown,includingtheaseismicCocos Ridge,whichhas effectivelychoked-offsubductionattheeasternendof theMiddle America Trench,andthenorthPanama thrust belt, whichconverges withthecoastnearPuertoLimon. Dashedlineoutlines extentoftheLimon-Bocas delToroback-arcbasin. Cone-shapedsymbolsarevolcanos. Therectangleoutlines theareaoftheLandsat imageshowninFigure1.2.WECARIBBEANPLATECOCOSPLATh‘‘:: : ,,. 1•_.Figure 1.2. Landsat satellite image of the area of western Panama and eastern Costa Rica outlined inFigure 1.1. The Changuinola River (CR), Chiriqui Lagoon (CL) and several coastal wetlands (CW) alongboth the Caribbean (top) and the Pacific (bottom) coast are visible. Note the contrast in coastalmorphology between wave-dominated (wd), river-dominated (rd) and tide-dominated (td) shorelines. TheCaribbean coast is microtidal (.- 0.3 in) and the Pacific coast is macrotidal (—. 8 m). Prevailing winds arefrom the N (top) to NE.•f•. ..-.. I;.I J4sulphur coals. The second objective is to document the effects of earthquake-driven subsidence events onthe peat and the peat-forming vegetation, and to assess the implications of such events on the evolution ofthis deposit and on the currently developing model of coastal tropical coal deposition.In order to address the first of these objectives it is necessary to utilize techniques which lendthemselves to interpretation in terms of coal geology. Environmental studies of coal use a variety ofapproaches, including palynology and palaeobotany of the peat-forming vegetation, geochemistry andisotopic composition of the coal, distribution of coal macerals and mineral matter, and studies ofassociated clastic depositional environments (the rocks which surround the coal). At the onset of thestudy, it was evident that certain large-scale aspects of the Changuinola deposit needed to be addressed inorder to establish a context in which detailed analysis would have any meaning. The geometry, internalstratigraphy, and evolutionary history of the deposit had to be established, and the absence of anyinformation on the nature of the peat-forming vegetation addressed, in order to determine the applicabilityof the deposit to coal studies. An initial study by Dr. A.D. Cohen and others (Cohen et al., 1989, 1990)laid the groundwork, and a fundamental series of papers on the coastal peat deposits of tropical Malesia(Anderson, 1964, 1983; Anderson and Mueller, 1975; Bruenig, 1976, 1990; Esterle, 1990; Cobb andCecil, 1993) provided a model of mire evolution through increasing oligotrophy. These bodies of workformed the starting point for this study. It was decided that phyteral analysis and detailed petrography ofthe peat would be more appropriate tools for a second phase of analysis. This study utilizes a survey ofthe modern vegetation, palynological profiles, the results of particle-size analysis (a quantitative measureof the degree of humification of the peat), sulphur chemistry, ash (mineral matter) content, and pH andsalinity measurements to characterize the peat. In addition, it looks at the nature of the barrier sand bodywhich underlies much of the deposit, and forms the seaward margin of the modern mire.)The second objective is to assess the effects of coseismic coastal subsidence and sudden sea levelchange on the mire, and the implications for the evolution of the deposit of a tectonic setting whichincluded punctuated coastal subsidence. The study commenced 10 weeks after a major (M =7.5)earthquake shook the region. Many lives were lost, roads, bridges and airstrips were destroyed, and theentire infrastructure of the region was damaged. Measurements of sea level change were made along theaffected coastline, and leveling surveys used to establish the amount of local subsidence. Over the courseof three years, detailed salinity and pH measurements were made in transects across the marine margin ofthe deposit, and an analysis of sulphur forms used to detect changes in geochemistry and microbial activitybrought on by sudden subsidence and marine inundation. Evidence of former subsidence events wassought in the stratigraphy of long submerged peat beneath and on the shores of Almirante Bay. This lineof inquiry suggests a subject for further, much more detailed research, as a datable record of earlierearthquakes would help to establish the cyclicity of such events, and possibly mitigate potential disastersin the future.1,2 PRESENTATIONThe four central chapters of this thesis each represent an individual manuscript, prepared forpublication in refereed journals, which together comprise a unified body of work, in accordance withuniversity guidelines. It is for this reason that each main chapter has its own Abstract, Conclusions andReferences, and thus forms a self-contained unit for readers interested in only certain aspects of this multifaceted study. A certain amount of repetition is unavoidable in such circumstances, particularly whendiscussing methods and providing background, but the goal of making the results of this research availableto the interested public in as accessible and comprehensive a form as possible, I hope, outweighs the riskof boring readers of the entire thesis. Some internal references to appropriate chapters of the thesis havebeen inserted, in parentheses, along with external references. This is intended to benefit the reader of thethesis, without compromising the integrity of individual manuscripts.6The second chapter presents an overview of the geological and geomorphological setting oftheChanguinola mire system, and discusses the geometry, internal structure and hydrology of the peat deposit.It presents an interpretation of the response of organic and elastic sedimentary processes to the regionaltectonic setting, and proposes a model for the structural evolution of the deposit within the context ofpunctuated coastal subsidence.Chapter Three presents the results of a botanical survey of the modern mire system, and discussesthe floral composition of the peat-forming vegetation throughout the history of peat deposition, based onpalynological analysis of the peat and of surface litter. By this means, a history of floral succession,which in turn reflects the fundamental hydrological evolution of the mire, is described. The chaptercompares the results of this study with the Anderson (1964) model of mire evolution developed fromobservations of peat deposits in the Old World tropics.The fourth chapter discusses some geochemical characteristics of the peat. It deals specificallywith the relationships between sulphur content, pH and salinity of the peat, and defines the manner inwhich the climate, biology and tectonic setting are expressed in the peat geochemistry. It then suggestssome implications of the data for the environmental interpretation of coal deposits.Chapter Five addresses the effects of the 1991 earthquake on the geochemistry of peat along thenewly submerged marine margin. It compares pH and salinity values onshore and offshore, and describeschanges in the distribution of forms of sulphur which have taken place as a result of the sudden rise in sealevel. In this way it lays a groundwork for analysis of forms of sulphur in coals which bear atransgressive signature as a result of rapid, tectonically-driven rather than gradual, eustatically-drivensubsidence.7The work concludes in Chapter Six by addressing the broader goals of the investigation, andsuggests some directions for future research which have come out of this study. There is then a series ofAppendices which contain the raw data from which the conclusions are drawn, and a brief explanation ofterminology and procedures used.1.3 REFERENCES CITEDAnderson, J.A.R., 1964. The structure and development of the peat swamps of Sarawak and Brunei.Journal of Tropical Geography, v.18, p.7-16.Anderson, J.A.R., 1983. The tropical peat swamps of Western Malesia. j : A.J.P. Gore (ed.),Ecosystems of the World 4B - Mires: Swamp, Bog, Fen and Moor; Chapter 6. Elsevier,AmsterdamAnderson, J.A.R. and Muller, J., 1975. Palynological study of a Holocene peat and a Miocene coaldeposit from NW Borneo. Review of Paleobotany and Palynology, v.19, p.29 1-351.Bruenig, E.F., 1990. Oligotrophic forested wetlands in Borneo. j: A.E. Lugo, M. Brinson and S.Brown, (eds.), Ecosystems of the World 15: Forested Wetlands, Elsevier 1990, Chapter 13.Bruenig, E.F., 1976. Classifying for mapping of peat swamp forest examples of primary forest types inSarawak, Borneo. : P.S. Ashton (ed.), The Classification and Mapping of Southeast AsianEcosystems, Univ. of Hull Misc. Ser. 17:57-75.Cobb, J.C. and Cecil, C.B., 1993. (eds.) Modem and Ancient Coal-forming environments. G.S.A.Special Paper 286, Boulder. pp.198.Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce,F., 1989. The Changuinola peatdeposit of northwestern Panama: A tropical back-barrier peat (coal)-forming environment. In:P.C. Lyons and B.Alpern (eds.), Peat and Coal: Origin, Facies and Depositional Environments.mt. J. Coal Geol., v.12, p. 157-192Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce, F., 1990. Changuinola Peat Depositof Northwestern Panama, 3 vols. Los Alamos National Laboratory pub. LA- 11211, July 1990.Esterle, 3. S., 1990. Trends in petrographic and chemical characteristics of tropical domed peat depositsin Indonesia and Malaysia as analogues for coal formation. Unpublished PhD thesis, University ofKentucky, Lexington, pp.27O8CHAPTER 2SEDIMENTOLOGY OF THE CHANGUINOLA PEAT DEPOSIT: ORGAMC ANDCLASTIC SEDIMENTARY RESPONSE TO PUNCTUATED COASTAL SUBSIDENCE9CHAPTER 2: SEDIMENTOLOGY OF THE CHANGUNOLA PEAT DEPOSIT: ORGANICAND CLASTIC SEDIMENTARY RESPONSE TO PUNCTUATED COASTAL SUBSIDENCE2.1 ABSTRACTAn extensive peat deposit on the Caribbean coast near Changuinola, Panama has developedin an area subject to periodic earthquake-driven coseismic subsidence. Thick, low-ash, low-sulphurpeat is accumulating immediately behind an aggrading and prograding barrier system, and adjacent toa flood-prone, sediment laden river. Measurements of changes in local sea level as a result of a recent(April, 1991) earthquake reveal 30 to 50 cm of subsidence, greatest at the southeastern extent of thestudy area, where the peat surface is submerged to a depth of 3 metres beneath the shallow waters ofAlmirante Bay. In the eastern part of the deposit, the effects of sea-level rise are evident in the degreeof humification, ash and sulphur content of mangrove and back-mangove peats offshore orimmediately adjacent to the marine margin, and in peats associated with brackish tidal channels whichdrain the deposit. However, most of the deposit shows no indications of marine influence, even thoughapproximately 40% of the deposit is below present sea level. The western section of the deposit hasevolved from low-lying, Raphia palm swamps originating in swales on the barrier bar, into anoligotrophic bog-plain with a water table elevated 6.75 m above sea level. As the mire evolved,transitions in vegetation resulted in transitions in peat types. Highly humified forest-swamp and palm-swamp peats underlie and surround well-preserved, fibric sedge peats, and create a partialhydrological bounding surface which restricts subsurface drainage from the central bog. The highwater table and elevated topography of the mire, and the low permeability and erosion-resistance ofthe dense, woody peat effectively insulate the deposit from both elastic influx and the extensive10intrusion of marine waters. It is evident that thick peat, and hence coal, deposits can accumulate dueto tectonically driven, punctuated subsidence, rather than gradual eustatic sea level rise, withoutleaving a record of high elastic input within the peat, even immediately adjacent to environments ofactive elastic deposition.112.2 TNTRODUCTIONThe usefulness of depositional models in coal exploration is well established. Detaileddescriptions of depositional environments and tectonic settings can be generalised into predictivemodels that are of economic value in both exploration and mine planning (Home et al., 1978).However, correlation of compositional variations in coals with paleoenvironments of peat depositionrequires detailed descriptions of like environments (McCabe, 1987). Thick coal beds require verythick accumulations of peat, and provide science with one of the most certain, yet most puzzling,records of environmental conditions at particular locations over geological time spans. Certainbecause we know, or think we know, what kind of conditions of climate and hydrology must prevail inorder to allow peat to accumulate, and somewhat less certainly, for how long. Puzzling because, giventhe dictum of uniformitaranism, we see few if any modern day examples of analogous peataccumulation. Early analogues were sought in the boreal and temperate peatlands of Eurasia andCanada, and the subtropical swamps of southeastern United States, but recent attention has focused onthick coastal peats accumulating in the Old World tropics. Tropical coastal depositional environmentsare believed to have been the settings for many known coal deposits (Wanless et al., 1969; Andersonand Muller, 1975; Cobb and Cecil, 1994). Tropical coastal peats, however, are still relatively poorlyunderstood, and differ in a number of ways from the temperate and sub-tropical studies on which earlydepositional models are based. Tropical climate affects air and water temperatures, water table andhydroperiod, seasonality, growth and decomposition rates, and the composition of the peat-formingfloral community. The focus of this study is the relationship between organic and elasticsedimentation in a large peat deposit near the town of Changuinola in northwest Panama. The objectis to determine the effects of periodic coseismic coastal subsidence on the character and geometry ofpeat accumulation and preservation, and by extension, on analogous coals.12Extensive coastal peats are currently being deposited in a variety of sedimentaryenvironments along part of the seismically active Caribbean coast of Panama. Like the vast peatswamps of Indonesia and Malaysia (Anderson, 1963, 1984; Cobb and Cecil, 1993) peat has beenaccumulating in coastal Panama at least since the stabilization of sea level some 4500 years ago(Pirazzoli 1991), and possibly for much longer: lignites of probable Miocene age have been collectedfrom coastal outcrops but never studied (Bohnenberger and Dengo, 1978). The Changuinola depositoccupies the strand plain behind an actively prograding barrier beach on a microtidal (range of 30 cm,diurnal) shoreline at 9°20’ N latitude, 82° 20’ W longitude on the Caribbean coast (Fig. 2-1). Thiscomplex mire system has developed within a few hundred metres of the wave-dominated coast, in atectonically active area that has experienced local coseismic subsidence. The peat deposit coversabout 60 km2 onshore, and another 20 km2 offshore beneath the shallow marine sediments ofAlmirante Bay. Cohen et al. (1989, 1990) first described the deposit, studied jointly by the U.S. LosAlamos National Laboratories and the Instituto de Recursos Hidraulicos y Electrificacion of theRepublic of Panama (IRHE), with the objective of assessing its resource potential as fuel for a peat-and coal-fired thermoelectric facility. They found the peat to be up to 9 metres thick, and the denseforest vegetation surrounding a plain of sedges, grasses and stunted vegetation suggested that thedeposit may be domed in a manner similar to the Malaysian peat deposits described in detail byAnderson (1964). Accumulation of organic sediments (peat) in this setting reflects an imbalancebetween organic productivity and decomposition. Plant growth and peat accumulation have exceedednet subsidence in part of the mire complex, such that 40% of the volume of the deposit is now belowsea level, unlike the alluvial Malaysian mires, which are almost all above mean sea level. The presentstudy commenced 10 weeks after a M 7.5 (surface magnitude) earthquake on April 22, 1991,epicentred 90 km to the west in the valley of the Rio Estrella, Costa Rica, which resulted in variable13coastal subsidence in the area, and the flooding of parts of the marine margin of the deposit along theshore of Almirante Bay (Fig. 2-2).Peat deposition has occured in many geological settings, including montane and alluvialsettings in which burial and preservation, and hence the development of significant coal deposits, isunlikely. As with other sediments, peat accumulation and preservation requires accommodation space,subsiding sedimentary basins. There are both similarities and significant differences between thesetting of the now well-known coastal Malaysian peats and this Neo-tropical deposit in Panama. Asin Malaysia, the Panamanian climate is humid-tropical, and temperature averages 2 6°C, with a ± 3°range. However, there is no dry season on the Caribbean coast of Panama; the annual precipitation of3000 mm is unifonnly distributed throughout the year (IRHE, 1988). Tidal effects are minimal, andstrong tradewind-driven longshore currents transport sediments to the southeast. Tropical storms andhurricanes pass to the north, but associated flood events are frequent. This extensive peat depositowes its existence to the ever-wet climate, and its morphology to the interaction between tectonicallydriven coastal subsidence, river sedimentation, and coastal processes related to waves, tide andcurrent.2.3 METhODSa) Measures ofEarthquake-induced SubsidenceThe most recent coseismic subsidence in the study area occurred during the April 22, 1991event. The very small (30 cm) normal tide range, and the presence of a tidal station at Almirantepermitted measurement of post-earthquake sea levels along about 30 km of shoreline bordering twosides of the peat deposit. Liquefaction of soils (fine-grained sands) was widespread on the alluvialplain as a result of the 1991 earthquake. The problem of distinguishing between liquefaction effectsand extensive preserved peat (coal) deposits are frequently associated with coastal margins of actively14Figure 2-1. Insert shows the location of the study area on the Caribbean side of the Talamanca Cordillera, oppositethe northeast moving Cocos Ridge. Large map shows general geology of part of the Limon-Bocas del Toro Basin,and bathymetry of the continental shelf off the study are in northwestern Panama.Legend:QR-Ala Quatemaly Alajuela Fm. -alluvium, peat (vegetation pattern), coralsTm Miocene Venado? Fm. -calcarenites, lutites, sstTM-GAus Miocene Gatun Group: Gatun-Uscari Fm. -mdst, 1st, cgl, pyrociTM-CAvi Miocene Canazas Group: Virigua Fm. -ands, basalts, breccias, dike swarmsTO-SEus Oligocene Senosri-Uscari Group and Fm. -mdst, cgl, sst, tuffsK-CHA Cretaceous Changuinola Group and Fm. -1st, mdst, sst, lavas, tiffs, andesites.15Generalized profile of elevaflon change *g0CmetresE m&re2 cr2t-+1 U +1-0T-_p _L_L_1I II II II I II IL2 i -2-I II I II IIII II II I II II II II I II II I II II I II II II I100001—----S—— I-I---.•--.- —:“‘,‘II I,_1 [I,_J I,_—I I——I(I,/ IFigure 2-2. Areas of observed and modelled uplift and subsidence along the Caribbean coast of Panama andCosta Rica as a result of the April 22nd, 1991 Ms=7.5 earthquake. The model is adapted from Plafker and Ward(1992), with additional subsidence data from this study. The profile of elevation change is generalized frompublished and unpublished data by Astorga (1991), Denyer and Arias (1992), OVSICORI (1991), Plafker andWard (1992), Phillips (1992) and Soulas (unpub. data).8300’I II II I I•:5SSS-SCARIB44AN5*4——-I-.I —5-..r-..1-.— I16and true structural subsidence was addressed by measuring subsidence at sites where bedrock isexposed in the intertidal zone at the northeast and northwest corners of Pta. Serrabata, two locationson Isla Carenero, and at Hospital Point on Isla Bastimentos in the Bocas del Toro archipelago. Theextent to which the (non-bedrock) margins of the peat deposit were affected by subsidence wasestablished through conversation with knowledgeable local inhabitants, and by the degree to whichshoreline vegetation showed the effects of drowning and saline intrusion into the groundwater (Phillipset al., 1994). Low-level colour infrared air photographs of the margin 2 years after the earthquakeshow a die-back zone varying from 5 to 50 m along the shores of Almirante Bay (Fig 2-3).b) Levelling SurveysVertical control and topography across the deposit and the barrier beachwas established bythe surveying of levelling lines (Fig. 2-4). Elevations on the landward side of the deposit are based onbench marks established by IRHE along the Almirante railway line. These bench marks, and datum atthe Almirante tide station were corrected by IRHE in 1992. Elevation at the barrier coast is tied toestimates of coastal subsidence based on changes in the swash zone, trenching across the beach, andon the drowning of shoreline vegetation. Elevation change for the affected outer coastline is shown inFigure 2-2.c) Remote SensingDrainage patterns and hydrology of the deposit, and present vegetation zonation wereestablished on the basis of SPOT multispectral satellite imagery, high altitude black and white, andlow altitude colour infrared photography. Non-surveyed sampling sites were located using aMagellan® GPS (Global Positioning System) receiver.17*-% 4—— 0-1. —‘ .SS’w-4.-iCj$ -— •.*$n4,rr4— 4-r4— —,-.4 .*.t— -4.•s?•4•:..‘4-4--$.-?“ :-0. •2•b. 3.er1 -‘4*‘ tr-1 .S’Th!” .,pt2*rr 4‘ 2e•——- 2- --.. 4-4t•_2Figure 2.3. Low level, oblique, colour infrared air photograph of part of the subsided margin of the peat depositalong the shore of Almirante Bay. The mangrove fringe forest, salt-tolerant sawgrass, and hardwood swamp areall healthy shades of pink and red. In the centre foreground, marine water has permeated 100 rn into the swamp,as shown by the dying vegetation (green tint). The peat deposit is 6 m thick at this site, and it extends 1 kmoffshore beneath the bay. The white spot in the centre foreground is a boat at drill site BDD 22D on Figure 2.4.18d) HydrologyCross-sections of the major drainage channels (blackwater creeks) were constructed, andflow rates during high and low-precipitation periods in 4 major creeks were used to estimatefluctuations in discharge .(Fig.2-5). Discharge rates are difficult to measure confidently, however, dueto the tidal nature of the drainage in the eastern section, and the dispersed western drainage pattern.Water table fluctuations are estimated from field observations, but no quantitative data on hydropenodare available.e) Clastic SedimentsSediment samples were taken at a point bar 2 km upstream from the mouth of theChanguinola River, at the mouth, and at 4 sites along the length of the barrier beach (Fig. 2-4), Beachsamples are from the upper 5 cm of the beach, at the top of the normal swash zone. Grains wereexamined in thin section for size, shape, sorting, and mineralogy, and sieved into the following grainsizes: -id), 0d, +1d1, +2c, +3d and +4(I). Similar procedures were followed for 4 samples takenfrom the base of the peat deposit using a Macauley-type hand operated corer, and samples recoveredfrom rotaly drill holes at 2 additional sites, from depths of 10 and 20 m. Characteristics of thesamples were compared in order to establish the nature of the basal sediments. Sediment from 78cores was recovered and described in the field as to colour, grain size and estimated organic mattercontent. Sedimentary structures in the barrier beach were examined in three shallow trenches dug atright angles to the coast, 2 across the beach berm and one in the back beach, 13 months after the 1991subsidence event (Fig. 2-6).f) Vegetation SurveyVegetation distribution was determined using multispectral SPOT satellite imagery (Fig. 2-7), high and low level air photographs, and by ground surveys. Major peat-forming plant species werecollected with the assistance of Sr. A. Hemandez and staff of the herbaria of the Smithsonian19Figuie 2-4. Map of the Changuinola peat deposit, showing sample sites referred to in Chapter 2. The heavySW-NE line is the surveyed transect cross-section (Fig. 2-9), and the lighter NW-SE line is that of thelongitudinal cross-section (Fig. 2-12). Heavy dashed line separates Eastern and Western sections.The EasternSection of the deposit is that part influenced by the 4 major blackwater creeks shown (see also the satelliteimage Fig. 7). Arrows show sites at which subsidence was measured or estimated.Large capitals are sites at which detailed physical or chemical analyses were performed. Other sample sitesare shown for easy reference to cross-sections.20Tropical Research Institute in Balboa and the University of Panama, Panama City. Pollen slides wereprepared from surface samples from sites representative of the 7 vegetation zones (phasiccommunities) identified in the vegetation survey, and from 2 cores, one in the central part of thedeposit and the second near the eastern margin. Pollen was concentrated, using standard palynologicaltechniques, from the fme fraction of the peat (< 0.25 mm).g) Peat SamplingPeat cores from the surface to the base of the deposit were taken at 78 sites, using handoperated Hiller- and Macauley-type coring devices, and a 3.5 cm diameter vibracore. Samples wererecovered in 25 or 30 cm increments whenever possible, described in the field, double-bagged andstored at room temperature until they could be refrigerated or frozen. Salinity and pH for mostsamples were measured at the time of collection, and verified in the laboratory using a Cardy® ModelPHi digital pH meter, and a Cardy® Model C 121 digital salt meter. Surface litter samples werecollected (to 5 cm depth) and 25 cm cubes cut with a machete.h) Peat characterizationA variety of methods was used to characterize the peat recovered in cores. In addition to pHand salinity, total sulphur content (dry weight percent) of 203 samples (dried at 50°C, crushed to 100mesh) was determined using a Leco® SC-l32 Sulphur Analyzer (see Tabatabai, 1992, p.313 for adescription of this instrument) and verified using wet chemical methods (Appendix G). Mineral mattercontent (wt % ash) of 137 samples was determined by weight loss on ignition in a muffle furnace at550°C (ASTM-D 2974: Jarret, 1982). Peat is defined according to ash content using the OrganicSediments Research Centre, University of South Carolina, standard (Andrejko et al., 1983). By thisstandard, peat is defined as Low (<5 wt%), Medium (5-15 wt%) and High Ash (15-25 wt%). Above25 wt% is carbonaceous sediment. Moisture content of wet peat, drained of superficial water, was21Figure 2-5. Cross sections, and high- and low-discharge rates measured in the three largest blackwatercreeks. Salinity and pH of bottom and surface water is shown. The banks of the channels are peat, and thechannel floors are clean medium grained sands. Measurements were taken 1 km upstream from AlmiranteBay (see Fig. 2-4).ojCaño SucioDischarge ratesm%ecLow HighRio Banano9.2 21.2SaL = U.2 pH =il= 1.90 pH = 7.37sal = 0.92 pH = 7.2111 = 2.10 pH = 7.32sal = 0.86 pH = 6.883.9Canal Viejo= I °3.2Om 5 104.8J15m22Figure 2-6. Cartoon showing the northwest part of the barrier beach, and a cross-section of the barrier at siteBEACH 1. Sediment from the Changuinola River is consistently carried to the southeast by longshore drift.The sediment plume is visible in Figure 2-8.Inserts show sedimentary structures revealed in trenches dug across the beach berm at the sites indicated. Thesurveyed elevation of the barrier is shown, as is the site ofpeat accumulation in the swales. The canal, dredgedin 1908, is visible as a straight line in the satellite imagery (Figs. 2-7 and 2-8). Vertical exaggeration is about500x.Raphiapalm swamp Beach 1 trenchBeach 2 trenchSite of flood channelCross-section across the barrier at BEACH 1177 rr—23Figure 2.7. False colour spot satellite image of the area of the detail map (Fig. 2.4). Concentric zoning ofvegetation is most obvious in thc western part of the deposit, but is also visible around 3 lesser domes in theeastern section. A larger scale SPOT image is reproduced on page 93. Black is sediment laden water, and whiteis clear water (see Fig. 2.8).Figure 2.8. False colour SPOT satellite image of the narrow coastal plain and barrier shoreline. The sedimentplume of the Changuinola River is clearly visible trailing off to the east. A plume of sediment is enteringAlmirante Bay on the flood tide (right centre), but most of the bay margin of the peat deposit is free of elasticinput, and is an area of carbonate deposition (dark blue). Bananas are in pink on the floodplain.24measured by air drying at 50°C (wt % moisture lost), and is used in plots as an approximation of thedensity of the peat. Degree of humification of the peat was established by particle-size distribution ofeach sample. Degree of humification of the peats is based on the relative proportions of coarse,medium and fine constituents as determined using a wet-sieving procedure modified from Staneck andSilo (1977), according to the following scheme (Esterle et al., 1987):Coarse (C) >25% >20 mm <30% <0.25 mm ( von Post fibric to coarse hemic)Medium (M) <25% >2.0 mm <30% <0.25 mm (= hemic)Fine (F) <25% >2.0 mm >30% <0.25 mm (= hernic to fine hemic)These three categories are used for the plotting of core data. In this study we chose to usewet sieving as a means of measuring tissue preservation because the technique lends itself well to thelarge incremental samples recovered by Macauley and Hillyer corers, usually 25 or 30 cm of core at atime. We report the results of sieving as percentages of total by dry weight, as the methods ofmeasuring volume are less accurate, and do not lend themselves well to woody or fibric peats. Thedegree of stratigraphic correlation possible using the bulk technique is necessarily limited, but thespacing of sample sites (500 to 1500 m at best), and the absence of continuous marker beds such as avolcanic ash fall, make precise correlation impossible by any of the means at our disposal. Themethod thoroughly homogenizes the increment, and thus generalizes the interpretation to a 25 or 30cm scale. However, there is little literature on the discriminating ability of wet sieving of peat,particularly woody tropical peat (Staneck and Silc, 1977), thus only by comparison to other methodscould we judge the effectiveness of the technique. Particle size by wet sieving is compared withpalynological analysis for a site in the central deposit (ED 3) and one from the eastern section (BDD23).Mineral matter was separated-out by gravity, and the results of sieving with 2 0 mm and0.25 mm sieves were dried in a 50°C oven, weighed, and plotted as the proportion of C, M and F25components for 13 core sites (Fig. 2-4). Tissue preservation (degree of humification) is described infield observations, and in the written descriptions in this report using a modified von PostHumification Scale adapted to tropical peats by Esterle (1990). There are 5 field categories: sapric,fine hemic, hemic, coarse hemic and fibric. The traditional use of field-determined peat types has beenused only sparingly in the study of these tropical peats, as recent work (Esterle, 1990) suggests lowcorrespondence between field classifications and the actual particle-size distributions as determined bypoint counting or sieving methods. Reference is made to the field categories, alongside particle-sizedistributions, because peat workers are familiar with this index of tissue preservation (von Post,1922). No sapric peat was encountered in any core. Peat Classification is based on the identificationof macroscopic plant parts and palynomorphs in the peat, compared to plant and pollen associationsidentified in the surface samples, and uses botanical (e.g. Rhizophora peat, sedge peat) nomenclature.2.4 GEOLOGICAL SETFINGThe Changuinola deposit is situated at the easternmost onshore extent of the Lirnón-Bocasdel Toro sedimentary basin (Fig. 1-1). The Bocas del Toro Basin in Panama, and its westwardextension in Costa Rica, the Limón Basin, together make up the Tertiary and Quaternary back-arcbasin behind the volcanic ranges of the Guanacaste and Central Cordilleras in Costa Rica, and theuplifted Tertiary marine sediments and intercalated Upper Miocene volcanics and plutonic rocks oftheTalamanca Cordillera of western Panama. The onshore part of the basin extends about 400 kmfrom the Costa Rica-Nicaragua border southeastward into Panama, narrowing eastward and becomingan offshore feature east of Laguna Chiriqui (Escalante, 1990).The Costa Rica - Panama island arc was created by subduction of the Cocos plate beneaththe western edge of the Caribbean plate. Andesitic arc building proceeded from latest Cretaceousthrough Eocene, and uplift and expansion of the emergent island arc continued through the Miocene.26The oldest sedimentary rocks are of the Cretaceous Changuinola Formation, interbedded foramimferallimestones, tuffs and lava flows, known from a single exposure in the valley of the Rio Changuinola(Fisher and Pessagno, 1965). Overlying Tertiary and Quaternary sequences, approximately 7000metres of predominantly marine elastic deposits, are known from exploratory drilling carried out byseveral oil companies since the 1920’s. The major island arc sedimentary sequences were depositedfrom the Oligocene to the middle Miocene as basaltic talus deposits, pro-delta and shallow offshoresandstones, limestones and shales. During Middle Miocene uplift of the Talamanca Cordillera,subsidence in the Limón Basin resulted in deposition of the Gatun Formation. This is a succession ofvolcaniclastics and carbonates that also includes isolated lignite lenses. Lignites and lignitic siltstonesare interbedded with argillaceous and sandy sedimentary rocks (Bohnenberger and Dengo, 1978).From Miocene to late Pliocene, the basin was the site of gradually shoaling marine deposition. Thetop of the sequence is the La Gruta Limestone Member, approximately 450 metres thick. From earlyPleistocene to the present, the region has experienced a significant rate of emergence (72 rn/Ma.;Coates and Obando, in press), likely in response to shallow subduction of the Cocos Ridge (Collins etal., in press).The aseismic Cocos Ridge has effectively blocked subduction at the eastern end of theMiddle America Trench, leading to deformation in an overall SW-NE compressional regime across theIsthmus of Panama. The structural geology of the Caribbean coastal area adjacent to the Ridge isdominated by a series of southwest dipping thrust faults, striking NW-SE sub-parallel to the coast(Camacho and Viquez, 1992; Denyer et a!., 1992). The general emergent trend, related by Collins andothers (in press), and Coates and Obando (in press) to crustal doming is punctuated by occasionalcoseismic vertical movements; uplift and coseismic folding near LimOn (Denyer et al., 1991), andsubsidence and localised marine transgression in the area of Almirante Bay. The M=7.5 earthquakein the Valle de la Estrella, Costa Rica, on April 22, 1991, and subsequent aftershocks affected 150 km27of the Caribbean coastline of Costa Rica and Panama (Fig. 2-2). Part of the affected coastline, fromthe mouth of the Rio Changuinola southeast 12 km to the Boca del Drago channel, consists of abarrier beach behind which the large Changuinola peat swamp has developed (Fig. 2-4). Studiesconducted since the earthquake of April 1991 suggest that the break in the trend of the coastlinemarked by Almirante Bay may be related to as-yet unmapped faults which segment the coast into aseries of tilted blocks (E. Camacho, personal communication),Offshore in the Panama Basin, the North Panama Deformed Belt (Fig. 2-1) is an extensive,thick (to 7000 m) accretionary wedge which displays recent compression-related thrust faulting andsoft-sediment deformation. The Belt approaches the coast in the near-offshore between Puerto Limónand Laguna Chiriqui, where continuous seismic reflection profiles show faulting, upthrusting anddeformation in the most recent sediments (Clowes, 1987). These profiles also reveal erosion, in theform of canyons and coast-parallel channels (near-shore profile CSP3-4; Clowes Fig. 5 .2a).2.5 GEOMORPHOLOGYThe northwest coast of Panama is a site of active elastic and organic sediment deposition.Much of the coastal plain is under a variable cover of Quaternary and Recent alluvium, and the coastis characterised by a series of barrier beaches behind which lagoons and extensive paralic swamps aredeveloped. The uplifted and folded sedimentary rocks of the Talamanca Range are deeply incised by anumber of large rivers, which transport sediment northward off the Cordillera to the Caribbean coast.The Changuinola peat deposit has developed behind a barrier beach extending 12 km southeast fromthe mouth of the Changuinola River, which drains an area of approximately 3200 km2. Wind drivencurrents move sediment consistently alongshore to the southeast, as shown in Figure 2-8, a satelliteview of the sediment plume of the Changuinola River. Sedimentation rates are high along this28microtidal (± 30 cm) coast, creating the ridge-and-swale morphology typical of prograding barriersystems (Reinson, 1984). Cores behind the shoreline show well-sorted barrier sands (Table 2-2)underlying Raphia palm peat 4 km landward of the present coastline and 673 cm below sea level.Thus the barrier coastline is both aggrading in response to subsidence and prograding seaward acrossthe continental shelf. Nearshore bathymetry reveals a shelf widest to the northwest off the coast ofNicaragua (Fig. 2-1), which becomes narrower and more frequently incised by submarine canyons tothe southeast near Limón. Canyons are present off the narrow shelf northwest of Limón, wheresubsidence was recorded as a result of the 1991 earthquake (Plafker and Ward, 1992). Southeast ofthe point at Linión, the shelf widens somewhat, and is free of canyons until past the Panama border atthe Sixaola River, beyond which it again narrows rapidly, and a canyon is present at the 100 fathomcontour 12 km off the Changuinola River mouth. Another, larger canyon is present seaward of Bocadel Toro (Fig. 2-1). The shelf seaward of the barrier is narrow (12.5 km) and has a gradient of 4.2ni/lan out to the 20 fathom line about 9 km offshore. A surface current of 1 to 2 knots movessediment consistently to the southeast of the river mouth (personal observations). The size and shapeof the sediment plume, which sweeps to the east and away from the remnant canyon, is revealed infalse-colour SPOT multispectral satellite imagery (Fig. 2-8). The barrier coastline terminates at PuntaSerrabata, where a tombolo connects the barrier beach to a rocky outcrop which was formerly anisland.Shelf bathymetry reflects erosion of the exposed coastal plain during the last glaciallowstand. In particular, the distribution of submarine canyons gives some indication of the underlyingstructural control on coastal morphology, and of the lateral migration of the major rivers during theHolocene. Major drainage across the coastal plain is directed towards those segments of the coastwhich are experiencing coseismic subsidence, such as the region around the mouth of the Rio Matinanorthwest of the point at Puerto Lirnón, and the section of coast to the southeast of the Rio Sixaola in29Panama, which is the focus of this study. Subsiding coastal regions are distinguished by extensiveparalic swamps, and in some cases embayments, whereas those parts of the coast which recordeduplift in the most recent earthquake are typically emergent sandy or rocky shores with dismpted andirregular drainage of the coastal plain. To the southeast of the Changuinola mire, Almirante Bay is arecessive coastal feature which is likely related to the tectonic regime described above.2.6 VEGETATIONFloral diversity in the present peat deposit is low and distinctly zoned (Cohen et al, 1989;Phillips et a!., in review) into a sequence of ‘phasic communities’ similar to those described for theoligotrophic peat swamps of Western Malesia (Anderson, 1964; 1984) and Borneo (Bruenig, 1990).Factors which account for floral zonation on peat are not fully understood, but are related tomicrotopography, nutrient levels, water table and pH of the groundwater, and to variations in theporosity and permeability of the peat itself. Concentric zonation in surface vegetation is echoed in avertical floral succession from the base to the top of the deposit resulting from peat accumulation,elevated water table, and increasing oligotrophy of the mire. The ‘Anderson model’ of domed peatdevelopment has been summarized by numerous authors (Bruenig, 1990; Phillips et al., 1994) andrelated to coal depositional environments by others (Esterle, 1990; Cobb and Cecil, 1993). TheChanguinola deposit includes 7 phasic communities, 6 of which contribute a distinctive peat type tothe accumulating deposit. These phasic communities are: i) Rhizophora mangrove fringe swamp; ii)mixed back-mangrove swamp; iii) Raphia taedigera palm swamp; iv) Carnpnosperniapanamensisforest swamp; v) mixed forest-swamp; vi) sawgrass ± stunted forest swamp; vii) Myrica-C’yrilla bog-plain. Distinctive communities of peat-forming plants, along with their associated groundwaterenvironments, determine much about the character of the peat which develops, and result in astratigraphic succession which reflects the trophic evolution of the mire through time. Consequently,30schemes which identify and classify peat (and coal) according to the principal floral componentsinherently imply much about the environment of deposition.False-colour satellite imagery shows the deposit to have two regions with differingvegetation patterns and surface drainage (Fig. 2-7). The western part displays a radial drainagepattern and concentrically zoned vegetation which reflects a domed topography attributed in theAnderson model to an evolution toward increasing oligotrophy with increasing peat accumulation. Nolarge streams flow west or north out of the mire; surface drainage off the central dome is principallyas a sluggish sheet-flow in the low-gradient areas from the central bog plain towards the Talamancas.Channelized flow, in blackwater creeks which extend to the base of the peat, drains the higher-gradientmargins towards the barrier and Almirante Bay. Creeks draining into the bay are stratified andbrackish up to 3 km upstream, despite the low tidal range (Fig. 2-5). These creeks dissect the easternpart of the deposit into hydrologically distinct units visible in the satellite imagery. In this study werelate peat characteristics in the western and eastern parts of the deposit to their respective states ofombrotrophy and rheotrophy, which in turn are related to the factors controlling height of the watertable, and of sea level.2.7 AGE AND GEOMETRY OF THE DEPOSITThe main body of the peat deposit is roughly rectangular in shape with long axis parallel tothe coast and occupies the entire 8 km width of the coastal plain between the Talarnanca hills and theouter barrier. Peat extends from the lower alluvial plain of the Changuinola river 12.5 km southeastto the marine margin of Almirante Bay, and at least another 1 km beneath the bay. Greatest thicknessof peat sampled was 950 cm (Cohen at al., 1989) and 833 cm (this study), and overall average peatthickness is estimated at 650 cm. Maximum elevation of peat measured is +667 cm above datum, and31the greatest depth -673 cm. Approximately 40% of the volume of the deposit is below present sealevel. Datum (mean sea level) is considered accurate to within 10 cm.Table 2-1: Radiocarbon Ages of Peat Samples(Beta Analytic Inc.)No. ID Sample Depth C14 C131C12 C13 adjo/oo age1 24/92 BDD 19A 475 2160+/-60 -28.2 2110+/-602 72024 MILE 5 810 3040+/-80 -25.0 3040+/-803 72025 MILE 5 475 850+/-80 -25.0 850+/-804 71 989* BDD 34 475 1910+/-60 -27.1 1880+/-60Note: * denotes AMS dating, CAMS-i 3316, L. Livermore Natl LaboratoryEstimates of the age of the peat deposit are based on the 4 radiocarbon dates listed in Table2-1. These dates give the following accumulation rates for different peat types:- mangrove/back mangrove peat (No. 1) 2.25 mm/a (from -4.75 m below present SL to present SL),- mangrove/back mangrove/forest peat (No.4) 2.52 mm/a (from -4.75 m to present SL),- palm/forest peat (No.2) 1.48 mm/a (from 8 m to 4.75 m below peat surface), and- sedge peat (No.3) 5.58 mm/a (from 4.75 m below peat surface to peat surface).The oldest peat tested, from 54 cm above the base of core MILE 5 in the central western section, gavea corrected age of 3040 ± 80 BP in woody, hemic Raphialmixed-forest peat over grey sand. The baseof the deepest peat core sampled (LAKE 9 - not dated) is topographically some 2 rn below the datedsample, and thus may represent an additional 1000 to 1500 years of peat accumulation, giving anapproximate age of 4000 to 4500 years for the deposit. This age is similar to that of many Malesiancoastal peat deposits (Anderson, 1983). It is possible that older peat may exist at lower elevations,most likely close to the Talamanca foothills, but none were dated in this study.The surface of the peat in the western section is an asymmetrical ‘dome’, only slightly raisedon the shore side, and sloping more steeply close to the barrier (Fig. 2-9). The maximum measuredgradient, in palm forest along the northern margin, is 1 rn in 330 m, comparable to gradients in most32— Percent of Medium particlesL-. Percent of Fine particles<E> ash spikeFigure 2-9. Top: SW-NE cross-section of the peat deposit, showing location of cores and distribution of Coarse,Medium and Fine peat in relation to the generalized peat stratigraphy. At ED 3, near the centre, a schematic coreshows results of pollen analysis (see Fig. 2-14).Bottom: Results of the particle size analysis (wet sieving) of 8 cores along the transect, showing uniformlymoderate- to highly hunilfied peat at the margins and base. The greatest contrast in humification is between theupper and lower parts of LAKE 10, ED3 and to a lesser extent MILE 5. Black shading, and dotted spikes indicatemineral matter> 4%. No correlation between mineral layers was possible.3310kmLegendUpper panelEl coarse peat sandmedium peat [[J clay partingfine peat IC14:3O4O’J-8riioca,on[] wood> 2cmLJLLower panelshading represents resultsof wet-seiving, in weight percentPercent of Mineral AshPercent of Coarse particlesMalesian mires, which commonly grade at 1 m in 300 to 500 m (Anderson, 1983), but in contrast tothe river-margin gradients of 1 in 25 (McCabe, 1987) to 1 in 65 (Esterle, 1990) in some Sarawakmires, probably influenced by erosion of the banks. The margin bordering the Talamanca hills gradesat about 1 m in 1 km. No topographic survey data are available for the flood-plain margin, butanalysis of SPOT imagery using the width of concentric vegetation zones to approximate topographiccontours suggests a gradient no steeper than 1 in 500.Topographic control of the underlying basal sands is provided by two transects normal tothe coast, and by isolated sample sites parallel to the barrier. The barrier has a ridge-and-swalearchitecture characteristic of prograding barrier coastlines (Reinson, 1984), that can be traced at thebase of the peat (Figs. 2-6 and 2-9). Cores are spaced at around 500 m intervals, but probes weremade at closer spacing, as the modem, subaerial barrier has a ridge ‘wavelength’ of between 50 and100 m. Traces of the barrier stmcture are also visible up to 2.5 km shoreward of the barrier in false-colour satellite imagery (Fig. 2-7). Peat originates in the swale immediately behind the beach bermwhere, despite the high porosity of the barrier sand, a linear marsh or swamp is established, due tohigh growth rates, dense vegetative cover, and precipitation which nonnally