Citation
Phylogeography and Conservation Genetics of the Yacare Caiman (Caiman yacare) of South America

Material Information

Title:
Phylogeography and Conservation Genetics of the Yacare Caiman (Caiman yacare) of South America
Creator:
GODSHALK, ROBERT E.
Copyright Date:
2008

Subjects

Subjects / Keywords:
Alligators ( jstor )
Crocodiles ( jstor )
Cytochromes ( jstor )
Haplotypes ( jstor )
Lowlands ( jstor )
Microsatellites ( jstor )
Polymerase chain reaction ( jstor )
Rivers ( jstor )
Species ( jstor )
Wildlife ( jstor )
City of Gainesville ( local )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Robert E. Godshalk. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Embargo Date:
5/31/2007
Resource Identifier:
444828045 ( OCLC )

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

PHYLOGEOGRAPHY AND CONSERVATION GENETICS OF THE YACARE CAIMAN ( Caiman yacare ) OF SOUTH AMERICA By ROBERT E. GODSHALK A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Robert Godshalk

PAGE 3

This document is dedicated to Dr. Federico Medem – a true renaissance scholar with a keen interest in knowledge, the supreme ‘ caimanologo’, friend. He spearheaded detailed studies of Neotropical crocodi lians, an inspired legacy that continues today in many countries. He encouraged me when I first be gan the study of these fascinating creatures. With his support, I undertook the first surveys in Venezuela of the endangered Orinoco crocodile, a cause dear to his heart. Our en thusiastic discussions on his patio at the Instituto Roberto Franco remain fond memories. His wonderful spirit and love of life and all living things was truly inspirational.

PAGE 4

iv ACKNOWLEDGMENTS First I want to thank the current and fo rmer members of my committee. Dr. F. Wayne King, committee chair, made it possible for me to attend the University of Florida. His broad knowledge of caimans a nd position in the world of crocodilian conservation assisted me in many ways. Fra nklin Percival and Ra y Carthy posed thought provoking questions on wildlife ecology and c onservation. Charlie Baer generously became a committee member at a late date and, with his phylogeographic experience, guided me through the complex genetic ananly sis. Clyde Kiker, my outside committee member, continues to stimulate my interest in natural resource economics. My gratitude goes to former members Richard Bodmer, currently in England, and John Eisenberg (deceased) for their time and interest in my project. The nature of fieldwork in a foreign c ountry placed me in constant need of assistance. Warm thanks go to the many indivi duals that helped me during my fieldwork and sample collection. Inclusion of all is im possible; I apologize to anyone inadvertently overlooked. Gracias to Tito and Rosario Ortiz for hospitality in Trinidad, Bolivia. My appreciation goes to the many ranchers who spent time with me, including Eduardo Añez, Rodolfo Brükner, Ron Larsen Dr. Romu lo Lozada, Osvaldo Monasterios, Andres Parra, Armando Pedacassé, Cristina Quaino de Paz, Rubens Rivero, Ernesto Salas and Hugo Serrate. The indigenous people were very open to new ideas and keen on developing their resources in a sustainable fashion. Valent in Muiba Guaji and Wigberto Rivero took

PAGE 5

v particular interest in the CITES project. The various nongovernmental organizations and individuals involved with Indian issues were also enthusia stic and helped at various times, thanks in part to Sylvia Chalukiá n, Cristian Gallegos, Jürgen Riester, Allyn Stearman, and Wendy Townsend. My warmest thanks go to so many governme nt employees in various offices that extended their time and data after only a br ief introduction: Ing. Fr eddy Arteaga, Teresa Centruión, Oscar Méndez, Ivan Pinaya, Luis Alberto Vaca, Lilian Villalba, and Jorge Montaño. Special thanks go to the Wildlife De partment employees in Beni who assisted during my trips: Gerardo Coimbra, En rique Cuellar, and Sergio Salazar. The various biologists I met, both Bolivia n and expatriate, offered opinions, advice and help: James Aparicio, Juan Pablo Arce, Claudio Barra, Stephen Beck, Paolo Bertella (deceased), Werner Honnegger, René Ibañ ez, Tim Killeen, Carmen Miranda, Luis Pacheco, Damian Ruiz, Andrew Tabor, and Marcelo Zalles. Thanks go to Dr. Mario Baudoin for finding time in his busy schedule to talk with me. Arturo Moscoso gave vigorous support and financial assistance from the Netherlands embassy. And last, but certainly not least, I thank Dante Videz R., who was my assistant for nearly all of the fieldwork. He had also accompanied Dr . King on surveys in 1986-87, and proved invaluable during our projects. My sample pool was substantially enhanced with the help of Crocodile Specialist Group members: Alej andro Larriera sent yacare samples from Argentina, Alvaro Velasco in Venezuela, Perran Ross arranged use of the Honduras specimens, and Kent Vliet provided samples for both Paleosuchus species from the St. Augustine Alligator Farm in Florida.

PAGE 6

vi My entrance into the new labyrinth of molecular science was eased by the mentoring by Dan Brazeau, Ginger Clark, Bill Farmerie and Ch arlie Baer. Highest thanks and praise go to Ginger Cl ark, head of the ICBR Gene tic Analysis Laboratory, who initially allowed me to have bench space, mentored me continuously, and later employed me. Special thanks go to the ICBR people, especially David Moraga and Sharon Norton of the Education Core for all the informa tive workshops on essential techniques. Kevin Holland and Ernesto Almira of the ICBR-GSSL assisted me with the bulk of the sequence data generation on the Amersham MegaBACE 1000. Savita Shankar of the ICBR Sequencing Core performed her seque ncing magic on stubborn templates and for haplotype verification. Tara Pa ton of the Sick Children’s Hospital in Toronto generated most of the raw microsatellite data. Naomi Phillips gave me much needed assistance, especially with her thorough analysis, advanced computer skills and moral support. Ed Braun helped me get through the myriad of molecular clock models and David Reed assisted with the phylogeographic analysis. Financial support for this project was supplied initia lly by a Tropical Conservation and Development (TCD) research grant, Programs for Studies on Tropical Conservation (PSTC) research grant, and by my father, Ernest L. Godshalk. Funding for the extensive population surveys and sample collection came from the CITES central office in Switzerland. Additional funding towards laboratory processing costs were supplied by my father and my wife, Vick i McGrath. Without Vicki’s organizational support this project might still be “in progress.” Tireless effort was spent by many individuals in editing drafts. I specifically want to thank F. Wayne King, Charlie Baer, Frankl in Percival, Ray Carthy, Clyde Kiker, Paul

PAGE 7

vii Moler, John Thorbjarnarson, Kent Vliet and Ginger Clark for comments on the manuscript. A special note of thanks goes to my parents, Ernest and Trudy, for their constant moral support and encouragement during my unconventional life. My warmest gratitude and love go to my wonderful wife, Vicki McGrath, who reviewed manuscripts, organized details for deadlines, endured endless piles of literature cl uttering her life and was very instrumental in the final product. My gratitude goes to anyone else unnamed that contributed in some part to the final product. The journey was a long time in th e making and I had contact with many people in several countries; thanks go to all.

PAGE 8

viii TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................xi LIST OF FIGURES.........................................................................................................xiv ABSTRACT.....................................................................................................................xi x CHAPTER 1 INTRODUCTION........................................................................................................1 Background...................................................................................................................2 Evolving Conservation.................................................................................................4 Current Study................................................................................................................8 2 YACARE CAIMAN DISTRIBUTION – RIVER SYSTEMS, HABITATS AND DETAILS ON THE BOLIVIAN LOWLANDS........................................................11 Introduction.................................................................................................................11 Amazon River Basin...................................................................................................15 Hydrology............................................................................................................15 Environment........................................................................................................23 Temperature.................................................................................................23 Precipitation.................................................................................................24 Soils..............................................................................................................26 Biota....................................................................................................................26 Flora.............................................................................................................26 Fauna............................................................................................................28 Human Dimension...............................................................................................30 Paraguay-Paraná River Basin.....................................................................................32 Hydrology............................................................................................................32 Environment........................................................................................................36 Temperature.................................................................................................36 Precipitation.................................................................................................37 Soils..............................................................................................................37 Biota....................................................................................................................38 Flora.............................................................................................................38