exceedsevapotranspiration rates (Fig. 2-6). Along a surveyed transect, the distance from the top of the swashzone to the nearest measured thickness of ‘swale’ peat (40 cm of peat, under 50 cm standing water inRaphia swamp) is 94 metres.2.8 SEA LEVEL CHANGE AND PUNCTUATED SUBSIDENCE“No lives were lost, here {Punta Mono}or at the other Indian settlements, in the neighbourhood, butthe ground appeared rent in various places, the sand on the beach was either raised in ridges, ordepressed in furrows; a place which, in the evening had been a small lagoon, or pond, in whichseveral canoes were floating, was now become quite dry; most of the huts were violently cracked andtwisted; and the effects, of the earthquake, were everywhere visible.”from Voyages and Excursions in Central America, by Orlando Roberts, 1827.34There exists a popular and historical record of seismic actitvity on the coast of Panama andCosta Rica dating back to at least 1798, and reports of such events back to the late 1500’s (Camachoand Viquez, 1993). Earthquakes are known to have occurred in 1798, 1822, 1912 (2) and 1991, allestimated (by Guendel, 1991) or recorded as greater than magnitude 7 on the Richter scale (Roberts,1827; Gonzalez-Viquez, 1910; Guendel, 1991). Roberts’ observations at Punta Mono in 1822 couldhave been repeated almost verbatim in 1991 when the Point was again uplifted, a nearby river mouthclosed, and parts of the back-barrier lagoon were left high and dry.a) Subsidence in 1991Coseismic subsidence occurred along 30 km of the coast between the mouth of the RioChanguinola and Hospital Point on the northwest tip of Isla Bastimentos during the April 22, 1991earthquake. Liquefaction effects were widespread and spectacular throughout the alluvial plain(Astorga, 1991; Denyer et a!., 1991; Guendel, 1991). Describing the scene a few days after the event,Guendel (1991) wrote that Changuinola “perhaps represents one of the most dramatic cases of local toregional soil liquefaction ever documented in Central America”. Mobilized sediment was consistentlyvery fme-grained grey sand, which erupted as sand volcanoes and sheet flows from crevasses up to500 m in length throughout the lower flood plain. Liquefaction-related subsidence was widespread,but uplift may also have occurred, leading to shoaling of dredged channels. Areas of coarser-grainedsediment, however, such as the beach and river-mouth bars, were free of any signs of liquefaction, andwe estimate a maximum 35 cm subsidence at the river mouth, based on measurements of the degree ofburial or submergence of trees rooted in beach sand (Fig. 2-10), and on sedimentary structuresdescribed in a later section.35*Figure 2.10. Photograph of drowned tree on the shore side of the beach berm at site Beach 2. The tree wasgrowing at a spot that is now 33 cm below present mean sea level.36The problem of distinguishing between liquefaction effects and true structural subsidenceoccupied several investigators after the 1991 event (Astorga, 1991; Camacho et a!., 1991; Denyer etal., 1991; Guendell, 1991; Sherstobitoff 1991). Greatest reported subsidence, and greatestearthquake damage, was in the communities of Changuinola, Almirante and Bocas del Toro.The former is on the Changuinola flood plain, and experienced extensive liquefaction and surfacefissuring, as well as local subsidence which left irrigation pumps standing suspended on their pipes 50cm above ground level. The latter two towns are built in part on dredged fill in former mangroveswamps, and variable subsidence in the filled areas can be attributed to liquefaction and settling.Bocas del Toro, however, experienced a uniform subsidence which unhappily left the entire sewersystem below sea level (personal observations, and personal communications with Ing. E. Reyes andothers, IRHE). Extensive shoreline observations were needed to establish the degree and extent ofsubsidence directly affecting the peat deposit and barrier beach. Destruction of the tide gauge atAlmirante necessitated re-surveying of bench marks used for topographic control in this study, and weconsider the elevations published here to be internally consistent, but with the possibility of anabsolute error of a few centimetres.Plafker et al. (1992) studied elevation change from the 1991 event along the affected coastfrom the Matina River in Costa Rica southeast to Punta Mona, and one site in Panama. They foundthe effects more or less consistent with movement on a 25° to 40° SW dipping thrust fault at a depthof 40 km beneath the valley of the Rio Estrella in Costa Rica (Fig. 2-2). Southeast of the ChanguinolaRiver, the amount of subsidence increases. At the easternmost extent of the barrier beacha tomboloconnects the beach to Punta Serrabata (Fig. 2-1). the exposed coast of which consists of limestonesand shelly sandstones of the Late Miocene Water Cay Formation and the Pliocene La Gruta Member.We estimate subsidence of 50 cm at Boca del Drago, and about 70 cm at Isla Carenero, the mostprofoundly affected location on the coast, based on measurements on dock pilings and rockyV 37shorelines. Across the Boca del Toro channel, 500 rn southeast of Carenero at Hospital Point and atBastimentos, no subsidence was detected along the rocky shore.b) Long-term Sea Level ChangeRegional Holocene estimates of sea level change are much more poorly constrained thanlocal historical data. Three published studies of Holocene sea level change in the western Caribbean,based on radiocarbon dating of mangrove peats (Bartlett and Barghoorn, 1973; Woodroffe, 1988;Berger, 1983), and a predicted sea level curve for eastern Panama (Peltier, 1988) give a poorlyconstrained estimate for regional sea level change in the study area (Pirazzoli, 1991). Since thePleistocene sea level low stand of about 18,000 years ago, Caribbean Central America experienced agenerally transgressive regime until about 4500 years ago, when sea level more or less stabilized andthe modern Changuinola peat deposit began to accumulate. The Bartlett and Barghoom curve, locatedclosest to the area now under consideration, is based on a single radiocarbon date of 35,500 yr.BPwhich Bush et a!. (1992) consider to be of questionable value, since it implies considerable uplift ofthe Gatun Lowlands (Canal Zone of central Panama). Ample evidence exists for Holocene sea levelchange in the area of the Changuinola peat deposit, but no data on local long-term sea level change hasbeen published. Regional sea level estimates for 4000 yr.BP are between -7.5 m and -1 m, and for2000 yr.BP, less than 1 m below present sea level (Pirazzoli, 1991). In this study a maximum 1 meustatic rise in regional sea level over the past 2000 years is assumed, due to the high degree ofuncertainty associated with the higher estimates (Bush et al., 1992). This estimate is compared to 2sub-sea level dates (a 14C standard date and an AMS date) of mangrove and forest peat from sitesnow lying offshore in Almirante Bay (Fig. 2-1; Table 2-1). This shallow body of salt-water forms theeastern termination of the barrier coastline, and terrigenous sediments extend a considerable distancebeneath the bay. Submerged peat was sampled at water depths of more than three metres. The peat isunderlain by medium grained grey sand, and in places overlain by a thin layer of silty sands, lime mud,38and coral. Beneath the grey sands, uncemented calcareous mud with abundant coral fragments occursfrom around 10 m depth down to at least 50 metres.Radiocarbon dates show that local relative sea-level rise has been proceeding for more than2000 years (Table 2-1; Figure 2-4; dating by Beta Analytic mc). Two samples of peat from 475-478cm below present mean sea level were dated by radiocarbon and AMS methods with a corrected ageof 2010 ± 60 (BDD 19A) and 1880 ± 60 yr. BP (BDD 34). BDD19A is a combination vibracoreMacauley series taken in 195 cm water depth at a site 600 m offshore in Almirante Bay. The coreconsisted of mangrove (Rhizophora) and back-mangrove (Rhizophora - Laguncularia - Acrostichum)peat from the base to the top of the core (-510 cm to -195 cm; sample salinity 17 %, pH 7,07),showing that the site was within the mangrove fringe - back mangrove zone throughout the period forwhich peat is preserved, and thus that the shoreline during that period was some 600 m farther eastthan at present. Taking into account the 30 cm tidal range, local sea level in Almirante Bay has risenby between 350 and 450 cm in 2000 yrs, a maximum rate of about 2.2 mm/a. at the drowned site,where accumulation did not keep pace with subsidence. BDD 34 is a series of 50 cm long half-corestaken with a Macauley corer in 60 cm water depth at the outer limit of the pre-earthquake mangrovezone, on a sheltered shore essentially free of clastic sediment input. The base of the core at 673 cm isin fine hemic mangrove peat over rooted medium grained grey sand containing large wood fragments.Above the base, the peat alternates between Raphia and forest-swamp peat, and suiphurous mangroveand back-mangrove peat, representing periods when the site was alternately at and above sea level.The dated Raphia palm peat, now at -475 cm (sample salinity 7%, pH 5.76), suggests average (net)subsidence of 2.1 to 2.6 mm/a. at a site at which peat accumulation evidently kept pace withsubsidence until fairly recently. Autocompaction of the peat has not been factored-in to theseestimates, due to the absence of any apparent density increases with depth in the extremely compactfine hemic woody peats.39Evidence from the Changuinola peat deposit suggests 4000 or more years of subsidence, at anet rate of 2.2 to 2.6 mm/a. for the past 2000 years. Historically, this subsidence has occurred as aseries of coseismic events, and the stratigraphy of the marine-marginal peats suggests that this style ofpunctuated subsidence has been the norm during that period. The upper rate is evidently slightlyhigher than that at which mangrove peat can accumulate in the eastern part of the deposit, and thegreater lateral extent of the deposit in the past suggests a slower rate of subsidence during the earlyhistory of the mire.2.9 ORGANIC AND CLASTIC SEDIMENTOLOGY2.9.1 CLASTIC SEDIMENTSa) Changuinola River FloodplainSediments of the floodplain occupy 60 km2, between the SanSan mire on the northwest andthe Changuinola mire to the southeast (Fig. 2-1), and interfinger with the northwestern margin of thepeat deposit. Sediments consist of boulder and cobble- to silt-size sediments originating in the foldedsedimentary rocks of the northern Talamanca range. River bar clastics are dominated by feldspar andvolcanic rock fragments. To the southeast of the floodplain, the lateral transition from alluvium topeat occurs over about 1000 m, the intervening sediments being interfingered floodplain silts andsands, cm-thick leaf beds, and thin peats (Fig. 2-1 1). Cohen et al. (1990) reported the ash content oftwo cores (BDT 5 and 4) located about 500 m and 1000 rn from the nearest active channel on thefloodplain (Fig. 2-12). The cores are not peat, but rather are carbonaceous sediment with ash (mineralmatter) content varying from 30.3 to 61.6 wt% at the site closest to the river, and 27.2 to 43.8 wt%closer to the mire. In the same study, a core 2.4 km from the river (BDT 13) consisted of low- tomedium-ash peat (range 0.9 to 12.9 wt% ash) throughout the entire 945 cm depth.40*Figure 2 11. Lanunated silts, sands, peat and cm-thick leaf beds of the alluvial plain. Deposits aremoderately bio-turbated; entrances of two rodent burrows can be seen, centre and right of centre.41o—CD-ot;.)CD CD——z—-t o°.o C)CDCD.D)•!fl -CD-•CDt.pipçCDC)0..0.. 000cnSediment Legend16kmI : I’,’,—’\“,bog-plainsedgepeatsawgrassIstuntedforestpeatback-mangrovepeatRhizophorapeatCampnospermapeatmixedforestpeatRaphiapeatmedium-finesand(rooted)carbonatesedimentsshellycarbonatesI-icoarsepeatmediumpeatfinepeat[]wood>2cmCoreLegend[]sand clayparting1C14:3040+/-8OIradiocarbonageb) Barrier SystemMost studies of barrier beach and strandplain evolution have recognized the importance oftidal range on the morphology of barrier beaches (Glaeser, 1978). However, data for much of thedevelopment of facies models in the ‘60’s and ‘70’s came from the temperate eastern seaboard of NorthAmerica, where seasonal storms are the norm and much of the event-driven Holocene stratigraphicrecord is written by hurricanes. The Changuinola barrier beach is microtidal and wave-dominated,unbroken by tidal channels and seldom overtopped by storm waves.Morphology of the barrier is shown in Figure 2-6. The exposed barrier consists of 3 parallelsand ridges, the highest of which is the most seaward (1.68 m, 1.44 m, 0.6 m at Beach 1). Thelandward decrease in elevation is a consequence of ongoing subsidence of the shoreline. The‘wavelength’ of 40-5 0 m between ridges reflects the relation between rates of subsidence andprogradation.I) Sedimentary structures - The effects of rapid subsidence and subsequent aggradation andprogradation are evident in the sedimentary structures revealed by trenching on the barrier. Trencheswere made at two sites: Beach 1, at a central point on the barrier 7.5 km from the Changuinola River,and Beach 2, close to the river mouth (Fig. 2-6). At Beach 1, landward dipping (6°) planar washoverbeds of the back-berm overlie 9° seaward-dipping beds of the former berm-face to a depth of about 30cm. At site Beach 2 the present shoreface is medium grained sand, steeply dipping (20°) seaward,with a wave cut notch above the top of the swash zone. A few weeks after the earthquake-inducedsubsidence, the barrier had resumed its original height arid was not experiencing washover duringnormal wave conditions. Trenching across the berm revealed steep (20°) medium grained seawarddipping beds, corresponding to a remnant beach face, overlying and truncating shallow landwarddipping fine sands of the pre-earthquake back-beach. The medium sands are in turn truncated andoverlain by finer grained landward-dipping beds of the post-earthquake (present) back-beach, to a43Figure 2.13-b. Photograph of bedding structures in a trench across the beach berm at Beach 2, shownschematically in Figure 2.6. In the photo, the sea is to the left, and flat-lying washover beds rest on steeplyseaward-dipping beach face sands. The barrier aggraded about 30 cm, and prograded several metres in the 13months that elapsed between the subsidence of the beach and the trenching44Figure 2.13-a. Planar laminated back beach sands at site Beach 2.thickness of 30 to 50 cm (Figs. 2-13-a and 2.13-b). This sequence is interpreted as a record of thelandward-stepping of the beach face, caused by coseismic subsidence, followed by rapid aggradationand progadation of the beach face and back beach.ii) Grain-size andMineralogy - Table 2-2 shows the distribution of grain size of sediments from 12sites shown in Figure 2-4, including a point bar in the Changuinola River, proximal and distal sandsalong the length of the barrier, and basal samples from below the peat deposit. Mineralogy, sorting,grain size and rounding from 12 sites are listed in Table 2-3. Grain size of the proximal barrier sandschanges little along 9 km of barrier shoreline: well-sorted medium to fine grained, angular to sub-angular sands, with an immature mineralogy including pyroxenes, feldspar and volcanic glass and lessthan 10% quartz grains. Heavy minerals are rapidly winnowed out down-coast (Tables 2-2 and 2-3).Basal sand landward of Beach 1 (halfway along the modem bar, site Lake 3) plots in the same grainsize range as the barrier sands. At the distal end of the barrier (Beach 4, 11 km), siliciclastics arepredominantly very fine sand (> 4 1) dominated by felspathic grains, to which a coarse to very finecarbonate component originating in the offshore barrier reef is added. The distal barrier experienced agreater degree of subsidence (an estimated 50 cm); the beach berm has not regained its formerelevation, and the beach face is submerged and impassable due to fallen trees.c} Peat Basal SedimentsSediments underlying the peat deposit can be divided into clean medium to fine grainedbarrier sands, fine to very fine sands with palm and sedge roots associated with drainage channels andthin palm-swamp peat, and silts and clays associated with drainage from the Talamancan slopes alongthe southwest margin of the deposit. Those from greater depth (-10 m, well below the peat-sandcontact) are similar in grain size to the barrier sands, and display the same trend of decreasing heavymineral content and increasing feldspars and volcanic rock fragments with increasing distance from45TABLE2-2:ResultsofThinSectionAnalysisofBarrier andBasalSandsNotes:Sz=estimatedmeangrainsizeinmm,Ab=abundance,Ro=roundness=rounded,s=sub-,a=angular;volcrf=volcanicrockfragments.feldsparisbothplagioclaseandkspar;pyroxeneisalmostallclino-,possiblywithatraceofolivineandamphibole.carbonatesareshells, coralfragments,andlimemud.D=DepthofsampleinmetresC-DistancefromopaquesquartzfeldsparvolcanicrfpyroxeneampholivinemicacarbonatesSampleRiverMouthSortingAbRoSzAbRoSzAbRoSzAbRoSzAbRoSzAbRoSzAbRoSzAbRoSzAbRoSzPointbar2kmpoor5sr.07trr.1355a.1335r.13trr.07upstreamBeach00kmmod40sr-.20trsa.2625a.20tra.2630a.20trsa.07saBeach21kmwell75r.13trsr.1310a.2610a.33Beach17.5kmmod15sa.135sa.2630a.2635a.2615a.26LakelO8.5kmwell10a.2010a.265a.2670a.46tra.46tra.46tra.13D=8mOuterbeach9kmmod5sa.075sa.1330sa.2050sa.205sa.13tra.07trsa.13Beach411kmvpoortrsa.075a.1320a.13tra.132a.07tra.1370a1.3BDD2313.5kmpoortrr.075r.0750a.2645a.33tra.07D=6mCS313.5kmmod5sr.075a.1380a-sr.785sr.135sa.13trsa.13tr.13D=3.5mBDD22D14kmwelltrsr.135a.2670sr-a.265r.2615a.26tra.13D=10mBDD22D14kmwelltra.26100a.39D=20mBDD3318kmmod5a.0745a.3350sr.39tra.07tra.13D=10m46>14)>24) >34) >44)5.01 22.41 44.307.64 72.71 19.6310.94 67.56 17.615.61 76.37 17.945.32 78.60 16.1615.77 26.26 33.4433.14 51.96 8.9010.11 9.46 10.2831.09 53.77 8.0811.66 47.37 15.1812.53 41.77 16.9915.97 65.48 9.67Distance4+4) from River25.83 2 kmupstream0.03 0 km0.15 1 km0.05 7.5 km0.06 9 km1.40 11km1.50 14km39.33 14km0.96 18km22.51 13.5 km19.89 13.5 km7.40 7.5 kmTABLE 2-3 Grain Size Distribution of SandsSeive Sizes (Phi scale)Sample >—14) >04)Point bar % 0.00 0.08Beach 0 % 0.00 0.00Beach 2 % 3.61 0.01shell&pebbleBeach 1 % 0.00 0.00Outer % 0.00 0.01beachBeach 4 % 1.97 9.77shell and coralBDD 22D % 0.08 1.27D=lOmBDD 22D % 5.05 15.29D=20mBDD 33 % 0.11 0.47D=lOmBDD23-b % 0.21 1.01D=5mBDD 23-b % 1.23 3.99D=6mLake 3 % 0.05 0.69D=3.3m47the main river channel. Basal sands from sites associated with long-lived channels (eg. BDD 23) areless well-sorted than the beach samples, although lithologically similar. To the southeast beneath thebasal sands at 20 m depth is poorly sorted carbonate with a slightly bimodal distribution (Tables 2-2and 2-3).d) Airnirante BayFine-grained clastic sediments enter the Boca del Drago channel on the flood tide as thedistal sediment plume of the Changuinola River (Fig. 2-8) and are deposited along the eastern marginof the deposit and in the deeper north-central part of the bay. Fine clastics sourced on the flanks of theTalamancas also enter the bay via the Rio Banano, from the southwest corner of the bay and aretransported toward Pta. Pondsock. Most of the bay margin bordering the swamp is essentially free ofclastic sediments, however, and active carbonate sedimentation is occurring. Sediments are an algalcarbonate mud, and numerous coralline forms. Two offshore patch reefs were drilled, with limitedsuccess, to determine the nature of underlying sediments. Carbonate sediments with a bimodal grainsize distribution (lime mud, and coral debris) were found to at least 6 in depth at both sites, and to atleast 20 m beneath the basal sands of core BDD31 (Table 2-3).2.9.2 ORGANIC SEDIMENTATIONa)Peat characterizationThe physical and chemical characteristics of the peat are sunnnarized in Table 2-4. Thetable shows high, low and mean values and standard deviation for pH, total sulphur, wt % moisture,salinity and ash for the eastern and western sections, and for selected mangrove peat samples, whichrepresent a special case in the east. Significant differences in all parameters are evident in theaveraged values from the domed, ombrotrophic west and the dissected, and in part rheotrophic easternsection.48wt%Moisture 92.74 5.1high - low 99.2-67.1%Ash 3.29 6.39high - low 38% -0%TABLE 2-4: Summary of Some Physical and Chemical Characteristics of Changuinola PeatsWestern Section Eastern section Rhizophora mangrove peatsmean s n mean s n= mean spH 4.38 0.59 321 5.74 1.09 389 6.5 0.58 70high - low 6.29-2.82 7.89-2.67 7.6 - 5.4wt% Sulphur 0.23 0.18 227 2.24 2.55 143 3.52 1.09 36high-low 0.51-0.08 13.7-0.08 5.9-0.4wt%Salinity <0.01 n.d. n.d. 0.69 0.87 413 1.42 0.96 104high-low 2.7-0 2.7-0Notes: s.d. = one standard deviation; n = number of samples in data set76 84.28 7.66 134 84.49 5.4595.7-60 94.1-67.585 10.49 9.19 43 18.57 19.7254%-1% 92%-4%452149b) Geochemical ParametersIn previous reports (Phillips et al., 1994; Phillips and Bustin, in press) the relationshipbetween pH, salinity and total sulphur content of the peats has been explored in detail. High sulphurcontent is spatially related to marine or brackish influence. Coastal mangrove and back-mangrovepeats with moderately high S content (1 to 5 wt% S) and high salinity (> 0.5 wt%) dominate theeastern margin and extend beneath the salt water and shallow marine sediments of the adjoining bay.Marine influence extends only a short distance onshore, except in the vicinity of brackish blackwatercreeks which drain the swamp into Almirante Bay. Peats associated with these channels are low insalinity (< 0.5 wt%) but very high in S (5 to —14 wt% S), around 90% of which is carbon-bondedsulphides, apparently the result of an ongoing biogeochemical chain of S reactions leading to theconcentration of C-S fonns. The western part of the deposit is domed and oligotrophic, and thevegetation and the peat are concentrically zoned. Stunted, sawgrass-dominated vegetation producesfibric, very low S (<0.25 wt% S) peat found in the upper few metres of the central bog plain. Aroundand below the bog plain peats, dense hemic and fine hemic peat, the product of mixed-forest andpalm-forest swamp vegetation, has consistently higher S content, averaging between 0.25 and 0.5 wt%S.The highest sulphur content (>5 wt%) is found in peats near the brackish influence ofblackwater drainage channels up to several km from the coast. Very high sulphur was also found inthe basal peats in the deepest parts of the deposit (Cohen et al., 1989), which are interpreted to havebeen associated with channels. Very high sulphur content is found even where salinity is undetectablein peat, porewater or basal sediments, and despite the low pH of these peats that normally inhibitsbacterial sulphate reduction.50c) Degree ofHunufication, Ash, and Moisture ContentTissue preservation is a standard parameter in environmental studies of both peat and coal(Stach et al., 1982). Diessel (1986) devised a Tissue Preservation Index (TPI), based on the relativevolumes of structured (i.e. well-preserved cell walls) and unstructured (finely comniinuted, orgelifled) tissue which he then used in interpreting the degree of wetness of the mires in which someAustralian coals originated. Numerous authors (Lamberson et al., 1991; Obaje et al., 1994) havesince applied this principle to other coals, with various qualifications, but with the commonassumption that highly degraded humic coals originate in a relatively oxygenated environment, andcoals with a high proportion of structured tissue in a waterlogged, acidic and poorly oxygenated mire.In coal studies, the TPI is determined by the point counting of macerals, a volumetric measure. It ispossible to apply volumetric estimates to peat samples, by measuring the volume of sieving results(Stanec and Silc, 1971), or point counting of thin sections (Cohen, 1968) or polished epoxy-impregnated blocks (Esterle et al., 1990). However, direct volumetric comparisons of peat to coalcannot be made, due to the variable compactability of different peat types, and also the selective lossof matrix material during coalification (Shearer, 1994), thus proportions by weight are used in thisstudy. The method, although not directly comparable to coal methodology, is accurate and allows forthe analysis of large samples.Ash content is also used as an environmental indicator in coal studies, and has been relatedto particle size or degree of humification in low ash peats (Davis et al., 1984). It is generally assumedthat: a) oxygenated environments which produce highly degraded peat will also selectively concentratewhatever mineral matter happens to be present; and b) that peat deposited in mires close to sites ofactive clastic sedimentation will contain higher proportions of mineral matter than more distant sites.51TABLE 2-5A: COPE DESCRIPTIONS OF PEATS FROM THE WESTERN SECTION OF THEDEPOSITCORE MILE 1 MILE 1.5 MILE 3 MILE 5 ED 3Peat Elevation: Top ± 617cm + 637cm + 655cm + 655cm + 624cmand Base + 148 cm ÷ 7 cm - 165 cm - 209 cm - 186 cm(+ above SL; - below SLModem Vegetation Raphia palm-swamp Mixed forest-swamp Sawgrass / stunted Myrica rn..Cynlla r., Mynca ni. -Cyrilla r.,forest sedges, grasses: bog- sedges, grasses: bog.plain plainHumilIcation 49.9% to 21.4% n.d. 58.8% to 19.6% 66.2% to 4.0% 60.6% to 5.40%(percent coarse fibres) avg. 32.5% avg. 39.6% avg. 36.1% avg. 39.9%SalinityRangc(wt%) <0.01 <0.01 <0.01 <0.01 <0.01Total Sulphur Range (wt%) n.d. 0.1610 0.51 nd. 0.l4toO.3 nd.pH Range 3.6 (top) to 49 (base) 3.3 (top) 104.8 (base) 3.8 (top) to 5.1 (base) 3.6 (top) to 5.5 (base) 3.6 (top) to 5.6 (base)MoistureRange n.d 91.3%to94,9% nd. 90.3%to97.l% n.d.avg. 93.1% avg. 94.4%Ash Range n.d. 14% 10< 1%; spikes 1.0% or less except 10% to < 1%: spikes 6% to <1%; spikes atat 120cm (4%) and for a single spike at at 90(10%), 390 60cm (6%), 240 cm510cm (14%) 300cm (14%) (4%), 480(6%) and (4%) and the basal570 cm (4%) metre (8%)Basal Sediments Rooted grey clay Peaty grey clay Peaty silt and VF Yellowish-grey peaty Rooted F grey sandsgrey sand sand, over F greyChannel sandsPeat Classification 0-124 F Hemic 0-270 C Hemjc-Fjbric 0-270 Fibric-C Hernic 0-330 Fibric-C Hemic 0-630 Fibric-C Hemic(depths in cm) 124-304 Hemic 270-510 Hernic, 300-810 Hemic; much 330-840 Hemic-F 630-750 C Hemic304-394 C Hemic- woody at 270-330 wood 690-750 Hernic; much wood Hemic, woodyFibric; Woody Hemic 510-570 C Hemic from 600 to base 750-8 10 F Hemicto 454Table 2-5A: ContinuedCORE LAKE 10 LAKE 8 LAKE 6.5 LAKE 2 NL IPeat Elevation : Top + 500cm + 397 cm + 342 cm + 254 cm + 150 cmand Base -320cm -295cm - 193cm -76cm - 130cm(+=aboveSL;-=belowSL)Modem Vegetation Myrica m. - Cyi-illa Myrica in. - Cyrilla Sawgrass I stunted Mixed forest-swamp Raphia palm-swampr, sedges. grasses: r.., sedges, grasses: forestbog-plain tree hammockHumification 73.7% to 5.60% 39.8% Tb 18.1% 40.2% to 15.0% 22.4% to 2.8% 26.4% 105.10%(percent coarse fibres) avg. 32.0% avg. 29.7% avg. 29.2% avg. 10.8% avg. I6.9%Salinity Range (wt%) <0.01 throughout <0.01 throughout <0.01 throughout 0.02 (top) to <0.01 0.01 (top) to <0.01TotalSulphurRange(wt%) nd. n.d. nd. 0.23to0.33 n.d.pH Range 3.6 (top) to 5.8 (base) 3.7 (top) to 5.4 (base) 3.2 (top) to 5.2 (base) 2.8 (top) to 4.9 (base) 3.9 (top) to 4.9 (base)Moisture Range nd. n.d. n.d. 78.6% to 95.2% n.d.avg. 88.8%AshRange <1%throughout <l%throughout lO%to<1%: spikesat 8%to<1%: spilceat <l%throughout275cm(10%), 425 150cm (8%)(4%)and 525 (I0%)BasalSediments PeatyFandVF CleanFandVFgrey PeatyFandVF CleanFandVFgrey Peaty,rootedgreyyellowish grey sand sand yellowish grey sand sand (Barrier) sand (Barrier-swale)over clean F grey sand over clean F grey sandPeat Classification 0-300 Fibric-C Hemic 0-225 C Bernie 0-475 Hemic F Hemic throughout F Bernie throughout(depths in cm) 300-550 Hemic 225-550 Hemic 475-500 C Hemic550-62sWoodyHemic 550-675 F Hernic 500-550 F Hernic625-750 F HemicNotes: C = coarse, F fine, n.d. as not determined52TABLE 2-5B: CORE DESCRIPTIONS FOR THE EASTERN SECTIONCORE BDD 23 BDD 22 BDD 31Peat Elevation Top + 10 cm + 5 cm sea level (high tide line)and Base -450cm -590cm - 400cm(+ above SL; - = below SL)Modem Vegetation Campnosperma Rhizophora mangle fringe Rhizophora mangle fringepanamensis forest-swamp forest forestHumiuication 30% to 16% 27% to 3.7% 22% to 2.5%(percent coarse fibres) avg. 19% avg. 12% avg. 12%Salinity Range (wt%) <0.01 throughout 0.59 to 2 20 0.11 to 0.41Total SulphurRange (wt%) 0.27 to 13.7 n.d. l.O7to 8.43pH Range 3.6 (top) to 5.9 (base) 6.7 (top) to 7.9 (base) 4.3 to 7.4Moisture Range 76% to 95% n.d. 75.4% to 85.3%avg. 90% avg. 81.7%Ash Range 16% to <1% 65% (shells, diatoms, 57% to <1%: spikes atspike at 100cm (16%; no ostracodes, forams, sponge 100cm (57%; no carbonate)carbonate) spicules) to 5% (mineral and 150cm (38%; shelly)grains, mixed litholon’)Basal Sediments Rooted F to VF sand, over Carbonate: rounded coral Clean F to VF sandclean F to VF sand, and shell fragments, plus (Banier)(Channel sands) dispersed mineral grainsPeat Classification 0 - 285 Coarse Hemic 0 - 150 Hemic 0. 165 Hemic and Fine(depths in cm) 285 -450 Bernie and Fine 150- 590 Fine Hernic Hernic, WoodyHemic 165- 400 Fine HemicNotes: C coarse, F fine, n.d. as not determined53The proportions of coarse, medium and fine organic particles in bulk peat samplesrepresenting 25 or 30 cm core increments are used to generate plots depicting the degree ofhumification of the peat at 3 sites in the eastern section of the deposit and 8 sites along a surveyedtransect in the west. All proportions are given as per cent of the total dry weight of the sample on amineral-matter free basis. Moisture content, also recorded for some samples, is related to the degreeof humification and the density of the peat. The deposit is divided into eastern and western sections onthe basis of vegetation zonation, and hydrological characteristics, and the results are presentedseparately in Table 2-5A and 2-5B.1) Western Section:--- Given the permanently high water table associated with everwet climate, thepreservation of structured plant parts is good in oligotrophic mires, where low pH of the groundwaterand low nutrient availability discourages both bacterial activity and luxuriant subaerial biomassproduction. Plants thriving in such conditions tend to have stunted above-ground components, butextensive root systems, that contribute to the fibrous component of the resultant peat. Figure 2-9(upper panel) shows a cross section through the western part of the deposit illustrating the distributionof well-preserved fibric peat and more humified hemic and fine hernic peat along the SW-NE cross-section shown in Figure 2-4. Stratigraphy is based on the degree of tissue preservation (particle sizedistribution) in 8 cores (Fig. 2-9 lower panel) along a surveyed transect from the foothill margin to thebarrier. Characteristics of the peat are summarized in Table 2-5A, and peat type is correlated withpollen stratigraphy in core ED 3, near the centre of the transect (Fig. 2-14). As can be seen in thelower panel (Fig. 2-9) particle size and humification in the marginal peats (Mile 1 near the southwestmargin, NL1 and Lake 6.5 toward the barrier bar) tend to be uniform for the full depth of the peat. Inthe central deposit ( Lake 10, Ed 3, Mile 5 and Mile 3 in the central bog plain, and Lake 8 in aMyrica-Cyrilla tree ‘hammock’) there is a contrast between the upper fibric and coarse hemic peats,54Myrica + Cyrilla PalmaeF: ::::.IBog PlainII111ii Sawgrass -U Stunted ForestU U, UICampnospermaForest SwampU RaphiaI I I I I PalmSwamppercent of total percent of totalcm 0 5% 10 0 2(Y340_________Ash content Results of wet sieving Succession interpreted from pollen analysesFigure 2-14. Comparison of the results of wet sieving (particle size distribution) with palynological analysis of coreED 3 from the central part of the bog-plain. Myrica + Cyrilla, and Palmae (Raphia pollen + palm phytoliths)represent the trend clearly. The wet sieving results distinguish basal hemic and fine hemic forest-swamp peats fromupper fibric and coarse hemic sedge peats, reflecting vegetation changes and increasing oligotrophy in the mire.The sawgrass/stunted forest-swamp pollen zone is transitional, and variable both in palynomorph assemblages andin degree ofhumiflcatiom of the peat. Pollen data is discussed in detail in Chapter 3.Particle sizes are as follows:Coarse >2.0 mm Medium > 0.25 mm Fine <0.25 mmVonPost classes are defined on page 21, and described in the Appendices.I I I I I1O%5% 0%20 40 6055and underlying hemic and fine hemic peats. Degree of humification is greatest in areas of greatestsurface gradient.ii) Eastern Section: --- The complex vegetation zonation of the eastern part of the deposit, visible inFigure 2-7, reflects an equally complex topography, hydrology, and peat stratigraphy in that part ofthe deposit which is drained by the brackish blackwater creeks emptying into Almirante Bay.Sedimentation within the mangrove fringe varies with location, from sandy peat to limey peat.Rhizophora mangle (the red mangrove) is the only mangrove colonizing the marine margin- the whitemangrove species Languncularia raceniosa is common to dominant in the immediate back-mangroveswamp and bordering brackish blackwater creeks. Rhizophora does not favour either silty orcarbonate substrates, and isolated individuals are found rooted in coral as much as a kilometreoffshore. A species of thin-shelled oyster (Crcissostrea sp.?), easily recognizable in core, is normallyfound attached to Rhizophora roots from the upper limit to about 30 cm below the lower limit of thetidal zone, and when found stratified within otherwise shell-free peat suggest possible past subsidenceevents. In addition to the mangrove fringe and back-mangrove forest swamps along the bay and creekmargins, there are three areas of domed, oligotrophic sawgrass marsh and stunted forest-swamp, eachsurrounded by a complex ofRaphia and hardwood forest-swamps. Figure 2-15 shows particle sizevariation in three cores in a cross section from Almirante Bay through one of these lesser domes to athe blackwater channel, Canal Viejo. Particle size in the forest-swamp and mangrove swamp peats isuniformly fine; at the coastal sites the <2.0 mm component averages 87% to 93%, and in only 2samples was there a fibric (>2.0 mm) component greater than 25%. All samples were classified ashemic or fine hemic; there was no sapric peat encountered, despite the high degree of humification.Peat characteristics are summarized in Table 2-5 B.56large wood fragments[:. : mineral ash layershelly layerOrganic particle sizes:coarse, fibrousmedium; <2.0 mmflne;<0.25mm ::::____Figure 2-15. A generalized view of wet sieving results of cores from across the marine margin of AlmiranteBay and up Canal Viejo, a brackish blackwater creek. The section approximates the SE end ofthe cross sectionin Figure 2-12, although BDD 31 is displaced. BDD 31 is a typical mangrove peat core, highly humifiedthroughout, and having both carbonate-rich and elastic ash layera. BDD 22 represents a saline estuarine site,also in the mangrove fringe-forest. The sediments are not strictly speaking peat, as they contain >25%non-humic materials. The ash includes scattered grains of feldspar, mica and quartz, but is dominated by theremains ofmarine organisms. The peat is very highly humifled, fine hemic throughout. BDD 23 is also highlyhumifled throughout, and is associated at the base with brackish drainage. Near the base the sulphur content isalmost 14 wt%. The basal peat was deposited in a back-mangrove zone. Figure 2-16 relates the results ofparticle size analysis to the palynology ofcore BDD 23.Peat Core Legend Sediment LegendI Campnospermapeatmixed forestpeatRaphia peatback-mangrovepeatbog-plain sedgepeatF, ‘ ‘I Rhizophora peatmedium - fine sand(rooted)carbonate sedimentsshelly carbonatessawgrass / stuntedforest peat57- MixedForest SwampIiII Raphiaj Palm SwampIIFigure 2-16. A comparison of the wet sieving results for core BDD 23 with palynological analysis (Chapter 3).Raphia palm peat tends to be less humified than woody angiosperm forest-swamp peats. The source of themineral ash spike at 100 cm is unknown, but may be related in some way to an attempt at dredging Canal Viejoin 1899. The site is about 50 m north of the present channel.Rhizophora # Laguncularia ÷Acrostichum a.Campnospermapariamensis] L CampnospermaForest Swamp60 40 20I I I I20 40 60 80nrnnt nf In IAsh content Results of wet sievingBack-mangroveSwamprwnt if IntISuccession interpreted from pollen analyses58BDD 31 is an example of an intertidal mangrove site (Fig. 2-15). Between 100 and 150 cmdepth there is a shelly layer and a silty layer suggesting a possible subsidence event, followed by aresumption of mineral-free peat accumulation. The BDD 22 core (upper part) represents a modemestuarine site, taken at the outlet of a tidal blackwater creek. Due to the high content of bivalve shellfragments, diatom frustules, ostracode valves, foram tests, fish scales and teeth and sponge spicules(24% to 72% by weight), the sediment is not true peat. Estuarine microfauna are not present in thebasal 160 cm of the peat. Silt-sized quartz, mica and other grains were present throughout the core.Below the silty base of the peat is carbonate sand consisting of well-rounded shell and coralfragments. These deposits are considered to represent shallow offshore (back-reef) sediments overwhich the barrier and peat deposit have prograded. The base of site BDD 23 represents an earlierchannel margin site influenced by the brackish waters of a blackwater creek. Transitions in peat tefrom top to base of the core echo lateral transitions in surface vegetation (Phillips et al., in review),from Campnosperma forest-swamp peat, to mixed-forest, Raphia palm, and back-mangrovepeat. The history of floral succession, interpreted from the palynology of the core, is plotted alongsideparticle size distribution and ash content of the peat in Figure 2-16.2.10 DISCUSSION2.10.1 CLASTIC SEDIMENTARY RESPONSE TO RECENT TECTONISMa) Changuinola River FloodplainRivers draining the Caribbean side of the Cordillera are prone to flooding, (2 such eventsoccurred in the Changuinola flood plain during the 3 years of this study), and the sedimentologicalrecord is written by the interelationship between flood events and tectonic events. The lowerChanguinola flood plain has been affected by agricultural activity, but certain inferences can be maderegarding the Holocene evolution of the floodplain based on sedimentary response to recentsubsidence. In the early Holocene it appears that the Changuinola river may have flowed southeast59from the defile at which it exits the foothills, along or parallel to the course of the present Rio Banano(see Fig. 2-4), and crossed the continental shelf via the deep canyon of Boca del Toro (Fig. 2-1), A180 m deep incised channel extends from near the mouth of the Rio Banano across Almirante Bay(much of which is less than 5 m deep) and through the Boca del Toro passage, indicating erosion to avery low base level. Local records and locally produced maps of the lower Changuinola floodplain(unpublished data of the Chiriqui Land Company; C. Stephens, pers. comm.) since the turn of thecentury document numerous changes in the main channel of the Changuinola River. Before 1920 theChangumola drained to the northeast during floods, through the mouth of the neighbouring SanSanRiver. A channel switch, which occurred at the defile, and dam building by the banana companies,brought an end to this, and the Changuinola now absorbs all the floodwaters.In July 1991, 10 weeks after the earthquake, a major storm struck the stretch of coastaffected by the earthquake, with very different effects in regions of uplift and of subsidence. On thecoast 60 km to the northwest, 10,000 hectares of the coastal plain shoreward of the mouth of the RioEstrella, the major river in the area, were devastated by flooding, the result of uplift at the coast whichpartially dammed the river mouth and caused floodwaters to back-up onto the alluvial plain. Theareas of coastal uplift are not prone to the development of extensive mires, despite their susceptibilityto flooding, as the rivers deposit much of their sediment load on the coastal plain. The floodingChanguinola River did not, however, inundate the coastal plain nor the immediately adjacent peatswamp to the southeast, the surface of which is somewhat elevated above normal flood levels andheavily forested with Raphia palm and hardwoods. Earthquake-induced subsidence lowered theelevation of the alluvial plain; the flooding river broached a levee, but remained channelized, enteringan abandoned oxbow and breaching the lowered barrier beachovernight. During the flood, theintroduction of overbank sediments into the peat swamp was minimal, and has been restricted to areaswithin 500 - 1000 m of the river during the entire period of peat accumulation, as evident from the low60ash content of peats 2 km away. Subsidence events evidently resulted in channelized stream flow, stillallowing the accumulation of low ash strand plain peats. The areas of peat deposition are betterinsulated from flood influence than areas of alluvium, by virtue of their elevation and the lowerodability of the peat.b) Barrier BeachThe tectonic setting dictates the scale of evolution of the barrier, in both time and space.Punctuated subsidence provides accommodation space for the aggradation of the sand body,apparently in 30 to 50 cm increments. The amount of subsidence, which determines whether thebarrier will step seaward or landward, increases to the southeast, although it has not yet beenestablished whether the thickness of the barrier sand body is greater in that direction. The cumulativerate of subsidence also determines whether organic sedimentation is terrigenous peat, or shallowmarine carbonates, such as are present at the south-eastern extreme of the barrier system. Regionaltectonics also detennine the effective width of the continental shelf, which eventually may limit theseaward migration of the barrier. Shelf gradient may not be a limiting factor; Glaeser (1978)calculated that 24% of the world’s barrier coastlines are associated with seas marginal to arcs (i.e.fore-arc or back-arc settings), and that there is no correlation between shelf gradient and the presenceof barrier islands or barrier beaches. A supply of sand-size sediment, and tidal flux are the controllingfactors in generating barrier coastlines, and tides and storms dictate barrier morphology and backbarrier sedimentation.Four factors defme the depositional environment of the barrier beach, working in consort todetennine its morphology and composition: tectonic setting, sediment source, climate and tidal range.Each has geological significance, and would leave a signature on the confining rocks of a strand-plaincoal. Barrier sands display structures and mineralogy which reflect tectonic influence. Bedding61structures and beach morphology reflect the style of subsidence, which occurs as discrete earthquake-driven events, and which is greatest to the southeast. Three sediment sources contribute to the barrierbeach. Immature sands of the Changuinola River are the principal source, originating at the northwestend of the barrier. The second source is an offshore coral reef, present at the extreme southeast end ofthe barrier, which both donates carbonate sands and debris to the barrier and creates a low-energybeach and a sediment trap in the shallow back-reef environment seaward of the barrier. The thirdsediment source is the lush tropical vegetation, which responds to subsidence through changes inphasic community along the marine margins, stabilizes beach ridges and deposits peat in swales.2.10.2 HYDROLOGICAL CONTROLS ON PEAT ACCUMULATIONHydrological controls on peat accumulation and peat character include source of recharge,rate of recharge, hydroperiod, rate of discharge from the mire, nature and morphology of the confininglayer, and penneability and heterogeneity of the peat (Clymo, 1983). In the case of ombrotrophicmires, recharge is limited to atmospheric sources, and water enters the mire from above and flows out.In rheotrophic mires, some water flows in, entering from the bottom or edges of the mire. Rate ofrecharge is the amount of water entering, and hydroperiod, which is the period of fluctuation of thewater table, is a measure of the regularity of recharge (Myers, 1990). The confining layer is the sealwhich prevents the mire from draining. Peat accumulates when the water table is above the miresurface, and deteriorates when it dries out.Peat permeability is highly variable and very poorly constrained, but generally variesdirectly with peat density, and thus with particle size and compaction. We do not have a great deal ofquantitative data on peat permeability in the Changuinola deposit, and substitute moisture content andparticle size distributions as proxy data for degree of humification and thus peat density. Two studiesof peat permeability based on the rate of intrusion of saline waters into freshwater peat at sites in62Almirante Bay, although not complete, indicate very low permeability for mangrove peat, which tendsto be dense and compact (85% moisture), and slightly higher permeabilities for Raphia and forest-swamp peats (85-90% moisture). The bog-plain sedge peats and stunted forest peats sampled in thisstudy average 95% moisture content, and have much higher permeability.a) Western Section: — Hydrology dictates the fundamental physical and chemical characteristics ofthe organic sediments in the western section of the deposit. At the present stage of evolution of thedeposit, hydrological controls in the ombrotrophic western region are internally driven, rather thanbeing a reflection of tectonic forces. As there is no evidence that recharge occurs from below themodem mire is assumed to be effectively ombrotrophic. At the Talamancan margin and theChanguinola River margin, discharge water flows out of the deposit as surface sheet flow, and inshallow (<50 cm) streams. Along the back-barrier margin, which has the steepest gradient, drainagehas been disrupted by the dredging of a canal into which five small streams discharge. One of thesecuts to the base of the peat, and connects four small sandy bottomed lakes which are remnants ofback-barrier lagoons.There is some variation in the height of the water table across the western part of thedeposit, and a corresponding variation in the degree of humification of the peat. The surface isslightly domed, and the vegetation and the peat are concentrically zoned. The Anderson model offloral succession and increasing oligotrophy in the evolution of raised forested mires in SE Asia hasbeen shown to be applicable to the Changuinola mire system as well (see Chapter 3; Phillips et al., inreview). Tissue preservation as estimated by particle size reveals the same overall stratigraphicpattern seen in sulphur and pH profiles and moisture content, and generally correlates with moredetailed palyno-stratigraphic analyses (Fig. 2-14). The margins of the deposit are generally well63decomposed hemic and fine hemic peat from top to base, whereas in most of the central bog plain 3 to5 m of fibric to coarse hemic peat overlie more decomposed hemic and fine hemic peat (Fig. 2-9).Stunted, sawgrass (sedge)-dominated vegetation produces fibric, low pH and very low S (<0.25 wt% S) peat found in the upper few metres of the central bog plain. Around and below the bogplain peats, mixed-forest and palm-forest swamps produce hernic and fine hemic peat, still low in pH,but with higher S content, between 0.25 and 0.5 wt% S. The sulphur