PAGE 9

ix Fauna............................................................................................................39 Human Dimension...............................................................................................39 The Caiman Species Complex....................................................................................40 Caiman crocodilus ...............................................................................................40 Caiman latirostris ................................................................................................43 Melanosuchus niger .............................................................................................44 Paleosuchus .........................................................................................................45 3 NATURAL HISTORY AND BASIC BIOL OGY OF THE YACARE CAIMAN, Caiman yacare ............................................................................................................48 Introduction.................................................................................................................48 The Yacare Caiman....................................................................................................50 Movement............................................................................................................53 Reproduction.......................................................................................................57 Prey......................................................................................................................60 Body Condition...................................................................................................63 Predation of Caiman............................................................................................64 Population Parameters.........................................................................................66 Ecological Role...................................................................................................68 Use of Caiman.....................................................................................................70 Commercial Hide Trade......................................................................................71 4 MATERIALS AND METHODS...............................................................................75 Introduction.................................................................................................................75 Sample Collection.......................................................................................................76 Fresh Samples......................................................................................................76 Dried Samples.....................................................................................................80 Bone Samples......................................................................................................81 Extraction of DNA......................................................................................................83 Phenol-Chloroform-Isoamyl Alcohol Extraction................................................83 Extractions from Bone.........................................................................................85 Quantity and Quality...........................................................................................89 Additional Isolations...........................................................................................90 Molecular Investigations............................................................................................91 16S rRNA Gene...................................................................................................91 The Cytochrome b gene.......................................................................................94 Microsatellite Loci...............................................................................................97 5 MOLECULAR PHYLOGENY OF EX TANT ALLIGATOROIDEA, WITH DETAILS ON THE CAIMANINAE.......................................................................106 Introduction...............................................................................................................106 Materials and Methods.............................................................................................108 Molecular Investigations..........................................................................................110 16S rRNA Sequences........................................................................................111

PAGE 10

x Cytochrome b Sequences..................................................................................114 Data Analysis............................................................................................................116 Results and Discussion.............................................................................................117 Alligatoroidea....................................................................................................117 Caimaninae........................................................................................................120 Northern South America and caiman migrations..............................................127 Evolution of the Isthmus...................................................................................131 Conclusions.......................................................................................................134 6 PHYLOGEOGRAPHY AND CONSERVATION OF Caiman yacare ...................136 Introduction...............................................................................................................136 Sample Preparation...................................................................................................137 Sequence Data..........................................................................................................141 Microsatellite Loci....................................................................................................144 Results.......................................................................................................................1 51 Cytochrome b ....................................................................................................151 Microsatellites...................................................................................................161 Discussion.................................................................................................................169 Conservation Genetics..............................................................................................175 Conclusions...............................................................................................................181 APPENDIX A LOCALITY INFORMATION FOR SA MPLES USED IN THIS STUDY.............183 B MICROSATELLITE ALLELE FREQUE NCY COMPARISON OVER POPULATIONS.......................................................................................................193 LITERATURE CITED....................................................................................................202 BIOGRAPHICAL SKETCH...........................................................................................227

PAGE 11

xi LIST OF TABLES Table page 3-1 Vertebrate predators of Caiman spp. listed from reports throughout the Neotropics. Most are found within C. yacare distribution (Modified from Gorzula and Seijas 1989).......................................................................................65 3-2 Abundance of caiman (ind./km) in Neotr opical habitats from selected sources (Modified from Pacheco 1993)..............................................................................67 4-1 Oligonucleotide primers for PCR am plification of m itochondrial 16S rRNA gene Fragments A and B for sequence anal ysis. Position refers to 5Â’ nucleotide of forward primers (A or F) and 3Â’ nu cleotide of reverse primers (B or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001)............93 4-2 Oligonucleotide primers for PCR amp lification of mitoc hondrial cytochrome b gene fragments for sequence analysis. Position refers to 5Â’ nucleotide of forward primers (H or F) and 3Â’ nucleoti de of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001).........................96 4-3 Oligonucleotide primers and paramete rs for PCR amplification of genomic microsatellite loci in Caiman yacare ...................................................................103 5-1 Data on species and material used for analyses of Alligatoroidea.......................110 5-2 Oligonucleotide primers for PCR am plification of m itochondrial 16S rRNA gene Fragments A and B used for sequence analysis. Position refers to 5Â’ nucleotide of forward primers (A or F) a nd 3Â’ nucleotide of reverse primers (B or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001).112 5-3 Oligonucleotide primers for PCR amp lification of mitoc hondrial cytochrome b gene fragments for sequence analysis. Position refers to 5Â’ nucleotide of forward primers (H or F) and 3Â’ nucleoti de of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001).......................115 5-4 Information for 1120 base cytochrome b sequence data set for 40 Caiman haplotypes used in analyses.................................................................................124 5-5 Pairwise Fst test results, all significant at an level of 0.05................................124