in these peats is principallyassimilatory, and the high level of humification concentrates resistant compounds, including those thatincorporate sulphur. It is expected that the same process of concentration by digestion would serve to2mcrease ash content as well. However m this study no significant correlation (r = 0.04) is foundbetween wt% total sulphur and wt% ash (HTA) for all low ash (<10%) and low S (<0.5%) peats forwhich data are available. Thus either the ash content is not directly related to degree of humification,or another factor such as the sulphur content of the original phyterals must be sought in explainingsulphur concentrations. Again, however, beyond the broadest distinction between herbaceous andwoody peats, we find no significant difference in the total sulphur content of the different types ofwoody peat (Raphia, Campnosperma, mixed forest-swamp), and the variable sulphur content of theoriginal plant material, although not thoroughly studied, still are less significant than the degree ofhumification in low-sulphur peats (Phillips and Bustin, in press). We conclude that some aspect ofthe biotic processes involved in the degradation of woody debris leads to approximately double theconcentration of organic sulphur compounds in woody peat than those processes which produce coarsehemic and fibric peat from bog-plain vegetation.It is apparent that degree of humification is much more dependent on hydroperiod and heightof the water table than on the measured pH of the peat in cores. High levels of humification occur invery low pH sites if they are subject to fluctuating water table, or if the plant community has a large64subaerial biomass. Water table fluctuations are greatest where surface gradients are steepest (betweenNL 1 and Lake 6.5 in Figure 2-9), corresponding to the zone of dense forest-swamp. These are thesites of most uniform hemic peat accumulation, and also presumably sites of high evapo-transpirationfrom the luxuriant foliage. Highest consistent water table is in Raphia swamp on shallow peat aroundthe margins of the deposit, and in the central bog-plain and the lesser domes to the east (Fig. 2-7),which are vegetated with algal mats and a sparse and stunted flora, and where evapo-transpirationrates are accordingly low. Raphia peats tend to be coarse hemic to hemic, and sedge peats fibric tocoarse hemic.The most distinctive features of the bog-plain peats are their fibric nature, reddish colourand extremely high moisture content. Drained of superficial water, western peat samples averaged93% moisture content (average includes the deeper forest-swamp peats: range is 67% to 99%),compared to an average of 84% for the eastern peats (Tables 2-5A and B). Equally distinctive is thecalculated rate of accumulation of 5.58 mm/a., based on a radiocarbon age of 850 ± 80 years forcoarse hemic peat (45% fibrous) 475 cm below the surface. This high rate contrasts sharply withrates published by Anderson (1983) for peats in the upper part of Malaysian domed deposits (2.22mm/a for the upperS m of peat in his phasic community 6). Anderson found that accumulation rateswere lower in the later-stage (his phasic community 6 and 7), oligotrophic swamps, possibly due toreduced biomass production. However, the central domes of the Malaysian deposits are stunted bog-forests, not sawgrass swamps , and the peat surface of the bog-forests frequently dries out. TheChanguinola deposit is apparently quite different in this respect; the bog plain surface is not onlynormally submerged, but along much of the surveyed transect, the upper root-mat is effectivelyfloating on a very watery ‘soup’ of fine and medium-sized granular debris, and up to 60% wellpreserved roots and other fibres. In the light of the wetness of the bog-plain peats, the in situ65accumulation rate uncorrected for moisture content means little, if the water table were to drop in thecentral plain, the peat surface would follow it down, possibly to the extent of creating a depression.The effective elevation of the central bog plain along the surveyed transect is I m in 7 km,with margin gradients a maximum 1 in 330. Winston (1994) has successfully fitted theoretical curvesto existing peat dome profiles from Malaysian mires, assuming constant hydraulic conductivity(permeability) of all peat in the profile. However, his models do not account for the significantvertical variations in conductivity that are inherent in deposits fitting the Anderson model (as describedin detail by Esterle, 1990), despite his recognition that vegetation type has an effect on hydraulicconductivity. In citing the case of a German bog in which later peat accumulation rates increased overearly rates, he ascribes it to the effects of ongoing anaerobic decay at depth which creates an apparentincrease in accumulation rate with time, without accommodating the implied increase in density andthus decrease in permeability. In situ accumulation rates mean little without factoring-in densityvariations. It may be that stacked hydraulic models can accommodate density changes, whether theybe due to autocompaction, ongoing anaerobic decay, or floral succession. In the case of theChanguinola deposit, the profile is felt to be the result of ponding within a basin of low-permeabilityhemic and fine hemic peat.The densest peats sampled, mangrove and forest peats, accumulate at rates comparable toAnderson’s late-stage peat (mangrove peat at ca. 2.25 mm/a and back mangrove peat at ca. 2.52 mm/a(this study), vs. 2.22 mm/a of Anderson, 1983). The very high in situ rate of accumulation of sedgepeats reflects lack of compaction. The domed central bog plain is essentially a mass of well-preserved root material floating in an acidic, nutrient-poor bath, contained within a basin of dense,woody peat of extremely low permeability. Discharge from the central mire is by surficial drainage,subsurface flow being effectively dammed by the surrounding forest-swamp peat, and66evapotranspiration is limited by the stunted vegetation. Only toward the steeper NE margin does thesurface dry out and water table drop to 10-20 cm below the peat surface. The disposition of hemicand fine hemic peat at the base and margins of the deposits, which form a type of internal boundingsurface, suggests that these dense, low permeability peats form a bowl within which the floating bog-plain has developed. The rate of accumulation of these peats is in inverse proportion to peat density,and suggests a process in which the variable hydrological characteristics of the peat dictate the heightof the water table, and hence the surface topography. The differential in accumulation rates thatallows for the evolution of a lip’ of dense, low-permeability peat is a function of the difference inbiomass production of forest peat vs sedge peat - i.e. it is created by increasing oligotrophy, asAnderson concluded. Increased degradation due to occasional drying-out of the surface serves toincrease density, and decrease permeability, but does not appear to be enough to counterbalance thehigh rate of biomass production. This is undoubtedly due to the high rainfall and relatively smallfluctuations in water table.b) Eastern Section: — In the western part of the deposit, the major transitions in peat characteristicsare the stratigraphic boundaries created by the upward evolution of the mire. Sea level, which almostbisects the western cross section of the deposit, does not appear to form an hydraulic boundingsurface, as there is no indication of saline influence at the base of the peat. In contrast, the complexhydrology of the eastern section has created a complex stratigraphy, over which in places atransgressive marine signature is superimposed. The overriding hydrological control throughout is theamount, and trend, of punctuated subsidence. The complexity is evident in palynological analyses, butvery much generalized in analysis based on degree of humification at the resolution attempted in thisstudy (Fig. 2-16). Plots of changes in humification, and of variations in the geochemicalcharacteristics, allow the construction of a generalized stratigraphy in which controlling trends areevident (Figs. 2-9 and 2-15).67There are 4 major tidal blackwater creeks draining the eastern section of the mire intoAlmirante Bay (two have been modified by dredging, and one of those now forms part of a canal).The mire is discharging down-dip, the streams flowing parallel to the barrier in the direction ofmaximum subsidence. All are long-lived, extend right to the base of the peat, have sandy or siltysediments at the channel floors, and have a stratified salinity profile, with fresh water at the surfaceand salinities of 19 to 23 ppt in the bottom waters. Discharge rates fluctuate considerably betweendrier and wetter periods (Fig. 2-5). A few smaller creeks are eroded into the peat, but the majorchannels effectively divide the eastern section of the deposit into hydrologically distinct sectorsbounded by channel-margin peats with a distinctive geochemical signature which has low salinity andvery high (5 to 14 wt%) sulphur. Each of these sectors has its own discharge pattern. SPOT satelliteimagery reveals small domes in 3 of the eastern sectors, based on vegetation zoning and surface waterdrainage patterns (Fig. 2-7). There are two possible explanations for these domes: firstly, they may beincipient bog plains experiencing elevation and increasing oligotrophy; secondly, they may beremnants of a more widespread bog plain which is ‘deflating’ due to incision by blackwater creeks(Fig. 2-17). Present surface vegetation zoning, which is roughly concentric about the domes, suggeststhe former. However, coring has revealed fibric peat at sites in the east which are now heavilyforested, but may at one time have borne a stunted vegetation. Spacing of coring sites is inadequate tomake a final judgement. The implications are profound; either the mire is evolving toward a moreextensive bog plain, or it is drowning.The high rate of basement subsidence along the eastern margin of the mire has resulted in asecond form of hydrological boundary within the eastern section of the deposit, the sea level boundary.This boundary is diachronous, trending approximately at right angles to the above-mentioned drainage68Freshwater palm-swamp and forest swamp peat onalluvial plain, and In swale on barrier.Drainage of the deposit Is represented by the symbols:,= Surface drainage(thick arrow= high flow or channelized flow)• = Flow out of deposit (toward viewer)+ = Flow Into deposit (away from viewer)Mangrove and back-mangrove peataround tidal channels;Mixed forest-swamp peat expands overalluvial and swale peats;Barrier agarades and progrades.Continued smallsubsidence eventsBog-plain develops on partial hydrologicalbounding surfaces of dense peatAggmdahon and progradation of barrier and peatdome continue.Figure 2-17. Proposed model ofpeat development on the barrier coast as a result ofperiodic punctuatedsubsidence. Black shading indicates mangrove and back-mangrove peats. Solid greys are forest-swamp and palniswamp peats. Patterned areas are fibric sawgrass/stunted forest-swamp and bog plain peats.r\jsubsidence —\ 2\..fl% 4__.mj•_•4—.-- periodic subsidence events3.•-- periodic subsidence events —-River channel shifts due to subsidence;Peat accumulates faster than subsidenceForest swamp expands laterally andvertically with rising water table;Barrier aggrades and progrades.Sawgrass - stunted forest-swamp peataccumulates with increasing oligotrophyIn central area;Barrier aggrades and progrades, followedseaward by forest-swamp.1k 1k 1k 1k 111 1k1k 1k 1k III 1k * 1k 1k 1k 1k,-1kW 1111k 1k_1l—Largersubsidence eventSea water Intrudes Into deposit via blaclcwater creeks -back mangrove and mangrove peat accumulates;More direct drainage into the sea leads to higherdlschai a via creeks, and partial collapse of perchedwater table In dome;I Part of deposit drowns;ork1k1k1k1k69channels and normal to the trend of the outer coastline, defined by the northwest-migrating margin ofAlmirante Bay. This boundary steps eastward with each subsidence event, driving biological andgeochemical responses within the mire system in time with the tectonic clock, and thus can be expectedto overprint those areas susceptible to flooding with its transgressive signature, and drive the stratified,brackish headwaters of the blackwater creeks deeper into the deposit.The highest sulphur content (>5 wt%) is found in peats proximal to the brackish influence ofblackwater drainage channels up to several km from the coast. Very high (-‘14 wt%) sulphur was alsofound in the basal peats in the deepest parts of the deposit (Cohen et al., 1990), which are thusinterpreted to have been associated with earlier channels. Very high sulphur content is found evenwhere salinity is undetectable in peat, porewater or basal sediments, and despite the low pH of thesepeats which inhibits bacterial sulphate reduction (Fig. 4-6). Peats associated with these channels arelow in salinity (< 0.5 wt%) and very high in S (5 to -‘14 wt% S), around 90% of which is stablecarbon-bonded suiphides, apparently the result of a biogeochemical chain of S reactions leading to theconcentration of carbon-bonded sulphur forms (Chapter 4). The peats are formed in a back-mangroveswamp, from a complex vegetative community that includes salt-tolerant (Raphia, Laguncularia,Acrostichum) and non-salt tolerant tree species (including Symphonia globuhjèra, Campnospermapanamensis and flex gulanensis) and shrubs (Miconia curWpetulata, Cespedezia niacrophylla andOuratea sp.). Sedges, grasses and ferns are common. The resultant peats are medium- to fine-granularhemic to fine hemic), consisting almost entirely of degraded wood, with occasional wood fragmentsand roots. They are lighter in colour, and slightly less dense and higher in moisture content thanRhizophora peat.702.10.3: IMPLICATIONS FOR COAL GEOLOGYa) Transgressive and regressive signatures, and the relation ofpeat deposition to active ciasticsedimentationCoal geologists often assume that low-lying coastal peat deposits must be highly susceptibleto storm generated clastic ‘contamination’, as is common today in North America. The inter-tropicalconvergence zone, which is free of cyclonic storms, is relatively narrow, but hurricane-free coastlinesmay well have been more extensive in the past, with a different distribution of continental masses andoceanic-atmospheric circulation. Hence the development of extensive wave- rather than storm-dominated coasts should not be discounted. Such coastlines are ideal for the formation of low ash,low sulphur strand-plain coals immediately adjacent to environments of active clastic deposition.In the Changuinola deposit thick, low-ash peat is accumulating on a rapidly subsidingcoastline within less than a kilometre of both a flood-prone, sediment-laden river, and an activelyprograding and aggrading barrier beach. Both climate and tectonics find eloquent and subtleexpression in this seemingly unlikely combination. Estimates vary, but most agree that compactionratios of 7:1 to 10:1 are not unreasonable for peat to bituminous coal (Ryer and Langer, 1980); thefactor may be less for dense woody peat (< 85% moisture), and considerably more for very wet (>90% moisture) peat. The Changuinola deposit, preserved in its present configuration, potentiallyrepresents a single coal bed some 8 km in width and 12 km in length, oriented NW-SE with the longaxis parallel to the coastline. On the northwest and southwest margins, the hypothetical bedinterfingers with alluvium and clays, and thickens quickly from carbonaceous sediments to 60 to 100cm of low ash coal (700 cm of peat), over a distance of 500 m (Figs. 2-9 and 2-12). To the northeast,behind the shoreline, the bed tapers-out more gradually, over 2500 m, cut by occasional channels. Tothe southeast, with more complex geometry, the bed tapers out beneath shallow marine carbonates,71again over about 2500 m, and in places ends abruptly in vertically walled channels, 3 to 10 rn wideand up to 3 km in length, oriented roughly parallel to the barrier sand body.The cessation of peat accumulation, and the burial and preservation of the deposit, couldresult from both tectonic and climatic causes. The observed trend of drowning and burial beneathshallow marine sediments in the southeast, as a result of an increase in the net rate of subsidence, mayescalate until mangrove growth is impossible. This has occurred already in the southeast, as shown bythe presence of drowned peat surfaces several kilometres away in Almirante Bay, and the focus of peatdeposition has migrated northwest with ongoing subsidence. Thus a transgressive termination of peatdeposition is considered the most likely end. A second possible cause is a change toward a drier andmore seasonal climate, which could lead to increased degradation and partial ‘deflation’ of the deposit,and also to an increase in siliciclastic sediment influx (Cecil et al., 1993). This may eventually resultin aggradation of the coastal plain and burial of the deposit beneath fluvial sediments.At present, the Caribbean coastline southeast of Puerto Limén is in an overall regressiveregime, despite the fact that it is subsiding rather than being uplifted, from about the Sixaola River toPunta Serrabata (Collins et al., 1994; Miyamura, 1975). The northwest section of the peat depositbears what, in terms of coal geology, would be considered a regressive signature. Well-sorted barrierbeachsands are aggrading in response to subsidence and prograding seaward over shallow marinecarbonates and silts of the continental shelf. The peat behind the barrier sands is prograding alongwith the barrier, is aggrading as the water table is elevated by internal hydrological processes, and issuccessionally zoned both vertically and horizontally. Cores from near the centre of the deposit carryno indications of transgression throughout perhaps 4000 years of peat deposition. A few km to thesoutheast, and in places at the base of the deposit, back-mangrove peat is overlain by a succession offreshwater forest-swamp peats. Coal resulting from this part of the deposit would be low in sulphur72and ash, and would be judged regressive from in-seam evidence. Evidence of the influence of rapidlyrising sea-level lies in the aggradation and the somewhat ambiguous bedding structures of the barriersand body, and in the presence of large numbers of buried logs and stumps, some in growth position,in beach deposits behind the berm crest, rather than within the peat.To the southeast, marine transgression, marked by the superposition of seaward on landwarddeposits, has resulted in the burial of a large area of drowned peat under carbonates and silts. About40% of the peat deposit is below present sea level, despite having been more extensive during much ofits history. If preserved as coal, the environment would be interpreted as transgressive; i.e.the depositformed during tectonically-driven local transgression. The net rate of subsidence has determinedwhether the peat accumulation rate has kept ahead of or fallen behind relative sea level rise, which isgreatest to the southeast. Hence the transgression is proceeding from southeast to northwest, parallelto the trend of the coast and the long axes of both the peat deposit and the major sand bodies. Internalexpressions of the transgressive signature include the presence of shelly layers within cores ofotherwise carbonate-free and relatively low-ash mangrove and back-mangrove peats, and indicationsof a redistribution of sulphur forms that occurs as sea water periodically inundates previously freshand brackish marginal pests (Chapter 5). At the flooding surface, and along the margins of thebrackish drainage channels, the peat is coherent and resists erosion. Were it preserved, the coal roofwould be in the form of an irregular xylic (peat) hardground with abundant teredo-bored logs (theteredolites ichnofacies; Savrda, 1991) beneath fine grained carbonates and corals.There remains some question as to how profoundly transgressing marine waters canoverprint freshwater peats with a marine signature. In Chapter 5 the short-term effects of earthquake-induced flooding (Phillips et al., 1994), and in Chapter 4, the distribution of total sulphur and forms ofsulphur in the marine- and brackish-influenced peats (Phillips and Bustin, in press) are examined. The73difficulty of distinguishing primary sulphur from secondary, ‘transgressive’ sulphur in mangrove andback-mangrove peats has been approached but not yet fully resolved. In most samples described inthose studies marine penetration was limited to 1 to 2 rn, and sulphur content did not increase withflooding. However, it is difficult to extend short term observations of transgressed peats totransgressive coals. It is likely that the rate of subsidence and subsequent sedimentation may affectthe degree to which marine waters ultimately affect the coal. The short term observations, however,point to the importance of the progenitor vegetation, which determines peat density and permeability,in admitting or limiting marine influence.Lithotype variations in a hypothetical coal bed would preserve a record of syndepositionalvariations in degree of huniification, and thus reflect variations in vegetation, peat density and hencethe hydrology and trophic state of the progenitor mire. Highly humified woody peats from the marginsand base of the central deposit would yield massive, dull, fine-grained coal, high in huminite. Fibricand coarse hemic sedge peats from the central bog plain and lesser domes in the southeast would formfinely banded bright coal with abundant structured macerals. Because of the variations in peat densitybetween forest peats and sedge peats it is likely that upon burial and dehydration the relativeproportions of woody coal over herbaceous coal would shift in the direction of a relatively greatervolume of woody coal. The large volume of herbaceous bog plain peat would become a relativelysmall volume of bright banded coal, fulfilling observations of Phillips and DiMichele (1981) that mosteconomic coals have formed from large woody vegetation.2.11 CONCLUSIONSThe Anderson model of floral succession and increasing oligotrophy with increasing peataccumulation appears to account for the evolution of the Changuinola peat deposit, with twoimportant differences related to the depositional environment which manifest themselves in the74hydrological controls on peat accumulation. The environmental factors that distinguish this depositfrom the Malaysian and Indonesian prototypes are: i) tectonically-driven punctuated subsidence, thatresults in periodic rapid rise in relative sea level rather than gradual eustatic rise; and ii) absence ofadry season and hence of the regular drying-out of the peat surface which strongly influences thesoutheast Asian bog forest peats. Thus climate and tectonic setting result in a variant of the model thatmay be applied to the interpretation of environments of coal deposition.On the Panama coast thick, low-ash peats have accumulated in an environment in whichsubsidence is earthquake-driven, and occurs instantaneously and incrementally. The net rate ofpunctuated subsidence has been very close to peat accumulation rates during the last 2000 years ofpeat deposition. The amount of subsidence in the most recent event was greatest to the southeast, anarea where the peat deposit has drowned beneath marine waters. However, although approximately40% ofthe deposit is below sea level, marine influence is not apparent in most of the deposit. Rather,it is restricted to that part of the deposit in the southeast where the subsidence rate is greatest and theswamp has drowned due to the inability of the mangrove vegetation to keep-up with the subsidence, orwhere the swamp is influenced by tidal blackwater creeks.The western part of the deposit is a domed ombrogenous bog with an elevated water tablewhich displays no internal evidence of the influence of punctuated coastal subsidence. In the ever-wetclimate the bog plain surface never dries out, but is commonly submerged beneath 10-20 cm of water.Discharge from the central bog is restricted to surface flow, probably as a result of low hydraulicconductivity in the dense marginal forest peals. Underlying the bog plain are stratified palm-swampand forest-swamp peals which act as low-permeability hydrological bounding surfaces. Rainwaterdrains slowly in a radial pattern, and the resulting oligotrophic bog produces a body of peat which iseffectively a suspension of well-preserved root material in a dilute granular matrix. Only in the clastic75sediments of the base and the barrier, and in the eastern, transgressed peats, is a record of coseismicsubsidence preserved. Thus it is evident that thick peat, and hence coal, deposits can accumulate dueto tectonically driven punctuated subsidence, rather than gradual eustatic sea level rise, withoutleaving a record of increased elastic input within the peat, even immediately adjacent to environmentsof active elastic deposition.The development of partial hydrological bounding surfaces is an effect of changes invegetation and peat type through time, and an indication of changes in the trophic state of the miredriven by internal hydrological controls. Rapid punctuated subsidence imposes an externalhydrological control, that of rising sea level. At low net subsidence rates, vegetation responds; dense,low-permeability mangrove and back-mangrove peats accumulate, and marine influence is excludedfrom all but a narrow marine margin, and areas in close proximity to brackish drainage channels. Inthese laterally restricted parts of the Changuinola deposit the degree of humification of peat isuniformly high, and the marine hydrological influence is in the form of a front which migrates to thenorthwest with each subsidence event. At higher net subsidence rates, it is likely that increaseddischarge into blackwater creeks would lead to a lowering of the peat surface and eventual drowningof the deposit and burial beneath shallow marine carbonates. A small reduction in the net rate ofsubsidence of the bottom of Chiriqui Lagoon and Almirante Bay could result in the infilling of theselarge areas by extensive peat swamps, starting with mangrove islands.ACKNOWLEDGEMENTSThe research was supported by NSERC PGS3 fellowship and IDRC Young CanadianResearch Grant 92-1201-18 (Phillips) and NSERC grant 5-87337 (Bustin) and The University ofBritish Columbia. The Smithsonian Tropical Research Institute provided laboratory and logisticalsupport, as well as expertise and encouragement, as did IRHE (the Instituto de Recursos Hidraulicos yElectrificacion) of the Government of Panama. The work would not have been possible without theunhesitating support of the Chiriqui Land Company. The authors would like to express their gratitudeto Eduardo Reyes, Andres Hernandez, Enrique Moreno, Paul Colinvaux, the Serracin family, theSanchez family, Bibi, Lulu, and Sammy Sanchez.762.12 REFERENCES CITEDBerger, R., 1983. Sea levels and tree-ring calibrated radio-carbon dates. In: P.M. Masters and M.C.Flemming (eds.), Quaternary Coastlines and Marine Archaeology. Academic Press, London.p.51-61.Anderson, J.A.R., 1964. The structure and development of the peat swamps of Sarawak and Brunei.Journal of Tropical Geography, v.18, Singapore, p.7-16.Anderson, J.A.R., 1983. The tropical peat swamps of Western Malesia. In: A.J.P. Gore (ed.),Ecosystems of the World 4B - Mires: Swamp, Bog, Fen and Moor; Chapter 6. Elsevier,Amsterdam.Anderson, J.A.R. and Muller, J., 1975. Palynological study of a Holocene peat and a Miocene coaldeposit from NW Borneo. Review of Paleobotany and Palynology, v.19, p.29 1-351.Astorga, G.A., 1991. Informe tecnico sobre el levantamiento de la costa Caribe de Costa Rica, comeconsecuencia del terremoto del 22 de Abril de 1991. University of Costa Rica.Bartlett, A.S. and Barghoom, E.S., 1973. Phytogeographic history of the Isthmus of Panama duringthe past 12,000 years (A history of vegetation, climate and sea-level change), In: Graham,Alan, 1973. Vegetation and Vegetational History of Northern Latin America. Elsevier,Amsterdam, 1973,p.203-3 1.Bobnenberger, O.H. and Dengo, G., 1978. Coal resources in Central America. In: F.E. Kottlowski,A.T. Cross and A.A. Meyerhoff, eds., Coal Resources of the Americas, G.S.A. SpecialPaper 179, 1978,p.65-72.Bruenig, E.F., 1990. Oligotrophic forested wetlands in Borneo. In: A.E. Lugo, M. Brinson and S.Brown, (eds.), Ecosystems of the World 15: Forested Wetlands, Elsevier 1990, Chapter 13.Bush, M.B., Pipemo, D.R., Colinvaux, P.A., De Oliveira, P.E. Krissek, L.A., Miller, M.C. andRowe, W.E., 1992. A 14,300-yr palaeoecological profile of a lowland tropical lake inPanama. Ecological Monographs 62(2), p.25 1-275.Camacho, E. and Viquez, V., 1992. Historical and instrumental seismicity of the Caribbean coast ofPanama. U. of Panama. In press.Camacho, E., Viquez, V. and Espinosa, A., 1992. Ground deformation and liquefaction distributionin the Panama Caribbean coastal region. University of Panama, in preparation.Cecil, C.B., Dulong, F.T., Cobb, J.C., and Supardi, 1993. Allogenic and autogenic controls onsedimentation in the central Sumatra Basin as an analogue for Pennsylvanian coal-bearingstrata in the Appalachian basin. In: Cobb, J.C. and Cecil, C.B., 1993. (Eds.) Modem andAncient Coal-forming environments. G.S.A. Special Paper 286, Boulder, p.3-22.Clowes, R.M., 1987. Geophysical reconnaissance study of the continental margin of Costa Rica forpetroleum exploration. UBC-RECOPE Project Final Report, The University of BritishColumbia Dept. of Geophysics and Astronomy 1987. pp.287.77Clymo, R.S., 1983. Peat. In: A.J.P. Gore (ed), Ecosystems of tlie World Vol.4A - Mires: Swamp,Bog, Fen and Moor. Elsevier, Amsterdam, p.159-224.Coates, A.G., and Obando, J., in press (1994). The geological evolution of the Central Americanisthmus. Smithsonian Tropical Research Institute, Panama. in Mann. P. ed.. Geologic andtectonic development of the Caribbean plate boundary in southern Central America.Geological Society of America Special Paper.Cobb, J.C. and Cecil, C.B., 1993. (Eds.) Modern and Ancient Coal-forming environments. G.S.A.Special Paper 286, Boulder. pp.198.Cohen, A.D., 1968. The petrology of some peats of southern Florida (with special reference to theorigin of coal) 3 vols. Unpublished PhD thesis, The Pennsylvania State University, 1968.pp.351.Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce, F., 1990. Changuinola PeatDeposit ofNorthwestern Panama, 3 vols. Los Alamos National Laboratory pub. LA-i 1211,July 1990.Collins, L.C., Coates, A.G. and Obamdo, J., 1994. Rates of uplift of the Burica and southern LimónBasins; Caribbean effects of Cocos subduction? In: Mann, P. ed., Geologic and tectonicdevelopment ofthe Caribbean plate boundary in southern Central America. GeologicalSociety ofAmerica Special Paper.Davis, A., Russel, S.J., Rinimer, S.M. and Yeakel, J.D., 1984. Some genetic implications of silicaand aluminosilicates in peat and coal. hit. J. Coal Geology, v.3, p.293-314.Denyer, P., Personius, S.F., Arias, 0., Rojas, W. and Alvarado, J., 1992. Tectonic effects of theearthquake of April 22, (1991). In prep.Diessel, C.F.K., 1982. An appraisal of coal facies based on maceral characteristics. In: C.W. Mallet,ed., Coal Resources: Origin, Exploration and Utilization in Australia. Proc. Symp. Geol.Soc. Aust. Inc. Coal Group, p.474-483.Diessel, C.F.K., 1992. Coal-bearing Depositional Systems. Springer-Verlag, Berlin. pp.721.Escalante, G., 1990. The geology of southern Central America and western Columbia, In: Dengo, G.and Case, J.E., eds., The Caribbean Region: Boulder, Colorado, Geological Society ofAmerica, The Geology of North America, v.H, p.20 I-230.Esterle, J.S., 1990. Trends in petrographic and chemical characteristics of tropical domed peatdeposits in Indonesia and Malaysia as analogues for coal formation. Unpublished PhDthesis, University of Kentucky, Lexington, pp.270.Esterle, J.S., Moore, T.A. and Hower, J.C., 1990. A modified method for the microscopicinvestigation of peat. Technical Report, Inst, for Mining and Minerals Research, Univ. ofKentucky.Fisher, S.P. and Pessagno, E.A., 1965. Upper Cretaceous strata of northwestern Panama. AmericanAssociation of Petroleum Geologists Bulletin, v.49, p.333-344.78Friedman, G.M., Sanders, J.E. and Kopaska-Merkel, D.C., 1992. Principles of Sedimentary Deposits:Stratigraphy and Sedimentology, Macmillan, New York. pp.717.Gonzalez-Viquez, C., 1910. Temblores, terremotos, inundaciones y erupciones volcanicas en CostaRica, 1608-1910. Tipografia de Avelino Alsina, San Jose, Costa Rica. 1910. pp.200.Guendel, F., and others, (En preparacion, 1991). Informe preliminar sobre los aspectos sismologicosdel terremoto del Valle de la Estrella Ms=7.4 del 22 de Abril de 1991.Home, J.C. and Ferm, J.C., 1978. Carboniferous depositional environments in the Pocahontas Basin,eastern Kentucky and southern West Virginia: a field guide. Dept. of Geology, University ofSouth Carolina, Columbia.IRHE, 1988. Atlas Nacional de la Republica de Panama, Instituto de Recursos Hidraulicos yElectrificacion (IRHE), Depto. de Hidrometeorologia, Government of the Republic ofPanama.Jarret, P.M., 1983. Ed., Testing of Peats and Organic Soils, American Society for Testing andMaterials ASTM STP 820, Philadelphia, Pa., 1983, pp.24 1.Lamberson, M.N., Bustin, R.M. and Kalkreuth, W., 1991. Lithotype (maceral) composition andvariation as correlated with palaeo-wetland environments, Gates Formation, northeasternBritish Columbia, Canada. mt. J. Coal Geol., 18, p 87-124.McCabe, P.J., 1987. Facies studies of coal and coal-bearing strata. In: Scott, A.C., (ed.), 1987,Coal and Coal-bearing Strata: Recent Advances. Geological Society of London SpecialPublication No. 32,p.51-66.Miyamura, S., 1975. Recent crustal movements in Costa Rica disclosed by relevelling surveys. In:N. Pavoni and R. Green, eds., Recent Crustal Movements. Tectonophysics 29, (1-4), p.191-198.Myers, R.L., 1990. Palm swamps. In: A.E. Lugo, M. Brinson and S. Brown, (eds.), Ecosystems ofthe World 15: Forested Wetlands, Chapter 11, p.267-286. Elsevier, Amsterdam.Obaje, N.G., Ligouis, B. and Abaa, S.I., 1994. Petrographic composition and depositionalenvironments of Cretaceous coals and coal measures in the Middle Benue Trough ofNigeria. mt. J. Coal Geol., 26,p.233-60.OVSICORA-UNI, 1991. Preliminary Report on the April 22 1991 Ms=7.4 Valle de Ia Estrellaearthquake. National University of Costa Rica, San Jose.Peltier, W.R., 1988. Lithospheric thickness, Antarctic deglaciation history, and ocean basindiscretization effects in a global model of post-glacial sea level change: a summary of somesources of non-uniqueness. Quatemary Research, v.29, p.93-112.Phillips, S., Bustin, R. M. and Lowe, L. E., 1994. Earthquake-induced flooding of a tropical coastalpeat swamp: a modem analogue for high-sulphur coals? Geology, v. 22, No.10. p.929-932.79Phillips, S. and Bustin, R. M., (in review) Tectonic, climatic and biological controls on sulphurdistribution in a coastal tropical peat deposit: implications for environmental interpretationof coal deposits. Journal of Sedimentary Research.Phillips, S., Rouse, G.E. and Bustin, R.M., (in review). Vegetation and diagnostic palynology of atropical Caribbean peatland, Changuinola, Panama. Palaeo-3.Phillips, T.L. and DeMichele, W.A., 1981. Paleoecology of a Middle Pennsylvanian age coal swampin southern Illinois: Herrin Coal Member at Sahara Mine No.6. In: K.J. Niklas, ed.,Paleobotany, Paleoecology, and Evolution. v.1, Praeger, New York. p.233-297.Pirazzoli, P.A., 1991. World Atlas of Holocene Sea Level Changes. Elsevier Oceanography Series, v.58. Amsterdam. pp.300.Plaficer, G. and Ward, S.N., 1992. Backarc thrust faulting and tectonic uplift along the Caribbeancoast during the April 22, 1991 Costa Rica earthquake. Tectonics, v.11, No.4, p.709-718.Reinson, G.E., 1984. Barrier island and associated strand-plain systems: In: R.G. Walker (ed),Facies Models, Geoscience Canada Reprint Series 1, p.1 19-140.Roberts, O.W., 1827. Narrative of Voyages and Excursions ion the East Coast and in the Interior ofCentral America. University of Florida Press, Gainesville, Fla., 1965. pp.31 1.Ryer, T.A and Langer, A.W., 1980. Thickness changes involved in the peat to coal transformation fora bituminous coal of Cretaceous age in central Utah. Journal of Sedimentary Petrology,v.50,p.987-92.Savrda, C.E., 1991. Teredolites, wood substrates, and sea-level dynamics. Geology, v.19,p.905-8.Shearer, J.C., 1994. Thoughts on the origin of inertinite-rich coals. TSOP Newsletter, Vol.11, No.3/4.Sherstobitoff, J., 1991. Talamanca earthquake - Costa Rica, April 1991. BC Professional Engineer,Nov. 1991, p.5-10.Staneck, W. and Sue, T., 1977. Comparisons of four methods for determination of degree ofhumification (decomposition) with emphasis on the von Post method. Canadian Journal ofSoil Science, v. 57,p.109-17.Tabatabai, M.A., 1992. Methods of measurement of sulphur in soils. In: Howarth, R.W., Stewart,J.W.B. and Ivanov, M.V., eds. 1992. Sulphur Cycling on the Continents; Wetlands,Terrestrial Ecosystems and Associated Water Bodies. SCOPE 48. Wiley and Sons,Chichester, p307-344.von Post, L., 1922. Sveriges geologiska undersoknings torvinventering och nâgre av des hittills vunnaresultat. Sv. MosskulturfOr. Tidskr. 1:1-27.Wanless, H.R., Barofflo, J.R. and Trescott, P.C., 1969. Conditions of deposition of Pennsylvaniancoal beds. In: Dapples, E.C. and Hopkins, M.E., 1969. eds., Environments of CoalDeposition. G.S.A. Special Paper 114, Boulder, Colorado, 1969. p.105-4280Weyl, R., 1980. Geology of Central America, 2nd edition. Berlin, Gebnider Borntraeger. pp.371.Winston, R.B., 1994. Models of the geomorphology, hydrology and development of domed peatbodies. G.S.A. Bulletin, v.106, p.1594-1604.Woodroffe, C.D., 1988. Mangroves and sedimentation in reef environments: indicators of past sealevel changes, and present sea level trends? Proc. 6th mt. Coral Reef Symp., Australia. v.3,p.535-9.81CHAPTER 3VEGETATION ZONES AND DIAGNOSTIC POLLEN PROFILES OF A COASTAL PEATSWAMP, BOCAS DEL TORO, PANAMA82CHAPTER 3: VEGETATION ZONES AND DIAGNOSTIC POLLEN PROFILES OF A COASTALPEAT SWAMP, BOCAS DEL TORO, PANAMA3.1 ABSTRACTA survey of the dominant vegetative cover of a large domed coastal swamp near Changuinola inthe Province of Bocas del Toro, Panama, has been undertaken as an initial step in reconstructing theHolocene history of peat accumulation on this coast. Seven phasic communities of peat-formingvegetation are defmed and mapped: 1) Rhizophora mangle mangrove swamp; 2) mixed back-mangroveswamp; 3) Raphia taedigera palm swamp; 4) mixed forest-swamp; 5) Campnosperma panamensisforest-swamp; 6) Sawgrass ± stunted forest-swamp; 7) Myrica-Cyrilla bog-plain. Pollen extracted fromsurface peat samples and collected from dominant vegetation at representative sites is used to prepare apollen profile of each phasic community. These profiles are then compared to pollen distribution in 2peat cores, one from the deep central part of the deposit and the second from a site near the marinemargin, in order to construct a history, by floral succession, of the 4000 year evolution of the deposit.The Changuinola mire originated as freshwater palm swamps that developed in close proximity toboth the Changuinola River mouth, probably behind a barrier bar and freshwater lagoon system, and alow energy, mangrove-dominated bay. The early swamp was likely drained to the southeast by brackishblackwater creeks much as it is today, and formerly extended considerably farther in the direction ofAlmirante Bay. The Raphia swamp was succeeded by hardwood forest-swamp dominated by a verylimited number of specialised species, only one of which (Campnosperma panamensis), is prone toforming monospecific stands. Increasing accumulation of woody peat promoted by the everwet climateimpeded drainage of the mire, leading to doming, increased oligotrophy, and establishment of bog-plain83conditions in the oldest, central regions of the mire. Mire development did not require the initial mangrovephase which is common to the peat swamps of southeast Asia, as the palm Rciphia tciedigera is able tocolonize and institute peat accumulation in a variety of freshwater and brackish environments.843.2 INTRODUCTIONPeat-forming mires are dynamic organo-chemical systems which evolve through time in concertwith changing conditions of growth and preservation. Stratification within Holocene peat deposits hasbeen used to construct histories of environmental change over periods of several thousand years (e.g.Barber 1981), and the stratigraphy of coal measures is used to reconstruct environmental changes on ascale of hundreds of thousands to millions of years (volumes edited by Scott, 1987; Lyons and Alpem1989; others). Thick extensive peat deposits may develop wherever precipitation exceedsevapotranspiration throughout most of the year, allowing vegetated wetlands with high water tables topersist year-round. Today the most extensive, and most studied, peatlands are found in the middle andhigh latitudes of the northern hemisphere. These formed as an effect of continental glaciation, and may,in fact, be moribund and in retreat. In the geological past, however, the large coal deposits of theCarboniferous, as well as some Cretaceous and Tertiary coals, bear witness to the existence in earliertimes of vast, long-lived forested peatlands in tropical and subtropical climes.Recently, attention has turned to the study of Holocene tropical peat swamps, in part due toincreasing awareness of the importance of, and growing threat to, tropical forest ecosystems, and partlyas a result of the emergence of a process model for deposition of coals derived from primarily arborescentvegetation, in tropical or sub-tropical conditions (Anderson and Muller, 1975; Esterle et al., 1989; Cobband Cecil, 1994). Pioneering work on the evolution of forest-swamps in southeast Asia by J.A.R.Anderson has lead to the development of what is now termed the Anderson model of domed peatdevelopment in the Old World tropics, described in detail in Anderson’s own publications (Anderson1961,1964, 1983; Anderson and Muller 1975), and summarized and to some extent formalized byBruenig (1990) and Esterle and Ferm (1994).85Anderson’s work concentrates on the vegetation, soil chemistry, topography and stratigraphy ofdomed, alluvial forest-swamps of vast size, and peat thicknesses up to 17 m, which have developed in thelast 3000-4000 years along the lower reaches of the Baram and Rejang Rivers of northern Borneo.Anderson also describes the coastal swamps of peninsular Malaysia, and Sumatra, which are planarrather than domed. Anderson constructs a depositional model based on vegetational succession, in whicha colonizing halophyte (mangrove) community stabilizes a silty prograding shoreline, trapping sedimentsand creating a protected environment amongst the mass of mangrove roots. The new environmentbecomes available to a succession of increasingly less salt-tolerant plant species as the shoreline continuesto prograde, and a perched, non-saline, acidic water table develops within organic detritus accumulatingover mangrove clays. Critical to this process is an ever-wet climate, in which abundant (-.3000 mm/a)rainfall throughout the year maintains the water table and allows the accumulation of thick peat, thesurface ofwhich becomes increasingly elevated. As it builds above the reach ofnutrient-rich tidal andriver flood waters, the peat swamp takes the form of a flattened dome or inverted pie-plate, the centre ofwhich becomes increasingly oligotrophic. Shortage of nutrients and greater acidity result in a sequence ofincreasingly more specialized plant associations occupying the raised central portions of these domedmires. The vegetation develops in a series of stages, in which recognizable associations of plants, termed‘phasic conmiunities’, succeed each other according to the degree of oligotrophy and doming. The planarforest-swamps of peninsular Malaysia are less-developed examples, which are elevated to a lesser degree,and in which the final stage in the succession has not been reached.Anderson (1964) defines a lateral sequence of six identifiable forest types, or ‘phasiccommunities’, from the perimeter to the centre of the fully developed Marudi domed deposit on the BaramRiver delta in Sarawak. The surface expression of this succession can be clearly seen in aerialphotographs of fully developed domes (reproduced in Anderson, 1983) as a series of roughly concentricforested zones radiating out from a central bog-plain occupied by a stunted, low diversity community86representing the final phase in the model of dome development, his phasic communit (PC) 6. Anderson’sphasic communites PC5 through PCI represent distinctive associations of arboreal species, generally ofincreasing size, towards the margins of the deposit. The zones approximate topographic contours, withrapid transitions occurring on the relatively steep edges of the dome, where drainage is better and nutrientconcentrations are higher (Bruenig 1990, referencing unpublished work by Salfeld, 1974 and Waughman,1974), and where in some cases the peat is thinner.Palyno-stratigraphic analysis of peat cores by Muller (1963, 1964) and Anderson and Muller(1975) show that a vertical sequence of pollen facies is present from the base to the surface of the peat.Correlation of pollen in the peat cores to pollen of surface vegetation reveals that vertical transitionsgenerally echo the lateral sequence of phasic communities. From these observations a model ofvegetational succession, reflecting increasing oligotrophy through time, was constructed. The authorspoint out that palynological analysis of the Marudi peats has some limitations, principal among thembeing lack of information about the local floral biology, pollen production and dispersal mechanisms ofmany of the plants. In addition, pollen cannot generally be differentiated at the species level, and specificforest-types could not be distinguished due to the presence of several species of the genus Shorea, thepollen of which are indistinguishable. On the other hand, the principal tree species are believed to bemainly insect pollinated, and this in conjunction with the density of the foliage, and the generally weakwinds prevalent in that area result in restricted pollen transport. This factor reduces overlap betweenphasic communities, and makes correlation between stratification in the peat and zonation in the surfacevegetation more certain.Until very recently, overviews of world peat distribution have made little mention of extensivepeatlands in the Neo-tropics, and their existence was rarely mentioned in the literature (exceptions includeChristen, 1973; Cohen et al., 1985, 1989, 1990; bibliography in Lottes and Ziegler, 1994). Cohen et al.87(1989) describe the large coastal peat deposit near Changuinola in western Panama, studied jointly by theU.S. Los Alamos National Laboratories and the Instituto de Recursos Hidraulicos y Electrificacion of theRepublic of Panama with the objective of assessing its resource potential as fuel for a peat-firedthermoelectric facility. They found the peat to be up to 9 metres thick, and the dense forest vegetationsurrounding a plain of sedges, grasses and stunted vegetation suggested that the deposit might be domed.The present study presents an overview of the dominant peat-forming vegetation of the Changuinola mire,defining 7 phasic communities according to their principal species. Pollen ‘fingerprints’ (sensu Cohen,1975) for each phase are then constructed using counts of the dominant identifiable pollen typesconcentrated from surface litter, and surface (upper 25 cm) peat samples. Finally, the pollen stratigraphyof two peat cores is recorded, one from the thick central region and a second from near the eastern marginof the deposit. By correlating transitions in the cores with the modern vegetation, the Anderson model istested and a history of the evolution of the mire proposed.3.3 GEOGRAPHY AND CLIMATEThe Changuinola peat deposit has developed in a back-barrier setting on the Caribbean coast ofPanama, at9020’ N latitude, 82° 20’ W longitude, on a microtidal (- 30 cm), wave dominated shoreline(Fig. 3.1). Although tidal effects in the western Caribbean are minimal, strong wind-driven longshorecurrents transport sediments to the southeast. Tropical storms and hurricanes pass well to the north, butassociated floods are frequent. The climate is humid-tropical and temperature averages 26°C, with a ± 30range. There is no dry season; the annual precipitation of 3000 mm is uniformly distributed throughoutthe year (IRHE, 1988). SurfIce winds are predominantly from the north to north-east throughout theyear, but strong south- westerlies blow down the slopes of the Talamancan Ranges one or two days ineach month. Normal windblown transport of pollen is from NE to SW, from the Caribbean barrier bartoward the centre of the mire. The NE margin is occupied by dense Raphia-swarnp and mixed forest-swamp. There is no mangrove fringe forest present along the windward margin of the mire.88Figure 3.1: Location map of the Changuinola peat deposit on the Caribbean coast of Panama.Sample sites which are mentioned in the text are shown, as are lines representing the NW-SEcross-section, and the NE-SW levelling traverse and cross-section. The shaded area includes thatpart of the deposit that has been sampled offshore of the mangrove fringe zone in Almirante BayEAN.c.iCOSTACU’ A••.....••)ocl- lainL41Rz*ç )‘V\44KR1o D8BDT14’-._...‘I. ) r D/ U/ILE 1.5 BDDBDD 258D029-- SOD 314Republic ofPanamaCHANCPOLLENCORESTFS* SURFACE POLLEN SiTESx * OThER SAMPL.E srrsEXTENT OF PEAT- -- iailwayALMIRANTE N O9 1BAYPONDSOCK89The NW coast of Panama is a region of active clastic and organic sediment deposition. Much ofthe coastal plain is covered by Quaternary and Recent alluvium, and the coast is characterized by a seriesof prograding barrier bars behind which occasional lagoons and extensive paralic swamps are developed.Sedimentation rates are high, and the peat deposit has developed adjacent to the Changuinola River, onthe back of a prograding barrier bar, the characteristic ridge-and-swale architecture of which can betraced at the base of the peat. (Fig. 3.2) The present-day mire occupies the narrow coastal plain betweenthe folded Miocene sediments of the Talamanca Range foothills and the exposed Caribbean coast. At theeastern terminus of this coastline, where the terrigenous sediments extend beneath Almirante Bay, peatoccurs submerged to depths of three metres or more, underlain by medium grained grey sand, and inplaces overlain by a thin layer of silty sands, lime mud, and coral. Approximately 40% of the deposit isbelow present sea level and the base of the peat is 670 cm below sea level at the lowest measured point.Most basal sediments are well-sorted medium grained grey sands similar to those currently beingdeposited on the coast by the Changuinola River.3.4 METhODSa) Problems:The coastal mires of Caribbean Central America are not well known. Access to most of the 1000km of coast between Cabo de Honduras in northeast Honduras, and Colon in eastern Panama is almostexclusively by boat and light aircraft. Botanical studies in the region are mostly limited to thoseconcerning the welfare ofplantation crops such as bananas and cacao, and thus have focused on thevegetation of the alluvial plains rather than that of the extensive and virtually impenetrable forest-swampsthat occupy much of the coastal plain. The Changuinola mire is used, but little known by the localinhabitants, who clear wet pastures and hunt for food around the margins, cut timber along the banks ofthe blackwater creeks which drain into Almirante Bay, and attempt the occasional rice crop between thestumps left by small scale logging. A systematic description of the vegetation of the area, indeed of the90Figure 3-2. NE-SW cross section of the deposit along levelling transect, showing surface vegetation zoned intoconcntric phasic communities (PCs). Vegetation zones correspond to those visible in satellite images (Fig. 3-3),verified by sampling. Vertical exaggeration is x 500.PC6PC4 PC7PC3 —:W10 km91province of Bocas del Toro, has yet to be attempted, although some reference material is available(notably by Hoidridge and Budowski, 1956, summarised in Porter, 1973). At the onset of the project,only local names were available for the majority of plants in the mire.b) Satellite ImageryThe mapping of vegetation zones in most of the 80 km2 Changuinola mire was accomplishedusing a combination of normal and oblique air photos (using black and white, normal colour and infraredfilms), and SPOT multispectral satellite imagery (Fig. 3.3; also Figs. 2.7 and 2.8). Detailed mapping ofdrainage channels and identification of many of the vegetation types required low-level aerialphotography, particularly in the complex eastern section of the deposit (see Fig. 2.3). The concentriczonation of the phasic communities in the raised western section did not emerge, however, until false-colour satellite images were generated. Using ‘groundtruth’ developed during the cutting of trails, and aMagellan® GPS (Global Positioning System) receiver for positioning, we have assigned 7 phasiccommunity descriptions to the categories in the digital images and photo-based maps.c) Levelling SurveysThree levelling surveys were conducted from the margins to the centre of the swamp in order todetermine if the surface is domed. A complication arose in that at the start of the study, in April 1991, aMs (surface magnitude) 7.5 earthquake, epicentred 90 km to the west in the Rio Estrella valley, resultedin approximately 50 cm of coseismic subsidence across the study area. Thus it was necessary to reestablish datum before levelling lines could be run.d) Coring ofPeat and Collection ofSurface SamplesCores were taken from the surface to the base of the peat at 78 sites. Samples were recovered in25 or 30 cm increments, wherever possible, using Hiller- or Macauley-type hand operated coring devices.92Figure 3.3. False colour SPOT satellite image of the Changuinola peat deposit. The black area (top and right) issilt-laden surface water in the Caribbean and Almirante Bay. The large white patch is clear shallow water in theBay. Areas of peat doming are in shades of blue and green (central western bog plain, and 4 smaller areas in theeastern section; PC 7). These bog-plains are surrounded by gold (sedge and stunted forest-swamp; PC 6). Theyellow areas are Raphia palm swamp (PC 3). Red and flecked white are mixed forest-swamp and Campno.sriermap. forest swamp (PC 4 and 5).93IIc)A‘ ‘‘1 % till$r ?\•:I_I,.-_...c-i i Oli OOOOOOO8’,s L—.. N ht,: v:.%w4i!.L”—.‘ / ‘—‘—-F N liFigure 3.4. Vegetation-zone map of the area shown in the SPOT image. Interpretation of the distinguishablePhasic Communities are shaded according to the legend.94Salinity and pH for each sample were measured at the time of collection, and verified in the laboratory(distilled water), using a Cardy® Model PH 1 digital pH metre, and a Cardy® Model Cl 21 digital saltmetre. In addition, surface litter samples were collected (to 5 cm depth) and 25 cm cubes cut with amachete, at one site in each of the 7 phasic communities. All samples were double-bagged and stored inas cool a location as possible until they could be refrigerated or frozen.e) Collection and Identf1cation ofPlantsLeaves and, where possible, flowers, of major species from each of the 7 phasic communitieswere collected with the assistance of Sr. Andres Hernandez, botanist with the Forest Dynamics Projectof the Smithsonian Tropical Research Institute (STRI), Barro Colorado Island, Panama. Arboreal pollenwas collected by Sr. Sammy Sanchez with a .22 calibre rifle. Identification was made by Sr. Hemandezand staffof the herbaria of the Smithsonian Tropical Research Institute in Balboa and the University ofPanama, Panama City. A list of the major species identified is included as Appendix F.J Peat Class/Ication, Pollen Preparation and Ident/1cationPollen slides were prepared from surface samples from sites representative of 6 of the 7 phasiccommunities, and from 2 cores, one in the central part of the deposit and the second near the easternmargin. Pollen was prepared and concentrated using standard palynological techniques from the finefraction of the peat (< 0.25 mm). Samples were first separated into coarse, medium and fine fractionsusing a wet-sieving procedure modified from Staneck and Silc (1977). By this method, the sample isthoroughly mixed before the pollen is extracted. This system was used in order to make a quantitativeassessment of the degree of humification of the peat. Degree of humification is traditionally describedusing a modified von Post Humification Scale, adapted to tropical peats by Esterle (1990). Thecategories are sapric, fine hemic, hemic, coarse hemic and fibric. The use of field-determined peat typeshas been used only sparingly in the study of these tropical peats, as recent work (Esterle, 1990) suggests95low correspondence between field categories and the actual particle-size distributions as determined bypoint counting or sieving methods. Peat classification is based on the identification of macroscopic plantparts and palynomorphs in the peat, compared to plant and pollen associations identified in the surfacesamples, and uses botanical (eg. Rhizophora peat, sedge peat) nomenclature (Cohen, 1968).Following identification of the dominant species present in the modem vegetation, descriptionsand photomicrographs of the pollen were prepared. Published descriptions for some Panamanian pollenspecies are available, notably in Bartlett and Barghoom (1973), Graham (1979, 1 988a and b) and in thesuperb publication of Roubik and Moreno (1991). Reference slides from the collections in thepaleoecology department of STRI, Balboa, were made avalable. Finally, as many gaps as possible werefilled using the herbarium collections at The University of Panama, and the Smithsonian TropicalResearch Institute, and by collection of fresh samples on-site in the mire. Descriptions of some species ofinterest, or for which no prior publication was found, are included here as Appendix E.g) Pollen Counts and Pollen DiagramsAs indicated by the title of this paper, the palynological analysis presented here is intended toestablish a stratigraphy of the deposit sufficient to reveal its broad evolutionary history. To accomplishthis it is necessary to relate the pollen preserved in the peat with modem plants, plant associations andcommunities. A complication encountered in attempting ecological interpretation from pollenstratigraphy is selective preservation. For example, the pollen of cerillo (Symphonia globulfera), adominant tree, was absent from surface samples and most of the cores. One can only count what ispreserved, and it is quite possible that what is preserved significantly misrepresents what actually was.To minimize this selective bias, pollen profiles were prepared that would be representative of thepreserved pollen record for each community. To establish the pollen ‘fingerprint’ for each phasiccommunity, a count of 1000 grains, where possible, was made from surface litter and surface peat96samples (3 sites produced less than 750 grains each from multiple slides). The results were combined,and the percentage of the total count plotted for the most numerous palynomorphs. Although it would bepreferable to compare counts from numerous similar sites to better represent each phasic community, itwas felt that including the upper 25 cm of the peat gave considerable range to the sample, and given thesampling interval in the cores (25 or 30 cm) finer discrimination was not necessary. No attempt is madeto relate the pollen found in the peat to the species which numerically dominate each community - theproblems associated with proportional representation of dominant plants in pollen profiles are wellknown. Rather it was hoped that recognizable ‘fingerprints’ would emerge which could then be related topollen recovered from cores.Counts of 200 grains for each 25-30 cm increment of two cores were used to construct the pollenstratigraphy of the central and eastern sections of the deposit (Appendix D). Pollen diagrams of the twocores were constructed using selected palynomorphs or combinations of palynomorphs that are consideredto be valid markers, as determined from the surface samples (fingerprints), and which best definetransitions in vegetation. In both surface and core samples, the following grains were omitted from thecounts: all single-celled fungal spores, all fungal hyphae, all fragments less than 2/3 complete, and anyambiguous palynomorph. In identifying palynomorphs, botanical names are used wherever possible.3.5 RESULTS AND INTERPRETATIONa) Vegetation DistributionAll of the vegetation described in this study is rooted in peat varying in depth from less than 1 m,in a narrow strip behind the barrier bar, to about 9 m in the central mire. Vegetation on mineral soiladjacent to the Changuinola mire complex is diverse and luxuriant. In a trial 20 m x 20 m quadrant (thebasic unit of a forest monitoring system developed by the Forest Dynamics Program at STRI), 164species of trees and shrubs > 2 cm dbh (diameter, breast height) were identified. This diversity is typical97of humid forested lowlands (A. Hernandez, pers. comm.). Floral diversity on peat soils is low (Myers1990): 54 major species were identified, along multiple transects through all phasic communities in themire complex (Appendix F). No detailed accounting of species present in forest sample plots (400 m2)along the transects was possible for this preliminary survey. Vegetation communities were identified bycomparing field notes taken during the cutting of trails for levelling surveys with low-level (300 m) aerialphotographs. Identification of the major phasic communities was then verified by the rather laboriousprocess of cutting trails into areas of interest visible on high altitude air photos in order to identifi thetrees.The domed western section of the mire is mapped as a set of phasic communities in a somewhatsimilar manner to that described for the oligotrophic peat swamps ofWestern Malesia and Borneo(Anderson 1964, 1984), and that tenninology has been adopted in this study. The eastern section is acomplex mosaic of forest types, and includes three areas in which incipient doming is evident. Largeareas of the mire are occupied by dense Raphia palm-swamp (locally called ‘Matomba’ (the plant) and‘matombal’ (the swamp) in Panama, ‘Yolillo’ and yolilal in Costa Rica) and a variety of sawgrasses(‘cortadera’ and cortaderal), on both shallow and deep peat.i} Mappable Vegetation Zones: --- Multispectral SPOTsatellite imagery with a resolution of about 15 mwas used to distinguish the distribution of phasic communities by isolating and enhancing the spectralcharacteristics of the various vegetation types present in the Changuinola mire. Factors like openness ofcanopy, reflectivity of foliage, and the extent to which standing water is present serve to distinguishvegetation zones, which can be mapped in false colour. A false colour vegetation map is reproduced asFigure 3.3 and interpreted in Figure 3.4. The concentric zonation of vegetation around the large centralbog plain can be clearly seen. In addition, incipient doming is evident in three areas to the east, two of98which may be coalescing. The narrow mangrove fringe forest and back-mangrove (PCi and PC2), andvariations in the mixed forest-swamp (PC4) types, cannot be reliably distinguished at this scale. Phasiccommunities 3, 6 and 7 are distinct and 4 is mapped as hardwood-dominated forest. Elevation of the sitesvaries from 0 m (intertidal) at BDD 25, to 6.24 m in the bog-plain (ED 3).ii) Description ofthe Phasic Communities: — The vegetation of the Changuinola mire has been dividedinto 7 phasic communities representing the sequence of zones from the margin to the centre of the deposit:1) Rhizophora mangle mangrove swamp; 2) mixed back-mangrove swamp; 3) Raphia taedigera palmswamp; 4) mixed forest-swamp; 5) Campnospermapanamensis forest-swamp; 6) Sawgrass ± stuntedforest-swamp; 7) Myrica-Cyrilla bog-plain. The pH and salinity of peat cores from each community areshown in the lower part of Figure 3.5.Phasic Community 1: Rhizophora mangle mangrove swampSite Description: BDD 25The mangrove fringe forest is a narrow zone restricted to the sheltered shoreline of Almirante Bayand extending eastward into shallow areas of Laguna Chiriqui. This community is nowhere present onthe outer coast, but dominates large areas of the shallow offshore in the lee of the Bocas del ToroArchipelago. In the shallow offshore, Rhizophora mangle and an epiphytic Clusia sp., are dominant, butin the fringe forest, the community is slightly more varied. Even in sheltered waters, where thecommunity thrives, its extent is limited to about 30 metres breadth by the small tidal range (normally 30cm).The dominant species are: Rhizophora mangle, Acrostichum aureum, Raphia taedigera, and afew salt-tolerant grasses and sedges, of which only Rhynchospora macrostachya (Cyperaceae) has beenidentified. Rhizophora mangle is the red mangrove, the dominant arborescent species and commonly the990Myrica-Cyrillamixedforest-swampCampnospermaRaphiapalm-swampback-mangroveRhizophorabog-plainforest-swampforestswampfringe-forestFigure3-5.Cartoonshowingtypicallateralsequence of‘hasicCommunities, frommangrovefringetobogplain.Representative coredatashows pHandsalinity(wt%)profilesfromthesurfacetothe baseofthepeat.Zonationinthe_—_----pHsallrflty—-pH!4. .Figure 3.6-a. Photograph of PC2 (Back-mangrove swamp): centre -a fallen Laguncularia racemosa,showing root mass and pneumatophores ; left and background - Rhizophora prop roots.Figure 3.6-b. Photograph of Campnosperma forest swamp, showing the distinctive, uniform canopy.Centre foregound, a small Euterpe precatoria palm; left and right of centre, the fine foliage of gavilan’(Pitheceilobiuni sp.).101only mangrove present in the fringe forest. Acrostichum aureum, known locally as helecho de manglar ormangrove fern, is a large (to 3 metres) and vigorously opportunistic colonizer of sheltered shorelines. Itis particularly in evidence along the recently subsided shoreline, where it is thriving amongst the debris ofsalt-killed mixed forest, and is even present as new growth in the intertidal zone. Raphia taedigera, theRaphia palm, is common immediately shoreward of the mangrove fringe, and occasionally present at thehigh tide line. Raphia standing offshore as a result of earthquake-induced subsidence appeared to beslowly succumbing to the high salinity (30 to 34 %o) 30 months after the seismic event.Phasic Community 2: Back-mangrove forest-swampSite Description: BDD 26A (Fig. 3.6-a)The back-mangrove swamp is one of the more complex communities evaluated in theChanguinola swamp complex. The observations here incorporate a number of sites, both coastal back-mangrove forest, and mixed forest marginal to the blackwater creeks which drain the eastern section ofthe peat swamp into Almirante Bay. Thus the sites have close associations with both true marine andbrackish environments. The forest community is dominated by Laguncularia racemosa and Raphiataedigera, but open sawgrass swamps (‘cortaderal) are also common behind the mangrove fringe. Alongthe banks of blackwater creeks, secondary species vary with distance upstream from the coast; thecommunity grades from an essentially Rhizophora fringe-forest near the coast, to a mixed freshwaterforest-swamp and floating sedge-grass swamp at about 2 km upstream. Figure 2.5 shows salinity profilesof three major creeks measured 1 km upstream, after one week of dry weather.In the lower stream reaches, Laguncularia racemosa , Rhizophora, and Acrostichum aureumdominate. Secondary species include a Clusia sp. epiphyte on the roots of the red mangrove. More than1 km inland, the following additional species may occur: Raphia. Euterpe precawria, a stilt palm;Pithecellobiurn sp. tree; Cassipourea eliptica, a small to medium sized tree; Chrvsobalanus icaco, a102shrub or small tree; C’yclopeltis semicordata, a fern common around the roots and bases of white and redmangroves, sedges, grasses and sawgrass. The fleshy-rooted, spiney arrowhead plant Dieffenbachialongispata is common. The above all show salt tolerance to some degree.Non-salt tolerant species include Symphonia globuhfera, an important prop-rooted upper storytree (Ca. 30 m); Campnosperma panamensis, an important buttressed tree; an Aichornea sp., a rare,buttressed tree; the shrubs Miconia curvipetulata, Cespedezia macrophylla, Ouratea sp. and Neea sp. 1.Bromeliad epiphytes are common. Rex guianensis, a medium to large tree (locally called ‘plomo’), ispresent, although more common in the mixed-forest zones. The occurrence ofRex guianensis haspreviously been associated with higher elevations, and with markedly seasonal rainfall (Johnston, 1949)although its presence in Bocas del Toro is noted by Bartlett and Barghoorn (1973).Phasie Community 3: Raphia faedigera Palm SwampSite Description: BDD 8 (Fig. 3.6-c)Raphia taedigera Martius (Calamaea) is a large palm having multiple trunks 5 0-80 cm indiameter, and enormous leaves 15 m or more in length, spreading widely from just above ground level.The plant is a significant contribttor to peat deposition, as it has an extensive root system withpneumatophores (“pneumatozones” - Kahn and de Granville, 1992). In-situ trunk bases have beenencountered at depths of three metres in peat cores, incidentally settling the question of whether Raphiamay be an introduced species in Central America (Otedoh, 1977). Raphia is one of only three specieswhich forms monospecific stands in the Changuinola swamp complex (the others are Campnospermapanamensis, and Rhizophora mangle). It forms large stands in all the marginal areas of the deposit, aswell as being a significant component in the back-mangrove and mixed forest-swanip communities. Themonospecific Raphia forest is dark and almost free of understory vegetation. Water table can be up to a103..Figure 3.6-c. Photograph of PC3 Raphia palm swamp, with sedge marsh in foreground: centre -dense Raphia taedigera forest; Symphonia globuflfera emergents.104metre above the peat surface. The two herbs Sagittaria and Dieffenbachia are common. In thinfreshwater peat, Saggittaria can reach a height of 3 metres, although both are normally a metre or less.The large prop-rooted Symphonia globu1fera (‘cerillo’) is the commonest emergent hardwood, found asisolated individuals or small groups. Isolated buttressed Campnosperma p. ernergents are also present, asare hex. Other shrubs, herbs and ferns are present but not identified in this community.Phasic Communtv 4: Mixed forest-swampSite Description: LAKE 2The mixed forest-swamp is not a single community, but a general description of the tree-dominated swamplands which are not monospecific stands. The various phases are defmed by the two orthree dominant tree species. Mixed forests mappable from low level air photos are: i) RaphiaCampnosperma; ii) Raphia-Symphonia-Euterpe; iii) Symphonia-Raphia ± other hardwoods iv)Symphonia-Campnosperma-sawgrass v) Raphia + hardwoods on thin peat. Symphonia is the mostcommon and distinctive dominant, observed in all the hardwood-dominated forests, but it does not formmonospecific stands. Allen (1956) reports that it does form such stands in Costa Rican fresh waterswamps on the Pacific coast. The loop-roots of this species are pervasive over large areas of the mixedforest-swamp surface. Scattered Rex (‘plomo’) is also common to all the forest-swamp typesencountered. The site LAKE 2 is typical of Type iii) Symphonia-Raphia-hardwood forest.Phasic Community 5: Campnosperma panamensis forest-swampSite Description: BDD 23 (Fig. 3.6-b)Campnosperma panamensis forms large monospecific stands on moderately deep peat. Thesestands are easily recognizable by the irregular, unbroken canopy, and by the slender mottled white trunks,branchless for 10 to 15 metres. Air photos from 1954, 1981 and 1992 reveal striking differences in thedistribution of the distinctive Campnosperma canopy within the complex region of mixed forest-swampin the eastern part of the mire. It thus appears that these dense stands develop rapidly and are quite short-lived. Secondary trees in the middle storey are: Euterpe precatoria, a stilt palm; Cassipourea eliptica105Figure 3.6-d. Photograph of PC6 Sawgrass / stunted forest-swamp: The medium-sized, lobateleafed tree in the foreground is Can2pnosperrna panamensis, Cyrilla race,nflora are visiblebeyond the workers, and a Symphonia globuflfera in the upper left corner.106and Pithecellobium sp., both of which attain greater size outside this community, and Ternstroemiatepazapote, which is rare in the swamp. The understory vegetation is quite sparse and open. Principalshrubs are Miconia curvipetulata, Tococa gulanensis and an Ardisia species, Sagittaria is the commonarrowhead plant, and Smilax is a common vine. The fern Trichomanes crinitum is found around thebases of Campnosperma.Phasic Community 6: Sawgrass / stunted forest-swamp (Fig. 3.6-cl)No single site represents this phasic community. Extensive open-canopied forest, grading intosparsely-treed sawgrass swamps (‘cortaderal’ in Spanish) can be seen from the air, and in satelliteimagery, as a transition zone between the closed mixed forest-swamps and the bog plain. However, onthe ground the floral composition is highly variable. Thus the PC 6 sawgrass / stunted forest-swamp isdistinct as a mapping unit, but is a transitional vegetation zone. Identification of the various grasses andsawgrasses (Gramineae, Cyperaceae) was not possible, with the exception of Rhynchosporamacrostachya, for which flowers were available. Some sawgrasses (Cyperus ligularis, Cyperusodoratus) were identified from open areas around the margins of the swamp, on shallow peat in runnelson the barrier bar, and in pastures, and may also be represented in the stunted forest and bog plain. PC 6is found both marginal to the bog plain community, where it forms a transition zone from forest-swamps,and as islands within the eastern forest-swamps. In the east, this community may be associated with bothold drainage channels, and ‘topographic’ highs (representing incipient peat doming, not basement highs).In all situations, high water table is common, as are ‘floating’ surfaces. Peat recovered from thisenvironment is very wet (difficult to sample) orange-red, fibrous to coarse hemic, and of low bulk density.Stunted Campnosperma and shrubby Syniphonia occur, but the dominant trees are Myrica mexicana andCyrilla racemflora. Bartlett and Barghoorn (1973) report that Myrica is found only along the marginsof residual patches of forest in high, dry windswept grasslands above 600-800 rn elevation. C’yrilla hasbeen reported from only one isolated site in Panama, at relatively high elevation at Cerro Jefe, near the107Canal Zone (A. Hernandez, pers. corn.). In the Okefenokee Swamp of Georgia, however, C’yrillaracemflora is the dominant species forming tree “hammocks”, isolated clumps of low-growing treeswithin open swamp areas (Cohen, 1975). Both of these trees are known locally as ‘mangle cimarron’ or‘bush mangrove’ because of the wet conditions in which they grow. A rare arborescent Clusia sp., a largetree with enormous prop roots originating 4 metres or more above the peat surface, has not been reportedelsewhere in Panama (A. Hernandez, pers. comm). ilex gulanensis occurs as a common shrub in thisphasic community. Chrysobalanus icaco is a dominant shrub, and Miconia curvipetulata and Myrsenepelucido-puntata are present. Ferns, including Salpichlaena sp. are found intergrown in the sawgrass.Mosses are also present.Phasic Community 7: Myrica-Cyrilla bog-plainSite Description: LAKE 10 (Fig. 3.6-e)The term bog-plain here uses bog in the sense of an ombrotrophic mire, a system dependingentirely on rainfall for its water and nutrients (Moore, 1989) rather than as a community dominated bymosses. Mosses and stunted sedges and grasses are present, as are algal mats in the open water betweenthe bases of the sawgrass and shrubs. Woody vegetation is stunted and shrubby. Myrica mexicana andCyrilla racemflora are dominant, both as shrubs in the open areas, and as low-growing trees (diameter to20 cm or more, b.h.) forming the ‘hammocks’ which are scattered across the plain. Chrysobalanus icacoand very stunted Campnosperma and Clusia sp., small Sagittaria, and ferns are present. Cohen et al.(1989) identified Sphagnum, and the insect-eating Drosera (sundew) and Utricularia (bladderworts).Walking across this terrain one must step on the bases of the sedge plants, as the moss- or algae-coveredsurface will not support weight. Water depth is a metre or more. The more ‘solid’ ground in the shrubhammocks is also at least partially floating, and if allowed to, the corer will drop several metres under it’sown weight.108Figure 3 .6-e. Photograph of PC7 Myrica-Cyrilla bog plain: Algal mats are visible in the rightforeground, along with stunted sedges and grasses and a small Carnpnosperrna p.; Myricamexicana along the horizon,(small leaves).[093.6 POLLEN DISTRIBUTIONThe aim of the palynological investigation was, in combination with particle size and geochemicaldata, to identify major trends in changing ecological conditions within the coastal swamp during itsapproximately 4000 year history. Pollen ‘fingerprint& representing 6 of the phasic communities wereconstructed, (Fig. 3.7-a to 3.7-f), from counts of surface samples in the hope that those communitiescould be recognized in the peat stratigraphy. Thus any sequential succession in the vegetation would berevealed, which could then be related to conditions of trophic state, pH, and pore water salinity associatedwith present vegetation. Selected palynomorphs referred-to in the following section are described andillustrated in Appendix E and in the Plates. PC 6, the sawgrass / stunted forest-swamp, is a variablecommunity that represents a transition; no single site was thought to be representative of this community,hence no pollen fingerprint was constructed.a) Surface Pollen DiagramsPC 1: Rhizophora fringe forest (Fig. 3.7-a)The most common palynomorphs from this site are an unidentified tricolporate pollen Type 1,Rhizophora mangle, monolete fern spores, Rex guianensis, and fungal spores. Rhizophora pollen hasbeen shown to be a good sea-level indicator (Muller, 1959; Van der Hammen, 1963), although in thisstudy it represents only about 9% of the total count in both PC 1 and PC 2. Thus the pollen recordsignificantly underrepresents the species. This is undoubtedly due to the narrowness of the mangrovefringe, as observed by Van der Hammen (1963), and the fact that the prevailing wind blows offshoreacross the mangrove fringe. Type 1, along with a second unidentified tricolporate, Type 2, wereencountered only in mangrove fringe and back-mangrove samples. Rex pollen, which is clearlyallochthonous, is present in almost all samples from the Changuinola mire. In addition, this is the onlysite at which Myrica mexicana pollen is absent. Acrostichuni aureun is also absent from this sample,although that fern is common in this phasic community and was observed in other niangove peat from110Figure 3.7-aPhasic community 1Sample:BDD25 Tot count % of total PC 1: RhizophoraRhizophora fringe forest 498 fringe forestSpecies1 Rhizophora mangle 43 8.60 5 10 15 20 252 Tricolpate type 1 115 23.1________________________________I I3 Tricolpate type 2 10 2.0 14 Campnosperma panamensis 1 0.2 2_____ ___5 Myncamexicana 0 0.046 Ilexguianensis 19 3.857 Cyperaceae 0 0.0 68 Gramineae 1 0.2 79 Trilete ferns 2 0.4 810 Monolete ferns 47 9.4 911 opaque fungal spore 29 5.8 1012 Other fusiform spores 25 5.0 11124 24.9 12Other 374 75.1Figure 3.7-bPhasic community 2Sample:BDD26A Tot count % of totalBack-mangrove swamp 279Species PC 2: Back-mangrove1 Rhizophora mangle 22 7.9 swamp2 Lagunculaña racemosa 15 5.43 Campnosperma panamensIs 4 1.40 5 104 Myncamexicana 8 2.9__ _ __5 Ilexguianensis 0 0.06 Raphia taedigera 5 1.87 Cyperaceae 23 8.28 Gramineae 26 9.359 Acrostichum aureum 3 1.110 Monolete ferns 27 9.711 Other Fusiform spores 0 0.012 Other fungal spores 22 7.9155 55.6Other 114 40.9 ii____12__ ___ ___Figure 3-7. Pollen ‘fingerprints’ from surface peat samples, PC I and PC 2.111cores. The pollen diagram reflects the low diversity of this community. The medium size fraction (<2.0mm and >0.25 mm) of this surface sample contained abundant calcareous foraminifera and ostracodes.PC 2: Back-mangrove forest-swamp (Fig. 3.7-b)Sampling at this site was by 3.5 cm diameter vibracore, the top 30 cm of which was processedfor pollen analysis, thus surface litter was not well represented. Palynomorphs were few and poorlypreserved, and fewer than 300 grains were counted from multiple slides. Rhizophora, Laguncularia andAcrostichum are present, as they are in the present vegetation, and both Cyperaceae and Gramineae arewell- and perhaps over-represented. Monolete fern spores represent about 10% of the total count.Tricolporates Type 1 and Type 2 are present but much reduced from PC 1. Microforaminifera tests arepresent, and dinoflagellate cysts are abundant, but were not counted. Many fragments, damaged andpitted grains were found. The absence ofRex pollen may be a function of the small number of totalgrains encountered.PC 3: Raphia palm swamp (Fig. 3.7-c)The pollen of Raphia taedigera, flingal spores, monolete fern spores and a diverse arborealpollen component characterize this site. An unidentified Chenopodipollis (7) -type of periporate pollen,Type 39, is the second most abundant grain (9%), and Rex is plentiful (6.5%). Also common is a spikeyspherical phytolith, superficially identical to those found in the epidermis of certain Old World palmleaves (Rosen, 1992) and here assumed to be associated with Raphia, rather than the much rarer palmEuterpe precatoria. An important palynomorph is the large fungal spore Fusiformisporites sp., which ispresent in Raphia swamp peats, but is much more common in the mixed forest and Campnospermaforest-swamp peats.112Figure 3.7-cPhasic community 3__________________________________________Sample: BDD8 otat coun % of totalRaphiaswamp 1010 PC 3: Raphia swampSpecies1 Campnosperrna panamensis 44 4.42 Mynca mexicana 28 2.8 0 5 103 Myrica type A 8 0.84 ilex guianensis 66 6.55 pericolporate C (39) 91 9.06 Raphia taedigera 101 10.07 Cyperaceae 42 4.28 Gramineae 4 0.49 Triletefems 0 0.010 Monolete ferns 75 7.4 911 Fungal spore type 14 58 5.712 Fusiformispolites 5 0.5 113 Other f-form spores 52 5.1 12574 56.8Other 436 43.2___ _______Figure 3.7-dPhasic community 4Sample: LAKE 2 Total count % of totalMixed forest-swamp 1039 PC 4: Mixed forestSpecies1 Carnpnospenna panamensis 20 1.9 0 5 10 15 202 Mynca mexicana 4 0.4__ ___3 flex guianensis 38 3.7 14 tncolporate 0 (40)(Tiliaceae?) 29 2.85 tricolporate E (48) 63 6.16 Raphia taedigera 12 1.27 Cyperaceae 1 0.1 68 Gramineae 0 0.09 Tnlete fern of Metaxya sp. 42 4.0 810 Monolete fern 1 45 4.3 9___ii Monolete fern 2 34 3.310_12 Cyclopeltiss. 104 10.013 other monolete ferns 163 15.7 1214 Fusiformispo,ites 13 1.3 1315 other f-form spores 32 3.1 14600 57.7 15Other 439 42.3_ ___Figure 3-7 cont’d. Pollen fingerprints from PC 3 and PC 4.113flex and Myrica pollen are common in both Raphia and mixed forest-swamp. These species werepreviously only reported together at higher elevations, associated with a marked dry season. Thiscombination is noted by Bartlett and Barghoorn (1973, p.245) and used as a basis for interpreting aperiod of cooler, drier climate in the Isthmus of Panama during the period 7300-4200 B.P. Clearly thecombination is also present within a few metres of sea level in an equable, everwet regime. Pollen of flexsp is found in association with Rhizophora at 35,500 BP (Bartlett and Barghoom, 1973).PC 4: Mixed forest-swamp (Fig. 3.7-d)High dicot pollen diversity, and a large proportion of both trilete and monolete fern spores (37%ofthe total count) characterize this community. Noteworthy is the absence of pollen of Symphoniaglobulfera, which is the dominant tree presently growing at this site, and throughout much of the mixedforest-swamp. Only 6 grains of Symphonia pollen were counted in the entire study, the paucity possiblyindicating entomophily. Despite the prolific flowering of this species, it is clearly not a viablepalynostratigraphic indicator. Sedge and grass pollen are virtually absent. Fusformisporites is common.PC 5: Campnospermq forest-swamp (Fig 3.7-c)Pollen of both Campnosperma panamensis and another Anacardiaceae-type pollen are abundantin this environment. The second species differs from Carnpnosperma in being more prolate (P/E of 1.6vs 1.2 to 1.4), but bears very similar exine structure. Raphia and Euterpe are both present in the pollenprofile, as they are in the modem vegetation. Fusformisporites sp.A and sp.C are about equallyrepresented, and fungal spores as a group are more common in this than any other phasic community.Fusformisporites is of particular interest, as it has been described in the Eocene (Fusformisporitescrabbii, Rouse, 1962; F pseudocrabbii, Elsik, 1968, Texas; F rugosus Sheffey and Dilcher, 1971,Tennessee and Kentucky); and Paleocene and Plio-Pleistocene (Elsik, 1980) of North America as well asfrom the Lower Eocene, and Holocene (possibly as re-worked grains)of Jamaica, (Germeraad, 1979) but114Figure 3.7-ePhasic community 5Sample:BDD23 Total count % of totalCampnosperma forest-swamp 1016 PC 5: CampnospermaSpecies forest—swamp1 Campnosperma panamensis 188 18.52 Campnosperma type 198 19.53 Mynca mexicana 6 0.64 Myrica type 20 2.0 0 5 10 15 205 flex guianensis 17 1.7 1 L6 tricolporate F (33) 10 1.0__________________________7 Raphia faedigera 62 6.18 Euterpe precatoria 10 1.09 Palm phytolith 9 0.910 Cyperaceae 53 5.211 Gramineae 14 1.412 Trilete ferns 33 3.2 913 Monolete ferns 30 3.01114 FusiformispodtesA 10 1.015 Fusiformispofites C 9 0.91316 Other f-form spores 37 3.6706 69.5 15Other 310 30.5Figure 3.7-fPhasic community 7Sample:LAKEIO Tot count % of total PC 7: Bog plainBog plain 741Species1 Campnosperme panamensis 11 1.5 0 5 10 15 202 Mynca mexicana 23 3.1 13 CytilIa racemiflora 14 1.9 24 flex guianensis 14 1.9 35 tricolporate A (26) 128 17.3 46 tricolporate B (54) 17 2.3 57 Cyperaceae 14 1.9 68 Gramineae 19 2.69 Tnlete ferns 36 4.9 810 Monolete ferns 13 1.8 g11 SporangiumA 54 7.3 1012 other fungal spores 40 5.4 ii383 51.7 12Other 358 48.3______Figure 3-7 contd. Pollen fingerprints from PC 5 and PC 7.115not, to our knowledge, from modern peat. Trilete and monolete fern spores each represent about 3% ofthe total count.PC 6: Sawgrass / stunted forest-swampNo pollen fingerprint was constructed for this phasic community, despite its utility as a mappingunit, as it is transitional between forest-swamp and bog plain communities, and is highly variable in floralcomposition.PC 7: Bog plain (Fig. 3.7f)The presence of Cyrilla racemflora, both in the modern vegetation and in the observed pollenrecord, is restricted to phasic community 7. Two unidentified tricolporate species (Types 26 and 54) arealso common and exclusive to the bog-plain sites. Tnletes dominate the spore count only in thisenvironment. Several species of trilete spores together account for 5% of the total. Monolete fern sporesare scarce.b) Pollen From CoresPollen analysis was performed on two cores, one from the domed centre (ED-3, Fig. 3.8) and asecond (BDD-23, Fig. 3.9) from the planar eastern part of the Changuinola mire. ED-3 is an 802 cmpeat core collected from a bog plain site of stunted shrubs, sawgrass and algal mats. The peat wassampled in 30 cm increments from the surface, at an elevation of 6.24 m, to the base in rooted, finegrained grey sand 178 cm below present sea level. BDD-23 is a 450 cm peat core over 40 cm of peaty,rooted sand, on a base of clean, grey, carbonate-free sand at 490 cm. The site is 1.5 km from the presentmarine margin, and all but the upper 15 cm of this core is below sea level. The BDD-23 site is presentlyat the edge of a monospecific stand of Campnosperma, adjoining an area ofRaphia / mixed hardwood116swamp. Sampling was attempted in 25 cm increments, but recovery was erratic due to the large amountsof wood encountered.ED-3 (PC 7) Pollen Diagram (Fig. 3.8)The pollen diagram for ED-3 was constructed using those palynomorphs or combinations thatbest illustrate stratigraphic variations in the core. The palynomorph assemblage from the earliest stage ofmire development (810-720 cm) at this site is consistent with a PC 3 Raphia palm-swamp, includingdominant monolete ferns and Pa[mae pollen. Raphia and Euterpe palms and the distinctive periporateType 39 are present. Palm phytoliths, included in Palmae, are abundant in the deepest sample (23% ofthe total grains counted). hex and many monolete ferns are also well represented in this earliest stage.Fusformisporites A and C are also present. Halophytes are absent, and there is no evidence of marineinfluence. Basal sediments are angular medium-grained grey sands containing abundant roots, andclosely resemble the sediments of the present barrier bar.The second stage, from 720 to 630 cm, is dominated by Campnosperrna, and fungal spores. l’hisassemblage is interpreted as PC 5 (Carnpnosperma forest-swamp) succeeding the palm swamp.Fusformisporites spp. disappear at 660 cm, but other dicellular spores persist. Both palms are present,and ferns become reduced in both numbers and diversity in the upper half of this stage. Sedges andgrasses occur at their lowest levels.In stage three, between 630 and 390 cm, there is first a decrease in hardwoods (except hex) andan increase in palms, followed by a decrease in palm pollen as well. Grasses and sedges increasesubstantially, and ferns are plentiful. Cyrilla makes its first appearance at 510 cm. and Raphiadisappears. The gradational change that is apparent throughout this section represents the PC 6transitional, open, somewhat stunted forest / sawgrass zone which occupies a large proportion of the117(cI:100western mire and indicates incipient doming. Types 26 and 54 make their appearance at the bottom ofthis zone, and become increasingly common.Stage four consists of the upper 390 cm of the core, dominated by a steady increase in the pollenof Cyrilla, Myrica, and Types 26 and 54. Grasses, sedges and fungal spores are common. hex andCampnosperma gradually disappear. The assemblage represents PC 7, and records the increasingoligotrophy of the maturing domed bog-plain.BDD-23 (PC 4) Pollen Diagram (Fig. 3.9)From 600 to 490 cm depth at this site, clean grey sand with occasional roots was recovered.From 490 - 450 cm, rooted, suiphurous peaty sand with abundant wood fragments occurs. Pollen in thebasal peats from 450 up to 285 cm is dominated by a PC 2 back-mangrove assemblage of Rhizophora,Laguncularia and Acrostichum, here plotted together, and the Types 1 and 2 tricolporates. flex is alsopresent between 450 and 350 cm, but is absent from the 300-350 cm sample. As this absence was alsoobserved in the PC 2 pollen fingerprint, it may still be a function of sample size, as the 300-350 cmsample had by far the least grains, and the most poorly preserved, of all those studied. The 450 cm peatsample is very high in sulphur (13.7 wt%). This has been interpreted as an indication of the influence ofbrackish waters, probably due to proximity to brackish blackwater creeks, rather than true marineconditions (Phillips Ct al., 1994; Phillips and Bustin, in review). The presence of Laguncularia andCampnosperma pollen supports this interpretation.From 285 to 150 cm the pollen assemblage indicates the succession of Raphia palm-swampvegetation. Raphia pollen and phytoliths, and Type 39 pollen are abundant. hex and Myrica increaseand the halophytes disappear. Fusiformisporues is abundant, as are fern spores, and there is a highdiversity of minor arboreal species. From 150 to 100 cm, a mixed forest-swamp assemblage dominates.1190RhzophoraLaguncularia+Campnosperma+Mynca+Cyperaceae+AcrostichumTypesI+2Camp.- typeType39°almaeMyncaAlaxGramineaeFuslformispoiitesPHASICCOIvTvIUNITYoI]ICAMPNOSPERMA5°IFOREST-SWAMPiooUIUfiMIXED125FOREST-SWAMP175.RAPHIA200IIPALMSWAMP225::I250flJ285I?5BACK-MANGROVE400JI]HIIISWAMP450[i1I!II0102000204060010200510010203005100204001020120Figure3.9.PollendiagramofcoreBDD23.DataisinAppendixD.Palms decrease as dicots increase, and there is a peak in Fuszformisporites abundance. Symphoniaglobuhfera pollen occurs at the top of this section (2 grains).The upper 100 cm is dominated by Campnosperma and Campnospern?a-tvpe pollen, and by adiverse assemblage of minor dicots. Palm pollen and phytoliths are also plentiful. Type 39 tapers-out,and Fusformisporites is present but much reduced. The assemblage is consistent with the present PC 5Campnosperma forest-swamp vegetation at this site.3.7 DISCUSSIONDescription of the floral succession from single or widely separated cores is necessarily restrictedto broadly general interpretations, and no serious attempt can be made to correlate between the two coresanalysed here, which are separated by 6.5 km of trackless swamp. Nonetheless, the two pollen recordsfrom the Changuinola deposit reveal successional changes in vegetation both in the domed central part ofthe deposit and in the eastern, planar section approaching the marine margin of Bahia Almirante (Fig.3.10). The central mire had its origins as a freshwater Raphia palm swamp upon sandy sediments verysimilar to those of the modern barrier bar. The comparable Raphia community presently exists on thethin peats in the runnels and directly behind the modern barrier. Further east at BDD 23, back-mangroveforest-swamp, likely close to brackish blackwater creeks draining into Almirante Bay, were succeeded byRaphia palm swamp. A similar circumstance is seen in the dense Raphia forest which crowds the back-mangrove zone along present blackwater creeks, and even behind mangrove fringe-forest, where Raphiaproves to be remarkably salt-tolerant. Swamp-forest succession based on early colonization by Raphiataedigera has been postulated in Costa Rican coastal swamps (Anderson and Mori, 1967; Myers, 1990)and in African palm swamps (Richards, 1952). Anderson and Mori (1967) describe a succession inwhich the early, closed Raphia forest canopy eventually matures and opens-up, allowing the succession of121I::—t,,,‘‘.-GENERALIZEDPEATPACIESbasedonpollenanalysisandsuifacevegetafioPC3:Raphiapalmswamp--jPC?:Bog-plainPC2:BackmangrovePC6:Sawgrass/stuntedforest-swampPCi:MangrovefxingePC4&5:Forest-swampUndifferentiatedsands,silt,clayFigure3-10.NW-SEcrosssectionof theChanguinola peatdeposit,asinterpretedfrompollenincoresandsurfacevegetationzones.Locationof thecrosssectionisshownonFig.3-1.Sitesmarkedwithanasterisk(*)coredatafromCohenetal.,1990.Verticalexaggerationisx500.IARIOTNICHA14*BAHIABDD20AALMIRANTE-‘°SEALEVELhardwood forest-swamp. A similar sequence is seen in both cores, one to a mixed forest (Fig. 3.10) andthe other to monospecific Campnosperma forest-swamp (Fig 3.9).The succession from forest-swamp, through a progressively less diverse, less arboreal and moreecologically specialized plant community, to oligotrophic bog-plain, is admittedly a broadly generalizedone. The limited number of samples and the resolution of the sampling program does not permit adetailed stratigraphy, and it is certain that many variations can occur within and between cores, but thegeneral vertical trend is clear. The phasic communities, at least in some sites in the Changuinola deposit,succeed each other in the order in which they are numbered. This is not to say that they must alwaysoccur sequentially, nor that they have equal roles in the development of the deposit. In the modern mire,Raphia-swamp occupies a far greater area, and is ecologically far more variable than is monospecificCampnosperma forest-swamp. It may not be reasonable to suggest that Canpnosperma forest representsa true phase in a succession, given its present limited distribution, and it is with some surprise that wefound the PC 5 pollen fingerprint occurring in both cores. The Asian species Carnpnosperma coriacea,and Campnosperma squamaa are reportedly prolific pollen producers (Anderson and Muller, 1975), andit is possible that the counterpart Campnosperina panarnensis is significantly overrepresented in someareas of the Changuinola mire. Nonetheless, the pollen of that species is decidedly not ubiquitous, nor doits numbers overshadow other species, so its presence is considered significant.Lateral relations between the phasic communities, moving from the margins toward the centre ofthe mire, are somewhat variable, but generally follow a similar sequence to that described in the verticalsequence. The mire complex is roughly rectangular in plan view. Each side is physiographically distinct:the SE is a low energy, mangrove-fringed marine shore, the NE a high energy, wave dominated barriercoastline, the NW a large, flood-prone alluvial plain, and the SW an abrupt transition to deeply erodedhill country. On all sides but one the same general sequence is seen: From mineral soil supporting a123diverse vegetation, peat depth increases through a belt of Raphia palm swamp. The Raphia gives way tomixed forest-swamp, to ampnosperma swamp, in some cases to sawgrass swamp (Fig. 3.10), andeventually the bog plain is reached. The exception to this sequence is in the SE along the shore ofAlmirante Bay. Here vegetation forms a complex mosaic which reflects the influence of marine andbrackish conditions, coastal subsidence and periodically rising sea level. In this area drainage is morecomplex, and doming and increasing oligotrophy have not proceeded to the same degree, even though thepeat depth is not significantly less. Incipient doming is occurring, however. If peat accumulation ratesexceed the rate of coseismic subsidence, the small domes may eventually coalesce. Moving from themangrove fringe forest toward the centre of one of these small domes, the sequence is variable, and any ofthe phasic communities may be absent, including the Raphia palm swamp.We have adopted the term ‘phasic community’ from Anderson, believing that it is not meant toimply an inevitable successional sequence, but rather is intended to describe the observed continuitybetween lateral and vertical relations within the Malaysian peat swamps, in just the same sense that thegeological concept of facies relations is intended to describe observable relations, rather than imply causeand effect in these relations. The concept of ‘facies association’ (defined in the geological sense byWalther’s Law) states that in an evolving depositional system, units which are laterally contiguous inspace will be vertically contiguous in time. This clearly applies to the Changuinola deposit (Fig. 3.10).Anderson (1964) also uses the concept of a “catenary sequence” of phasic communities to relatethe distribution of vegetation zones directly to the domed topography of the Malaysian peat deposits. Ourobservations of zoning in the Changuinola mire are not sufficiently well-controlled to justif’ the use ofthat term, and we are not able to state that the gradient of the peat surface controls the distribution of thevegetation. Anderson relates both gradient and doming to peat accumulation rates based on radiocarbondating of peat in cores. Describing a 12 m core, Anderson (1983) reports the following radiocarbon124dates: 5 m = 2255 ± 60; 10 m = 3850 ± 55; 12 m = 4270 ± 70. These give accumulation rates of 4.76mm per annum for the early (12-10 m) basal (mangrove) peat, 3.14 mm/a for the intermediate (forest-swamp) stage, and 2.22 mm/a for the upper (PC 6?) 5 m of peat. This slowing of accumulation accountsfor a deposit which, while generally domed, is quite steep on the edges and flat in the raised centre.Accumulation rates for the Changuinola deposit portray a somewhat different history,summarized for different peat types as follows. Dates are the radiocarbon age in years before present(BP) from three cores:4.75 m =2110 ± 60 BP in fine hemic mangrove peat throughout,4.75 m = 1880 ± 60 BP in woody, fme hemic mangrove/back mangrove peat,4.75 m = 850 ± 80 BP in fibric, bog-plain sedge peat,and 8.0 m = 3040 ± 80 BP in woody, hemic palm / mixed forest peat.These dates give the following accumulation rates for different peat types:mangrove/back mangrove peat 2.25 mm/a (from 4.75 m below present SL to present SL),back mangrove/ forest peat 2.52 mm/a (from 4.75 m to the present surface),palm/forest peat 1.48 mm/a (from 8 m to 4.75 m),and sedge peat 5.58 mm/a (from 4.75 m to surface)It can be seen from these results, particularly in the sedge peats, that there is a great deal of variation inrates of accumulation olin situ peat. The above values are as much an indication of variations in bulkdensity and compaction as they are of accumulation rate. The rate of accumulation is in inverseproportion to peat density (Chapter 2.), and a true comparison of preservation and accumulation indifferent environments would require factoring-in compaction and density variations. The above in situvalues, however, suggest that the hydrological characteristics of the peat dictate the height of the watertable, and hence the surface topography. The densest peats sampled, mangrove and forest peats,accumulate at rates comparable to Anderson’s late-stage peat. The very high rate of accumulation of125sedge peats reflects lack of compaction. (The flbric nature of these peats makes it difficult to removeyounger root material, and great care must be taken to avoid contamination.) The domed central bogplain is essentially a mass of well-preserved root material floating in an acidic, nutrient-poor humic soup,contained within a basin of dense, woody peat of extremely low-permeability. Drainage of the centralmire is superficial, subsurface flow being effectively dammed by the surrounding forest-swamp peat.3.8 SUMMARY AND CONCLUSIONSThe general trend of vegetational development and peat accumulation in the Changuinola depositcan be seen in Figure 3.10. The deposit originated as freshwater palm swamps that developed in closeproximity to both the Changuinola River mouth, probably behind a barrier bar and freshwater lagoonsystem, and a low energy, mangrove dominated bay. The early swamp was likely drained to the southeastby brackish blackwater creeks much as it is today, and formerly extended considerably farther in thedirection of Almirante Bay. The Raphia swamp was succeeded by hardwood forest-swamp dominated bya limited number of specialised species, only one of which, Campnosperma panarnensis, is prone toforming monospecific stands. Increasing accumulation of woody peat promoted by the everwet climateimpeded drainage of the mire, leading to increased oligotrophy, and establishment of bog-plain conditionsin the oldest, central regions of the mire. It is significant that peat development did not require themangrove phase common to the Sarawak peat swamps, as Raphia taedigera is able to colonize andinstitute peat accumulation in a variety of environments. Raphia is found thriving on both mineral soiland deep peat, with pH varying from 8 to 3, and salinity up to about 30 ppt.The Changuinola mire complex has several features in common with the extensive coastal miresof southeast Asia, as well as some fundamental differences. In surface morphology, and in the zoneddistribution of vegetation, this neo-tropical mire closely resembles the swamps of Malesia described byJ.A.R. Anderson (1961,1964), although the degree of doming is so slight that the term is more useful as a126category of mire than as a description of the topography: the maximum gradient recorded is im in 330m,in mixed forest swamp near the NE margin of the mire. Pollen analysis shows a succession of vegetationfrom the base to the surface of the peat deposit which echoes the lateral sequence of phasic communitiesfrom the margins to the centre of the mire. Thus the model of peat accumulation in an environment ofincreasing oligotrophy described by Anderson applies to the Changuinola mire complex, despitesignificant differences in geography, vegetation, and depositional environment compared to the southeastAsian deposits.Thick, extensive peat deposits can develop on a high energy shoreline in close association withactive elastic sedimentation if the sediment supply is sufficient to cause aggradation-progradation, and aneverwet climate maintains high water tables.In this region, mangroves are not essential to the development of extensive coastal peats. Raphiais a primary colonizer in the Changuinola mire, and represents the first stage in the succession in non-mangrove environments.Doming, sufficient to exclude floodwaters, can be related to factors other than differential peataccumulation rates, and is dependent on peat density and permeability. Peat density is associated with thepeat-forming vegetation and thus with ecological conditions. The late-stage bog-plain accumulation rateis very rapid, because the density is very low - the sedge peats are not dense and compacted like the forestpeats, and they are mostly floating.The concept of ‘facies association’ (in the geological sense, which states that in an evolvingdepositional system, units which are laterally contiguous in space will be vertically contiguous in time -Waither’s Law) applies to the organic sediments of the Changuinola mire, and in the broadest sense could127be expected to apply to petrographic variations in coals originating in domed peat swamps. This wouldapply whether dealing with palynofacies, or with lithofacies if it can be shown that there is a relationbetween coal-forming vegetation and microlithotypes.In the course of this study, certain practical considerations became apparent. The first is that,with sufficient work ‘on the ground’, satellite imagery can be used not only to determine the existence andextent of tropical peatlands, but also the vegetation types, some aspects of the drainage patterns, ‘degreeof wetness’, etc., all valuable elements in the monitoring and managing of wetland resources. This isparticularly important in remote areas, and in developing countries.Ecological and climatic interpretation based on pollen profiles is fraught with danger, asexemplified by the association of cooling, and increased seasonality with the Myrica-Rex pollenconnection in the Bartlett and Barghoom (1973) core from 7200-4300 B.P. Much more work is neededin the tropics before we can apply palynological principals with the same degree of confidence as in highlatitudes. In the present case, a more in-depth palynological study, with better representation of pollenfingerprints to their vegetation, would include multiple core and surface samples from each phasiccommunity in order to reduce the under- and over-representation inevitable from single-site sampling.ACKNOWLEDGEMENTSSupported by the International Development Research Centre, Ottawa, Y.C.R. Grant 92-1201-18(SP), the Natural Sciences and Engineering Research Council of Canada Grant 5-87337 (RMB), and TheUniversity of British Columbia. The authors are grateful to The Smithsonian Tropical Research Institute,Balboa, Panama, the Instituto de Recursos Hidraulicos y Electrificacion of the Republic of Panama, andthe Chiriqui Land Company, Changuinola, for their unhesitating support throughout the duration of thisproject. In particular, thanks to Dr. Paul Colinvaux and Sr. Enrique Moreno of STRI for their aid andthe use of their labs, Sr. Andres Hernandez of the Forest Dynamics Unit, Barro Colorado, for hiswillingness to cheerfully join us in the swamp, and Ing. Eduardo Reyes of IRHE for his surveying skillsand his continuing efforts on our behalf. Special thanks to the Serracin family and the Sanchez family ofBoca del Drago, and to the eagle-eyed Sr.Sammy Sanchez.1283.9 REFERENCES CITEDAllen, P.H., 1956. The Rainforests of Golfo Dulce. Univ. of Florida Press, Gainesville, Fla., 417 p.Anderson, J.A.R., 1961. The ecology and forest types of the peat swamp forests of Sarawak and Bruneiin relation to their silvicuhure. PhD Thesis, University of Edinburgh.Anderson, J.A.R,, 1964. The structure and development of the peat swamps of Sarawak and Brunei.Journal of Tropical Geography, v.18, Singapore, p. 7-16.Anderson, J.A.R., 1983. The tropical peat swamps of Western Malesia. i A.J.P. Gore, (ed.),Ecosystems of the World 4B - Mires: Swamp, Bog, Fen and Moor; Chapter 6. Elsevier,Amsterdam.Anderson, R. and Mori, S., 1967. A preliminary investigation of Raphia palm swamps, Puerto Viejo,Costa Rica. Turrialba, v.17, p. 22 1-224.Barber, K.E., 1981. Peat Stratigraphy and Climatic Change. A.A. Balkema, Rotterdam, 1981. 219 p.Bartlett, A.S. and Barghoorn, E.S., 1973. Phytogeographic history of the Isthmus of Panania during thepast 12,000 years (A history of vegetation, climate and sea-level change); I: Graham, Alan,(ed.), Vegetation and Vegetational History ofNorthern Latin America. Elsevier, Amsterdam,1973,p.203-3 1Bruenig, E.F., 1990. Oligotrophic forested wetlands in Borneo. A.E. Lugo, M. Brinson and S.Brown, (eds.), Ecosystems of the World 15: Forested Wetlands, Elsevier 1990, Chapter 13.Christen, H.V., 1973. Climatic classification and land use of the humid parts of Colombia, withparticular consideration of forestry. Unpublished manuscript, available at Chair ofWorldForestry, Hamburg.Cohen, A.D., 1968. The petrology of some peats of southern Florida (with special reference to the originof coal), unpublished PhD thesis, The Pennsylvania State University, 1968. 352 p.Cohen, A.D., 1975. Peats from the Okefenokee swamp-marsh complex. Geoscience and Man, vol.xi,Apr. 25, 1975, p.123-131.Cohen, A.D., Raymond, R. Jr., Mora, S., Alverado, A. and Malavassi, L., 1985. Economiccharacterization of the peat deposits of Costa Rica, preliminary study. In: B. Wade (ed.),Tropical Peat Resources, Prospects and Potential, mt. Peat Society, Helsinki, Finland, p. 146-169.Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce,F., 1989. The Changuinola peatdeposit of northwestern Panama: A tropical back-barrier peat (coal)-forming environment. :P.C. Lyons and B.Alpem (eds.), Peat and Coal: Origin, Facies and Depositional Environments.mt. J. Coal Geol., v.12, p. 157-192.129Cohen, A.D., Raymond, R.Jr., Raniirez, A., Morales, Z. and Ponce, F., 1990. Changuinola Peat Depositof Northwestern Panama, 3 vols. Los Alamos National Laboratory pub. LA- 11211, July 1990.D’arcy, W.G., 1980. Flora of Panama. Annals of the Missouri Botanical Gardens, v. 67.Elsik, W.C., 1968. Palynology of a Paleocene Rockdale lignite, Milarn Co., Texas 1: Morphology andtaxonomy. Pollen and Spores, v.10, No.2, p. 263-314.Elsik, W.C., 1980. The utility of fungal spores in marginal marine strata of the late Cenozoic, northernGulf of Mexico. Proceeds IV International Palynology Conference, Lucknow, 1976-77, v.2, p.436-443.Esterle, J.S., Ferm, J.C. and Yiu-Liong, T., 1989. A test for the analogy of tropical domed peat depositsto ‘dulling-up’ sequences in coal beds - preliminary results. Organic Geochemistry, v.14, No.3,p.333-342.Esterle, J.S., Ferm, J.C., Durig, D.T. and Supardi, 1987. Physical and chemical properties ofpeat nearJambi. Sumatra, Indonesia. International Peat Society Symposium on Tropical Peat, 1987.Esterle, J.S., 1990. Trends in petrographic and chemical characteristics of tropical domed peat depositsin Indonesia and Malaysia as analogues for coal formation. Unpublished PhD thesis, Universityof Kentucky, Lexington, 270 p..Esterle, J.S. and Ferm, J.C., 1994. Spatial variability in modern tropical peat deposits from Sarawak,Malaysia and Sumatra, Indonesia as analogues for coal. International Journal of Coal Geology,v.26, p.1-41.Fredericksen, N.O., 1980. Sporomorphs from the Jackson Group (Upper Eocene) and adjacent strata ofMississippi and Alabama. USGS Professional Paper 1084.Gomez, L. D., 1986. Vegetacion de Costa Rica. Vegetacion y Clima de Costa Rica, Vol. 1, University ofCosta Rica, San Jose.Graham, Alan, 1973. (Ed.) Vegetation and Vegetational History of Northern Latin America. Elsevier,Amsterdam, 1973, 393 p..Graham, A., 1979. Mortoniodendron (Tiliaceae) and Sphaeropteris / Trichipteris (Cyatheaceae) inCenozoic deposits of the Gulf-Caribbean region. Ann. Miss. Bot. Gard. 66, p. 572-576.Graham, A., 1 988a. Studies in neotropical paleobotany V: The Lower Miocene communities of Panama- the CuLebra Formation. Ann. Miss. Bot. Gard. 75, p. 1440-1466.Graham, A., 1 988b. Studies in neotropical paleobotany VI: The Lower Miocene communities of Panama- the Cucaracha Formation. Ann. Miss. Bot. Gard. 75,p.1467-79.Holdridge, L.R. and Budowski, G., 1956. Report of an ecological survey of the Republic of Panama.Caribbean Forestry, v.17, p. 92-110.IRHE, 1988. Precipitacion media anual de Ia Republica de Panama, Atlas Nacional de Ia Republicade Panama, Instituto Geografico Nacional “Tommy Guardia”, 1988, p.42-3.130Johnston, I.M., 1949. The Botany of San Jose Island. Sargentia, 8. The Arnold Arboretum of HarvardUniv., Jamaica Plain, Mass., 288 p.Kahn, F. and De Granville, J., 1992. Palms in forest ecosystems in Amazonia: Ecological StudiesVol.95, Springer-Verlag, Berlin, 226 p.Lottes, A.L. and Ziegler, A.M., 1994. World peat occurrence and the seasonality of climate andvegetation. Palaeo., Palaeo., Palaeo., 106, p.23-7.Muller, 3., 1963. Palynological study of Holocene peat in Sarawak. in: Proc. Symp. Ecol. Res. HumidTropics, UNESCO, Kuching, p.147-156.Muller, J., 1964. A palynological contribution to the history of the mangrove vegetation in Borneo. in:L.M. Cramwell (ed.), Ancient Pacific Floras, Univ. of Hawaii Press, p. 33-42.Mustard, P.S. and Rouse, G.E., 1994. Stratigraphy and evolution of Tertiary Georgia Basin, andsubjacent Upper Cretaceous sedimentary rocks. Bulletin of the Geological Survey of CanadaNo.481.Myers, R.L., 1990. Palm swamps. k: A.E. Lugo, M. Brinson and S. Brown, (eds.), Ecosystems of theWorld 15: Forested Wetlands, Chapter 11, p.267-286. Elsevier, Amsterdam.Otedoh, M.O., 1977. The African origin of Raphia taedigera - Palmae. Nigerian Field, 42: p.1 1-16.Phillips, S., Bustin, R. M. and Lowe, L. E., 1994. Earthquake-induced flooding of a tropical coastal peatswamp: a modem analogue for high-sulphur coals? Geology, v. 22, No.10.p.929-32Phillips, S. and Bustin, R.M., (in review). Sedimentology of the Changuinola peat deposit, Panama: ananalogue for some back-barrier coal deposits. G.S.A. Bulletin.Porter, D.M., 1973. The vegetation of Panama: a review. in: A. Graham (ed.), Vegetation andVegetational History in Northern Latin America. Elsevier, Amsterdam, p. 167-201.Roubik, D.W. and Moreno, J.E., 1991. Pollen and Spores of Barro Colorado Island. Monographs inSystematic Botany v.36, Missouri Botanical Gardens.Richards, P.W., 1952. The Tropical Rain Forest. Cambridge University Press, London. 450 p.Rosen, A.M., 1992. Palm phytoliths. in: G.R.Rapp Jr. and S.C. Mulholland, (eds.), PhytolithSystematics: Emerging Issues. Plenum Press, New York, 1992. 350 p.Rouse, G.E., 1962. Plant microfossils from the Burrard Formation of western British Columbia.Micropaleontology, v.8, No.2, p.187-218.Staneck, W. and Silc, T., 1977. Comparisons of four methods for determination of degree ofhumification (decomposition) with emphasis on the von Post method. Canadian Journal of SoilScience, v.57, p.109-117.Scott, A. C., 1991. An introduction to the applications of palaeobotany and palynology to coal geology.Bull. Soc. Geol. France, vol. 162, No. 2, p. 145-153.131Sheffey, M.W. and Dilcher, D.L., 1971. Morphology and taxonomy of fossil fungal spores.Palaeontographica, 133, 1-3, p.34-Si, plates 13-16.Sherwood-Pike, M. A., 1988. Freshwater fungi: fossil record and paleoecological potential.Palaeogeography, -climatology, -ecology, vol. 62, p. 271-285.von Post, L., 1922. Sveriges geologiska undersoknings torvinventering och nãgre av des hittills vunnaresultat. Sv. MosskulturfOr. Tidskr. 1:1-27.132CHAPTER 4SULPHUR IN THE CHANGUINOLA PEAT DEPOSIT, PANAMA, AS AN INDICATOROF THE ENVIRONMENTS OF DEPOSITION OF PEAT AND COAL133CHAPTER 4: SULPHUR IN THE CHANGUINOLA PEAT DEPOSIT, PANAMA, AS ANINDICATOR OF THE ENVIRONMENTS OF DEPOSITION OF PEAT AND COAL4.1 ABSTRACTThe sulphur (S) content of coal is frequently used to infer aspects of paleo-climate, trophicstate, and proximity to marine influence, of the mire in which it was deposited. In this study, the Scontent of peat in a large back-barrier mire on the Caribbean coast of Panama is related to climatic,biological and tectonic factors of the depositional environment. Earthquake-generated subsidence isgreatest to the southeast, leading to drowning of the deposit beneath Almirante Bay, and 40% of thepeat is now below sea level. Coastal mangrove peats with moderately high S content (1 to 5 wt% S)and high salinity (> 0.5 wt%) dominate the eastern margin and extend beneath the salt water andshallow marine sediments of the adjoining bay. Marine influence extends only a short distanceonshore, except in the vicinity of brackish blackwater creeks which drain the swamp. Peats associatedwith these tidal channels are low in salinity (< 0.5 wt%) and very high in S (5 to -.44 wt% S),apparently the result of a biogeochemical chain of S reactions leading to the concentration of C-Ssuiphides. The western part of the deposit is domed, and the vegetation and the peat are concentricallyzoned. Stunted, sawgrass-dominated vegetation which produces fibric, very low S (<0.25 wt% 5)peat, occupies thç central bog plain. Around the bog plain, mixed-forest and palm-forest swampsproduce dense hemic and fine hemic peat with higher S content, between 0.25 and 0.5 wt% S. The Scontent is in proportion to the degree of humification of the peat, and both are independent of the pHof the groundwater. The distribution of organic and inorganic sulphur forms in the tropical peats arefound to be comparable to published values for temperate and subtropical peats, despite differences invegetation and climate. The distribution of high sulphur peats in the eastern and low sulphur in thewestern part of the deposit, and the SE - NW transgression parallel to the trend of the coastlinereflects the regional structural trend of coseismic subsidence greatest to the southeast.1344.2 INTRODUCTIONSulphur content and distribution are important factors in the evaluation of peat and coaldeposits, and are frequently used in the reconstruction of paleoenvironments of coal measures. A largecoastal peat swamp near Changuinola in western Panama has been studied as a model for low-sulphurback-barrier coal deposition. The peat deposit, which averages 6 to 8 m in thickness, has an area of80 km2, about 20 km2 of which is offshore beneath the shallow marine sediments of Almirante Bay(Fig. 4.1). The association of a thick, laterally extensive peat deposit with alluvial, barrier-bar, andshallow-marine sediments makes it an excellent analogue for marine influenced tropical coal deposits.In addition, the deposit is in a tectonically active setting, in which earthquake-generated coastalsubsidence is leading to deposition of carbonate sediments over peat.Primary sulphur content of peat is dependent on the availability of aqueous sulphate to livingplants and bacteria. Aqueous sulphate (SO42-) in rain, groundwater and the oceans is the principalsource of sulphur found in peat (concentration of sulphur is from <ito 8 ppm in fresh water, 885ppm in seawater; Gross 1982). Sulphate is reduced by respiring plants and sulphur-reducing bacteria.Plants use the oxygen, as CO2. for respiration, and fix some of the sulphur, as carbon-bonded (C-S)sulphur, in a variety of amino acids and related compounds, and as so-called ester sulphates (C0S03)in polysaccharides, choline sulphate, phenols, and other compounds (Casagrande et al., 1976).Reduced sulphur, as suiphide, may be taken up by rootlets, and subsequently re-oxidized to a varietyof oxidation states, in the form of polysulphides, elemental, or pyritic sulphur. In strictlyombrotrophic bogs and swamps, the atmosphere is the sole source of primary sulphur, whereas inminerotrophic mires, and those with brackish and marine influence (ie. rheotrophic), groundwater andseawater are important sources. In brackish and marine-influenced peat, sulphur concentration can bevery high (>10% dry weight), much higher than can be accounted for by pore-water sulphate concentration (Luther and Church 1992; Giblin and Wieder 1992). In135c•:STUDYAREA.a.sv-•:-.-PANAM4-Figure 4.1. Site map showing the known extent of the Changuinola peat deposit. Inset map of CentralAmerica shows the location of the site on the northwest coast ofPanama. Sample sites for sulphurtesting are marked. Site name is followed by the mean total sulphur content of the core in brackets, andone standard deviation, in italics.______T5(Ol2*TU * T1(039)05(034).4____ _______________ _ _____(O.52 BDDI(37)26xLE9BDT2Q* tGZ),o7I20’_ISLAI COLON1)46.) r4 dqo.20(3 2)5 BI7, (14) XS)D19A(38)9 (236)1(O2*).o *BDT23C (O1 -Republic ofPanamaCHANGUINOLA PEAT DEPOSITx.* SAMPLE SITE(MEAN SULPHURCONTENT) STANDARDDEVIATiONEXTENT OF PEAT\JI 31(3 8)18ALMIRANTEBAYUNTA PONDSOCK_____________________25NOga 15’.136freshwater, particularly ombrotrophic, settings, sulphur content tends to be low (< 1% dry weight).Thus the sulphur content of a coal is frequently used to infer the trophic nature of the progenitor mire,and its proximity to marine influences (Williams and Keith 1963; Raymond and Davies 1979;Chandra et al. 1983; Shimoyama 1984; Hunt and Hobday 1984; Querol et al. 1991; others).In coals in which no marine influence is evident, sulphur content has been related to otheraspects of the geochemistry of the precursor peat (Cecil et al. 1979; Renton et al. 1979; Renton andBird 1991). These authors observe that acid peats (pH <4) are characterized by well-preserved plantmaterials, low degrees of humification, and low sulphur content. In more neutral peats (pH 4 to 8),the environment is more favourable for bacterial activity, hence humification rates are likely to behigh, preservation poor, and sulphur tends to become concentrated. Acidity in turn may reflect thetrophic state of the mire, the presence of buffers in the groundwater and associated sediments, thedegree of marine influence, and the composition of the peat-forming flora.Since the early 1960’s, work in modern peat-forming environments has contributed much tothe understanding of sources of sulphur in peat and coal. Numerous problems remain however,particularly with regard to rapid lateral and vertical changes in sulphur form and concentration, therelation between peat sulphur and overlying marine elastics and carbonates, and the signatures left byrapid subsidence and marine inundation. The unique setting of the Changuinola deposit provides anopportunity to address some of these longstanding problems, as well as take a first detailed look atNeo-tropical forest-swamp peats from the perspective of coal science.4.3 REGIONAL SETrINGa) Geography, Climate and BiologyTropical coastal depositional environments are believed to have been the settings for manyknown coal deposits. (Wanless et al. 1969; Anderson and Muller 1975) Tropical coastal peats,however, are still relatively poorly understood, and differ in a number of ways from the temperate and137sub-tropical deposits on which many current depositional models have been based. Tropical climateaffects air and water temperatures, seasonality, growth and decomposition rates, and the compositionof the peat-forming floral community.The Changuinola peat deposit has developed in a back-barrier setting on the Caribbean coastof Panama, at 9020’ N latitude, 82° 20’ W longitude (Fig. 4.1), on a microtidal (30 cm), wavedominated shoreline. Although tidal effects in the western Caribbean are minimal, strong wind-drivenlongshore currents transport sediments to the southeast. Tropical storms and hurricanes pass to thenorth, but associated flood events are frequent. The climate is humid-tropical and temperatureaverages 26°C, with a ±3° range. There is no dry season; annual precipitation is about 3000 mm,uniformly distributed throughout the year.Vegetation on mineral soil is diverse and luxuriant. In a 20m x 20m quadrant, 164 species oftrees and shrubs were identified (See Chapter 3, Appendices). This diversity is typical of humidforested lowlands (A. Hemandez, personal comm.). Floral diversity in the peat deposit, however, islow (54 major species representing the entire peat swamp) and distinctly zoned into a sequence of‘phasic communities’ (terminology of Anderson 1964, and Bruenig 1976) similar to those described forthe oligotrophic peat swamps of Western Malesia (Anderson 1964; 1983) and Borneo (Bruenig 1990).The factors which account for floral zonation on peat are not fully understood, but are generallyconsidered to be related to nutrient levels, water table and pH of the groundwater. Thus distinctivephasic communities of peat-forming flora, along with their associated groundwater environmentsdetermine the character of the peat which will develop. Consequently, schemes which identify andclassify peat according to its principal floral components inherently imply much about the environmentof deposition. The Changuinola deposit includes 7 phasie communities, some of which can be furthersubdivided according to the presence of secondary species. These are: i) Rhizophora manglemangrove-swamp; ii) mixed back-mangrove swamp; iii) Raphia taedigera palm-swamp; iv)138Campnospermapanamensis forest-swamp: v) mixed forest-swamp; vi) Cyperaceae (sawgrass) *stunted forest-swamp; vii) Cyperaceae bog-plain.b) GeomorphologyThe NW coast of Panama is a site of active clastic and organic sediment deposition. Much ofthe coastal plain is under a variable cover of Quaternary and Recent alluvium, and the coast ischaracterised by a series ofprograding barrier bars behind which a chain of lagoons and paralicswamps is developed. Several large rivers transport sediments northward off the TalamancaCordillera, and wind driven currents move sediments consistently alongshore to the southeast.Sedimentation rates are high, and the peat deposit has developed on the back of a prograding barrierbar, the characteristic ridge-and-swale architecture of which is described in Chapter 2. Basalsediments are well-sorted medium grained grey sands similar to those currently being deposited on thecoast by the Changuinola River. At the eastern termination of this coastline, where the terrigenoussediments extend beneath Almirante Bay, peat occurs submerged to depths of 3 m or more, underlainby medium grained grey sand, and in places overlain by a thin layer of silty sand, lime mud, and coral.Levelling lines surveyed across the peat deposit show a slightly domed surface that has amaximum elevation of 6.75 m in the centre of the swamp, a maximum gradient of 1:330 (the NEmargin) and a base that is from 4 to7 m below sea level. An estimated 40% of the volume of the peatis below present sea level. An offshore sample from 475 cm below sea level was radiocarbon dated at2110 ± 60 a. This mangrove peat is overlain by 265 cm of brackish-water peat, 15 cm of carbonatesediment, and 195 cm of salt water. Assuming deposition close to mean sea level, and only minorchanges in regional sea level over this period (Pira.zzoli 1991; Chapter 2), about 4.5 m of netsubsidence is estimated to have occured in the past 2000 years, including that generated by a M(surface magnitude) 7.5 earthquake in April 1991, 10 weeks before the start of this study.139c) Geological SettingThe study area is at the eastemmost onshore extent of the Limon-Bocas del Toro sedimentarybasin. The Bocas del Toro Basin in Panama, and its westward extension in Costa Rica, the LimonBasin, together make up the Tertiary and Quaternary back-arc basin behind the volcanic ranges of theTalamanca Cordillera of southeastern Central America. Chapter 2 gives an overview of the regionalgeology, and the tectonic setting of the area.Regional vertical movement along the coast has been established on the basis of sea levelchanges as observed in the degree of exposure of near-shore coral reefs, limits of intertidal organismson dock pilings, and changes in the swash zone on beaches at eighteen points between the mouth of theRio Matina in the northwest and Punta Mona (Gandoca) to the southeast (Astorga 1991; Plaflcer andWard 1992). These observations have been augmented (Camacho, unpublished data; See Chapter 2)to include the coastline of Panama from the Rio Sixaola at the Costa Rica border to the Boca del Torochannel and Cayo Solarte to the southeast. In addition to regional vertical movements are localliquefaction effects, on a scale of tens to hundreds of metres, both of subsidence and uplift (Camachoet al. 1992). It is concluded that 50 cm to 70 cm of regional structural subsidence occurred. Theseconclusions are based on aerial photographs, observations of the swash zone of beaches and of plantmortality along vegetated shorelines in the area, and on interviews with individuals familiar with thearea and with a working knowledge of sea and river levels before the event.4.4 SAMPLING AND EXPERIMENTALFigure 4.1 shows in plan view the extent of the peat deposits between the Changuinola Riverand Isla Colon. Marked sample sites represent cores taken to the base of the peat, by Cohen andothers (1989) and by Phillips and Bustin (1991 to 1993). Most samples were collected using Hilleror Macauley-type hand-operated coring devices. Samples were described in the field, and bagged in25 cm or 30 cm increments. Continuous cores were collected in 3.5 cm. diameter plastic sleeves usinga small vibracore, refrigerated and subsequently frozen for long term storage. Salinity and pH140were measured in the field and verified in the laboratory. A digital pH-meter (Cardy® Model PHi)was used both in the field and the laboratory to measure the pH of the wet peat and porewater. At thesame time, salinity was measured using a digital salt-meter (Cardy® Model C 121), which measuressodium ion concentration, and then applies a correction factor to calculate total salinity for values>0.01 wt%. Total sulphur content (dry weight percent) of 203 samples (dried at 50°C, crushed to 100mesh) was determined using a Leco® SC-132 Sulphur Analyzer (see Tabatabai 1992, p.3 13 for adescription of this instrument) and verified using wet chemical methods. For a coastal sample site,and for site BDD23 adjacent to a major drainage channel, total sulphur was sub-divided into organicsulphur, in both carbon-bonded sulphide (C-S) and ester sulphate (oxidized compounds reducible byHI) form, and inorganic forms expressed as mineral sulphate, elemental sulphur and pyritic (+marcasite) sulphur. The procedures for determining the proportions of the various forms of sulphurare described in detail by Lowe (1986) and Tabatabai (1992) (Appendix G), who also discuss some ofthe shortcomings of the methods. In addition, previously published total sulphur values of another 206samples (Cohen Ct al. 1990) are incorporated in this study.Degree ofhumification of the peats is based on the relative proportions of coarse, medium andfine constituents as determined using a wet-sieving procedure modified from Staneck and Silc (1977),according to the following scheme (Esterle et al 1987):Coarse >25% >2.0 mm <30% <0.25 mmMedium <25%>2.Omm <30% <0.25mmFine <25% >2.0 mm >30% <0.25 mmThe procedure is described in detail in Chapter 2. Degree of humification is described using amodified von Post Humification Scale adapted to tropical peats by Esterle (1990) (Appendix G). Thecategories are sapric, fine hemic, hemic, coarse hemic and fibric. Peat classification is based on theidentification of macroscopic plant parts and palynomorphs in the peat, compared to plant and pollenassociations present in the modem swamp. For some samples, the percent moisture content of141drained wet peat was measured by re-weighing after drying at 50°C until no further weight lossoccurred.4.5 RESULTSa) Geochemical Parametersi) Sulphur and Salinity.---Sulphur content of the Changuinola peat varies from less than .01wt% to in excess of 13 wt%. Although a full spectrum of sulphur values is present, the samples fallinto low-sulphur (.01 to 1.0 wt%) and high-sulphur (> 1.0 wt%) populations when plotted againstother parameters such as pH (Fig. 4.2). Average sulphur content of each core, shown on Figure 4.1,shows a geographical distribution, in which the western part of the deposit is made up of low sulphurpeats (mean of 0.23 wt%, standard deviation .18; Fig. 4.3), The eastern part, comprising peatwithin 3 km of Almirante Bay, has much higher sulphur content and much greater variability (mean of2.24 wt%, s = 2.55; Fig. 4.4). Figure 4.4 further reveals two populations of sulphur values in theeastern section, a lower- and a higher-sulphur group. Both plots suggest that there is little correspondence between sulphur content and depth of burial (and hence age) of the sample, but that lateraldistance from the south-east marine margin is significant.To claril’ the apparent relationship between average sulphur content and proximity to themarine influence ofAlmirante Bay, the salinity of all samples was plotted against sulphur. Salinity isreadily measurable in the field, and in the present study proved a useful indicator of the presence ofmarine influence in porewaters. Salinity cannot be related directly to aqueous sulphate contentwithout qualification: aqueous sulphate may not diffuse through peat pore water in the same manneras other mineral salts. Sulphate diffusion is biologically mediated (Brown 1985) and thus driven by adifferential that is heavily modified by an ongoing “sulphate reduction front”, whereas diffusion ofmineral salts is determined primarily by the ability of groundwater to circulate. This may createcomplication in using salinity to approximate the movement of sulphate-rich waters in rheotrophicmires, but is unlikely to be a problem in ombrotrophic situations, in which the supply of atmospheric142Figure 4.2: wt% Sulphur (logarithmic scale) vs p!-I. n216. 1431432.50 3.50 4.50100pH5.50 6.50 7.50 8.50TotalSulphur(wt%),,nu.UUU U:U U UUUUUUUUUnUIW.UUUUUU3UUU1010.10.01Depth in cm*AAi AA A LA*A aMLa aFigure 4.3. Scatter plot of wt% sulphur vs depth, western part of the deposit. n = 244.aamaa maIaaaaaaaFigure 4.4. Scatter plot of wt% sulphur vs depth for samples in the eastern part of the deposit.Note the distinct bimodal distribution in sulphur content. n = 144.10 100 100010010wt%sulphur__________________________ ___________________________(logarithmicscale)0.10.01Depth in cm10 100 100010010wt%sulphur 1(logarithmicscale)0.10.01144sulphur to the mire can be assumed to be relatively constant, and uniformly distributed. With theexception of three basal samples, the highest salinity found in the ombrotrophic western peats is 0.02wt%, the vast majority of the samples being below the sensitivity of the meter (0.01 wt%). This isalso true of the porewater in the basal sediments at all sites measured. In the eastern peats, some ofwhich are rheotrophic, salinity is highly variable (<0.01 to 2.7 wt%). For interpretive purposes threesample populations are differentiated based on a plot of S vs salinity (Fig. 4.5). These are:Group I: sulphur < 1.0 wt%; salinity < 1.0 wt%Group II: sulphur> 1.0 wt%; salinity < 1.0 wt%Group III: sulphur> 1.0 wt%; salinity> 1.0 wt%The three groups can be used to map porewater environments. Figure 4.6 (plan view and cross-section) approximates the three-dimensional distribution of the three groups. In the western section,all peat is Group I, except at the bases of the three deepest cores, which are placed in Group II (basedon low basal porewater salinity of nearby sites). In the eastern section, the distribution is morecomplex (reflecting the complex history of the marine margin). Some inland sites are entirely Group I,whereas all the marine-margin and offshore sites are Group ifi. Group II samples (medium to highsulphur, low salinity) are found only near major drainage features (blackwater creeks). Salinity ofsurface waters (upper 3 m) along the marine margins is between 2.9 and 3.2 wt%. Salinity profiles ofthe three major blackwater creek channels which drain the eastern half of the deposit are shown inFigure 2-5. Salinity and pH were measured at points 1 km into the swamp. Stream banks are linedwith dense stands ofmixed Rhizophora mangle (red mangrove) and Laguncularia racemosa (whitemangrove), and tidal effects and mixing are prominent despite there being only a 30 cm tidal range.Salinity of creek bottom water is 70% of the salinity of the bay, and at 1 m below the surface creekwater is 25%, of bay water salinity.ii) Sulphur and pH. —The pH of all samples collected was measured in the field, and pH oflocal waters was also tested at regular intervals: pH of rainwater is 6.9 to 7.0, and the sea water in theCaribbean and Almirante Bay between 6.9 and 7.4 in1451.5Salinity(wt%)1.00.5Total Sulphurwt%GROUELJILFigure 4.5. Scatter plot of total sulphur (logarithmic scale) vs salinity. The fields Group I, Group II andGroup ifi represent sulphur-salinity groups as defined in the text. n = 213.QI2.52.0 I----- --I. I— — - - - -— I- - - - -I II III—-i”..II————--I I IzLHIIfl IIIflI LI 111 I!!UII iz0GROUP I GROUP II146LAKE9:CHANGUINOLA::::::::2::::::::::::2:::::::W18kmLIGroupI-lowsulphur,lowsalinityGroupII-highsulphur,lowsalinityGroupifi-mediumsulphur,highsalinityFigure4.6.Arealanddepthdistributionofsulphur Groups 1,11andillintheChanguinola peatdepositAcross thetopofthediagram, variationsintotalsulphur andpHwithdepthalongaW-Ecrosssection.Below,planviewandcrosssectionfromChanguinolaRiver toAlmiranteBay.Asterisk()signifies datafromCohenetaL,1989.Wt %total sulphur0.102030.40.38the upper 3 m. The pH of water in the blackwater creeks, measured after about one week of dryweather, is near-neutral (Fig. 2.5). The near-neutrality of the drainage water in the blackwatercreeks contrasts with the commonly low pH of water draining from temperate peat bogs. Forexample, very acidic bog water results in fish kills in Norwegian streams at the end of dry periods,when low water tables resulted in the oxidation of suiphides in the acrotelm. The resulting sulphate isthen flushed out of the peat, with the onset of wet weather (Brown 1985). The ever-wet conditions inthe Changuinola swamp, combined with the buffering effect of admixed seawater, presumeablymaintain a high pH, and thus very favourable conditions for bacterial sulphate reduction, at least inthe immediate vicinity of the creeks.Figures 4.7-1 and 4.7-2 plot pH against depth for samples in the western section (n = 321)and the eastern section (n = 387) of the deposit respectively. In the west, pH is low (mean of 4.4) andfairly uniform (standard deviation of 0.6), and displays a moderate positive correlation with increasingdepth (r = 0.4). Eastern samples are on average less acidic (mean of 5.7) and more variable (s = 1.1).Correlation with increasing depth is very weak (r = 0.16). The higher degree of variability reflects thecomplex history of the eastern section and the presence of marine and brackish influences at varyinglevels in cores.The contrast in geochemical values between eastern and western data sets, as well asdifferences in vegetation, and the physical character of the peat, reflects differing hydrologicalconditions in the ombrotrophic western part of the deposit and the locally rheotrophic, marineinfluenced eastern section (See Chapter 2). In order to clarify the distinctions, the results will bedescribed under separate headings for east and west.b) Total Sulphur Distributioni) Sulphur andpH in the Western Section ofthe Deposit.---The western section of the depositis a slightly domed, roughly concentrically-zoned mire, the centre of which is an extensive bog-plain.148-I >0—C).C)‘0C).‘H 1.__—pH34567pH24U—flUI.U——.,•Depth——incm——__100————___-200-——____—•__“+:IIt_300-—•—•—_II__——•400--500--•——•—::600700Figure4.7-2.pHvsDepthintheeasternsectionofthedeposit;n=387,meanpH=5.7(s.d.=1.1), range2.67to7.89.r(correlation) =+.16Depthincm 100200300400--500-600--700--800--900--UFigure4.7-1.pHvsDepthinthewesternsectionof thedeposit;n=321,meanpH=4.4(s.d.’O.6),range2.82to6.29.r(correlation)The surface of this region is permanently submerged, and drainage is by surface run-off in a radialpattern. The bog plain vegetation is an impoverished community dominated by sawgrass (sedges),together with Sagittaria Iancfo1ia, Dieffenbachia longispata, ferns, grasses, mosses and algae.Stunted Myrica mexicana-Cyrilla racemflora tree hammocks are scattered sparsely across the plain.Sawgrass-stunted forest-swamp, Campnosperma panamensis and mixed forest-swamp, and Raphiapalm-swamp surround the central bog plain.Despite the variations in vegetation, the absolute sulphur content of all peat types is low.Sulphur is lowest in the bog plain peats, and slightly higher in the marginal zones. Towards thebarrier bar, windborne suphate may add to nutrient availability, and in shallow peat towards thesouthwest margin groundwater-borne nutrients may be more available. In all cores, pH (dashed linesin Fig. 4.6) shows an erratic but overall increase with depth, coinciding with increases in the degree ofhumification of the peat. This increase overshadows the effect of concentrated humic and fulvic acidsin the more degraded peats, which would tend to lower the pH at depth, but the same trend ofincreasing pH with depth is found in almost every core. The trend in pH does, however, echotransitions in peat type, and in floral association as detennined by pollen analysis, from principallyCampnospermapanamensis and mixed forest- and palm-swamp peat in the lower parts to bog-plainsawgrass peats in the upper parts of the cores. Utilizing particle size distribution by wet-seiving toquantify the degree of humification of the peat, a very close correspondence between increases inhumification and increased sulphur concentration is found (Figure 4.8).In the bog-plain cores, the pH of the underlying forest-swamp peats is higher than that of theoverlying sawgrass peats. Thus higher pH is associated with forest-swamp peat, increasing depth(age), and slightly higher total sulphur content. To further explore these three relationships, datafrom the bog plain (MILE 5 - Fig. 4.8) is compared to that from site LAKE 2, located in mixed forest-swamp 1200 m from the barrier beach. This core, Figure 4.9, is composed of fme hemic woody peatthroughout its entire depth, and the particle size distribution is consistent throughout. The sulphur150Figure 4.8. Variation inparticle size (humification)pH and sulphur in an 870cm core from the centralpart of the deposit (MILE5) showing the closecorrespondence betweenincreases in humificationand increases in sulphurcontent.The particle size categoriesare coarse (c >2.0 mm),medium (m> 0.25 mm),and fine (f< 0.25 mm).Figure 4-9. Variation inparticle size (humiflcation)pH and sulphurconcentration in aforest-swamp core fromnear the margin ofthewestern section (LAKE 2),showing an inverserelationship between pHand sulphur contentParticle sizes are as for Fig. Size Distribution Wt% Sulphur5.5 20 0 0.2 0.4—z1NCRE4S1NG INCRFASINGHUMIFIQ4TION SULPHURpH % Size Distribution Wt% Sulphur4 5 o o 6o 0.2 0.3 0.4iNCREASING INCREASINGHUMIFICATION SULPHUR-151concentration is higher than in sawgrass peat, and decreases slightly with depth independent ofhumification. The pH of the LAKE 2 forest-swamp peat is lower (avg. 3.7) than that of the MILE Score (avg. 4.2), and pH increases with depth. This low pH peat is both highly humifled, and relativelyhigh in sulphur content. Low sulphur peats originating from forest-swamp vegetation have highersulphur content than do sawgrass peats from the bog plain, regardless of pH. The range in sulphurvariation is considerable (0.16 wt%); thus fine hemic peats have about twice the sulphur content ofcoarse hemic peat. All samples had salinity below 0.02 wt% (Group I).ii) Sulphur, Salinity andpH in the Eastern Section ofthe Deposit.---The upper panel inFigure 4.6 shows sulphur and pH values for a series of cores along the NW-SE cross-section from theChanguinola River floodplain to the submerged peat beneath Almirante Bay. The eastern part of thedeposit is that part within the influence of tidal creeks which drain into Alnürante Bay, represented bycores BDD 8, 23, 20 and 19A in Figure 4.6.. Present vegetation is a complex mosaic of mixed forestswamp, monospecific stands of Campnosperma panamensis, Raphia palm and sawgrass(Cyperaceae), with mixed mangrove and back-mangrove communities bordering the creeks up to 2 kmfrom the coast. The bay margin is bordered by Rhizophora mangrove fringe forest. The complexityof the vegetation reflects the complex hydrology of this section, which is divided by the three majorchannels (Figs. 4.1 and 2.5) into areas of low salinity and sulphur (Groupl) forest and sedge peats,bordered by Group II back-mangrove peats, and Group III Rhizophora peats.Site BDD 8 is in mixed forest-swamp near the head of a blackwater creek 3 km inland, andBDD 23 is 1.5 km closer to the bay, in similar vegetation. The sites are about 50 m from the presentchannel of Canal Viejo, and salinity is <0.1%. In the upper 2.5 m at both sites the peat is forest-swamp and Raphia peat, pH is below 4.5, and sulphur content is less than 0.5 wt% (i.e. Group I).Lower in the cores, pH increases rapidly to near-neutrality. Sulphur remains low at the inland site,but increases to 13.7 wt% at BDD 23 in back-mangrove (Laguncularia-Acrostichum-Rhizophora)peat (Group II). At the coast (BDD 20) pH varies between 5.4 and 6.3. Sulphur averages 3.2 wt%,152pHWt% SulphurFigure 4.10. Total sulphur and pH vs Depth forcore CS 3.2 4 6 8255075100125150175200D 225(cm)250275300325350375— I ISpHI I I I I0 2 4 6 8 10153and increases with depth (Fig. 4.6). About 600 m offshore at BDD 19A, sulphur is slightly higher(mean 3.8) and pH is neutral to alkaline (7 to 7.6). Both are Group III, predominantly Rhizophora,peat.In the coastal mangrove fringe (Group III) peats, pH averages around 5.8 (s = 1.35; n =237). The peat is consistently low in moisture content and highly humified (Fine Hemic to Hemic).Group II forest-swamp peats associated with drainage channels also are highly humified woody peats,but are lower in pH. Figure 4.10 shows pH and total sulphur for core CS 3, a site located 25 m fromthe banks of Caflo Sucio (Black Creek) and 2 km upstream from the mouth. pH for peat in the coreaverages 3.4. Sulphur content varies from 0.52 to 8.09 wt% with no relation to pH. The basal sand25 cm below the contact has a pH of 7.46 and a sulphur content of 3.19 wt%. From such cores it isevident that highly humified, high sulphur peat can form in a low pH environment.c) Forms ofSulphurIn order to compare the sulphur distribution of Groups I, II, and III peats in the complexcoastal zone, the total sulphur content of a coastal mangrove peat (site BI 3) and forest-swamp - backmangrove peat (the upper and lower parts of core BDD 23 respectively) is sub-divided into organic(C-S or carbon-bonded 5, and ester sulphate), and inorganic forms (mineral sulphate, pyritic S andelemental S). The results of the analysis are summarized in Tables 4.1 and 4.2.I) BI 3: An Example ofa Group III Peat.---The pH, total S, salinity and the distribution offorms of sulphur in a 2.5 m mangrove peat core taken at site BI 3, across Almirante Bay on the shoreof Isla Colon are shown in Figure 4.11. The top of the core is above sea level, but the peat porewater becomes strongly saline with depth. There is a very close correspondence between increase insalinity of the peat and the sulphur content, most particularly in aqueous sulphate content, and also inester sulphate. It is not clear why the pH decreases at the level where salinity increases; dissitnilatory154TABLE 4.1: SULPHUR FRACTIONS IN BI 3 RHIZOPHORA PEATNo SPMPLESITE DEPTH TOTALSULPH %INORGANIC HI-S S04 PYRITIC ELEMENTAI %ORGANIC C-S ORGSO41 813-0-25 25 1.16 0.27 0.015 0.006 adpH=6.42 0.22 0.013 0.0040.014mean 0.24 0.014 0.005 0.916 0225means as % of Total 100% 2% 1.19% 0.04% nd 98% 79% 19.40%2 813-50-75 75 2.28 1.53 0.555 0.02 0.106pH=5.7 1.42 0.576 0.022 0.112mean 1.48 0.563 0.021 0.109 0.802 0.785means as % of Total 100% 30% 24.70% 0.93% 4.80% 70% 35% 34.40%3 813-100-125 125 4.33 1.81 0.86 0.07 0.106pH=5.63 1.66 0.778 0.013 0.261mean 1.74 0.819 0.042 0.184 2.59 0.692meansas% ofTotal 100% 24% 18.90% 0.10% 4.20% 76% 60% 16%4 813-150-175 175 3.78 1.31 0.713 0.016 adpH6.22 0.93 0.81 0.015mean 1.12 0.762 0.015 2.66 0.371means as % of Total 100% 20% 20.10% 0.40% nd 80% 70% 9.80%Table 4.1. Forms of sulphur at four levels in core BI 3.TABLE 42: SULPHUR FRACTiONS IN ODD 23 FOREST PEATand BACK- MANGROVE PEATNo SAMPLESITE DEPTh TOTALSULPH %INORGANIC HI-S S04 PYRITIC ELEMENTAI %ORGANIC C-S ORGSO41 BDD23-5 175 0.38 0.0718 0.001 ad adpH4.07 0.0572 0.0006Campnospema mean 0.0645 0.0008 0.316 0.064meansas%oftoIalS 100% 020% 0.20% ad ad 99.80% 83% 16.80%2 BD023-8 250 0.78 0.1301 0.008 ad adpH=4.38 0.1415 0.007forest-swamp mean 0.1358 0.008 0.644 0.128means as % of total S 100% 0.60% 0.60% nd ad 99.40% 83% 16.40%3 8DD23-11 400 13.70 2.9685 0.93 0255 0.11pH=4.96 3.1548 0.744 0.254 0.069back-mangrove mean 3.0617 0.837 0.255 0.09 10.64 1.88means as % of total S 100% 8.30% 6.10% 1.86% 0.70% 91.70% 78% 13.70%4 BDD23-13 450 1.43 0.8647 0.367 0.026 0.015pH5.94 0.9173 0.317 0.033 0.01back-mangrove 0.018mean 0.891 0.342 0.029 0.014 0.539 0.505means as % of total S 100% 27% 23.90% 2.08% 1.00% 73% 38% 35%Note: HI-S is sulphur reducible by HI; C-S is carbon-bonded sulphur.n.d. = not determinedTable 4.2. Forms of sulphur at 4 levels in core BDD 23155+I.++-h+Figure 4.11. Summary of the geochemical characteristics of a coastal mangrove - back mangrove peat core, takenjust above the high tide line at site BI 3. Variations in salinity, total sulphur, forms of sulphur and pH with depth areshown. The increase in the proportion of inorganic sulphate corresponds with an increase in salinity. Sulphur datafrom Table 7.0% of Total Sulphur20 40 60 80Hcm50100150200250++0 1:0 2:0 3:0 4:0 5.0 SulphurS Sal,)*Ô 0:4 1:2 1:6Wt% Salinitysulphate c_s1elementalpyriteForms of Sulphurdata from Table 4.1156sulphate reduction, the process carried on by sulphate-reducing bacteria, increases alkalinity, whichbuffers pore water pH (Giblin and Wieder 1992), but any buffering effect from this activity or fromthe intrusion of marine waters is not apparent here. The opposite occurs, as the pore water pH dropsfrom 6.4 to 5.4 in the upper 1 m.There is an added complication in analysing shoreline cores from the Changuinola deposit as aresult of recent seismic activity. When sampled, the peat surface at core B13 was 10 cm above hightide level, but 13 months earlier, this site was 50 cm higher in elevation and somewhat farther onshore.Thus there has been substantial modification of the porewater geochemistry as a result of earthquake-driven subsidence (See Chapter 5; Phillips et al., 1994). Samples from farther onshore show thatprior to subsidence, salinity in this core would have been lower. Consequently, the aqueous sulphatecontent in particular is higher than normal for mangrove peat. Additionally, the total sulphur is lowerthan in other mangrove peats tested. An analysis of means test (paired 2-sample t-test; n = 34)comparing the B13 peats with intertidal mangrove peats shows a significant difference in means (cc=05) of total sulphur content: 2.19 at BI 3, vs 3.45 for 34 other mangrove peats. Figure 4.12 plotssalinity profiles of five mangrove-peat cores, 3 onshore and 2 from> 500 m offshore, and shows thatthe general trend is to a decrease in salinity with depth in the onshore cores, as hydraulic head in thefresh groundwater system, combined with low permeability of the dense woody peat, resists marineintrusion. Offshore cores are more saline, particularly near the top. Again, the BI 3 core is notexactly typical, resembling neither the onshore nor the offshore curves. The relationship betweensalinity and aqueous sulphate is reasonable, however, and other aspects of the distribution,particularly from the uppermost sample, are considered to be applicable to mangrove-fringe peats.Casagrande and others (1977; 1986) found that ester sulphate normally comprises around24% of the sulphur in high sulphur peats. This agrees fairly well with the findings of Phillips et al.(1994) (ester sulphate averaged 19.9% of total S in Panamanian peat), and with Lowe and Bustin(1985) (ester sulphate averaged 18.9% of total S in Fraser delta peats). Those studies documented157Figure 4.12: Wt% salinity vs Depth inS mangrove cores. 2of the cores are now offshore and permanently submerged and3 are in the intertidal mangrove fringe forest.Wt% Salinity0 1 2 32575125175225275Depth(cm)325375425475525565r:c::.(offshore —core)BDD27BDD32I (offshorecore))BDD 31.I158marine-margin peats which have not been overprinted with a transgressive signature (see Chapter 5),and suggests that in this core both 34.4% (at 75 cm) and 9.8% (at 175 cm) are unusual values. Thisspatial variability in ester sulphate content likely reflects variable responses to the recenttransgression due to the heterogeneity of redox environments and bacterial populations within thepeat. The specific causes and direction of, changes in this fraction are unclear and difficult tointerpret. High ester sulphate, along with high (>4%) elemental sulphur content in the middle of coreBI 3, suggest that sulphate-reducing bacteria may be actively producing H2S stimulated by the influxof marine water. Low ester sulphate content can also be an effect of bacterial reduction of organicsulphates, but with the very high aqueous sulphate concentration in this sample (20.1%), bacteria arenot expected to favour ester sulphates for reduction. Pyrite content is very low.ii) BDD 23: An Example ofGroup land Group II Feats .---The sulphur profile of core BDD23 records a change in the environment of deposition at this site, from a back-mangrove forest near thebase, to Raphia - sedge, and then to Campnosperma - mixed forest in the upper 250 cm. Thetransition is reflected in the change from Group II at the base to Group I above. Group I is defined aslow salinity, low sulphur, and Group II as low salinity, high sulphur peat. Table 4.2 and Figure 4.13summarize the results of the analysis of forms of sulphur at four levels of core BDD 23. The samplesare from depths of 175, 250, 400 and 450 cm. In all samples organic sulphur is dominant, comprisingfrom 99.8% to 9 1.7% in the peat samples, and 73% of total sulphur in the sandy peat at 450 cm. Thetotal sulphur curve shows the remarkable contrast between the upper 250 cm and the lower 240 cm ofpeat in this core. The transition is evident in the % moisture curve; the forest-swamp and sedge pealsare wetter, less dense, and more acid (pH around 4.15 vs 5) than the back-mangrove peat.In the low-sulphur (Group I) samples from the upper part of the core the distribution ofsulphur forms is consistent with peat from an onibrotrophic mire: there is a small amount of aqueoussulphate present, but the bulk is split between carbon bonded (C-S) sulphur (83% of total S) andorganic (ester) sulphate (>16%). These values are similar to those recorded by Casagrande et al.159pH Wt% Sulphurcm51060710 % of Total Sulphur ‘ ‘20 40 60 80 (S %M100I200k J I+ [ j *I ...,....300 ...+ IE .400 lU.Ii. /.:i.•fll% sulphate C-S estersulphateelemental600_______________________‘pyrite Forms of Sulphur • • , ,data from Table 4.2 70 75 80 85 90 95% MoistureFigure 4.13. Summary of the geochemical characteristics of core BDD 23, located about 50 m from Canal Viejo, atidal blackwater creek which drains the eastern part of the deposit into Almirante Bay. Shown are variations informs of sulphur, total sulphur, moisture content and pH with depth. Moisture content is used as proxy data forpeat density, as described in Chapter 2. Sulphur data from Table 4.2.160(1977) for low sulphur peats from the Okefenokee Swamp. The source of this assimilatory sulphur ispredominantly the peat-forming vegetation itself, as discussed later. Variations in concentration mayreflect specific vegetation, or the degree to which atmospheric sulphate is available for assimilation.Periodic high water table and high surface runoff may lead to little sulphate entering the soil, whereasdrier conditions and ‘more permeable peat may increase available sulphur in the system. Atmosphericsourced sulphate is rapidly assimilated into the organic sulphur pool, in a matter of hours or days(Brown 1985) under anaerobic conditions through the medium of dissimilatory sulphate-reducingbacteria. However, the absence of detectable elemental sulphur and the uniform proportions of estersulphate and C-S suggest very low levels of bacterial sulphate reduction. The pH of these peats (meanof 4.15) is below normal tolerance for sulphate-reducing bacteria (Postgate 1984). Sources ofFe2+ inthis elastic-sediment starved ombrotrophic environment are limited to plant materials and pyritefonnation is likely Fe as well as pH-limited. Thus the sulphur content of peat at the upper levels hasnot been influenced by sulphate-rich water present in the nearby tidal creek or groundwater.The lower levels in core BDD 23 have high total sulphur content, particularly at the 400 cmlevel (13.7 wt% S), which is 90 cm above the base of the peat. Aqueous sulphate has been in amplesupply, although it is presently in much lower concentration (6.1% of total) than it is in the basalsandy peat (23.9%), or in the newly-flooded peats of BI 3 (24.7%). Aqueous sulphate is present in thesame proportion as in BI 3 mangrove peat that is not affected by secondary enrichment (i.e. normal fora marine-margin peat). Proportionally, the sulphur distribution most resembles that of the uppermostsample at B! 3, and others which have not been overprinted, although here the ester sulphate fractionis somewhat lower than in those samples. This may reflect a difference between mangrove fringevegetation (BI 3) and this mixed-mangrove vegetation. The higher aqueous sulphate content of thebasal sandy peat may be due to lower levels of sulphate reduction, or to its greater permeability andproximity to the underlying rooted sand, and brackish waters of the creek, but salinity is below .01wt%. Pyrite represents 2% of total S in the basal sediments, 1.9% at 400 cm., and was not detectedin the top 2.5 m. This low pyrite content differs from the BDT 3 site analysed by Cohen et al. (1990)161in which pyrite was interpreted to be the dominant sulphur form, based on SEM and petrographicanalysis. That sample is mangrove peat with 14.9 wt% total sulphur. Rhizophora root peat frequentlyhas high pyrite content (Cohen et al. 1984; Altschuler et al. 1983). However, at pH 4.9, the mixed-mangrove peat in this sample is somewhat more acidic than previously studied mangrove peats.4.6 DISCUSSIONa) Sulphur, pH and Marine InfluenceThere is a clear relationship between marine influence and sulphur content in the Changuinolapeats. In the western section of the deposit the peat is consistently low in sulphur. Cohen et al (1990)observed that sulphur content increased to the northeast, in the direction of the barrier bar, andsuggested a possible marine influence from the Caribbean Sea. This study found no evidence ofmarine influence (beyond windblown saline aerosols) in the peals developed behind the barrier.Cohen et al. (1990) also found the lowest sulphur values associated with sedge-grass-fern peats(0. 1%-0.3%), and somewhat higher values for the forest-swamp peats, as is confirmed by this study.Increase in sulphur coincides with higher degree of humification towards the margins of the deposit,just as it does towards the base, where pahn-forest and swamp-forest peats dominate. Sulphur inthese peats is principally assimilatory and high degrees of humification concentrate resistantcompounds, including those that incorporate sulphur.The variability evident in the eastern part of the cross-section (Figs. 4.1, 4.4) reflects theprofound effects, but limited extent, of marine influence. In the upper half of core BDD 23 (Fig.4.13), 50 m from a tidal channel and almost completely below sea level, sulphur, pH and salinitytrends resemble those of the ombrotrophic western samples (i.e. MILE 5, Fig. 4.8). Sites associatedwith channels are lower in pH, and higher in sulphur content, although more variable than the strictlymarine samples, and resemble the lower part of BDD 23. Total sulphur and pH of coastal mangrove-fringe sites at the shoreline resemble those of BDD 20 (Fig. 4.6), and to some extent BI 3 (Fig. 4.11),particularly the uppermost part of that core. In the inter-tidal environment total sulphur content tends162to increase in the upper metre or so, to 3% to 4% sulphur, and remain at those levels until near thebasal sediments, where it decreases. The relationship between assimilatory sulphur concentration anddegree of humification that is found in the low sulphur peats of the central part of the deposit is notevident in the marine- and channel-margin peats: the scale of variation in total sulphur is so great insamples associated with drainage channels that any relationship to degree of humiflcation isovershadowed.The expected relationship between high sulphur content and near-neutral pH, based on the pHpreferances of the sulphur-reducing bacteria, is not evident in the brackish peats. Channel marginpeats have pH around 5 and yet high sulphur content (5 - 14 wt%, 92% of which is organicallybound). This high reduced organic sulphur content is a most interesting aspect of these channel-margin peats. Carbon-bonded sulphur compounds in peat and soil are not well understood, butinclude thiols, organic sulphides and disulphides, and sulphur-containing hurnic and fulvic acids whichcould be diagenetic products of reactions under reducing conditions between lignin-type materials andinorganic reduced sulphur (SH-, Sf2) (Luther and Church 1992). Given anoxic conditions,microenvironments of tolerable pH, an ongoing supply of mineral sulphate, and sufficient organicmatter, these very high concentrations imply an enduring biogeochernical chain of sulphur reactions,with carbon-bonded forms as the end result.b) Sulphur and VegetationHigh degrees ofhumification are associated with woody forest-swamp peats and lesserdegrees with bog-plain peats. This is in part a reflection of environmental conditions, and secondarilya consequence of the floral response to these conditions. Preservation is best and humification lowestin the topographically high central region. Here, the floral community of sparse sawgrasses, stuntedarborescent species, shrubs, mosses and algae reflects the oligotrophic state of the mire. Competitionfor available nutrients contributes to the high degree of preservation: for example. Sphagnum spp.adapted to highly stressed environments, have the capacity to lower the pH of the growing163environment, reducing its viability for less adaptable taxa, and thus reducing species diversity. LowpH, combined with a large proportion of the biomass of the stunted vegetation being in the form ofroots and rootlets, means that biomass production is low but preservation is high. The fibric andcoarse hemic peats which are formed in this environment are of relatively low bulk density, high moisture content, and do not concentrate sulphur to high levels. Observed S levels for coarse hemicsawgrass peats are around 0.10 to 0.25 wt% sulphur.Woody forest peats form in environments dominated by a luxuriant arborescent vegetation,much ofthe biomass of which is above the peat surface. Intense degradation above and just below thesurface results in a high proportion of fine- to clay-size particles in the upper level of the peat. Tothese may be added more resistant lignin-rich elements of wood, bark and cuticle, resins, spores andpollen, ingrown roots, and a significant chitinous component from insects. Table 4.3 lists the totalsulpur content of a number of fresh plant parts from this and other studies. The fine hemic and hemicpeat thus formed is dense and compact, with relatively low moisture content and high bulk density andsulphur content. The most highly degraded forest-swamp peats (LAKE 2, Figure 4.9, for example)contain 0.25 to 0.5 wt%.sulphurHalophytes of the marine mangrove fringe forest and back-mangrove association colonize themargin ofAlmirante Bay and form a distinctive fine, dense hemic peat composed primarily of rootletsand root tissues. (Florida Rhizophora peat has been characterised by Cohen (1968) as root peat andsedimentary peat, the latter distinguished by containing more than 5% non-root material.) Dominantplants in this narrow fringe forest are Rhizophora mangle, Acrosticum aureum, Raphia taedigera,and salt-tolerant sedges and grasses. Sulphur content of 36 Rhizophora peat samples from shorelineand now-submerged offshore sites averaged 3.52 wt% (s = 1.1, range from 1.07 to 5.98).Behind the mangrove fringe, any of several sub-classes of mixed forest-swamp, monospecificCampnosperma panamensis (hardwood) or Raphia taedigera (palm) forest occur. The woody peats164TABLE 4.3 A: Total sulphur content of some plant parts from this study:Forest-swamp wt% sulphurCampnosperma panamensis wood/bark 0.093Heliconia latispatha leaf 0.48Symphonia globulfera resin 0.39Bog plainMyrica mexicana wood/bark 0.045Mangrove associationRhizophora mangle root 2.1Acrostichum aureum base 0.15Raphia taedigera base 1.33TABLE 4.3 B: Some published total S content of plants:wt% sulphur sourceRhizophoram. leaf 0.18 aRhizophora m. root 0.23 aLaguncularia r. leaf 0.25 aLaguncularia r. root 0.06 aplants (Little Shark River) 1.83 bplants (Minnie’s Lake) 0.078 bsurface litter (“) 0.028 bplants (Chesser Prairie) 0.24 bsurface litter (“) 0.093 bMariscusjamaicense leaves 0.058 cbasal calm 0.049 crootstock 0.166 cfine rootlets 0.072 cSource:a = Price and Casagrande, 1991b= Casagrandeetal., 1976c = Altschuler et al., 1983165associated with these forest-swamps are consistently in the upper end of the low-sulphur (0.25 to 1.0wt%) range, even though in many cases they extend to within 15 rn of the shoreline. Detailed study oftwo different coastal fringe sites which experienced coseismic subsidence and marine inundation hintat the range of possibilities of marine influence in very similar environments. Thirteen months afterthe subsidence event, saline intrusion into the peat shoreward of the high tide line at one site (B! 3)was found to be only about 175 cm (Chapter 5), while at a second site (BDD 35) heightened salinitywas measurable 80 m onshore, and is reflected in the dying-off of all non salt-tolerant species acrossthis breadth of shoreline. A new community of halophytic peat-forming species (Rhizophora,Acrostichum, etc) is currently establishing itself across the subsided marine margin.c) Sulphur and Climatic InfluenceA comparison of peats in this study with temperate high sulphur peats suggests that althoughclimate determines vegetation, and extensive woody forest-swamp peats do not seem to be forming intemperate climes at the present time, climate is not a strong influence on total sulphur concentrationsor on the distribution of sulphur fractions in coastal peats. The proportions of sulphur fractions foundin the onshore Changuinola peats unaffected by recent marine inundation (i.e. primary sulphur) arecomparable to those found in temperate brackish peats of the Fraser Delta (Lowe and Bustin, 1985).Mean total sulphur for 12 tropical samples is 3.02 & 0.9 wt%, compared to a mean of 3.03 ± 1.58 for8 temperate brackish sedge peats. Mean organic sulphur content for the tropical peats tested is 94%,compared to 94.2% for the Fraser peats, and organic sulphate represents 19.9% in the tropical peats,and 18.9% in the temperate. The inorganic fractions vary slightly more: mineral sulphate 4.7% vs3.6%; elemental sulphur 1.6% vs 1.1%; and pyritic sulphur 0.6% vs 1.2%, for tropical and temperatepeats respectively. Thus pyrite is lower in the tropical peat, probably limited by Fe availability in thisclastic sediment starved environment. Mineral sulphate and elemental sulphur are slightly moreprominent in the tropical peat.166d) Sulphur and Tectonic InfluenceTectonically-driven regional sea-level fluctuations are the ultimate control on peat depositionin the Changuinola area, and the tectonic setting of the Changuinola deposit is in some ways reflectedin the sulphur distribution in the peat. Since the late Holocene stabilization of global sea level, theCaribbean coast of southern Central America has been in an overall regressive mode, and uplift andhigh sedimentation rates have moved the shoreline seaward (Collins et a!., 1994). At the same timeregional structural movements have created both localized regressive and transgressive coastlines. Tothe northwest of this study area, in Costa Rica, coastal uplift has exposed nearshore reefs and causederosion of carbonates and redeposition of offshore bar sediments. There are no major coastal swampsalong this emergent coastline. However, 90 km to the east in Panama, seismic events, including theApril 1991 event, have a long recorded history of causing subsidence. This has resulted in thedevelopment of extensive low-lying freshwater swamps, localised replacement of forest swampvegetation by halophytic mangrove-fringe forest, and eventual drowning, as observed in the easternpart of the Changuinola peat deposit. In this regime, transgression is from SE to NW in the deposit,effectively parallel to the overall trend of the coastline. The distribution of medium and high sulphurpeats reflects this geometry: medium-sulphur peats now underlie Almirante Bay at all Sites sampled;very high sulphur peats are exclusively associated with present tidally influenced drainage channels,or topographic lows (old drainage) now below sea level; high-end low sulphur (0.5 to 1.0 wt% S)peats are found in forest swamp zones, including within metres of the mangrove fringe and centimetresof the high tide level. Farther west, very low S peats are found in developing bog-plains elevated adecimetre or more above sea level, and in the entire centre of the western part of the deposit. Organicsediments are thus recording regional vertical trends. Peat with both very high and very low primarysulphur content has accumulated due to rising sea level in a region which is experiencing overalluplift.1674.7 IMPLICATIONS FOR ENVIRONMENTAL STUDIES OF COALAs has long been known to coal geologists, medium to high total sulphur content in peat isspatially related to marine or brackish influence (Williams and Keith 1963). This study confirms therelation. In addition, despite the short-term nature of these observations compared to coal studies, thedistribution of sulphur in the Changuinola peat deposit provides some new insights into sulphur in coaland its usefulness as an environmental indicator.As with the peat in this study, it may be possible to distinguish between primarysyndepositional sulphur, secondary or overprinted sulphur due to transgression, and sulphurassociated with channels and brackish environments in coal, based on proportions of forms of sulphurpresent. Diagenetic changes to organic and inorganic components during coalification inevitablysmear depositional chemical signatures. However, clarification of the origins and proportions of labileand conservative S forms associated with particular conditions provides the starting point forenvironmental interpretation. For example, high stable C-S sulphur content in high sulphur coals islikely an indication of estuarine environments, or brackish drainage associated with channels, in whichprimary C-S tends to be high. Channels in peat tend to be long-lived and highly resistant to erosion,and restriction of brackish influence can be due to low permeability of the peat, doming of the peatsurface, or hydraulic pressure from the high water table. Thus high C-S sulphur coals may belaterally restricted, and adjacent coals a few metres away may be low in prry sulphur, even thoughdiagenetically enriched in pyrite.Coals which originated as distinctly marine peats (i.e. from halophytes) are likely to havesignificant pyrite, both primary and diagenetic, but organic forms still dominate the primarycomponent. The above characteristics can develop in both transgrcssivc and regressive depositionalregimes. The marine mangrove peat in this study has moderate total sulphur content (2 to 5 wt%),168principally in organic forms but with a significant inorganic component. Pyrite varies from 0.1 to 3.6% of total sulphur.Coals with a distinct transgressive signature are likely to have a large proportion of sulphur asdiagenetic pyrite resulting from the influx of secondary inorganic sulphur, and possibly from estersulphate. However, such coals are not necessarily high in total sulphur. Coastal peats that haveexperienced marine transgression and are now submerged display a transgressive signature whichincludes highly variable ester sulphate content and an increase in the proportion of inorganic sulphurforms (Phillips et a!., 1994; Chapter 5). The influx of marine waters has resulted in a largerproportion of mobile sulphur forms relative to stable C-S sulphur, but has not resulted in largeincreases in total sulphur content. Tectonically driven coseismic subsidence can lead to long termflooding of coastal peats by sulphate-rich sea water. Over the short term the variability in the estersulphate fraction, over cm-scale distances, is due to the heterogeneity of pore water environments andthe highly variable penneability of peat. In the longer term, this overprinting would likely result inpockets of secondary pyrite formation in coal.In oligotrophic settings with no groundwater sulphate sources, total sulphur content of peat islow. In coals in which no marine influence is evident, sulphur content has been related to the inferredpH of the precursor peat. In Changuinola peats with no marine influence, total sulphur content variesdirectly with the degree of humification of the peat, which in turn broadly reflects peat type, but bearslittle direct relation to pH. Peat-forming plant communities which have a high proportion of subaerialbiomass (forest-swamp peats) produce a woody, highly hurnifled hemic and fine hernic peat in whichsulphur may be concentrated at levels double that of more fibrous, root-dominated peat (sedge peats)of raised bog-plain environments. Beyond the broad distinction between herbaceous peats and woodypeats, there is no significant difference between the total sulphur content of the different types ofwoody peat (Raphia, ampnosperrna, mixed). The variable sulphur content of the original plantmaterial is less significant than the degree of humification in low-sulphur peats. In turn, degree of169humification is much more dependent on the height of the water table than on the pH of the peat orporewater. Thus it is risky to infer high paleo-pH of a coal swamp on the basis of highly humifiedpeat precursors alone. High levels of humification occur in very low pH sites if the water level drops,or if the plant community has a large subaerial biomass. The sulphur found in these low-sulphurpeats is overwhelmingly organic in form. Any pyrite in coal associated with peats from ombrotrophicbog environment is likely post-depositional.4.8 CONCLUSIONSThe following observations can be made regarding the distribution of sulphur in theChanguinola peat deposit:High sulphur content is spatially related to marine or brackish influence.The highest sulphur content (>5 wt%) is found in pêats proximal to the brackish influence ofblackwater drainage channels up to several km from the coast. Very high sulphur was also found inthe basal peats in the deepest parts of the deposit, which are interpreted to have been associated withchannels. Very high sulphur content is found even where salinity is undetectable in peat, porewater orbasal sediments, and despite the low pH of the peat, which inhibits bacterial sulphate reduction.High sulphur peat is not necessarily high in pyrite, axid large quantities of sulphur may bebound in unreactive organic compounds from which it is unlikely to be released biogenically.Peat-forming plant communities which have a high proportion of subaerial biomass (ie.ombrotrophie forest-swamp peats) produce a woody, highly humified hemic and fine hemic peat inwhich sulphur may be concentrated at levels double that of more fibrous, root-dominated peat ofraised bog-plain environments.In the absence of marine influence, total sulphur content varies directly with the degree ofhumification of the peat, which in turn broadly reflects peat t3pe. However, beyond the broadestdistinction, between herbaceous and woody peats. there is no significant difference between the totalsulphur content of the different types of woody peat (Raphia. C’ainpnosperma. mixed forest-swamp).170Thus the variable sulphur content of the original plant material is less significant than the degree ofhumification in low-sulphur peats.Degree of humification in peat is much more dependent on the height of the water table thanon the pH of the peat or porewater. High levels of humification can occur in very low pH sites if thewater level drops, or if the plant community has a large subaerial biomass. However, in anoligotrophic mire with no groundwater sulphate sources, total sulphur content remains low.ACKNOWLEDGEMENTSSupported by the Natural Sciences and Engineering Research Council of Canada grant 5-87337 (Bustin), the International Development Research Centre Young Canadian Researcher award92-1201-18 (Phillips), and The University of British Columbia. We thank L.E Lowe and his soilscience laboratory for analyzing forms of S, and Dr. Lowe and Dr. Bill Barnes for critical readings ofthe manuscript. We also thank the following individuals and organizations in the Republic of Panamafor their support: the Instituto de Recursos Hidraulicos y Electrificacion, the Chiriqui Land Company,the Smithsonian Tropical Research Institute, and in particular Ing. Eduardo Reyes, and the familiesSanchez and Serracin of Boca del Drago, Panama.1714.9 REFERENCES CITEDAltschuler, Z.S., Schnepfe, M.M., Silber, C.C. and Simon, F.O., 1983, Sulphur diagenesis inEverglades peat and origin of pyrite in coal: Science, v.221, p.221 -227.Anderson, J.A.R., 1964, The structure and development of the peat swamps of Sarawak and Brunei:Journal of Tropical Geography, v.18, Singapore, p.7-l6.Anderson, J.A.R., 1983, The tropical peat swamps of Western Malesia: in A.J.P. Gore (ed.),Ecosystems of the World 4B - Mires: Swamp, Bog, Fen and Moor; Chapter 6. Elsevier,Amsterdam.Anderson, J.A.R. and Muller, 3., 1975, Palynological study of a Holocene peat and a Miocene coaldeposit from NW Borneo: Review of Paleobotany and Palvnology, v.19, p.291-351.American Society for Testing and Materials (ASTM), 1969, D260769. Standard classification ofpeats, mosses, humus and related products: ASTM, Philadelphia, Pa., 1 p..Astorga 0., A., 1991, Informe tecnico sobre el levantamiento de la costa Caribe de Costa Rica, comcconsecuencia del terremoto del 22 de Abril de 1991: University of Costa Rica, in prep.Brown, K.A., 1985, Sulphur distribution and metabolism in waterlogged peat: Soil Biology andBiochemistry, vol.17, No.1, p.39-45.Bruenig, E.F., 1990, Oligotrophic forested wetlands in Borneo: in A.E. Lugo, M. Brinson and S.Brown, (eds.), Ecosystems of the World 15: Forested Wetlands, Elevier 1990, Chapter 13.Bruenig, E.F., 1976. Classifying for mapping of peat swamp forest examples of primary forest typesin Sarawak, Borneo. in P.S. Ashton (ed.), The Classification and Mapping of Southeast AsianEcosystems, Univ. of Hull Misc. Ser. 17:57-75.Camacho, E. and Viquez, V., 1992, Historical and instrumental seismicity of the Caribbean coast ofPanama: University of Panama. In press.Camacho, E., Viquez, V. and Espinosa, A., 1992. Ground deformation and liquefaction distributionin the Panama Caribbean coastal region: University of Panama, in preparation.Casagrande, D.J., Gronli, K. and Sutton,N., 1980, The distribution of sulphur and organic matter invarios fractions of peat: origins of sulphur in coal: Geochimica Cosinochimica Acta, v.44,p.25-32.Casagrande, D.J., Seiffert, K., Berschinski, C. and Sutton, N., 1977, Sulphur in peat formingsystems of the Okefenokee Swamp and Florida Everglades: origins of sulphur in coal:Geochimica CosmochimicaActa, v.41,p.161-7.Cecil, C.B, Stanton, R.W., Dulong, F.T. and Renton, J.J.. 1979, Some geologic factors that controlmineral matter in coal: W. Virg. Econ. Geol. Sun’. Bull. B/37: 1, p.43-56.172Chandra, D., Mazumdar, K. and Basumallick, S., 1983, Distribution of sulphur in the Tertiary coalsof Meghalaya, India: International Journal of Coal Geology, v.3, p.63-75.Coates, A.G., and Obando, J., in press (1994), The geological evolution of the Central Americanisthmus. Smithsonian Tropical Research Institute, Panama: in Mann, P. ed., Geologic andtectonic development of the Caribbean plate boundary in southern Central America.Geological Society ofAmerica Special Paper.Collins, L.C., Coates, A.G. and Obamdo, J., in press (1994), Rates of uplift of the Burica andsouthern Limon Basins; Caribbean effects of Cocos subductiön? : in Mann, P. ed., Geologicand tectonic development of the Caribbean plate boundary in southern Central America.Geological Society of America Special Paper.Cohen, A.D., Spackman, W. and Dolsen, P., 1984. Occurrence and distribution of sulphur in peatforming environments of southern Florida. International Journal of Coal Geology, v.4, p.73-96Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce,F., 1989., The Changuinola peatdeposit of northwestern Panama: A tropical back-barrier peat (coal)-forrning environment: InP.C. Lyons and B.Alpern (eds.), Peat and Coal: Origin, Facies and DepositionalEnvironments. International Journal of Coal Geology, v.12, p. 157-192.Cohen, A.D., Raymond, R.Jr., Ramirez, A., Morales, Z. and Ponce, F., 1990, Changuinola PeatDeposit ofNorthwestern Panama: 3 vols. Los Alamos National Laboratory pub. LA- 11211,July 1990.Dapples, E.C. and Hopkins, M.E., 1969 , eds., Environments of Coal Deposition: Geological Societyof America. Special Paper 114, Boulder, Colorado, 1969. pp.204Denyer, P., Personius, S.F., Arias, 0., Rojas, W. and Alvarado, J., 1992, Tectonic effects of theearthquake ofApril 22, 1991: In press.Esterle, J.S., 1990. Trends in petrographic and chemical characteristics of tropical domed peatdeposits in Indonesia and Malaysia as analogues for coal formation.: Unpublished PhD thesis,University of Kentucky, Lexington, p.270.Esterle, J.S., Ferm, J.C., Durig, D.T. and Supardi, 1987, Physical and chemical properties of peatnear Jambi. Swnatra., Indonesia. International Peat Society Symposium on Tropical Peat,1987.Giblin, A.E. and Wieder, R.K., 1992, Sulphur cycling in marine and freshwater wetlands: inHowarth, R.W., Stewart, J.W.B. and Ivanov, M.V., eds. 1992. Sulphur Cycling on theContinents; Wetlands, Terrestrial Ecosystems and Associated Water Bodies. SCOPE 48.Wiley and Sons, Chichester,p.85-ll7.Escalante, G., 1990, The geology of southern Central America and western Columbia: in Dengo, G.and Case, J.E., eds., The Caribbean Region: Boulder, Colorado, Geological Society ofAmerica, The Geology of North America, v.H.173Gross, M.G., 1982, Oceanography. 3rd.ed.: Prentice-Hall, Englewood Cliffs, N.J. 498pp.Hunt, J.W.and Hobday, D.K., 1984, Petrographic composisition and sulphur content of coalsassociated with alluvial fans in the Pennian Sydney and Gunnedah basins, Eastern Australia:In: Rhamani, R.A. and Flores, R.M., eds., Sedimentology of Coal and Coal-bearingSequences, International Association of Sedimentologists Special Publication No.7,p.43-60.Lowe, L.E., 1986, Application ofa sequential extraction procedure to the determination of thedistribution of sulphur forms in selected peat materials: Canadian Journal of Soil Science,v.66,p.337-45.Lowe, L.E. and Bustin, R.M., 1985, Distribution of sulphur forms in six facies of peats of the FraserRiver delta. Canadian Journal of Soil Science, v.65, p.53 1-541.Luther, G.W. and Church, T.M., 1992, An overview of the environmental chemistry of sulphur inwetland systems: In Howarth, R.W., Stewart, J.W.B. and h’anov, M.V., eds. 1992. SulphurCycling on the Continents; Wetlands, Terrestrial Ecosystems and Associated Water Bodies.SCOPE 48. Wiley and Sons, Chichester, Chapter 6,p.125-44.Plafker, 0. and Ward, S.N., 1992, Backarc thrust faulting and tectonic uplift along the Caribbeancoast during the April 22, 1991 Costa Rica earthquake: Tectonics, v.11, No.4,p.709-18Phillips, S., Bustin, R.M. and Lowe, L.E., 1994, Effects of earthquake-generated marinetransgression on sulphur distribution in a tropical peat: a modem analogue for marine-roofedcoals: Geology, v.22,p.929-32.Pirazzoli, P. 1991, World Atlas of Holocene Sea Level Changes: Elsevier, Amsterdam, Netherlands,300p.Postgate, J.R., 1984, The Sulphate-reducing Bacteria. 2nd Ed. :Cambridge University Press, 208 p.Querol, X., Femandez-Turiel, J.L., Lopez-Soler, A., Hagemann, H.W., Dehmer, J., Juan, R. andRuiz, C., 1991, Distribution of sulphur in coals of the Teruel Mining District, Spain:International Journal of Coal Geology, 18.p.327-46Renton, J.J. and Bird, D.S., 1991, Associaction of coal macerals, sulphur sulphur species and theiron suiphide minerals in three columns of the Pi Etsburg coal: International Journal of CoalGeology, 17,p.21-50.Renton, J.J., Cecil, C.B., Stanton, R. and Dulong. F., 1979, Compositional relationships of plants andpeats from modern peat swamps in support of a chemical coal model: In: A.C Donaldson,M.W. Presley and J.J. Renton, eds., Carboniferous Coal Guidebook, v. 3. W.V. Geol. Econ.Surv., Morgantown, w.v., p.57-100.Raymond, R.Jr. and Davies, T.D., 1979, Content and form of sulphur in coal: a reflection of peatdepositional environments: Geological Society of America Abstract with Programs, v.11(7),p.501.174Shimoyama, T., 1984, Sulphur concentration in the Japanese Paleogene coal: In; Sedimentology ofCoal and Coal-bearing Sequences. eds.R.A. Rhamani and R.M. Flores, InternationalAssociation of Sedimentologists Special Publication (1984) 7, p.36 1-372.Staneck, W. and SiLc, T., 1977. Comparisons of four methods for determination of degree ofhumification (decomposition) with emphasis on the von Post method. Canadian Journal ofSoil Science, v.57,p.109-17.Tabatabai, M.A., 1992, Methods of measurement of sulphur in soils: in Howarth, R.W., Stewart,J.W.B. and Ivanov, M.V., eds. 1992. Sulphur Cycling on the Continents; Wetlands,Terrestrial Ecosystems and Associated\1ater Bodies. SCOPE 48. Wiley and Sons,Chichester, p307-344.Wanless, H.R., Barofflo, J.R. and Trescott, P.C., I 969, Conditions of deposition of Pennsylvaniancoal beds: In: Dapples, E.C. and Hopkins.\l.E., 1969. eds., Environments of CoalDeposition. Geological Society of Am’ special Paper 114, Boulder, Colorado, 1969.p.105-142.Williams, E.G. and Keith, M.L., 1963, Relationship between sulphur in coals and the occurrence ofmarine roof beds: Economic Geology, v.5X. p.720-729.175 1CHAPTER 5EARTHQUAKE-INDUCED FLOODING OF A TROPICAL COASTAL PEAT SWAMP: AMODERN ANALOGUE FOR HIGH SULPHUR COALS176CHAPTER 5: EARThQUAKE-INDUCED FLOODING OF A TROPICAL COASTAL PEATSWAMP: A MODERN ANALOGUE FOR HIGH SULPHUR COALS5.1 ABSTRACTHigh sulphur content of coal is generally attributed to marine inundation during or shortlyfollowing peat formation. When a large peat deposit on the Caribbean coast of Panama subsidedduring a 1991 earthquake, sulphate-rich seawater inundated the margin of the deposit, providing theopportunity to evaluate the role of short-term marine flooding on sulphur abundance and chemistry.Salinity and pH measurements across the new marine margin indicate that penetration of salinewaters into the peat is restricted to < 2 m both vertically and horizontally. A comparison of forms ofsulphur in the newly flooded peat vs adjacent, subaerial peat reveals that total sulphur content is notincreased by flooding, but that the distribution of sulphur forms differs markedly after marineinundation: inorganic forms, particularly mineral sulphate, make up a higher proportion of the total;the organic sulphate fraction becomes highly variable; and the carbon-bonded sulphur contentremains seemingly unaffected. Heightened bacterial activity is seen as a likely mediator in thisredistribution of sulphur forms. Our results indicate that high sulphur content of coals and peatscannot form by short-term periodic flooding events such as storms but must reflect long terminfiltration of marine or brackish waters measured on time scales of hundreds to thousands of years.1775.2 INTRODUCTIONIt is a paradigm in coal geology that high sulphur (S) coals result from syn- or postdepositional inundation of peat by brackish or marine waters (e.g. Home et al., 1978). Theimportance of short-term inundation by marine waters on S content of peat and, by analogy, coalshas, however, never been tested. Aqueous sulphate (SO42-) in rain, groundwater and the oceans isthe principal source of S found in peat. Concentration of S is from <ito 8 ppm in fresh water and885 ppm in seawater. Sulphate is reduced by respiring plants and S-reducing bacteria. Plants usethe oxygen for respiration, and fix some S, as carbon-bonded (C-S), in a variety of amino acids andrelated compounds, and as so-called ester sulphates (C-0S03)in polysaccharides, choline sulphate,phenols, and other compounds. Reduced S may be subsequently reoxidized to a variety of oxidationstates, leading to the formation of polysuiphides and of elemental or pyritic S. The primary Scontent of peat is dependent on the availability of aqueous sulphate to living plants and bacteria. Inbrackish and marine-influenced peat, S concentration can be high, and is not limited by pore-watersulphate concentration.Part of the Caribbean coast of Panama is undergoing marine transgression as a result ofearthquake-induced subsidence, and thick peat beds are overlain by shallow-water carbonate andclastic sediments (Fig. 5.1). On April 22, 1991 a magnitude 7.4 earthquake (Camacho et al., 1993)near Changuinola, Panama, led to 50 — 70 cm of coseismic subsidence and resulted in marine waterspenetrating into the large back-barrier swamp. The effects of that rapid flooding on theconcentration and distribution of a variety of S forms are assessed by comparing long-submergedpeat, newly drowned peat, and immediately adjacent subaerial peat. By taking advantage of a178Earthquake-induced transgressionpeat BDD 19A site_________deposit_________BDD 29 tAlm;rantet_BayN15_____%44%\sites Pta. PondsokW08227 W08220Figure 5.1. Site of the study on the Caribbean coast ofPnma is shown, lower left, in relation to the epicentre ofthe 1991 earthquake in the Rio Estrella valley, Costa Rica. The block diagram illustrates the extent ofmarinetransgression at site BI 3 as a result ofearthquake-induced subsidence.lime mud--- pre-earthquake sea level--imangrove 0peat / 12msand & /,coral’___________________:.:_ ________ ________sample =0sites//<tl,I/ NIepicentre ofApril1991_earthquake sample4%5 kmLocationof179measurable subsidence event, this study attempts to provide insight into the processes that lead to theformation of very high sulphur coals (>5% S), particularly the role of short-term marine flooding.5.3 SAMPLING AND EXPERIMENTAL PROCEDURESSamples were collected across a sea level transect and from three submerged offshore sites(Fig. 5.1). At the transect site, red mangrove (Rhizophora mangle) and sawgrass (Rhyncospora sp.)dominate the peat fonning vegetation, both at the surface and throughout the peat cores. Thefollowing analyses were performed: pH, salinity and total S on all samples (Fig. 5.1, Table 5.1).On eight samples, forms of S were determined: organic S in both carbon-bonded suiphide (C—S) andester sulphate (oxidized compounds reducible by HI) forms, and inorganic forms expressed asmineral sulphate, elemental 5, and pyritic (+ marcasite) S (Table 5.2). The analytical methods usedare described by Lowe (1986) and Tabatabai (1992).5.4 RESULTSa) Total Sulphur, Salinity andpHShallow onshore samples have low total S (0.29—0.65 wt%), and low salinity (0.02—0.05 wt%)except adjacent to the high tide line, where higher salinity (0.94 wt%) reflects the extent of windblown salt and wave washover. Below sea-level, S values both onshore and offshore increase withdepth, to the range of 3.0—4.5 wt%. Although total S increases steadily with depth, there is alsosubstantial spatial variation (Fig. 5.2). Sulphur content of the basal carbonate sands is low (0.4wt%). Salinity onshore is consistently low (<0.08 wt%) and spatially variable. Offthore, salinity ishigh near the peat-water interface (maximum 2.1 wt%), and decreases irregularly down to the baseof the peat. Significantly, total S in the now inundated and much more saline part of the transect isnot elevated compared to onshore values.180TABLE 51. TOTAL SULPHUR, SALINITYAND pH OF 46 PEAT SAMPLESSite Depth pH Salinity Total S(cm) (wt%) (wt%)Samples from transectBIT 3-1 10 6.14 0.02 0.52BIT 3-2 42 5.90 0.06 3.59BIT 3-3 10 6.36 0.04 0.47BIT 3-4 45 5.80 0.04 2.68BIT 3-5 10 6.30 0.03 0.29BIT 3-6 36 5.71 0.08 3.79BIT 3-7 10 6.39 0.02 0.52BIT 3-8 53 6.44 0.04 1.75BIT 3-9 10 6.54 0.02 0.63BIT 3-10 39 6.25 0.03 1.34BIT3-11 45 5.99 0.03 1.70BIT 3-12 90 5.65 0.04 3.75BIT 3-13 10 6.35 0.02 0.61BIT 3-14 53 6.30 0.03 3.35BIT 3-15 10 6.20 0.03 0.54BIT 3-16 33 5.50 0.06 3.50BIT 3-17 10 6.45 0.05 0.52BIT 3-18 32 6.35 0.06 3.75BIT 3-19 10 6.41 0.03 0.65BIT 3-20 39 6.57 0.06 4.00BIT 3-21 10 6.09 0.94 0.53BIT 3-22 53 5.79 0.17 3.69BIT 3-23 10 6.42 0.04 1.16BIT 3-24 25 6.25 0.08 1.56BIT 3-25 50 5.70 1.70 2.28BIT 3-26 75 5.40 1.70 4.22BIT 3-27 100 5.63 1.50 4.33BIT 3-28 125 5.66 0.98 3.26BIT3-29 150 6.22 1.10 3.78BIT 3-30 195 6.87 1.20 1.24BIT 3-31 200 6.92 1.30 1.98BIT 3-32 70 6.90 2.10 1.98BIT 3-33 80 6.79 2.00 3.08BIT 3-34 113 6.59 1.80 3.78Submerged samples from offshore coresBDD 19A 350 7.48 2.20 4.01BDD 19A 425 7.57 2.50 3.02BDD 19A 475 7.06 2.10 3.21BDD 19A 510 7.07 1.70 5.08BDD29 25 5.58 2.10 1.94BDD 29 50 5.60 1.70 2.22BDD 29 75 5.19 1.40 2.08BDD29 100 5.21 1.30 4.17BDD29 425 6.60 1.30 3.19BDD 29 450 6.49 1.50 3.93PTA POND 575 3.57 1.40 3.53PTA POND 600 2.89 1.30 3.03Table 5.1. Total sulphur data, BI 3and offshore sites.181TABLE 5.2. SULPHUR FRACTIONS IN EIGHT PEAT SAMPLESNo. Sample Site Depth Total Inorganic HI-S ESS Pyritic Elemental Organic C-S Organic(cm) Sulphur Sulphur (wt%) (wt%) Sulphur Sulphur Sulphur Suiphu Sulphate(wt%) (%) (wt%) (wt%) (%) (wt%) (wt%)1 BIT3-23 10 1.16 2 0.24 0.014 0.005 n.d. 98 0.916 0.225% of total 100% 1.19% 0.04% 79% 19.4%2 BIT3-25 50 2.28 30 1.48 0.563 0.021 0.109 70 0.802 0.785% of total 100% 24.7% 0.93% 4.8% 35% 34.4%3 BIT3-27 100 4.33 24 1.74 0.819 0.042 0.184 76 2.59 0.692% of total 100% 18.9% 0.1% 4.2% 60% 16%4 BIT3-29 150 3.78 20 1.12 0.762 0.015 nd. 80 2.66 0.371% of total 100% 20.1% 0.4% 70% 9.8%5 BIT3-9 10 0.63 7 0.17 0.041 0.004 n.d. 93 0.461 0.123% of total 100% 6.6% 0.9% 73% 19.5%6 BIT3-12 90 3.75 9 1.12 0.237 0.037 0.061 91 2.63 0.788% of total 100% 6.3% 1% 1.6% 70% 21%7 BIT3-32a 55 1.98 32 1.44 0.319 0.072 0.233 68 0.543 0.812% of total 100% 16.1% 3.6% 11.8% 27% 41%8 BIT3-33a 85 3.78 37 1.65 1.109 0.072 0.208 63 2.126 0.264% of total 100% 29.4% 1.9% 5.5% 56% 7%Note: HI-S is sulphur reducible by HI; ESS is easily soluble mineral sulphate;C-S is carbon bonded sulphur. n.d. = not determined182Total S was also determined for 12 peat samples collected from three sites that have beensubmerged for much longer periods. One sample (BDD 19A) from 475 cm below present sea-level,radiocarbon dated at 2110 ± 60 yr., has a total S of 3.21 wt% (salinity of 2.0 wt%). It is overlam by265 cm of peat, 15 cm of carbonate sediments, and 195 cm of salt water. Total S for all the long-submerged peats averages 3.2 ± 1.04 wt% (Table 5.1).b) Forms ofSulphur1) Organic Forms --- Organic forms predominate in all samples (Table 5.2) ranging from 63 — 98%of total S. All onshore samples have> 90% organic S, whereas offshore samples have < 80%organic S. The two samples with the lowest proportion of organic S (samples 7 and 8) aresubmerged, and thus not surprisingly have the highest pore-water salinity and pH of all samplestested. Onshore, ester sulphate is uniform at 19 — 21% of total S. C-S ranges from 73 — 79%. Incontrast, offshore ester sulphate varies widely, from 7% to 41% oftotal 5, and generally decreaseswith depth. C-S varies from 27 — 70% of total S offshore, and increases with depth.ii) Inorganic Forms - In all samples, mineral sulphate is the dominant inorganic form. Onshore,mineral sulphate comprises 1.2 — 6.6% of total S. Elemental S was only detectable in one sample, at1.6%. Pyritic S represents 0.04 — 1% of total S. Offshore, mineral sulphate is found in much higherproportions, varying from 16.1 — 29.4% of total S. Elemental S constitutes 4.2— 11.8% of total S,and pyritic S 0.1 — 3.6%. The presence ofH2S could be detected by a faint smell in all samples butwas not measured.183OFFSHORE SUITE ONSHORE SUITEPRESENT SEALEVEL-20%.. ... ... •..0.03 0.08 3.0%PREEARTHQUAKE*/L______________SA LEVEL 10T[7U20’ SULPHUR1 m : 8 : smwr.:•.. Sample no. from Table 2:728 341 56influenced .::jhflHHTiflfl estersuIphatebyinundafion 4::entoISForms of Sulphur for 8 Samples2m 5mFigure 5.2. Spatial variation in total sulphur and salinity ofpeat across the newly-submerged marine margin.Sample sites in boxes are also shown in Figure 5.1. Inset columns I - 8 depict variation in sulphur fractions for theeight numbered samples, based on data in Table 5.2. The columns are arranged from left to right in increasingdistance from the peatfsalt water interface, illustrating the effects of inundation on the relative proportions of C-Sand ester sulphate, and the increase in aqueous sulphate. Iso-sulphur cuves (dotted lines) of 1,2 and 3% S revealthe spatial variability of these geochemical signatures: the depressed region is around the remains of a 1cmdiameter mangrove root, which provides a conduit for rainwater.Vertical exaggeration is xlO.1845.5 DISCUSSIONSulphur concentration in marine-roofed coal is commonly highest at the top and decreasesdown-seam. Such a distribution has been used to suggest that marine or brackish water inundationresults in S enrichment near the peat surface due to stimulation of anaerobic sulphate reduction. Thedecrease in S content with depth is considered to reflect the progressive bacterial depletion of S042as the water passes downward (Brown and Macqueen, 1984). Most high-S peats, however, do notshow the trend observed in coals. Rather, total S in marine-influenced peats (Altschuler et al., 1983;Bustin et a!., 1987) tends to increase with depth in the upper few cm, then remains relativelyconstant with increasing depth, as it does in this study.The primary S content of the peat in this study is moderately high but typical of peats havingformed in a marine-influenced environment by the progradation of a mangrove fringed shoreline(Given and Miller, 1985; Cohen et a!., 1989). Higher S concentration toward the base of the peat isattributable to ongoing assimilatory bacterial reduction, and increased decomposition of organicmatter with burial, leading to concentration of resistant compounds in the older, more degraded peat.Much higher S content, on the order of 10 wt%, was recorded in high-ash, brackish peats from theFlorida Everglades (Casagrande et al., 1977), and levels as high as 13.7 wt% (92% organic ) wasfound in samples collected near the site of this study (Cohen et al., 1989; Phillips et al., in press). Itis still not clear, however, what environmental conditions lead to such high S content. If it is theresult of secondaiy enrichment following marine inundation, it clearly requires more time than the 13month exposure to marine waters of peats of this study. Brown and Macqueen (1985) determinedthat metabolic S042 uptake from water inundating a peat surface may be delayed until thenecessary reducing environment (Eh - 100 my.) has been established near the peat surface. In theirlaboratory study this took only a few days. Once reducing conditions were established, molecular185diffusion driven by biologically mediated concentration gradients proceeded at a rate that, althoughdependent on sulphate concentration of the inundating waters, was generally comparable to thediffusion rate of H2S04in dilute aqueous solutions. In the present study it is not evident from totalS content of the inundated vs. onshore samples that any measurable increase occurred as a result ofthe sudden flooding of the peat with seawater. In fact, iso-S contours (Fig. 5.2) are depressed togreater depths in the newly inundated peat. Salinity and S values from the long-submerged offshorecores (Table 5.1) suggest minimal percolation of seawater down through the peat during the periodof submergence and no indications of total enrichment beyond that expected for mangrove peat.Pore water salinity and pH distribution indicate that the lateral extent of marine infiltration islimited to 2 m. Such restricted marine infiltration may be due to the ever-wet climate, whichmaintains a high water table and positive hydraulic head in the groundwater system year-round,together with variations in permeability of the peat. The higher salinity, and high mineral sulphatecontent of the offshore samples, (average 21% of total 5), suggests that inundation has indeedallowed the introduction of aqueous sulphate into the peat. However, mineral sulphate content doesnot decrease consistently with depth, either proportionately or in absolute terms. The proportion ofmineral sulphate in the onshore samples is low (1.19% to 6.6% of total S).The presence of a significant amount of elemental S in offshore samples may indicatesubstantial sulphate-reducing bacterial activity. Elemental S may result directly from the reductionprocess or from the subsequent oxidation of dissimilatory hydrogen sulphide, for example aroundroots of mangroves. Pyrite, linked to bacterial activity in mangrove peats (Cohen et al., 1984) is alsopresent to a greater extent in samples 7 and 8, where elemental S concentrations and pH (6.9 and1866.59) are highest. All of these factors indicate that a more favorable environment for heterotrophicsulphate reducing bacteria now exists in the offshore peat.Onshore, the organic S fraction is dominated by C-S forms similar to those found in high-Scoals, and as appears to be the norm for marine peats. Onshore ester sulphate is consistently around20% of total 5, comparable to values reported for other marine and brackish peats (Casagrande etal., 1977; Bustin et al., 1987). Both the organic S content, and the relative proportions of C-S toester sulphate are much more variable in the offshore than onshore samples. The C-S pool inorganic matter has been found to be more conservative than the ester sulphate fraction (Howarth andStewart, 1992); thus changes in the environment are more likely to be reflected in the ester sulphatefraction. The two samples closest to the marine influence display very high ester sulphate content.In sample 7, ester sulphate is the dominant form (41%). However, a few cm lower in the core, thatfraction represents only 7% of the total, and 75 cm below that, 9.8%. Assuming 20% to be areasonable proportion of ester sulphate in high-S peats unaffected by secondary enrichment, it seemsthat samples a few centimetres apart can undergo variable secondary effects. Enrichment of theester sulphate fraction may occur as the result of bacterially reduced inorganic S, in the form ofelemental S, recombining with humic compounds in the peat (Casagrande et al., 1977). Depletion ofthe ester sulphate fraction may be caused by bacterial digestion of organic matter and increasedpeatification, as it is in low S peats. Thus, it is possible to speculate on a bacterially mediated fluxin the ester sulphate fraction, and the results reported here may give some indication of the potentialmagnitude of that flux. Depletion of the ester sulphate fraction may also be affected directly bybacterial reduction, but in this case aqueous SO42 is in ample supply and is preferred over organicsulphate by bacteria (Altschuler et al, 1983). Thus, no decrease as a result of direct bacterialreduction is expected. Like anaerobic bacterial activity, increases in elemental S and pyrite are pH187and Eh dependent, and thus not inconsistent with variable ester sulphate content. In this study, thehighest pyritic and elemental S contents are found in the sample with the highest ester sulphate, thelowest C-S fraction, and the highest pH. Microenvironments of varying oxidation-reductionpotential associated with root penetration, or other permeability factors, may result in variablebacterial activity and may be a factor in the large variations in ester sulphate.5.6 CONCLUSIONSIn the study area on the Caribbean coast of western Panama, tectonically driven subsidence isleading to inundation of a thick peat deposit and burial beneath marine sediments. No evidence ofenhancement of total S concentration in peats inundated by normal marine water is evident. Spatialdistribution of S fractions and of elevated pore-water salinity suggests —175 cm of penetration ofmarine waters both horizontally and vertically in the 13 months that elapsed since inundation.Sulphur concentration increases with depth in the upper 175 cm, up to a level typical of, but nohigher than, other mangrove peats sampled in this area.The distribution of S forms differs markedly between onshore samples unaffected byinundation and those now under water. Inorganic S forms represent a much larger proportion of totalS in offshore peat. A higher proportion of mineral sulphate is found in the offshore samples.Significant quantities of elemental S are present in all but the deepest of the offshore samples,whereas it is undetectable in the onshore samples. Pyritic S is present in all samples but is only >1%of total S in one of the offshore samples. Pyrite is not a dominant S form, and is not seen to beforming at the expense of ester sulphate.188The short-term effects of marine inundation are to increase the relative proportion of inorganicover organic S in peat and to introduce considerable variability in the ester sulphate content. Thusthe proportion of ester sulphate may be the best indicator of the geochemical histoty of a peatdeposit. The geological implications remain to be determined by detailed analyses of forms of S inmarine-roofed coals.By analogy with peats ofthis study, the high sulphur content of coals cannot reflect short-termor periodic storm flooding events but must reflect prolonged infiltration, probably measured on timescales of hundreds to thousands of years.ACKNOWLEDGEMENTSSupported by the Natural Sciences and Engineering Research Council of Canada grant 5-87337, the International Development Research Centre Young Canadian Researcher award 92-120 1-18, and The University of British Columbia.1895.7 REFERENCES CITEDAltschuler, Z. S., Schnepfe, M. M., Silber, C. C. and Simon, F. 0., 1983, Sulphur diagenesis inEverglades peat and origin of pyrite on coal: Science, v. 221, P. 221-227.Brown, K. A. and Macqueen, J.F., 1985, Sulphate uptake from surface water by peat: Soil Biologyand Biochemistry, v. 17, p. 4 11-420.Bustin, R. M., Styan, W. S. and Lowe, L. E., 1987, Variability of sulphur and ash in humid-temperate peats of the Fraser River delta, British Columbia: Compte Rendu, 10thInternational Congress of Carboniferous Stratigraphy and Geology, Madrid, Spain; v. 3, p. 79-94.Camacho, E., Viquez, V. and Espinosa, A., 1993, Ground deformation and liquefaction distributionin the Panama Caribbean coastal region: University of Panama, Panama City in press).Casagrande, D. J., Seiffert, K., Berschinski, C. and Sutton, N., 1977, Sulphur in peat formingsystems of the Okefenokee Swamp and Florida Everglades: Origins of sulphur in coal:Geochinñca et Cosmochimica Acta: v. 41, p. 161-167.Cohen, A. D., Raymond, R. Jr., Raniirez, A., Morales, Z. and Ponce, F., 1989, The Changuinolapeat deposit of northwestern Panama: Los Alamos National Laboratory Report LA-i 1211, v.2,83 p.Cohen, A. D., Spackman, W. and Dolsen, P., 1984, Occurrence and distribution of sulphur in peatforming environments of southern Florida: Coal Geology, v.4, p. 73-96.Given, P. H. and Miller, R. N., 1985, Distribution of forms of sulphur in peats from salineenvironments in the Florida Everglades: International Journal of Coal Geology, v.5, p. 397-405.Home, 3., Ferm, J. C., Carnccio, F. T. and Baganz, B. P., 1978, Depositional models in coalexploration and mine planning: Bulletin of the American Society of Petroleum Geologists, v.62,p. 2379-2411.Howarth, R. W. and Stewart, 3. W. B., 1992, The interaction of sulphur with other elements inecosystems. in Howarth, R. W. et al., eds.: Sulphur cycling on the continents; wetlands,terrestrial ecosystems and associated water bodies. (SCOPE 48): Chichester, United Kingdom,Wiley& Sons. 350 p.Lowe, L. E., 1986, Application of a sequential extraction procedure to the determination of thedistribution of sulphur forms in selected peat materials: Canadian Journal of Soil Science, v.66, p. 3 37-345.Tabatabai, M. A., 1992, Methods of measurement of sulphur in soils, in Howarth, R.W. et al., eds.Sulphur cycling on the continents; wetlands, terrestrial ecosystems and associated water bodies,(SCOPE 48): Chichester, United Kingdom,Wiley & Sons, p. 307-344.190CHAPTER 66.1 CONCLUSIONS“Let us suppose an earthquake, possessing the characteristic undulatory movementof the ernst, in which I believe all earthquakes essentially to consist, suddenly to havedisturbed the level of the wide peat-morasses, and adjoining flat tracts of forest on the oneside, and the shallow sea on the other. The ocean, as usual in earthquakes, would drain offits waters for a moment, from the great Stigmaria marsh, and from all the swampy forestswhich skirted it, and, by its recession, stir up the muddy soil, and drift away the fronds, twigsand smaller plants, and spread these, and the mud, broadly over the surface of thebog Presently, however, the sea would roll in with impetuous force, and, reaching thefast land, prostrate every thing before it.”Henry Darwin Rogers, 1842Transactions ofthe Association ofAmerican Geologists and Naturalists, Vol.1, p. 433-474.When Henry Rogers, in the 1840’s, was seeking to understand the processes by which theAppalachian coal strata came about, he imagined the swampy, tropical coast of ancient Pennsylvaniawracked by periodic violent earthquakes, the sudden subsidence and subsequent tsunamis and floodingbringing a violent end to peat deposition and producing the widespread carbonaceous shale partings that hewas finding in his explorations. It is a faculty of the human mind to imagine the processes which accountfor the world around us, and is a particularly valuable ficulty for the geologist, who is often trying toreconstrnct events, much of the evidence of which is long obliterated.Detailed documentation of an event similar to that imagined by Henry Rogers may not support hisconjectures, as in the present study of the Changuinola peat deposit, but both the speculation and thedocumentation have their roles in the evolution of a model which fits processes to products. Rogersimagined his Pennsylvanian Sf1gmaria marsh as a vast, perfectly planar marine savannah. The closestknown modem analogues to coal-forming mires, the Changuinola mire system and the known thick peatdeposits of southeast Asia, do not conform to this pattern, but develop an elevated topography and an191internal structure that encourages continuing peat accumulation, and insulates them from the surroundingclastic sedimentaiy environments.. Several generations of geologists have since contributed to ourunderstanding of the palaeogeography and topography of coal-forming mires. In any event, it wouldrequire a far more destructive earthquake than the Ms 7.5 shock which struck the Panama coast in 1991 toleave the kind of geological evidence envisaged by Rogers.The goals of this investigation were to evaluate the Changuinola peat deposit as a possibleanalogue for the deposition of low ash, low sulphur coals, and to document the effects of earthquake-driven subsidence events on the peat and the peat-forming vegetation. The deposit differs in significantways from previously studied tropical peats, and thus provides some new insight into the process of thickpeat accumulation, and hence the nature of coal-fonning depositional environments. In addition, andperhaps surprisingly, it is evident that the impact of rapid coseismic subsidence on the peat deposit isconsiderably more subtle than Henry Rogers would have expected.This study shows that the development of elevated, oligotrophic bogs, as described in theAnderson model, and the accumulation of thick peat and hence coal deposits can occur in response totectonically-driven punctuated subsidence, as well as gradual eustatic sea level rise. The process canproceed without leaving a record of increased clastic input within the peat, even immediately adjacent toenvironments of active clastic deposition. The approximately 10 m of relief on the basal sands, a productof the mode of subsidence and sedimentary response with which the barrier sands aggrade and prograde, isa significant difference between the Changuinola deposit and the modern southeast Asian peats, whichhave planar bases and are almost entirely above sea level.Climatic influences have a profound effect on the nature of peat deposition. On this wave- ratherthan storm-dominated coastline, the back barrier environment is free of all but minor aeolian clastic input.192In the tropical climate peat accumulates on barrier sands in swales directly behind the beach, andprogrades seaward with the barrier shoreline, without the colonizing mangrove fringe common to theMalesian deposits. In fact, the greater part of the deposit displays no internal evidence of marineinfluence, despite being approximately 40 % below sea level. Although instantaneous subsidence of 30 -50 cm left its mark in structures in the barrier sand body, it is not detectable in the peat behind the barrier.The part of the deposit which does display marine influence is at the southeastern extent of the barriercoast, adjacent to and beneath the shallow waters of Almirante Bay. There, the presence of marineinfluence, evident in the dominance of halophyte (mangrove) peats, and in the distribution of sulphur,reflects the regional structural trend of greatest subsidence to the southeast. Evidence of the punctuatednature of that subsidence is found in the presence of shelly layers within otherwise carbonate-freemangrove and back-mangrove peats. In addition, at least in the short-term, the distribution of forms ofsulphur in newly-inundated peat is significantly different from the distribution in peat that has not beenflooded. This suggests a future line of inquiiy into longer-term geochemical signatures of rapid coseismicsubsidence and marine inundation of coastal peatlands.6.2 SUGGESTIONS FOR FURTHER RESEARCHThe Changuinola peat deposit is an excellent analogue for coal deposition, for the reasons statedabove. One of the most valuable aspects of the deposit for coal studies is its almost untouched state. Likethe European peats, many of the southeast Asian deposits are sites of intense human activity, have beendrained or mined, and the forests have been logged. Differences in structure and hydrology between thisdeposit and those of southeast Asia leads to the question of whether there are peat deposits in tropicalAmerica which are developed on a base of less relief, and which more closely fit the Asian model.There are numerous possibilities for more extensive and intensive study of the evolution and thedetailed hydrology of the Changuinola deposit. Of interest are the long term effects of sudden inundation193on coastal peats, and burial of peat beneath lime-mud and coral. Also, there is the possibility ofaddressing one of the chief problems inherent in the use of modern peats as analogues for coal; the changesin the nature of coal-forming vegetation through time. Although the problem is greatest for Paleozoiccoals, modem palm peat may well provide a very close analogue to Tertiary palm coals. The use of palmpeat in artificial coalification studies thus has potential to reduce the problems inherent in assessing theeffects of vegetation change. Finally, in addition to its potential for refining the evolving model of tropicalcoal development, the Changuinola deposit has interest because of the possibility of determining a historyof coseismic subsidence for this part of the Caribbean coast based on radiocarbon dating of earthquakerelated stratification in submerged peat.194:::::::::;;r;:;:;;:;::::::::c4c-eeeeC12JViL.JJ00000000000000-J--.J-J-.)-1—J—.1-J--J—)00—1i-——00-----------00—-.-——-.-‘3—.——.——.——_.)t)00CCi-.—CC‘3‘30000.-t.30000CO’C.I4.ViCC—a--——-—.—...-.——.-——-—.-—.—.—-.-.—-.-————————.—.—————-.-.————‘3‘3‘3‘3‘3‘3‘3‘3‘3‘3‘0D‘3‘3‘3‘3‘.0‘3‘.0‘8‘.0‘0‘8‘.0‘.0‘3‘0‘3‘0‘3‘.0‘.0‘.0‘0‘.0‘3‘3.)k)I’.J—.3——-‘I—I—I————I—-i.-.—I—I——-‘ei-—I—I—00m——04......4.....Vi(.IiViVi.4.ViViViViViViViVi‘.JiVi‘.JiUViViViViVi....V.ViViV.V.‘.1.ViVi(‘00000‘.0‘.0‘3CCC‘.0——C——————‘——CCCC0000CCCCC—I—-‘>..4c0’.V..JC’.CVi..0’.00’.0’.-..L.30’.—4.Vi.C’..C’.L.0‘3‘0.‘..o400i-’Vi—t.3‘.JCt’.)C0000O.00.JC.Vi0’.V.00Vi\0ViCL.)Vi‘0C00)—a’.SC’.ViC.)Vi-J.Ct..)CCt’-)t.J-C)‘.000‘3C00.CC).C00‘..JC’.—I——‘—l————,——-—-.——.-—1—I——.———————————..-.—.---CCCCCCCCCCCCCCCCCCCCCCCCCCOCCCC0C0CCCC‘.J)C))I.)ç.ç3...‘.jViVi0000000000CCCCQ0CCC0000CCC.-ZC-.001.3ViC’.000’.00ViVi—‘.3t’..)‘0%0’‘000ViC-1.3‘.3‘.0..l..300‘0-CCCi0000ViVi300JCVi’0Vi00—-.(.Vi1.3Vi00C%0CC’.Vi-.3‘.3ViVi00C00C1.3CC’0..—.-‘00i-’t-.3(.3004-‘-JCC-J‘0-00Vi-t.).(.3•tI1ViVi0’.Vi4..Vi0’.Vi.0’.—bViViViVi..b-.)k3(.3‘.31.3—(.3(.3..JC.fViIC’.--.-JVi——(.3-Vi0’.CC’...(.300C%0.—.‘Vi--.JViC’.CrnCViViViCViCViCCCCCV.CViC3CC0’.00ViViC(.3ViCViC- E___________(Thxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxcl)xxxxxcIx-,______________xxxxxxjx xxIx.n‘C’C—C’CC.)-C’C0ViViViCtn‘C‘C‘C‘CViViViViViViViViVi.ViViVi4.Vi.WL.)ViJViViVi——---ViVi---Vi---C’ViViViVi‘-—.—-.I-I--I-I-I-’II-e.—--.4....‘C‘C..(JJViViJVi000000ViViViC’Vi.00CViDVi‘-Vi-000000—-.--——————--—-.-———————.-————--—.————-.-————.——‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘C‘0’0‘C’0‘C‘C‘C‘C‘C‘C‘C‘0‘0‘C.o‘C‘C‘C‘C‘Ct.)t’.)J..JJ..J3(c)(JfC)L(.JI,.)ç)....Vi...........Vi3.4..P.......ViC’..44....ViViViVi.......4..eliC’CC00’C’CCCViViViViVicJC00‘CCC‘CC’..4CC’‘t.)‘0ViVi)C‘C-a.Vi00L.)-C-‘—I.C’Vi4..ViC’-‘‘COOViVi‘CViciC’C’-aC’‘.0—aC..)C00tC..)C’C’-aViC’C..)(...—IC-.)CCC..)-.CVit.)..00‘CCCC’-C-.)Vit.).C.-Vi-.-C’I-CCC’11‘111.’1C’CI...’1111I1‘-1—11.’-—11-.-1CCCCCCCCCCCCCCCCCCC000000000000CCCCCCCCCCCCCC0000‘C-00—LIC’.—Ii-.)-.—aC’C’ViViVi‘.0‘CCCCC-’C-t)C000000-a...-.‘CViI--C’t-.)Vit-C-.c...C..)C..).‘C‘C‘C‘.0C-.aVi-.ii-’‘C‘C-..C—.-J-.J-.)‘CC’C..)i-’-00Vic-a00.‘CC’Vit.)t’.)..Vi-0000C%0C-aCCCCViC’’.0OOVi’.0%0’C)C’C’C%0CCCCCQC..)VirnWViC’V.t.)00-J00-J0.C..)0000C’C’ViC-.)t..)C..)00ViC’..ViIIVi-aIC’I00IC’ViC.-.-‘C’‘CIib..)C..)‘COOC..)‘COOC..)iIICII.IIt)..IIII-C’CC’VienViC-aoC.C’C-.)C..).‘CCC-.)L-.)ViViVi-CCViC.)ViCCC.— I;xxx;.<,-xxxxx.<xxxxxxxxxxxxxxxtn ><%D00000\00i-—%O-J-J-CC.ViViViVi4-4-tt-3---HHHHHHH-3HHHH(00>flIJ>(Thb>OW>W>W>>-4-i-’04--‘H ‘-I>Cl)z>txiH tn00-—-—-—-—-—-—-—-—-—-—-CJ‘JJCJJJL.J‘I..3(J.JC..)C..)C..)C..)4-C..,4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-4-Vi4-4-en0\a00--—.100-.-00000000%.OOOViO4-4-ViC..)4-4-t—C..)0’.4-4-‘-00-JVi—J‘-0‘-3>0Vi-t-.4-0000i-%3t—C..)0’.0b-.)C..)0’.4-0’.%3Vi0—.0000’.00004-000%0-.)Vi00-0’ViC..)00000-.0--JViVi—J-J00’-’ViViVi0Vi0000-Vi000000000000000-,.,0’.00’-’0000Vi0ViVi00ViVi00Vi0000000000000000000000000t’C..)ViCl)—1C11111—1111’1‘‘)l‘))——C-l-C-.‘00000000000000000000000000000000000000000004-4-4-4-4-4-4-4-4-4-c.)4-004-4-4-C..)4-4-C..)C..)C..)C00—-..0’.000000C.C00000000-J--J-J0i-.t-.3t—)t-.)C..)C..)C..)4-000’.-J00C.000.)00-3t-.)%00C..)-J-.04-C..)-Vi00%0-J0C..)4--0C..)4--0000Vi%3i—Vi00C..)Vi0-JViL-.)ViVi000’.4-000004-004-0000Vi--J-..JViViViVi0000ViVi00ViVi.300000000Vi0000000‘.3C.)‘.300000ViViVi0’0’000000000000ViVi0000000000’0’00000-‘t’.)04-‘ (D>tllViCD•C.)C/)‘-rCI)‘-4‘ i Cl) :. C- 4- 0-—aTb.4...CC.JC..)C.)C..C.)C.)C..)C..)C..)-.i-i-i-i-.’I.-’I-’-—I-.-‘I-‘IIIIIIIIIIIIIIIIIIOQQ\Vii—IIIIIIIIC..J,)00-..C.)C..)C..)000- ViVi00VI000I0Vik)‘‘‘-C/))00Vi3C.)I0-.Vii’.)00Viii0i00-IIIa.•..00—000ViViCJVi00Vi0---___9_-—000Vi0Vi00Vi000ViVi0Vi000Vi0ViOViVi.r.0-av.00-.Vi00Vi00—OVi0a.t.)0000Vi000Cl)V.4.CVi.C..)C..)C.)4.....4.--.1C’ViViViViViViViViViViViViQViViViO’C.40eoooo0cn00000-.-c.)0Vi00t.)\O—0.).)0Vi-...00-.J0000t.)—.ViVi—0O000C..)000Vi%CCj0 C..);;--AAAAooooo00 00-VI0I-’04C.Jt.30000000IIIICl)II[ I III>IIci>0000000000000000000000-I—‘I‘IIiii-.O00-Cc,i.JI—a00-)C c00jI-avCV.C———c\V.C‘‘C>CV.V.CV.CCC‘r—‘1”...CCC)CV.CCV.V.V.CCC-(L3tJ—•00C,vivivIcchCoCCC,a’a’caa’a’a’a’-via’vivi.........V.V.viVi--a’0’.a’a’0’0’00-t’.)a’‘.0a’vi%O%0%0-vi0’..—CCC—.C‘-a’a’‘.0a’C‘.00’.-CCV.C-300t,...‘.0C‘.0ViC—4.ICl)CO)cCCCCCpCpC00Ca’Ct-J;;-t-.)AAAAAACCCCCCCCCCC22C2C°C0CCCC‘CCCCC——,_—,_——,_‘,_,_,_—2<IIIIIIIIIIIIIIIIIIIII_IIC)IC C———————————————————————————————————————acccVicIIIIIIILI‘‘•.I.-—‘‘‘‘‘‘‘‘‘‘,‘‘‘—————.‘—-——‘tn————‘.000Q’..Vi.C)‘.000-.)Q’Vi.‘.3i-.Vii-CCI..)——t.Ji-ViCC-Vi-Vi)C’.cc.),••ViCViCtViViCCViCCI-’ViC0ViCCViCViC,IViCViCViCCtT-Jit.JJtVit.C.JCCVi3ViCC—aCcSiCVC0CViCCCViCViCViCViCC..—.)—ViC-ViCCViCCCC—J00-ViCC-C00ViVi)C—.Vik)C%0Ck)CViCViCCCViViCViCCCCViCVi.CC’.CViCCCViCCViCViCViCViCC-.-a—L)‘..ViVi—C’.-CC’...4.CC-.JC‘CC’.ViVi‘..ViVi‘Vi004000000—JVi‘0.‘.0CCC‘.)Vi.C00--.0’.ViViI‘.0t’3C’.—.0’.00-’Vi-..-.CViViViCViViViC00 Io-—CCt’..).3—CCCCCC00t’..ViViI’..)t’..t.)..‘..)00CCt’.)CCCCCC‘.0C‘0o‘.)ViCC-.J....0’.L.)CCCCCCCCCCCCC.)CC0’.‘.Vi.‘‘.JCCCCCCCViCC-t’Jt3t.)--i--i--i-i--.)_i--i--IiCCCCCC0CCCCCCCD‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.0‘.04.0000000000000000000nIIIIIIIIIIIIIIIIIIIIII—————‘.o--Vi>00J:-t1.I—C‘.0Vi4.Ii.—oVi...).)Vi4.CViVIViViViCViViViVi.‘Vi.•CViCViCViC1i-Ct.)-.C00ViCCOOIt’.)IIiIIIIC—‘-.)IC\‘.0i—C‘.0a..—‘.oo’..‘—0000CZVi4CVioViCViCcVICViViCVICVICCViCVIVi2922292--9---9.299—---CViC)CViCViCCtViC%04.0O.O40000C00ViL.C00CuViCCCViC0CC0CViViCCViViViViCViCViCViCViCViViCCViCVi0CC00--C’0’ViViViViViViViaQ’.Vi-.a--0o0’.00Cai—CJCCViOVi4ViViVi0-O04c..JC‘.0-J00‘.04000C..3-0.-00‘--Vi4.0Vi00-‘c)00CC—00I-.Vi-W—Vi4.J..JL.3t.)c..)VIc.Jç.cc.o—CCCCC‘.04.04.0‘.0COOC---.?);;OCCCCCCCCCCC0’.-C’...-.CViVC’.C’.Vi.C’.—J-ViC’.ViVi._Vi-ViViVi..ViCCCC‘-i.)CViVi-.4.0.0’.C000CCCCCCCCCCCCCCCCCCCCCCCC-<IIIIII LIIIIIIIIIIIIII—..IoIIIIIIIIIiIIiI IIIIIII IUIii1111111I11011111111111111111111IIIIIiIIII:1:IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIillIlIllIlliIIIiIW‘HI11111111111IIlOllIlIlliCeekt-)tJ—--‘I—‘I-I‘——-CCCCCCCCCCCt1t’.——cC—,ç)——CacoCoCcCViC—Vi.3-ViViCCCC+CCCCC—ViCViCViCVi,ViC•CCCCCCCCCC•cLC-jc.i—ViViViVi.VViCCCI)OCCOCCViCViCCCCCCCViCViC——a.a’Vi‘——‘‘—‘-ViViViVi..‘.JcJ——c—.via,vik)—viC’viovio.CViC’ViCC’ViViViViC’C’C’CC’VC’ViCViViCC’CC’C’CViC’ViCoa0—.—0%0%ViViViViViViViVi0%0%..J0%0%0%-a-J0%00%0%0%ViVi..4.-00oC’ooC’oovi-c-a.%Vi00%0Vi0%0-C00Vi-I-ViC-.Vi-C’..Vi-J—C’00-tJ...;;-aaaaa-r-aaaa..aaaaaaaaaaaa‘.C’.C’CC’C’CC’C’CC000CC’C’C000——————————————————%__pci>)t’.)I.)t3aCViViViViViViViViViViViViViVitCrjtIIIIIIIIIIIIIIIIIIIIIIIk)O(Th((rr1IIIIIIIIIII.ViVitt3I-’000QVi.ViViVi..tVi.4.(JViViViViVi..-cViVi30ViViViViC-Vi)CICViViViViViViViViViCV.CCViCCViCV.CCViCViOViCCViVi)C-Vi-)CViCViCViOCC‘-.ViCViCViCViCViViCCCV.CVi00-—Vi3ViVICVitViV.Vi..I.0Vi‘.Vi0ViViVi..Vi....JJViCViCViCV.CViCCCV.CV.00ViC—Vit’3CC‘-Vi‘00’ViVit.)C-.Vit’J.ViCCCCCCCCCCCCCCCViCViCViCViCCViCViViCCCViCViCViVi-HVi0CViViViViViV.Vi.ViViVi00C0Vi--0000C—.)—--ViVi‘0CVi4ViC-0oo0-JViVi00C..)Vi00.CC-.IC00pppppp-—CCVi’0OO0O%0AAAAAAAAAAAAAACCCCCCCCCCCCCC4CC.J4ViC)c,30I-)---*6-6-6-6-6-6-6-6-6-6-CC().Vic.co.000CCCCCCCCCCCCCCIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIiIIIIIIIIIIIIIIIIIIICII4IIIIIIIIIIIIIIIIIIIIIIIiIiIIIIIIIIIIIIIIIIIIIIIII.I6IOIIIiIIIIIIIHIICwwwwwwt%)t.)t.)t)000000000000000000-_I-—--,.---—J——--—I-.--——I0iCC’,O’C’iCrn‘•ti•,,i•iet,ii.,.,..,eei,,trnI—.Q003-00-q.,ViWJ,)l._._.JC’C’C’CViC’C’JV.C’Vi.)ViC’)ViC’,.ViC’C’CCC’C’ViViC’ViCV.C’V.C’ViCCViCViC’ViC’CV.,V.C’C’C’CCC’C’C’CCcCVic.’(..JCCC’CViCC’V.C’—Vit3C—.)Vi.)-ViC—.1Vit’.C’ViC’..C’ViC’ViCViCVi—.1Vi‘C’V.CCC’CCCViC’ViCViCViC’ViViC’ViCViCViCCCCC’C’CCCViCViVi4ViViViVi..4..1_i-’•C’ViVi...J3ViI.)ViViVi-C’ViC’-Vi.JC—Vi-Vic-.) C’—)ViC’ViCbC’—-.3—t’.)-Vi-CViCC’C’ViViViCViC’ViC’ViCViCViViC’ViCViCViC’ViViViViViViViViCViViViViViC’iViVi3c.)4.ViVi--C’.C,ViViViViViViViViViCCC’CC’%DCC’C’—OViO0t.)-)ViC’C00Vi00O00Vit300O00t—)C’—Vi,-‘—0000.00I-——O.-l.JC’CC00Ci00.00—.)-I c-.)C’C’C’C’CCCC’C’C’CCC’C’C’C’CC’C’C’C’C’CCC’C’C’C’C’CC‘-‘C’C’C’CC’----—CC’—tt’—JVi-00C—\-400C...3Viit3C’C’,-C’..00QC—L’3.ViOC’..C’-CCC’CC’C’CCC’C’‘ViC Vi2C”Jc.Jc.Jc.JJL.Ji.-.i-.i,iICCCC‘0‘0‘0‘0D‘0‘0‘0‘0‘0‘0‘0‘0‘0‘0‘0‘0‘0’0-rIrIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIc’VI.i‘000-OViI-E’JI-’%0000’Vi-.t’.)UIL.C—p—.....C)V-.C-ViL.)C-Jk-Vi)CViCViO’Is.)Vit)—Vit.,)C-Vit’.)0-Vit’.)0-Vit-3CVi-)ViCCVIOUIVIViCViCVICViCOViCViCViVIOViCUIOUICUICVICUICVIC•CUIC,aiiIIIIIIIIIIIIIIII$I(.JIIIIIIIIIIIIIIIIIIN•0———VI.I—)t-.Jt.)b..)k)t——.—-.ViVi‘‘ViVi.1..(.•1.).)W-).-.——-U.V.kViCVi-CViCCUIC0000ViOViOViOViCOViCCVi0VICtT——‘-.t.-k)—1l..3ViVi4....La)t..).3I-tliViCCViCViCCCCCViCViCViCViCViCVaCCViCViCViCViCViCViCViCViCVicViVIVIUIVIC’0—C’0-00Vi‘000..00‘0C000ViC004..Ct-J—.—LaLa)0——C4.t-‘30’00a—t.)La—CQ00—.t-3.00Vi‘0-‘0La)t-.)’00..—a-b’000-.‘00—‘0CV..0a-00%0CaC-..Vit.)C‘0t-)00‘0I—-Pa-Pa-00a--0\00‘0Vi‘iCt’.)o‘0LaJk)00ViC0aCLa)-Pa0aC‘J.k--ViCCLa)‘0001’OC’.Pa-PaCa-0000.Pa-00cc————a-I-CCCCCCCCCCCCCC0CCCCCOCa-a-’CCCCCCCCCCCC—a—ac3a-a-,)00ViLaViVi00La)4aICCCCC000Vi0’,C0’a-PaC’,—J00.Paa-O’a-a-t’.)00C’iLa)0000OCC’a‘0ViCQLa)C’..Pa--CCC,-<Ca-—0000000000000000-J-00000’,00-J00000000‘0%0‘3‘0000000‘00000000000‘0‘.000Vi.PaLa).PaCLaa-Vi004-JLa)‘.000-J00‘3CC0000‘.3‘.0C‘3000’.0’.VithotHtHot—*-e,IVIJLJL.)c.J88ViViViViViViViViViViViViViViViViViViVi.4..LIIIIIIIIIIIIIii!-!‘‘(tC00CVi..)0a.—.--0-C’.t’0Vit.)-Vi,Vi‘‘‘0ViViVi0VI00Vi,Vi0Vi0c1111CC0000000C00—:i——3%OC’VitVi0Vi—-—‘0000-0Vi0Vi0Vi-L——————————————C’——————Vi0Vi0Vi0ViCVi0ViViCVi0Vi‘.CC’Vi.)00000000000000OOOOViViCVi0Vi0ViC-1jViViViViC’C’ViViViViViVi.C’CC’C’C’--)4.C%CVi000000C’CJ00Vik)0.)0-C’..c,-—C’0000C’C’)Q0C’C’I)p——00á0pCC-i—Cpp00000000-000%C%CO000000C’-i-C00000000CC0000C00.C’000CCC000CCC’-.-.000—.Vi-3k)C—a—t).<0000LJ0J00L0000000C’C’00IIIII—’IIIIIiIIIIIIIIIIi.I%IViIIIII-400IIC’It.3IOI00I00I00ICI•TTii00ij1iiiiiiijHJII101IIIIIIIIIIIII:rIIlIII101IIIIIIC -aI-I•—I-1-41-4—11-1-4Cd)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)Cl)1-—‘———1-I—1-—c)c.)Iaaaaaaal)1).(.3l..It1)1l..t)1ICJtJC.)C.)‘—oo-a--I-GO—a1-’aaaa•aaaaaaaaaaaaaaaaVi4C..—)—————‘—%00-t)Crjooc.—0C.)C.J‘)t%)t..)111-1lVit.)c—aVi0-Vi0‘°--++±00+±±±-+C.)+Vi÷—-aVit’.)—a-ViOViCIOCJICVIOViC1aViVi00aaIIIIIIIaa-aVit-.)C—.vt-.)o-aVi-)v—.)k)C.)CViC.)C.)t-)ViC..)—CViViC-aVi-_)ViCViCViCViCViCViCViCViCCVI‘ViCCCCCC-t—C-aS-aVIt’-CVit-CVit)C.)CVi-QC..)----—-——--aC.)4.C-.)..C-.)C-.)t-.)t—JC.)t-.)C-.)C-.)-QViCVi0.C’.0’.ViVi0’.0’.0’.C’.ViC’.0’.0’.ViC’.C’.C’.ViC’.0’.0’.0’.—aC’.0’.0’.C’.0C.oC’.-aViaViViViViCC-000C’.t.)CCViViVi)CCC00.)Cl)C-.)Vi000’..C-.).ViC’.-C’.Vi4-—Ct—)41-.)C‘-.3C—CC-.)I—I—CC..)Ct)Cc-..)Ct—)C.CC-.)CC-.)C-.)C———C-.)ViC00.--ViViViViViViViVi‘-CC-)Vi‘.00‘.0.Vi00t-.)t’.)C‘0a’-.)‘-0‘0ViViC.C-.)Vi—C.Vit)00CVi‘0‘-.)000000000..0’.00;;oCCCCCC000CC0CCC0%0GOC-—eI00.C’.t-300C-.).t—.)-C..)C-.)t..)C-.)taC’.C-.)C’.Vi..a’.C-.)-J.-CCC0’.00CC-———————————l-<LOUOJJOOUJOOUJ00IIIIIVi,C’.1ViI.JiViIViIViI—JIa0IC-.3I’000OOICjCIIlJ?lIIII>IC 00FTEit...J)C3IIIIIIIIIIIIIIIIIIIIIIIIIIViLchVi)0)-0t11IIIIIIIIIIIIIIIc0OViViC%QC0vCVivcuoooooooooooooooooo‘ro09—;IIIIIIIIIIIIIcç,,0ViVit)0Vi0Vit.)ViCi—00Vi0-4—00\C-cC-..-00ViCC00--JViViVi——.-Vi0000-00ViC000Vi‘C00000000000000000000000Vi......)ViViViViVi....4.4...4.4i.L.(.J---.-J.)000000,)QCO00\O%O-00ViQ0-..I--Vi--J00-Jci-00%OC0.3‘.01l.00‘.0‘.0JI-‘.0‘.0-‘.00‘.0-—4.Vi‘.0t-.)‘CC’Vic..-.Vi-00CPooo000000•0t.J.JC.J00‘.0Vi---AAAAAAAAAAAAA0000AAAAAAAAAAAAAAAAAAAAAAAAAAA0000CCC0000000000000000000000000000000000•0•0•0•0•00000000000•00000000000000•0•0•0•000———————————————————————————————————‘.<IIII—jIIIIIIIIIIIIIIIIIIIIIj0000U00IIIIflIIIiIIniCr-t-r-rt-rrrt-r-rrrr-r-rr-r-4-J-0’00%0%00’Q\0%0%00%0’0%00%0%0%0%0%0%0%0%0%0%0’0%0’0%0’0%0%0%0%ViViViV.ViViViVitn)-ViViV.V.ViViViViV.VIVIVi0-i_-C0’Vi-CViVi.....)t3‘,ViCViCViCViCViCViCViCV.CV.CViViViCViCViCViCViC,ViCViCViCViCi-ViViCViCViCViCViOViCViCViCViCViViViCViCViCViCViCVi0OVi—JViCViCVi0-‘-ViViViViCCViCVit-)C-ViCViC-ViCI—%0Vi)CViC-ViCC)00-.JViC-CCCViCViCViCViCViCViCViCViCViCViViCViCViCViCViCVi00Vi-JViCViCVi.i.L.J•C—J0C%Q0000‘.000—.-‘.00%311CC.‘.0Vi0%Vi0’ViI-’C‘.00%0%Vi.00—J—‘0C‘.000Ct.)Vi00c.JC0%—.0%‘.0(Vi——0%-Vi‘0-.tJ..C‘0CC0%l‘.0—ii’!•C•CCCCCC©CCOCCCCCCCCCCC——————————————————————‘—c--r-r-rrrrrrt-t--rrrtt4rr--——-000000000000coo.,.......•....•...rxi..oo-o’.v...)—cM.IJt)000Vi.c.00—0000--ViV....c.-..-aa’,o’VicMViVi.t.,.I-a’.0__V.ViViOViOViOVi0000Vi)CU.0ViCViCViCC0C00ViCViCVi0Vi000ViCViC0CViOCCV.VitiICiIIIiiiIIIIIIIIIIIIiIIIIIIIII—000000‘-C\ViVi.44..)0Q’,ViViVi.J.)t’t’‘0’C‘‘VIC-.ICJC0V.0Vi0‘CViC-Vit.)0ViC-Vi0U.U.U.CU.C’V.CVICU.OViOV.CCCCCU.0ViCViCViVi00ViCViCViOCCCC-QViViVi-.-e-CCVit’.)0-.0-.0Vi0Vi0Vi00cO—ViC—Vi-)0ViC-Vi0‘.00’-..V.CcMCOViOViOViCVi000000U.CVi0Vi0ViVi0CViCVi00Vi0ViU.U.CC....ViViV..Vi..)..CViV.ViViViViVi..Vi3c.)c.V)0Vi....c.Vi)V.-CJCcc00o000%0c00o0-00-—‘.0‘.0‘.0))V.-c.ccCCVc.)ViCk)c.)0’,.C\0—.‘.000t.)00‘.0V.‘.0—C-00)Cpppppppppppppc.Jt.)c.)c.)t.)c.)c.).)tc..c.c.)c..t——...‘.0‘.00-CI-.CC.-..0000COC000000000000C0000000000000000000000COC————————I-I-I-i-III)1‘I11—‘I—1—Acc00ccoccccccccccccc‘cccccccc.c.cca’,-c.-a’j.c.ccoocc-CIDU.U.4UHUILHH1tr1t11t11txrn00000000000000•..,••II-’-I-‘—I-I--‘r1r-—.D00-IC’Vi..(JViViViViViViViViViViViViViVi0000000000000000000——.)——0IIIIIIIIII4IIIIIIIIIIIIIIIIIIIIIII-’‘%D00ViI-II-I--D00-CVi.)-‘0-0%00Vic-0JI)t.)t.)0000-.—CViViViVi..4.-J0—..-.00Vi0-.-.I-’00ViO’I0t..)-0-.t.)0Vi\0.DVi..4.3...00000000000•00Vi0ViViViVi0Vi0Vi0Vi0Vi0000Vie.IIIIII..3tIJt3i-i-cojViViViVi.L.)t,)t&)I!IIIIIIIII.IIIIIIIIIOL30—...—00Vic..) .OCc.0-.1.00Vi000Vif..)Vit’.)0Vi0-Vib’.)0—Vi0Vi0%VioooVi00uocIoVioVioooooop.(.Ct.)t,)I-.-0000--.0’.0’.0’.ViViViVi4...)—C’.C0—X.—00Vit-300’.CJ0—..I-00ViViVit’0Vi0—Vi0—Vi0Vi0\DVi..44....00000000000000Vi00Vi0Vi0Vi0Vi000000-ViViViViViViViViVi44ViViViViViViViViaVi•Vi•Vi•00•\o•Vi•ViVi0000’00Vi00D00ViC’.)00O0a’.--a000%0C’.00-J)---\0000000ViViVi-C’.4)Cl);AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA0000000000000000000000000000000000000000000000000000•0•0•0000000000000000000000000—-II1‘———IIII—11)11——1—111—11111—‘.<IIIII IIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIIAPPENDIX BSAMPLE DEPTH PLOT pH TOTAL SALINiTY DRAINED 14C AGENUMBER RANGE DEPTH SULPHUR (%) MOISTURE (a. B.P.(cm) (wt%) (wt%) adjusted)MILE 1-13 364-394 394 4.08 <0.01MILE 1-14 394-424 424 4.27 <0.01MILE 1-15 424-454 454 447 <0.01MILE 1-16 454-469 469 4.46 <0.01MILE 1-17 469-484 484 485 <0.01MILE 1.5-1 0-30 0 3.79 0.22 <0.01 91.4MILE 1.5-2 30-60 60 3.33 0.18 <0.01 91.4MILE 1.5-3 60-90 90 3.44 0.21 <0.01 93.7MILE 1.5-4 90-120 120 3.25 0.18 <0.01 92.6MILE 1.5-5 120-150 150 3.40 0.19 <0.01 93.9MILE 1.5-6 150-180 180 3.52 0.16 <0.01 93.1MILE 1.5-7 180-210 210 3.58 0.16 <0.01 94.9MILE 1.5-8 210-240 240 3.71 0.17 <0.01 94.0MILE 1.5-9 240-270 270 3.84 0.18 <0.01 93.5MILE 1.5-10 270-300 300 3.99 0.22 <0.01 94.6MILE 1.5-11 300-330 330 4.33 0.26 <0.01 93.9MILE 1.5-12 330-360 360 432 0.22 <0.01 93.3MILE 1.5-13 360-390 390 4.52 0.22 <0.01 93.3MILE 1.5-14 390420 420 4.56 0.25 <0.01 94.6MILE 1.5-15 420450 450 449 0.23 <0.01 93.6MILE 1.5-16 450480 480 4.85 0.26 <0.01 92.6MILE 1.5-17 480-510 510 4.79 0.25 <0.01 91.3MILE 1.5-18 510-540 540 4.72 0.24 <0.01 92.4MILE 1.5-19 540-570 570 4.63 0.25 <0.01 92.2MILE 1.5-20 570-600 600 4.79 0.27 <0.01 92.2MILE 1.5-21 600-630 630 4.76 0.51 <0.01 84.8MILE 2-1 0-30 30 3.69 <0.01MILE 2-2 30-60 60 3.43 <0.01MILE 2-3 60-90 90 3.45 <0.01MILE 2-4 90-120 120 3.54 <0.01MILE 2-5 120-150 150 3.79 <0.01MILE 2-6 150-180 180 3.87 <0.01MILE 2-7 180-210 210 3.96 <0.01MiLE 2-8 2 10-240 240 4.09 <0.01MILE 2-9 240-270 270 3.90 <0.01MILE 2-10 270-300 300 4.37 <0.01MILE 2-11 300-330 330 414 <0.01MILE 2-12 330-360 360 4.21 <0.01MILE 2-13 360-390 390 4.34 <0.01MILE 2-14 390420 420 4.36 <0.01MILE 2-15 420450 450 4.33 <0.01MILE 2-16 450480 480 4.45 <0.01MILE 2-17 480-510 510 438 0.01MILE 2-18 510-540 540 4.37 <0.01MILE 2-19 540-570 570 4.48 <0.01MILE 2-20 570-600 600 — <0.01212APPENDIX BSAMPLE DEPTH PLOT p11 TOTAL SALIMTY DRAINED 14C AGENUMBER RANGE DEPTH SULPHUR (wt%) MOISTURE (a. B.P.(cm) — (wt%) (wt%) adjusted)MILE 2-21 600-630 630 <0.01MILE 2-22 630-660 660 <0.01MILE 2-23 660-690 690 <0.01MILE 2-24 690-720 720 <0.01MILE 3-0 Surface 0MILE 3-1 0-30 30 3.91 <0.01MILE 3-2 30-60 60 3.83 <0.01MILE 3-3 60-90 90 3.80 <0.01MILE 3-4 90-120 120 3.91 <0.01MILE 3-5 120-150 150 3.84 <0.01MILE 3-6 150-180 180 3.83 <0.01MILE 3-7 180-210 210 4.01 <0.01MILE 3-8 210-240 240 407 <0.01MILE 3-9 240-270 270 429 <0.01MILE 3-10 270-300 300 4.34 <0.01MILE 3-11 300-330 330 4.15 <0.01MILE 3-12 330-360 360 3.97 <0,01MILE 3-13 360-390 390 4.03 <0.01MILE 3-14 390-420 420 4.06 <0.01MiLE 3-15 420-450 450 3.93 <0.01MILE 3-16 450480 480 4.35 <0.01MILE 3-17 480-510 510 4.27 <0.01MILE 3-18 510-540 540 406 <0.01MILE 3-19 540-570 570 4.36 <0.01MILE 3-20 570-600 600 4.63 <0.01MILE 3-21 600-630 630 4.55 <0.01MILE 3-22 630-660 660 4.79 <0.01MILE 3-23 660-690 690 4.79 <0.01MILE 3-24 690-720 720 4.79 <0.01MILE 3-25 720-750 750 4.86 <0.01MILE 3-26 750-780 780 4.64 <0.01MILE 3-27 780-8 10 810 5.15 <0.01MILE 4-1 0-30 30 3.98 <0.01MILE 4-2 30-60 60 3.78 <0.01MILE 4-3 60-90 90 3.84 <0.01MILE 4-4 90-120 120 3.67 <0.01MILE 4-5 120-150 150 3.95 <0.01MILE 4-6 150-180 180 3.99 <0.01MILE 4-7 180-210 210 4.12 <0.01MILE 4-8 210-240 240 4.21 <0.01MILE 4-9 240-270 270 3.93 <0.01MILE 4-10 270-300 300 4.44 <0.01MILE 4-11 300-330 330 4.18 <0.01MILE 4-12 330-360 360 4.20 <0.01MILE 4-13 360-390 390 4.06 <0.01MILE 4-14 390-420 420 416 <0.01213APPENDIX BSAMPLE DEPTH PLOT pH TOTAL SALINiTY DRAINED 14C AGENUMBER RANGE DEPTH SULPHUR (wt%) MOISTURE (a. B.P.(cm) — (wt%) (wt%) adjusted)MILE 4-15 420-450 450 4.22 <0.01MILE 4-16 450-480 480 4.36 <0.01MILE 4-17 480-510 510 4.39 <0.01MILE 4-18 510-540 540 4.48 <0.01MILE 4-19 540-570 570 4.39 <0.01MILE 4-20 570-600 600 4.18 <0.01MILE 4-21 600-630 630 4.63 <0.01MILE 4-22 630-660 660 467 <0.01MILE 4-23 660-669 669 4.67 <0.01MILE 4-24 669-690 690 4.43 <0.01MILE 4-25 690-720 720 4.89 <0.01MILE 4-26 720-750 750 5.06 <0.01MILE 4-27 750-756 756 5.06 <0.01MILE 4-28 756-780 780 5.39 <0.01MILE 5-1 0-30 30 3.89 0.14 <0.01 92.4MILE 5-2 30-60 60 3.79 0.20 <0.01 93.1MILE 5-3 60-90 90 3.69 0.21 <0.01 93.1MILE 5-4 90-120 120 3.60 0.20 <0.01 94.8MILE 5-5 120-150 150 3.89 0.19 <0.01 93.0MILE 5-6 150-180 180 3.79 0.21 <0.01 95.5MILE 5-7 180-210 210 3.82 0.20 <0.01 94.5MILE 5-8 210-240 240 4.05 0.18 <0.01 95.0MILE 5-9 240-270 270 4.02 0.20 <0.01 95.6MILE 5-10 270-300 300 3.77 0.20 <0.01 95.6MILE 5-11 300-330 330 4.10 0.17 <0.01 95.3MILE 5-12 330-360 360 4.30 0.18 <0.01 96.0MILE 5-13 360-390 390 4.12 0.20 <0.01 94.0MILE 5-14 390-420 420 4.29 0.26 <0.01 93.1MILE 5-15 420-450 450 3.92 0.17 <0.01 94.5 850+/-80MILE 5-16 450-480 480 4.28 0.15 <0.01 97.1MILE 5-17 480-5 10 510 4.50 0.20 <0.01 95.3MILE 5-18 510-540 540 4.43 0.23 <0.01 94.8MILE 5-19 540-570 570 4.45 0.26 <0.01 96.2MILE 5-20 570-600 600 4.25 0.23 <0.01 94.9MILE 5-21 600-630 630 428 0.17 <0.01 93.8MILE 5-22 630-660 660 .4.54 0.24 <0.01 96.0MILE 5-23 660-690 690 4.53 0.23 <0.01 90.3MILE 5-24 690-720 720 4.44 0.19 <0.01 92.7MILE 5-25 720-750 750 4.63 0.21 <0.01 93.7MILE 5-26 750-780 780 4.83 0.24 <0.01 95.7 3040+/-80MILE 5-27 780-810 810 5.29 0.29 <0.01 93.6MILE 5-28 810-840 840 5.40 0.30 <0.01 93.3MILE 5-29 840-870 870 5.52 0.22 <0.01 93.3NL1 0 0 3.87 0.01NL 1-1 0-25 25 3.87 0.01NL 1-2 25-50 50 3.96 <0.01214LI.>-I-I-——•l-)IIIIIItIII———.J! VIVI Vivi3CVI)CLIio.VI.CVICCCVI>VICVIV.I-1C—.00-..ViCV.CCV.CCCViCViCViC4........00VI00VI-a-c’.—oo—c,—‘-tOCV.3 !!CCC•C——————————‘—I-<ea VICJIAPPENDIX CIAPPENDIX C: RESULTS OF WET SIEViNG:: II%f = Moisture content of drained peat, in wt%.Z = Plotted depth of sample, in cm.%Ash = wt% of mineral grains, separated by flotation.% C = dry wt% of particles> 2.0 mm, by sieving.%M=diywt%ofparticles<2.Ommand>025mm.% F = dry wt% of particles < 0.25 mm.SAMPLE Depth %M Z %Ash % C % M % F CommentsBDD 22 l3ase= Minerals of mixed lithology and no coral frags: note ostracodes and forains disappear at base.1 BDD 22-1 0 Hemic 0 24.4 10.7 30.7 34.2 diatoms,bugs,fragmental2 BDD 22-2 100 — C Hemic 100 32.8 20.6 20.4 26.2 ‘picles,tooth,diatoms,etc3 BDD 22-3 150-200 Hemic 175 52.1 6.4 18.7 22.8 fishscales,forams,frags4 BDD 22-4 200-250 F Hemic 225 50.8 1.8 7.1 40.3 stracods,forains,diatoms,spicu1es5 BDD 22-5 300-350 Hemic 325 59.0 4.6 18.1 18.3 OStX4spp,forX2spp,wOOd, shells6 BDD 22-6 400-450 — Hemic 425 65.4 2.7 19.2 12.7 C:shells : diat,spic,ost,for,shell7 BDD 22D-1 560-590 F Hemic 560 5.0 9.6 32.4 53.0 flOOO8 BDD 22D-2 600+BASE not peat 600 72.0 2.5 10.5 15.0 filmS, pollen,no coral9 BDD 22-7 600-650 not peat 625 rounded corai&shefls,ost,for,minsNL110 NL 1-1 0-25 F: Hemic 25 0.0 26.4 17.6 56.0 clear water11 NL 1-2 25-50 F: Hemic 50 0.0 25.9 29.7 44.0 clear water12 NL 1-3 50-75 F: Hemic 75 0.0 24.0 33.9 42.113 NL 1-4 75-100 F: Hemic 100 0.0 17.0 35.5 47.514 NL 1-5 100-125 F: Hemic 125 0.0 19.0 40.8 40.215 NL 1-6 125-150 F: Hemic 150 0.0 19.6 36.4 44.016 NL 1-7 150-175 F: Hemic 175 0.0 13.9 35.6 50.5 Campnospermaseed”a17 NL 1-8 175-200 — F: Hemic 200 0.0 12.5 44.7 42.818 NL 1-9 200-225 F: Hemic 225 0.0 11.4 32.9 55.719 NL 1-10 225-250 F: Heinic 250 0.0 10.9 41.1 48.020 NL 1-11 250-270 F: Hemic 270 0.1 5.1 43.3 51.5 fewangularsandgrains21 NL 1-12 270-300 280LAKE 6.5 Disseminated mins throughout in Evaporate fraction; occ more concentrated.22 LAKE 6.5-1 0-50 F Hemic 25 0.0 27.8 29.2 43.0 s5y,WOOdfrS.S: xtals23 LAKE 6.5-2 75-100 CHemic 100 0.0 37.1 26.3 36.6 Camp seeda,roots: xtals24 LAKE 6.5-3 100-125 Hemic 125 0.0 28.9 33.2 33.9 grassy: XtalS25 LAKE 6.54 125-150 Hemic 150 0.0 31.5 34.7 33.8 grassy: xtals26 LAKE 6.5-5 150-175 Hemic 175 0.0 30.0 40.0 30.0 (Estimateduetoloss).grassy:xtals27 LAKE 6.5-6 175-200 ‘ Fibric 200 0.0 40.2 36.4 23.4 cork Palm wood, woodfrags: xtals28 LAKE 6.5-7 200.225 — Hemic 225 0.0 28.8 38.9 32.3 grassy.woodfrags29 LAKE 6.5-8 225-250 — C Hemic 250 0.0 35.0 37.1 27.9 smwoodfrag,grass30 LAKE 6.5-9 250-275 Hemic 275 9.9 22.9 32.4 34.8 numerous seed d,grassy: LTA31 LAKE 6.5-10 275-300 Hemic 300 2.3 25.9 45.0 26.8 twig, minerals,grassy: LTA32 LAKE 6.5-111300-325 FHemic 325 0.0 20.2 50.9 28.9 whitemould33 LAKE 6.5-1j25.350 Hemic 350 0.0 34.0 43.8 22.2 gt’’i’ags34 LAKE 6.5-13 350-375 — Hemic 375 0.0 27.0 46.5 26.5 grassfrags35 LAKE 6,5-14 375-400 — Hemic 400 0.0 32.3 48.4 19.3)’216APPENDIX CSAMPLE Depth %M Z %Ash % C % M % F Comments36 LAKE 6.5-15 400-425 Hemic 425 4.1 29.3 36.6 30.0 detritalmins37 LAKE 6.5-16 425-450 CHemic 450 0.1 35.5 40.0 24.4 ‘Y38 LAKE 6.5-17 450475 Hemic 475 0.1 25.9 47.0 26.0 gIy39 LAKE 6.5-18 475-500 C Hemic 500 5.7 36.3 38.0 20.040 LAKE 6.5-19 500-525 F Hemic 525 10.0 15.0 35.0 40.0 wood,barkfrags LTAonfmes41 LAKE 6.5-20 525-550 notpeat 550 76.8 2.5 8.4 12.3 fotpeat HTALTAonfmesLAKE 8 Base= frosted white, grey, and pale green silt, no carbonate, much wood with embedded mineral grains.42 LAKE 8-1 0-50 — FHemic 50 0.0 18.1 46.6 35.3 gry,r0otu00stt43 LAKE 8-2 50-75 C Hemic 75 0.0 38.9 38.2 22.9 gry,roots44 LAKE 8-3 75-100 — Hemic 100 0.0 21.5 41.1 37.4 H2Oalmost clear45 LAKE 84 100-150 Fibric 150 0.0 55.3 29.4 15.3 Cyrillaleaf,wood,bark46 LAKE 8-5 150-175 — Fibric 175 0.0 39.8 32.2 28.0 peanut-shaped root tubes47 LAKE 8-6 175-200 Hemic 200 0.0 30.6 43.5 25.9 peanuts, grassy, H2Oclear48 LAKE 8-7 200-225 Fibric 225 0.0 38.4 38.4 23.2 fewerpeanuts, more flatfibres49 LAKE 8-8 225-250 — Hemic 250 0.0 24.0 43.4 32.6 H20 still clear50 LAKE 8-9 250-300 Hemic 300 0.0 30.9 51.5 17.6 M:sniall seed51 LAKE 8-10 400425 Hemic 425 0.0 33.2 36.6 30.2 C:mass ofs. fibres-mossy?52 LAKE 8-11 500-525 — Hemic 525 0.0 29.0 47.3 23.7 C:rootlet-penetiated root ortwig53 LAKE 8-12 525-550 Hemic 550 0.0 32.7 40.7 26.6 H2Ostillck5r54 LAKE 8-13 550-575 Hemic 575 0.0 23.9 44.2 31.9 clear55 LAKE 8-14 575-600 — Hemic 600 0.0 23.4 49.3 27.3 F:axmuli,M:sheath-likesheetsmoss?56 LAKE 8-15 600-625 — Hemic 625 0.0 23.2 47.7 29.1 F:annuli M:seedhead (moss?)57 LAKE 8-16 625-650 Hemic 650 0.0 20.6 47.9 31.4 seed”g’: less annuli58 LAKE 8-17 650-675 FHemic 675 0.0 18.6 52.9 28.5 g”x2nowoodoinerals59 LAKE 8-18 675-692 not peat 692 91.0 3.0 3.0 3.0 lunitedlithxta]sembedinwoodLAKE 10 Colourless elongate xtals below 600; volc glass? + detrital xtals at base.6OLAKE1O 0 061 LAKE 10-1 0-25 C: Fibric 25 0.0 46.9 28.0 25.1 gr,toots62 LAKE 10-2 75-100 F: Hemic 100 0.0 22.2 36.5 41.3 g’63 LAKE 10-3 100-125 C: Fibric 125 0.0 73.7 21.1 5.2 Campnospemed,fruita64 LAKE 10-4 125-150 C: Fibric 150 0.0 50.0 35.0 15.0 small!65 LAKE 10-5 150-190 C: CHemic 190 0.0 47.1 22.3 30.6 Campnospermaseed66 LAKE 10-6 190-200 C: CHemi 200 0.0 48.8 31.4 19.8 wood, seedc67 LAKE 10-7 200-250 M: Hemic 250 0.0 27.5 53.3 20.268 LAKE 10-8 290-300 C: Fibric 300 0.0 62.5 23.2 14.3 woody roots69 LAKE 10-9 425450 M: Hemic 450 0.0 9.7 57.7 32.670 LAKE 10-10 450475 M: Hemic 475 0.0 13.1 55.5 31.4 ‘Y71 LAKE 10-11 475-500 — M: Hemic 500 0.0 18.1 52.9 29.072 LAKE 10-12 500-525 — M:Hemic 525 0.0 5.6 52.5 41.9 gy73 LAKE 10-13 525-550 M:Hemic 550 0.0 10.6 55.9 33.574 LAKE 10-14 560-575 — C Hemic 575 0.0 41.5 34.9 23.6 barlcblackfrags75 LAKE 10-15 575-600—CHemic 600 0.0 42.6 32.2 25.2 wood,bark,grass76 LAKE 10-16 600-625—C: Hemic 625 0.0 36.3 37.6 26.1 Wood,bS ss: many long xtals77 LAKE 10-17 625-650 — FHemic 650 0.0 19.8 38.5 41.7 Y”78 LAKE 10-18 675-700 — FHemic 700 0.1 26.8 35.6 375 many xtals79 LAKE 10-19 715-725—M: Hemic 725 0.1 20.2 42.3 37.4 blaekvolcglass,otherxtals80 LAKE 10-20 725-750 M: Hemic 750 0.1 17.3 45.7 37.0 glyfranents217APPENDIX CSAMPLE Depth O/,4 Z %Ash % C % M % F Comments81 LAKE 10-21 800-820 820MILE 182 MILE 1-1 4-34 — F Hemic 34 0.0 23.2 27.7 49.1 palm phytoliths, diatoms83 MILE 1-2 34-64 C Hemic 64 0.0 33.3 36.6 30.1 murky water, phytoliths,diatoms84 MILE 1-3 64-94 — Hemic 94 0.0 25.2 29.7 45.1 twigs,murky85 MILE 1-4 94-124 Hemic 124 0.0 21.4 39.9 38.7 rooty,twiggy86 MILE 1-5 124-154 CHemic 154 0.0 33.2 37.6 29.2 Iog(reinoved), wood87 MILE 1-6 154-184 C Hemic 184 0.0 31.8 41.2 27.0 wood, twigs, sedge?grass88 MILE 1-7 184-2 14 C Hemic 214 0.0 37.7 37.7 24.6 WoOdy,rootS,graSS89 MILE 1-8 2 14-244 Hemic 244 0.0 27.0 43.1 29.9 g’90 MILE 1-9 244-274 Hemic 274 0.0 29.5 46.7 23.8 barlç twigs, grass91 MILE 1-10 274.304 — CHemic 304 0.0 34.9 38.0 27.1 bark, grass,twigs92 MILE 1-11 304-334 Fibric 334 0.0 43.3 33.7 23.0 bulbous root, twigs93 MILE 1-12 334-364 Fibric 364 0.0 49.9 30.1 20.0 Campnospermaseedx6,bark94 MILE 1-13 364-394 Fibric 394 0.0 38.8 34.6 26.6 seed, bark, grass, wood95 MILE 1-14 394424 424 0.1 25.2 43.7 31.0 palm?wood, wood96 MILE 1-15 424454 454 0.1 33.0 41.3 25.6 wood, granulartexture97 MILE 1-16 454469 46998 MILE 1-17 469484 484MILE 3 330-630, White frosted mm grains disseminated throughout, earthy yellow magnetic mm at 300, 33099 MILE 3-0 Surface— C:Fibnc 0 0.0 47.7 24.2 28.1 nioss?,roots,wood100 MILE 3-1 0-30 C:CHemic 30 0.0 44.3 31.0 24.7 moSS?,roots101 MILE 3-2 30-60 C:Fibric 60 0.0 47.9 25.1 27.0 Charcoal in M! moss?102 MILE 3-3 60-90 — C:Fibric 90 0.0 58.1 18.3 23.6 “conns”,moss?, wood103 MILE 34 90-120 C:Fibric 120 0.0 52.2 26.9 20.9 “coflfls”,WOod104 MILE 3-5 120-150 C:Hemic 150 0.0 43.1 34.9 22.0 twif105 MILE 3-6 150-180 C:Hemic 180 0.0 43.4 34.7 21.9 large red wood(excluded)106 MILE 3-7 180-2 10 C:Fibric 210 0.0 48.5 24.6 26.9107 MILE 3-8 2 10-240 C:Fibric 240 0.0 47.7 29.2 22.9 wood108 MILE 3-9 240-270 — C:Hemic 270 0.0 41.7 25.7 32.6 WOOd, black H20, large spore?109 MILE 3-10 270-300— F:FHemic 300 14.4 28.5 22.5 34.6 white, earthy yellowmagmins110 MILE 3-11 300-330 F:Hemic 330 1.0 33.5 31.0 345 mixgr,white,magmms,Campseed111 MILE 3-12 330-360— C:Fibric 360 00 41.9 29.2 28.9 ‘‘‘112 MILE 3-13 360-390— C:Fibric 390 0.0 58.8 25.3 15.9 longfibres,clearwater113 MILE 3-14 390-420 C:Hemic 420 0.0 36.0 36.0 28.0 l.fibres114 MILE 3-15 420-450 C:Hemic 450 0.0 42.0 26.8 31.2115 MILE 3-16 450480 C:Hemic 480 0.0 35.7 35.1 29.2 Camp seed, woodfrags116 MILE 3-17 480-510 — C:Hemic 510 0.0 44.5 32.2 23.3 lg.wood.bark(included),noss?117 MILE 3-18 510-540 — C:Hemic 540 0.0 39.6 35.0 25.4 grassy, long fibres, less wood118 MILE 3-19 540-570 — F:Hemic 570 0.0 28.9 40.0 31.1 Seed e,fibres119 MILE 3-20 570-600 C:Hemic 600 0.0 35.3 35.6 29.1 twigs,’tubes’,Fruitb120 MILE 3-21 600-630 M:Hemic 630 0.0 30.6 40.2 29.2121 MILE 3-22 630-660 660 0.0 0.0 0.0 0.0122 MILE 3-23 660-690— C:Hemic 690 0.0 39.0 31.0 30.0 lgwooth2123 MILE 3-24 690-720—M:Hemic 720 0.0 21.6 48.0 30.4 V00’124 MILE 3-25 720-750 C:Hemic 750 0.0 37.8 37.8 24.4 Woody125 MILE 3-26 750-780 M:Hemic 780 0.0 22.9 50.0 27.1 woodfrags218APPENDIX CSAMPLE Depth %M Z %Ash % C % M % F Comments126 MILE 3-27 780-810 M:Hemic 810 1.0 19.6 49.1 30.3ED 3 POLLEN DONE______127 ED 3-1 0-30 Fibric 30 0.0 45.8 22.9 31.3 mogrOOtmass128 ED 3-2 30-60 — Fibric 60 0.0 43.5 23.9 32.6 COflhlSSY129 ED 3-3 60-90 Fibric 90 0.0 50.9 20.8 28.3 Y,10O130 ED 3-4 90-120 C Hemic 120 0.0 35.9 17.9 46.2 ‘Y131 ED 3-5 120-150 Fibnc 150 0.0 46.9 23.2 29.9132 ED 3-6 150-180 — Fibric 180 0.0 43.4 24.6 32.0133 ED 3-7 180-2 10 — Fibric 210 0.0 53.1 23.3 23.6 WOOd134 ED 3-8 2 10-240 Fibric 240 0.0 57.1 29.4 13.5135 ED 3-9 240-270 Fibric 270 0.0 57.1 23.7 19.2136 ED 3-10 270-300 Fibric 300 0.0 60.3 27.1 12.6 C POOSP ma seed137 ED 3-11 300-330 Fibric 330 0.0 60.6 24.8 14.6138 ED 3-12 330-360 Fibric 360 0.0 60.5 29.3 10.2139 ED 3-13 360-390 Fibric 390 0.0 56.4 27.5 l6.l°°140 ED 3-14 390-420 F Hemic 420 0.0 20.7 33.7 45.6 BIackI-120, woody granularpeat141 ED 3-15 420-450 F Hemic 450 0.0 23.4 28.0 48.6 BlackH2O, woody granularpeat142 ED 3-16 450-480 — Fibric 480 0.0 44.9 29.3 25.8 Fibrous, grassy, clearH2O143 ED 3-17 480-510 Fibric 510 0.0 48.6 28.4 23.0 Fibrous, grassy, clearH2O144 ED 3-18 510-540 — F Hemic 540 0.0 22.8 34.6 42.6 BlaekH2O,woodygranularpeat145 ED 3-19 540-570 Hemic 570 0.0 34.8 37.6 27.6 Fibrous, grassy, clearH2O146 ED 3-20 570-600 Hemic 600 0.0 31.9 40.3 27.8 Fibrous, grassy, clearH2O147 ED 3-21 600-630 Hemic 630 0.0 34.5 38.1 27.4 H20 grey148 ED 3-22 630-660 F Hemic 660 0.0 18.7 32.3 49.0 Black(burnt?)pithyfrags,blackH2O149 ED 3-21 660-690 Heniic 690 0.0 34.0 40.0 26.0 Black(burnt?)pithyfrags,blackH2O150 ED 3-22 690-720 Fibric 720 0.0 52.8 26.2 21.0 WoOd frags151 ED 3-23 720-750 Fllemic 750 0.0 21.7 49.2 29.1 woodfrags152 ED 3-24 750-780 FHemic 780 0.0 5.4 58.0 36.6 H2Oslighilymurky153 ED 3-25 780-8 10 F Hemic 810 0.0 10.8 36.5 52.7 tf5.C ofmineralBDD 23 POLLEN DONE Base=frosted sand, mixed lithol., mica, mins embedded in woodfrags: Charcoal at top and base154 BDD 23-1 0-50 92.2 F Hemic 50 0.0 18.7 26.5 54.8 root5,o51,bk.Ck155 BDD 23-2 100 91.7 Hemic 100 18.2 24.6 24.9 32.3 flUfls,iHTAM&F156 BDD 23-3 125 91.9 FHemic 125 0.0 15.8 39.6 44.6 H2Oslmurky157 BDD 23-4 150 949 FHemic 150 0.0 16.0 42.4 41.6 Fusiformisporites,seed”i”158 BDD 23-5 175 93.8 FHemic 175 0.0 17.8 39.3 42.9159 BDD 23-6 200 94.2 FHemic 200 0.0 21.7 46.2 32.1 Fsporites160 BDD 23-7 225 94.2 FHemic 225 0.0 26.1 42.3 31.6 lVOOfra5,”161 BDD 23-8 250 95.0 F Heniic 250 0.0 21.2 46.5 32.3 same162 BDD 23-9 250-285 92.7 F Hemic 285 0.0 15.9 34.1 50.0 palmfrags?,seed,H2Oblack163 BDD 23-10 300-350 81.0 F Hemic 350 0.0 21.7 32.7 45.6 wood/barkfrags,seeds,rootlets164 BDD 23-11 350-400 84.6 FHemic 400 0.0 16.0 46.0 38.0 Campseed,annuli;H2Oblack:LTA165 BDD 23-13 415-450 760 notpeat 450 76.6 2.5 8.3 12.6 mixedmins,frosted,charcoalH,LTA166 BDD 23-14 450-500 72.2 notpeat 500 woodfrags, fibres167 BDD23-15 550-600 699 notpeat 600BDD 31 Sedimentary peat like BDD23 base; mins frosted_grey&white,_mica, silty. Resin balls and fuic granules.168 BDD 3 1-1 0-25 82.4 F Hemic 25 0.0 20.9 29.9 49.2 IWOOd,leaflitter,blackH2O169 BDD3L-2 25-50 84.1 Fllemic 50 0.0 16.9 27.7 434I”f’,20bl5219APPENDIX CSAMPLE Depth O/ Z %Ash % C % M % F Comments170 BDD 31-3 50-75 781 FHemic 75 5.0 14.8 27.9 573 SpiCules, silty171 BDD 31-4 75-100 75.4 FHemic 100 56.9 11.7 13.5 18.0 spicules,silty,mica,nocarb172 BDD 31-5 100-125 81.8 F Hemic 125 5.0 3.4 23.7 68.6 silty, resins?, fine fibres173 BDD 31-6 125-150 83.5 FHemic 150 37.8 10.3 23.3 28.6 j,sheh1s”P’174 BDD 31-7 150-175 80.7 F Hemic 175 2.0 12.9 30.0 55.1 asabove,H2Oblack175 BDD 3 1-8 175-200 821 F Hemic 200 1.0 12.0 37.2 50.0 seed,fmef,bres,mica176 BDD31-9 200-225 84.4 FHemic 225 0.0 9.2 31.8 59.0177 BDD 31-10 225-250 83.0 F Hemic 250 0.0 8.4 35.5 55.8 Resin balls,H20 black,bark178 BDD 31-11 250-275 84.4 FHemic 275 2.0 10.3 38.5 49.2 worootlfrostedinins179 BDD 31-12 275-300 85.3 FHemic 300 0.0 6.7 45.0 48.3180 BDD 31-13 325-350 784 FHemic 350 0.0 7.8 35.7 56.5181 BDD 3 1-14 375-400 79.8 F Hemic 400 0.0 2.5 45.4 52.1 nul&t182 BDD 3 1-15 400-440 69.2 notpeat 440 940 2.0 2.0 2.0 not peatnot carbonateLAKE 2183 LAKE 2-1 0-20 84.2 Hemic 20 0.0 17.4 41.1 41.5 Fusiforanisporites184 LAKE 2-2 25-50 92.1 FHemic 50 0.0 2.8 27.4 69.8185 LAKE 2-3 50-75 786 F Hemic 75 0.0 12.0 33.8 54.1186 LAKE 2-4 100-125 90.6 F Hemic 125 0.0 5.1 43.4 51.4 Seed187 LAKE 2-5 125-150 93.4 Hemic 150 0.0 22.4 39.3 38.3188 LAKE 2-6 175-200 95.2 F Hemic 200 0.0 7.4 47.4 45.2189 LAKE 2-7 225-250 83.7 FHemic 250 0.0 11.8 45.4 42.8190 LAKE 2-8 275-300 94.1 F Hemic 300 0.0 9.4 43.9 46.7 SCCdS191 LAKE 2-9 300-325 87.7 FHemic 325 0.0 9.3 34.6 56.1 traceofsandgrains192 LAKE 2-10 325-350 754 notpeat 350 98.0 LTAMILES193 MILE 5-1 0-30 924 C: Fibric 30 0.0 56.5 23.4 20.1 Y194 MILE 5-2 30-60 93.1 C: CHenaic 60 0.0 51.9 30.3 17.8 SSY195 MILE 5-3 60-90 93.1 F: Hemic 90 0.0 21.2 21.2 57.6 bark?, fern annuli196 MILE 5-4 90-120 94.8 C: C Hemic 120 0.0 47.4 21.1 31.5 gy197 MILE 5-5 120-150 93.0 C: Fibric 150 0.0 56.5 17.3 26.2 ‘Y198 MILE 5-6 150-180 95.5 C: CHemic 180 0.0 42.5 23.6 33.9 7cmwood,twisedgeleaf199 MILE 5-7 180-2 10 94.5 C: C Hemic 210 0.0 40.4 23.2 36.4 yelloW Wood lxO.5 cm200 MILE 5-8 210-240 950 F: Hemic 240 0.0 25.5 28.9 45.6201 MILE 5-9 240-270 95.6 C: C Hemic 270 0.0 44.2 24.3 31.5202 MILE 5-10 270-300 95.6 * 300 0.0203 MILE 5-11 300-330 95.3 C: Fibnc 330 0.0 58.2 23.7 18.1 2typesofwood204 MILE 5-12 330-360 96.0 F: Hemic 360 0.0 34.4 27.4 38.2 blakfrags, wood, fiujitb205 MiLE 5-13 360-390 94.0 F:FHemic 390 0.0 24.8 28.1 47.1 fruitb206 MiLE 5-14 390420 93.1 F: FHemic 420 0.0 18.6 29.9 51.5 lgwoodandpalinfrags207 MILE 5-15 C 420450 945 F: Hemic 450 0.0 33.7 29.5 36.8 seed, large wood208 MILE 5-16 450480 97.1 C: C Hemic 480 0.0 45.0 27.7 27.3209 MILE 5-17 480-5 10 95.3 F: Hemic 510 0.0 26.2 30.0 43.8 large wood, bark210 MILE 5-18 5 10-540 94.8 F: Hemic 540 0.0 22.8 27.4 49.8211 MILE 5-19 540-570 96.2 F Hemic 570 0.0 11.4 30.3 58.3 Seed212 MILE 5-20 C 570-600 949 C: Wood 600 0.0 66.2 16.2 17.6 gugpo213 MILE 5-21 600-630 93.8 C: Fibric 630 0.0 48.3 24.4 27.3 wood, long roots214 MILE 5-22 630-660 96.0 C Hemic 660 0.0 30.6 28.9 40.5 mostly wood, black (palm) frags220APPENDIX CSAMPLE Depth %M Z %Ash % C % M % F Comments215 MILE 5-23 660-690 90.3 F: Hemic 690 0.0 18.5 31.0 50.5 mostly Wood,Cainp seed, blaekH2O216 MILE 5-24 690-720 92.7 C: Wood 720 0.0 76.1 14.2 97 ALL WOOD!217 MILE 5-25 720-750 93.7 C: Wood 750 0.0 45.5 25.7 28.8 mostly wood218 MILE 5-26 C 750-780 95.7 F Hemic 780 0.0 11.4 21.9 66.7 WOOd,tWig219 MILE 5-27 780-810 93.6 F Hemic 810 0.0 11.9 31.2 56.9 red wood, Ig root, black H2O220 MILE 5-28 810-840 93.3 FHemic 840 0.0 4.0 26.2 69.8 Black H20221 MILE 5-29 840-870 93.3 not peat 870 50.0 3.4 31.1 65.5 1Y LTAonwlioleandfinesBI 3222 BI 3-1 0-25 84.0 F Hemic 25 0.0 7.9 22.6 69.5 black H20, fine223 BI 3-2 25-50 85.9 F Hemic 50 7.9 4.9 26.0 61.1 J4TA224 BI 3-3 50-75 81.8 F Hemic 75 0.0 5.4 36.9 57.7 blackwater,blackfrags225 BI 3-4 75-100 83.6 F Hemic 100 7.6 3.9 34.5 54.0 seed: HTA226 BI 3-5 100-125 81.8 F Hemic 125 0.0 14.7 34.5 50.8227 B13-6 125-150 83.1 FHemic 150 0.0 10.3 23.6 66.1228 BI 3-7 150-175 79.2 FHemic 175 2.8 2.2 15.3 79.7 muchshell,sandandsilt229 BI 3-8 170-195 86.6 F Hemic 195 0.0 6.3 27.4 66.3 black water230 BI 3-10 175-200 77.9 notpeat 200 7.9 0.7 22.9 68.5221Jugbndaceae oooooeooooooeoee00000000eCompostae 00OOOOOO00OC0O0OOO0O000 000000000000ci. Mlconlasp. 0 00000000 0 00 00 0 00 00 000000000000IlexgulanensIs ‘00UCyrilla iacemifl.ora U) N 0 — N — 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0MyricatypeB 000000000000000000000 00000 000000000000MyricatypeA 00000000r)000000000000000 0000CO00U) .. U) U) C U) U) I- C lb l. ID U) U) U) U) lb U)Mydca mexicana— e 0 N — — C) 0 lb 0 0 0 N - c 0 0 0 0 — ,..: —tncolporate F (33) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 — 0 0 0 0 0 0 0 0tncolporate E (48) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N .— 0 0 0 0 0 0 0 0 0 0 0 0 0tricolporate D (40) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 — 0 0 0tricolporate C (39) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 — 0 0 0 0 0 0 0 0 — — N N (0 c. ci — COtricolporate 8 (54) ( — 0 — 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0tricolporateA(26) 0 00000000 000000000000!Campnosperma type 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 lb U) 0 0 0 — 0 — 0 0 0 0,Campnospermap. 0000000N0 N’ U)000!. tricolpateType2 00000000000000000000000000 0000000000co.tricolpateTypel 00000000000000000000000000 000000000Lagunculadar. 00000000000000000000000000 00000000N0O. Rhizophoiram. 00000000000000000000000000 00000000)0N8<e__.!I. ipNCO’U)(0-U)O)(.4cf Verrucatospotites oe——Salplchlaenasp. 000000’Cyclopeitiss.Sw. 00OtN000!Monoletefem8 000OOO0O0000ON)0000D.N 0000O0N000’Monolete fern 3 0 0 0 0 0 0 0 0 0 0 0 0 0 — 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 CMonoleteferri2 0000000000000e00000e’t 00000000000CMonoIetefem1 0000000000000 O0O0O0Lfl0cf.Laevlgatospontes (D’tlOAcrostichum a. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 —cf.Mefaxyasp. O0000000000000r..ee!OO000 000000000000Unknowntnletespores* U0N_UPaimphytohthA 0 000000000000000000000000 ON0t000OEuterpeprecatorta 0 0 0 0 0 0 0 0 0 0 0 0 N — 0 0 0 0 0 — N 0 0 0 0 0Raphia taedigera 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 N D 0 N N —PA pe 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 — 0 0 0 0 0 0 0 0 0 0 0GraminiaeCyperacea.,C0z223TOTALCOUNT 8888888C’4C4C4 C4C1 .C1’ (‘4%OtherDicotsU) (DO) (‘4 U) (0 U) U) U) U) U) at) U) U)% None-arboreal an d Z a’ ‘i ‘ (‘I C’) C) — Co C) Co (0 O (‘iit) (0 0) an o U) CO U) U) U) an an C’ C’) 4) an an an an an U)Au Others an o o an an a.. r- a — o ci ! r d Co an an r- 2 .an — an an (‘1 an ag an an an a-.. a-.-. an an an in C) an in-‘ %identifiedDm0-cysts 00000000000000000000000000 oeooooo’rZeOFuslfomiisporifesC 00O00000000000000000000Fusifonni spodtes B 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0FusifonnisporitesA 0000000000000000000000 0CoCo0SporangiumA 0000000000000000000000000 000000000000a’) a’) U) a’) U) a’) a’)opaquetetrahedron 0 0 0 0 0 0 0 0 0 0—0 0 0 0 0 (‘4 0 C) 0 (‘4 C’) 0 0 0FungalsporesC..’c1APPENDIX E: POLLEN DESCRIPTIONS(Plates 1 to 3)PLATE 1-1Trichomanes crinhtum (Hymenophyllaceae); Monolete fern spore, irregular ovoid, 41-51 jim x 3 0-40 jim; scierine plicate, echinate.Ref. PC 5, fresh material.PLATE 1-2Cyclopeltis semicordata (Polypodiaceae - Tectariaceae): Monolete fern spore, kidney-shaped, 41gm X 25gm; laesura 22gm long, thickened margo, 1.5gm thick. Scierineirregularly verrucate.Ref. PC 4, fresh material, LAKE 2 type 4. GL 30.6,121.1; 31.3,126.4. Photos 19,20.PLATE 1-3Salpichlaena sp. (Blechnaceae): Monolete fern spore, kidney-shaped, 33-43gm x 29gm,laesura Ca. 22gm long, exine thin, finely rugulate; sclerine frequently has one to severalplicae paralleling or crossing the laesura.Ref. fresh material. GL 29.2,121.4. Photos 23,30.PLATE 1-4Cyathea(?) rnultflora(?) (Cyatheaceae): Trilete fern spore, amb triangular, 40-45gm x3 2gm, laesurae Ca. 10gm long. Exine stratified, scierine thin, psilate.Ref. fresh material (immature sample?), GL - Photos 2 1,22.PLATE 1-5Lindsaea sp. (Dennstaetiaceae): Monolete fern spore, 36gm x 29gm. oblate, often pointedat ends, with laesura extending into points. Laesura irregular, extends length of spore;ends of laesura sometimes branching; margo entire, >1gm thick. Exine scabrate.Ref pc 7, fresh material. GL 30.1,121.2. Photos 3 1,32.PLATE 1-6Monolete fern type 1: Kidney-shaped spore, 57gm x 27gm, laesura 2/3 length of spore,distinct margo 2gm wide, 3 jim thick; scierine psilate, perine psilate with 1 to severalprominent folds (plicae), often merging or intersecting.Ref pc 4, LAKE 2 type 1. GL 34.3,110.7 Photo 24.PLATE 1-7Monolete fern type 2: Kidney-shaped spore, 63gm x 4 1gm, laesura 35gm long, scierinepsilate.Ref pc4, LAKE2type2.GL31.9,113.3. Photo 27PLATE 1-8225cf. Verrucatosporites sp., monolete fern type 3: Kidney-shaped spore, 41.im x 25iim,laesura 25iim long, sclerine weakly verrucate, perine psilate.Ref. pc4, LAKE2type3. GL32.8,113.5. Photo 26.PLATE 1-9Monolete fern type 8: Kidney-shaped spore, 35i.tm x 25iim, laesura 19im longRef. pc 1, BDD 25 type A. GL 28.9,127.3. Photo 2.PLATE 1-10cf. Metaxya sp. (Cyatheaceae) (Roubik & Moreno), trilete fern spore: amb rounded-triangular, eq. dia. 44-50jim, laesurae 18-22j.im long, 2-3j.tm wide, scierine thin,unstratified, scabrate to clavate.Ref. PC 4, LAKE 2 type 29. GL 29.6,112.8. Photo 25.PLATE 1-11Raphia taedigera (Palmae): Monocolpate, shape irregular-reniform to sub-spherical, 23-27i.tm x 17-19i.tm, colpus straight, 15-l9iim long x Ca. 1.8j.tm wide. Exine reticulateRef. Pc 3, fresh sample. Photos 33,34.PLATE 1-12Palm phytolith, probably Raphia taedigera. spikey sphere, 10-30 imin diameter.PLATE 1-13Euterpe precatoria (Palmae): (Euterpe precatoria is the only palm, other than Raphia t.,present in the swamp, but attempts to secure fresh pollen of this species failed): prolate,25.tm x15.tm, monolete; laesura straight, 22j.tm long x 1.5-2.3i.tm wide, with distinctmargo; exine thin, scabrate.Ref. PC 5, BDD 23 type 32.GL 24.8,119.1; 24.2,119.2. Photos 9,12.22620jLPLATE 1: Fern spores: 1) Trichomanes criniturn; 2) C’yclopeltis seniicordata Sw.; 3)Salpichiaena sp.; 4) Cyathea n;ultf1ora(?); 5) Lindsaea sp.; 6) Monolete type 1;7) Monolete type 2; 8) Monolete type 3; 9) Monolete type 8; 10) cf. Meiaxyasp.; Palmae: 11) Raphia taedigera; 12) Raphia phytolith; 13) Euteipeprecatoria.•12[3546879111012 13227PLATES 2-1; 3-1 and 3-2 (details)Campnospermapanamensis (Anacardiaceae): tricolporate, prolate, 27-30 gm x 17-21gm.Exine ca 1.8gm thick, nexine thicker than sexine; sexine striate, 0.7gm thick, striae 0.5gmwide, absent near pores. Colpi ca. 22gm long, evenly spaced, costae colpi small.Ref. pc 5, fresh material; BDD 23-23, 23-36. GL 43.5,1 19.6;43.5,120.3 Photos 35-38PLATE 2-3Tricolporate type 1: Prolate, 28gm x 15gm, prominent transverse colpiRef pc 1, BDD 25-1. GL27.5,112.8;24.3,117.8 Photos 1,3.PLATE 2-4Tricolporate type 2: Prolate, 24.tm x 16gm, well-developed transverse colpi.Ref. pc 1, BDD 25-36. GL 30.2,111.2; 36.4,118.3 Photo 4.PLATE 2-5Tricolporate type 26: Oblate-spheroidal, 20gm x 22gm, exine 2gm thick, stratified, sexinepsilate, pores (ora?) open, with well developed vestibulum 4gm wide x 4-6gm deep.Ref. pc 7, LAKE 10-type 26.GL 37.7,124.5 Photo 13.PLATE 2-6Tricolporate type 54: Oblate, 23gm x 16gm, exine 1.5gm thick, stratified, sexinerugulate, separated from nexine at pores, pores (ora?) open, 6gm wide, with vestibulum,and membrane across oral floor; colpi about 3 gm deep, with margos about 1.5 gm wide.Ref. pc 7, LAKE 10-type 54. GL 38.7,119.5 Photo 39.PLATE 2-7Myrica mexicana: triporate, sub-oblate, amb rounded-triangular, 26gm, exine scabrate,ca 1.2gm thick; annulus arched, with fine teeth on lower surface.Ref pc 3, BDD 8-28 GL 39.4,111.3. Photo 16.PLATE 2-8Myrica-type A: triporate, amb rounded-triangular with a distinct 3-armed fold around thepole; dia. 27gm, exine finely scabrate, ca 1.2gm thick; pores open, 2.2gm wide,vestibulum rounded, 6gm x 3gm deep; annulus arched, with prominent inward notch(lower left) and teeth-like projections on inner surface (arrow).Ref pc 3, BDD 8-41 GL 3 1.4,118. Photo 19.PLATE 2-9Tricolporate type 33: 3-brevicolporate, amb circular, 23-27gm dia, exine 1.75gm thick,baculate, semi- tectate; endexinous thickenings beneath colporoids show as dark areas(arrows). (cf. Mortoniodendron?) (Graham, 1979).Ref pc 5, BDD 23 type 33 GL23.8,118.9, Photo 10.228PLATES 2-10, 3-3 (details)Unknown type 39: cf. Chenopodipollis sp., amb circular-hexagonal, 21-24j.tm, periporate;pores irregular ovoids without margo, Ca. 4x2gm; exine stratified, >2im thick, sexinebacculate-echinate.Ref. PC 5, BDD 8-39, 8-40. GL 26.5,112.4 Photos 17,18.PLATE 2-11Tricolporate type 40 (Tiliaceae?): amb circular, 3-brevicolpate, dia. 32-36iim, colpiequatorial; sexine clavate, semi-tectate; exine has 1 to several plicae Ca. 4im wide and 22-25.tm long.Ref. pc 4, LAKE 2-40 GL 35.6,115.8. Photo 28.PLATE 2-12Tricolporate type 48 (Tiliaceae?): amb rounded, 3-brevicolpate, dia 27-31 jim; costa colpipresent, weak margo, atrium present but reduced, with vestibulum; exine reticulate, Ca.2.5j.tm thick, reticulum ca 1.5j.tm thick.Ref. LAKE 2-48 GL 35.8,116.2; 27. 1,121.4. Photo 29.22911PLATE 2: 1) Carnpnosperrna panarnensis; 2) Carnpnosperrna-type; 3) Tricoiporatetype 1; 4) Tricolporate type 2; 5) Tricolporate type 26; 6) Tricolporate type 54; 7)Myrica mexicana; 8)Myrica type A; 9) Triporate type 33; 10) Periporate type 39; 11)Triporate type 40; 12) Tnporate type 48.48121 2 35 6 72OI910230PLATE 3-1Campnospermapanamensis low and high focus equatorial (left, top and bottom), andpolar (right, top and bottom) views.PLATE 3-2Campnospermapanamensis exine detail.PLATE 3-3Unknown type 39: cf. Chenopodipollis sp. details of pore (top) and exine (bottom).PLATE 3-4Fusformisporites sp. A: Fusiform dicellate fungal spore,73- 80j.tm x40- 50j.tm, ovoid,striate; striations >3.tm wide running length of hemisphere, branching in a few cases;central divider a thickened ring rather than a solid wall.Ref. pc 4, LAKE 2-41 GL42.5,117.0 Photo 14.PLATEFusfonnisporites sp. B: Fusiform dicellate fungal spore, 15tm x lOj.tm, blunt-ovoid,striate; striations Ca. him wide, Ca. 10-12 per hemisphere, unbranched, running length ofhemisphere.Ref. pc 4, LAKE 2-57 GL 30.1,125.6 Photo 15.PLATE 3-5Fusfonnisporites sp.C: Fusiform dicellate fungal spore, 3611m x 9jim, striate; striationsCa. 1 urn wide, 14-16 per hemisphere, running length of hemisphere, occ. branched. Poleswith internal thickenings. Central divider is apparently a solid wall.Ref. pc 5, BDD 23 GL 20.4,118.5. Photos 7,8.231:-1LPLATE 3 (10 im scale bar): 1) Carnpnosperniapananiensis low and high focus in bothequatorial and polar views; 2) Detail of exine decoration of Carnpnosperma (x1780); 3) Periporate type 39, low and high focus. (20 im scale bar): 4)Fusforrnisporites sp. A; 5) Fusiformisporites sp. C.232I—A4APPENDIX F: PLANTS IDENTIFIED iN THE CHANGUINOLA MIREAPPENDIX F: List of species. Dom = dominant: C = common: R = rare.FAMILY SPECIES TYPE/COMMONNAME/NOTESAlismataceae Sagittaria 1ancfolia C herb! Otoye de lagartoAnacardiaceae Campnosperma panamensis Dom tree!Orey, OnAnnonaceae Guatteria inuncta C shrub!Apocynacae Thevetia ahouai C shrub! Huevo de gatoAquifoliaciae Rex guianensis C tree! PlomoAraceae Dieffenbachia longispata C herb! Otoye de lagarto(spiney)Blechnaceae Salpichlaena sp. fern! ostrich 1Bignoniaceae Tabebuiea rosea R tree/ Roble de sabanaBurseraceae Protium panamense Dom tree!Chrysobalanaceae Chrysobalanus icaco C tree! “Round-leaf’Combretaceae Laguncularia racemosa C tree! White mangroveCyatheaceae Cyathea? multzjlora? fern!Cyperaceae Cyperus ligularis sawgrass/Cyperaceae Cyperus odoratus sawgrass!Cyperaceae Cyperus sp. sawgrass/Cyperaceae Rhynchospora macrostachya. sawgrass!Cynillaceae Cyrilla racemzflora tree! mangle cimarron “a”(red bark)Dennstaetiaceae Lindsaea sp. fern! tree hammockEuphorbiaceae Aichornea larifolia C shrub!Euphorbiaceae Alchornea sp. R tree!Guttiferae Calophyllurn belizense C tree! MariaGuttiferae Clusia sp. R tree! HigoGuttiferae Symphonia globu4fera Dom tree! Cerillo, barilloHymenophyllaceae Trichomanes crinitum fern! base of On (mfilamentous)Lauraceae Ocotea pyraniidata R tree!233Leguminosae Cae Cassia reticulata R shrub!Leguminosae Mim Inga spectabilis R tree! Guaba macheteLeguminosae Mim Pithecollobium gp. C tree! Gavilan; Rain treeLeguminosae Pap ErythrinaJisca R tree! GallitoLiliaceae Smilax sp. C vineMalvaceae Hibiscus pernambucensis R shrub!Melastomataceae Miconia curvipetiolata C shrub!Melastomataceae Tococa guianensis C shrub!Moraceae Cecropia insignins R tree! GuarunioMoraceae Cecropia peltata R tree! Guarumo; trumpettreeMusaceae Heliconia latispatha C herb!Oja de BijaoMyrsinaceae Ardisia sp. C shrub!Myrsinaceae Myrsine pellucido-puntata C tree!Nyctaginaceae Neea sp. 1 R shrub!Nyctaginaceae Neea sp. 2 R shrub!Ochnaceae Cespedezia macrophylla Dom tree!Ochnaceae Ouratea sp. C shrubPalmae Euterpe precatoria C palm!Palmae Raphia taedigera Dom palm! MatombaPolypodiaceae Acrostichum aureum Dom fern! helecho demanglarRhizophoraceae Cassipourea eliplica C tree!Rhizophoraceae Rhizophora mangle Dom tree! red mangroveRubiaceae Cephaelis tomentosa C herb! labios ardientesRubiaceae Palicourea triphylla C shrub!Sterculiaceae Herrania purpurea R shrub! cacao cimarronSterculiaceae Sterculia sp. R shrub!Tectariaceae Cyclopeltis semicordata (s. w.) C fern! ostrich 2Theaceae Ternstroemia tepazapote Dom shrub; R tree!Miguelario; Manglillo234APPENDIX G: TERMS AND ANALYTICAL PROCEDURESThe American Society for Testing and Materials (ASTM, 1969) sets standards for thecomposition of organic-rich sediments: by definition, peat contains less than 25% dry weightinorganic material (ash); clayey peat (or sandy peat) has 25% to 50% ash, and peaty clay (or peatysand) has greater than 50% ash, dry weight. Less than 10% mineral matter is ‘low-ash’ peat, and10% to 25% mineral matter is ‘high ash’. Fresh peat is frequently 90% water by weight (Cameronet a!., 1989). Peat is defined according to ash content using the Organic Sediments ResearchCentre, University of South Carolina, standard (Andrejko et a!., 1983). By this standard, peat isdefined as Low (<5 wt%), Medium (5-15 wt%) and High Ash (15-25 wt%). Above 25 wt% iscarbonaceous sediment.Peat may form in virtually any humid climate, if the groundwater table remains for asufficiently long time at a level such that the organic material near the surface remainswaterlogged, and yet the growing surface is not drowned, or buried by inorganic sediments. Thatpart of the peat which is commonly above the groundwater table, aerated and permeable, is termedthe acroteim, while the denser, less permeable, waterlogged part is the catoteim (Moore, 1989).Peat deposits can be classified according to their source of water, the nature of the drainagepatterns, and the underlying topography on which they accumulate. The term mire is a general oneused in Europe to refer to any non-saline peat-forming wetland system (Moore, 1989). This usagehas now been expanded into the concept of a mire complex in acknowledgment of theheterogeneity and evolutionary potential of wetland systems. In the present study, the marginalmangrove swamp, a saline wetland, would strictly speaking be excluded from the definition ofmire. It is, however, a fundamental element in the evolution of many tropical mire complexes, andthus encourages the expansion of the definition.Rheotrophic (Moore, 1989) or minerotrophic (Cameron et al., 1989) mires are fed byboth rainwater and groundwater. Ombrotrophic (or ombrogenous -Stach et al., 1982) miresreceive water only as rainfall. In both types of mires, the water table can be either above or belowthe peat surface. Rheotrophic mires tend to be nutrient-rich (eutrophic) due to the influx ofgroundwater. They may have complex internal drainage systems, a diverse and luxuriant flora,and significant inorganic sedimentation. Ombrotrophic mires tend to be nutrient-poor(oligotrophic), the flora relatively poorer in species and stunted in its growth, and the resultingpeat low in mineral matter (ash).The terms ‘bog’, ‘moor’, ‘marsh’ and ‘swamp’ are all applied to peat-forming wetlands, withvarying degrees of consistency. To most authors, raised bogs or ‘high moors’ are ombrotrophic235mires, with perched water tables fed by rainfall, which form in closed depressions that result froma variety of geomorphological processes. They are commonly lens-shaped, with domed or convexsurfaces (hence ‘raised’), and the relative impermeability of the catotelm results in a domed watertable, and drainage away from the centre, which tends to be particularly nutrient-poor. Near themargins, where the peat is thinner, the substrate may provide plant nutrients, and the vegetationbecome progressively more luxuriant. Vegetation ranges from moss (Sphagnum) and stunted treesin higher latitudes, to dense forests in the tropics. Raised bogs may contain areas of swamp ormarsh (Cameron et al., 1989).Types of mires (from Moore, 1989):‘Bog’ is land that is wet, soft and spongy, and underlain by peat. Bogs may occur withinmarshes and swamps (Cameron et al., 1989).‘Marsh’ is relatively open, saturated or shallow-water wetland dominated by herbaceousvegetation - sedges, rushes, grasses. It is normally rheotrophic.‘Swamp’ is shallow-water wetland dominated by trees. It too is normally rheotrophic.Stach et al. (1982) tend to use the terms ‘reed-swamp’, ‘forest-swamp’ and ‘moss-swamp’for some of the above tenns, and many authors use generic identifiers (Mangrove swamp) on theassumption that the context will clarify the hydrologic nature of the site. In the case of rheotrophicpeats, they may be freshwater (telmatic or limnic), or experience brackish or marine influences,with profound effects on the biology and chemistry of the deposit.Tissue preservation (degree of humification) is described in field observations, and in thewritten descriptions in this report using a modified von Post Humification Scale adapted to tropicalpeats by Esterle (1990). There are 5 field categories, described by Esterle (1990) as follows:Fibric: Coarse and fine fibrous particles with rare occurrence of large root or wood particles.Coarse Hemic: Abundant coarse fibrous particles with particulate matrix and occasional. largeroot or wood fragments.Hemic: Dominant fine particulate matrix with minor amounts of coarse fibrous particles andcommon large root and wood fragments.Fine Hemic: Dominant fine particulate matrix with minor amounts of coarse fibrous particles andfewer large root and wood fragments than hemic peat.Sapric: Primarily fine particulate matrix material with occasional large roots or wood fragmentsand few coarse fibrous particles.In this study, reference is made to the field categories and the vonPost system, alongside particle-size distributions determined by wet sieving, because peat workers are familiar with this index of236tissue preservation (von Post, 1922). A scale often is used to grade the peat by degree ofdecomposition in the von Post system:Hi: Completely unhumified and muck-free peat; upon pressing in the hand, gives off onlycolourless, clear water.H2: Almost completely unhumifled and muck-free peat; upon pressing, gives off almost clear butyellow-brown water.H3: Little humifled and little muck-containing peat; upon pressing, gives off distinctly turbidwater, no peat substances pass between the fingers and the residue is not mushy.H4: Poorly humifled or some muck-containing peat; upon pressing, gives off strongly turbidwater. The residue is somewhat mushy.H5: Peat partially huniifled or with considerable muck content. The plant remains arerecognizable but not distinct. Upon pressing, some of the substance passes between thefingers together with mucky water. The residue in the hand is strongly mushy.H6: Peat partially humifled or with considerable muck content. The plant remains are not distinct.Upon pressing, at the most, one third of the peat passes between the fingers together withmucky water. The residue in the hand is strongly mushy, but the plant residue stands outmore distinctly than in the unpressed peat.H7: Peat quite well humifled or with considerable muck content, in which much of the plantremains can still be seen. Upon pressing about half of the peat passes between thefingers. if water separates it is soupy and very dark in colour.H8: Peat well humifled or with considerable muck content. The plant remains are notrecognizable.. Upon pressing about two thirds of the peat passes between the fingers. Ifit gives off water at all, it is soupy and very dark in colour. The remains consist mainlyof more resistant root fibres, etc.H9: Peat very well humified or muck-like, in which hardly any plant remains are apparent. Uponpressing, nearly all of the peat passes between the fingers like a homogenous mush.HiO: Peat completely humified or muck-like, in which no plant remains are apparent. Uponpressing all of the peat passes between the fingers.The traditional use of field-determined peat types has been used only sparingly in the study of thesetropical peats, as recent work (Esterle, 1990) suggests low correspondence between the traditionalfield classifications and the actual particle-size distribution as determined by point counting orsieving methods.237Degree of humification of the peat was established by particle-size distribution of eachsample. Degree of humification of the peats is based on the relative proportions of coarse, mediumand fine constituents as determined using a wet-sieving procedure modified from Staneck and Sue(1977), according to the following scheme (Esterle eta!, 1987):Coarse >25% >2.0 mm <30% <0.25 mm (= fibric to coarse hemic)Medium <25% >2.0 mm <30% <0.25 mm (= hemic)Fine <25% >2.0 mm >30% <0.25 mm (= hemic to fine hemic)The results of sieving are recorded as percentages of total by dry weight, as the methods ofmeasuring volume are less accurate, and do not lend themselves well to woody or fibric peats.Mineral matter was separated-out by flotation, and the results of sieving with 2.0 mm and 0.25 mmsieves were dried in a 50°C oven to constant weight.Peat Classification is based on the identification of macroscopic plant parts andpalynomorphs in the peat, compared to plant and pollen associations identified in the surfacesamples, and uses botanical (e.g. Rhizophora peat, sedge peat) nomenclature.Mineral matter content (wt % ash) of 137 samples was determined by weight loss onignition in a muffle furnace at 550°C (ASTM-D 2974: Jarret, 1983).Moisture content of wet peat, drained of superficial water, was measured by air drying at50°C (wt % moisture lost), and is used in plots as an approximation of the density of the peat.Total sulphur content (dry weight percent) of 203 samples (dried at 5 0°C, crushed to 100mesh) was determined using a Leco® SC-132 Sulphur Analyzer (see Tabatabai, 1992, p.3 13 for adescription of this instrument). Procedures for the determination of sulphur forms have beendiscussed by Lowe (1986, 1992) and Lowe and Bustin (1985) and are summarized here. Sulphatesulphur was determined by extraction with a 0.1 M CaCI2 solution followed by sulphatedetermination by HI reduction and Bi-colorimetric determination ofH2S (Kowalenko and Lowe,1972). Pyritic sulphur was determined by Zn-HC1 reduction of the whole peat followed by H2Sdetermination, using the methods outlined by Williams and Steinberg (1959). Elemental sulphurwas determined by soxhlet extraction with chloroform for three hours followed by Zn-HC1reduction and colorimetric determination ofH2S. The total organic sulphur was determined bydifference between total sulphur and the sum of elemental sulphur, sulphate sulphur and pyriticsulphur.238For determination of carbon-bonded sulphur and organic sulphate sulphur (C-O-S formof sulphur) the entire peat was reduced by HI and H2S was determined by colorimetry. HIreduction reduces elemental sulphur, sulphate sulphur, organic sulphate sulphur and pyriticsulphur to H2S (Freney, 1961). Organic sulphate sulphur was estimated by subtracting elementalsulphur, sulphate sulphur and pyritic sulphur from the HI reducible sulphur. Carbon-bondedsulphur was estimated by the difference between total sulphur and HI reduced sulphur. If any acidsoluble suiphides are present, they are removed by Zn-HCI reduction and thus lumped with pyriticsulphur.Pollen slides were prepared from surface litter and shallow peat samples, and from 2 cores,one in the central part of the deposit and the second near the eastern margin. Pollen was preparedand concentrated using standard acetolysis techniques from the fine fraction of the peat (< 0.25mm). The use of HF was avoided to enable phytoliths to be counted. Samples were first separatedinto coarse, medium and fine fractions using a wet-sieving procedure modified from Staneck andSilc (1977). Pollen profiles were prepared for each phasic community. To establish the pollen‘fingerprint’ for each phasic community, a count of 1000 grains, where possible, was made fromsurface litter and surface peat samples (3 sites produced less than 750 grains each from multipleslides). The results were combined, and the percentage of the total count plotted for the mostnumerous palynomorphs. Although it would be preferable to compare counts from numeroussimilar sites to better represent each phasic community, it was felt that including the upper 25 cmof the peat gave considerable range to the sample, and given the sampling interval in the cores (25or 30 cm) finer discrimination was not necessary. No attempt is made to relate the pollen found inthe peat to the actual species which dominate each community - the problems associated withproportional representativeness of dominant plants in pollen profiles are well known. Rather it washoped that an easily recognizable ‘fingerprint’ would emerge which could then be related to pollenrecovered from cores.Counts of 200 grains for each 25-30 cm increment of two cores were used to construct thepollen stratigraphy of the central and eastern sections of the deposit (Appendix D). The pollendiagrams of the two cores were constructed using selected palynomorphs or combinations ofpalynomorphs which are considered to be valid markers, as determined from the surface samples,and which best define transitions in vegetation. In both surface and core samples, the followinggrains were omitted from the counts: all single-celled fungal spores, all fungal hyphae, all239fragments less than 2/3 complete, and any ambiguous grain. In identifying palynomorphs,botanical names are used wherever possible, and pollen and spore genera where necessary.240REFERENCESAmerican Society for Testing and Materials (ASTM), 1987. D4427-84. Standard classificationof peat samples by laboratory testing, in Annual Book of ASTM Standards, V.4.08.ASTM, Philadelphia.American Society for Testing and Materials (ASTM), 1969. D2607-69. Standard classification ofpeats, mosses, humus and related products. ASTM, Philadelphia, p1.Andrejko, M.J., Fiene, F. and Cohen, A.D., 1983. Comparison of ashing techniques fordetermination of the inorganic content of peats. j P.M. Jarret, Ed., Testing of Peats andOrganic Soils, American Society for Testing and Materials ASTM STP 820, 1983, p.5-20.Cameron, C.C., Esterle, J.S. and Palmer, C.A., 1989. The geology, botany and chemistry ofselected peat-forming environments from temperate and tropical latitudes. : P.C. Lyonsand B.Alpern (eds.), Peat and Coal: Origin, Facies and Depositional Environments, Tnt. J.Coal Geol., v.12, P. 105-156.Esterle, J.S., Ferm, J.C., Durig, D.T. and Supardi, 1987. Physical and chemical properties of peatnear Jambi. Sumatra, Indonesia. International Peat Society Symposium on Tropical Peat,1987.Esterle, J.S., 1990. Trends in petrographic and chemical characteristics of tropical domed peatdeposits in Indonesia and Malaysia as analogues for coal formation. Unpublished PhDthesis, University of Kentucky, Lexington, p.270.Freney, J.R., 1961. Some observations on the nature of organic sulphur compounds in soil.Australian Journal of Agriculture Research, v. 12, p. 424-432.Jarret, P.M., 1983. Ed., Testing of Peats and Organic Soils, American Society for Testing andMaterials ASTM STP 820, Philadelphia, Pa., 1983, 241 p.Kowalenko, C.W. and Lowe, L.E., 1972. Observations on the bismuth sulphide colorimetricprocedure for sulphate analysis in soil science and plant analysis. Communications in soilscience and plant analisis, v. 3, p. 79-86.Lowe, L.E., 1986. Application of a sequential extraction procedure to the determination of thedistribution of sulphur forms in selected peat materials. Canadian Journal of Soil Science,v.66,p.337-45.Lowe, L.E., 1992. Studies on the nature of sulphur in peat humic acids from the Fraser Riverdelta, British Columbia. The Science of the Total Environment, 113, p.133-145. Elsevier,1992.Lowe, L.E. and Bustin, R.M., 1985. Distribution of sulphur forms in six facies of peats of theFraser River delta. Canadian Journal of Soil Science, v.65, p.531-541.2$ IMoore, P.D., 1989. The ecology of peat-forming processes: a review. : P.C. Lyons andB.Alpern (eds.), Peat and Coal: Origin, Facies and Depositional Enviromnents. mt. J. CoalGeol., v.12, p. 89-104.Stach, E., Mackowsky, M-Th., Teichmuller, M., Taylor, G.H., Chandra, D. and Teichmuller, R,1982, Coal Petrology, 3rd edition, 535 pp., Gerbruder Borntraeger, Berlin-Stuttgart.Staneck, W. and Silo, T., 1977. Comparisons of four methods for determination of degree ofhwniflcation (decomposition) with emphasis on the von Post method. Canadian Journal ofSoil Science, v. 57,p.109-17.Tabatabai, M.A., 1992. Methods of measurement of sulphur in soils, in Howarth, R.W., Stewart,J.W.B. and Ivanov, M.V., eds. 1992. Sulphur Cycling on the Continents; Wetlands,Terrestrial Ecosystems and Associated Water Bodies. SCOPE 48. Wiley and Sons,Chichester, p307-344von Post, L., 1922. Sveriges geologiska undersoknings torvinventering och nâgre av des hittillsvunna resultat. Sv. MosskultuifOr. Tidskr. 1:1-27.Williams, C.H. and Steinberg, A., 1959. Soil sulphur fractions as chemical indexes of availablesulphur in some Australian soils.. Australian Journal of Agriculture Research, v. 10, p.340-352.242APPENDIX H REMOTE SENSINGAccess to the central areas of the Changuinola mire is extremely difficult, and necessitatedthe use of multiple remote sensing techniques. Digital satellite imagery from both Landsat andSPOT® satellites was utilized in this study. Landsat imagery of western Panama is available fromthe image archives of the US Geological Survey.Eros Data Centre. The image used in this studywas acquired in January 1979 by the Landsat 3 satellite.SPOT satellite imagery was acquired and interpreted by the author specifically for thisstudy. The image, which has a ground resolution of 20 m, was obtained by the SPOT 3 satellite onJanuary 7, 1994. The images published herein were generated from 3-band multispectral (Green,Red and Near-infrared) digital data using E. R. Mapper® image analysis software. Digitalmanipulations were performed to enhance contrasts in reflectivity of vegetation types, and tohighlight the presence of standing water. The digital data, and a large number of interpretedimages are on file at the University of British Columbia. and images are available to interestedparties on diskette as .TIF files.Low level, oblique normal colour and colour infrared photographs were taken by theauthor and used in the mapping of vegetation zones and drainage channels. Colour infrared isparticularly useful in assessing the health of vegetation and thus is useful in monitoring the extentof saline intrusion into freshwater wetlands. High altitude air photograph stereo pairs are availablefor part of the deposit from the Instituto Geografico ‘Tommy Guardia’, Panama City, and wereused to map detailed vegetation zones.243APPENDIX I LEVELLING SURVEYSA total of 9 km of survey lines were run by the author and Ing. Eduardo Reyes of theInstituto de Recursos Hidraulicos y Electrificacion. Accuracy is considered to be within 10 cmelevation in the marginal areas of the peat deposit, and 20 cm in the central regions. The use of thetheodolite became problematic in areas where the peat surface was submerged, extremely soft, andirregular, as is the case in much of the central region. In such areas, the tripod was placed on 10cm wooden pads which supported it at what was deemed the peat ‘surface’. The deepest standingwater encountered by this method was 40 cm, near Lake 10. It was not possible to traverse thesection between Lake 10 and Ed 3. Thus this part of the transect is represented by a straight lineon all cross sections. The SE ends of the levelling lines were tied to benchmarks along theAlmirante - Changuinola railway. The NW end was tied to sea level using tidal data from theInstituto Geografico ‘Tommy Guardia’, Panama City. The survey data and field notes areextensive and thus have not been included in the text. They are, however, archived at TheUniversity of British Columbia and will be made available to interested parties on request.Site locations not surveyed were determined by Magellan® Model 5000 Pro OPS receiver.Attempts to use GPS for elevation control were abandoned due to the accuracy limits, about 15 m,of single-receiver GPS data. All positions recorded in Appendix A are thus limited to at best I Smaccuracy, and in some cases may be as much as 50 m out. This inaccuracy can be due to thegeometry ofthe satellites at the time of acquisition, but is more frequently the result of theinterference of overhead foliage. At the time of writing, the use of GPS in surveying is consideredimpractical in treed areas.244


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items