PAGE 12

xii 5-6 Population average pairwise distan ces. Above diagonal: Average number of pairwise differences between populations ( xy). Diagonal elements: Average number of pairwise diffe rences within population ( x); Below diagonal: Corrected average pairwise difference ( xy-( x+ y)/2), all significant at an level of 0.05.........................................................................................................124 5-7 Distances between popula tions (kilometers) of Caiman samples used in Mantel analysis.................................................................................................................126 6-1 Oligonucleotide primers for PCR amplification of Caiman yacare mitochondrial cytochrome b gene fragments for sequence analysis. Position refers to 5’ nucleotide of forward primers (H or F) and 3’ nucleotide of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al 2001).............................................................................................................142 6-2 Oligonucleotide primers, repeat type and Tm used for PCR amplification of genomic microsatellite loci in Caiman yacare ....................................................149 6-3 Distribution of 25 cytochrome b haplotypes for Caiman yacare . No haplotypes detected are shared among basins. Localit y for original haplotype specimens: Bo=Bolivia, Pa=Paraguay....................................................................................155 6-4 Analysis of Fst values for populations of Caiman yacare . Top set: Population pairwise distance Fst values, Bottom set: Matrix of significant Fst values, p <0.05....................................................................................................................159 6-5 Average pairwise distances for populations of Caiman yacare . Above the diagonal: average number of pairwise differences between populations ( XY). Diagonal elements: average number of pairwise differences within population ( X). Below diagonal: corrected average pairwise difference ( XY-( X + Y)/2).160 6-6 Measures of genetic diversity within populations of the río Amazon (north) and río Paraguay-Paraná (s outh) distributions............................................................160 6-7 Microsatellite loci, allele number and fragment size data used for analyses in Caiman yacare. Test of allele frequencies for Hardy-Weinberg Equilibrium (HWE). (+ = frequencies in HWE, = frequencies out of HWE).......................162 6-8 Average pairwise distances for populations of Caiman yacare . Above the diagonal: average number of pairwise differences between populations ( XY). Diagonal elements: average number of pairwise differences within population ( X). Below diagonal: corrected average pairwise difference ( XY-( X + Y)/2).163 6-9 Summary statistics for all loci across all sampled populations for Caiman yacare . North = Beni, Mamoré, San Migue l and Paraguá, South = Paraguay and Paraná. N=number of individuals sampled, AN = number of alleles per locus, HO = observed heterozygosity, and HE = expected heterozygosity...........164

PAGE 13

xiii 6-10 Analyses of Molecular Variance (AMOVA) of sampled Caiman yacare populations for Fst (upper panel) and Rst (lower panel) values. Results in panels are presented for populations analyzed se parately and grouped into Northern (río Amazon) and Southern (río Pa raguay-Paraná) distributions........................165 6-11 Analysis of Fst and Rst values for microsatellite alleles in populations of Caiman yacare . Top set: Population allele Rst values (sum of square differences) above diagonal; pairwise distance (number of alleles) Fst values below diagonal, Bottom set: Matrix of significant values: Rst, p 0.05 above diagonal; Fst, p <0.05 below diagonal..................................................................166 6-12 Matrix of Rst and geographic distances used fo r Mantel test. Upper diagonal contains riverine distances (kilometers) among the sampled Caiman yacare populations. Lower diagonal contains the Rst values (Slatkin 1995)...................166 6-13 Comparison of microsatellite he trozygosities for crocodilian species................174

PAGE 14

xiv LIST OF FIGURES Figure page 1-1 A robust adult male yacare caiman ( Caiman yacare ) from the rio Itonamas drainage in central Bolivia.........................................................................................4 2-1 Map of the distribut ion of the Yacare caiman, Caiman yacare in South America, with political boundaries shown. Distri bution includes portions of Bolivia, Brazil, Paraguay and Argentina...............................................................................12 2-2 South America with details of th e two main river systems within the Caiman yacare distribution, the Amazon (north) a nd the Paraguy-Paraná (south). Most tributaries outside the distribution ar e not shown for map simplification. Red arrows indicate approximate latitudinal range limits...............................................13 2-3 The northern distribution of Caiman yacare , found within tributaries of the upper río Madeira system of the Amazon basin. Dashed line marks the southern limit of the Amazon drainage, dotted line is 500m elevation contour.....................16 2-4 Rio Mamoré, central Beni Department. Repeated change of course creates many isolated bodies of water off the main ch annel, which provid e excellent habitat for Caiman yacare (left). Smaller caños meander across open savannas and, along with associated water bodies, also provide excellent habitat (right)..............20 2-5 Localities for Caiman yacare recorded during surveys of the northern distribution ( -Medem 1983; King and Videz 1989; Brazaitis et al . 1990; Godshalk and King 2002 sample localities for this study)....................22 2-6 Mean annual temperatures for the area of distribution of Caiman yacare . Isotherms in û C ........................................................................................................24 2-7 Mean annual precipitation fo r the area of distribution of Caiman yacare . Isohyets in mm.........................................................................................................25 2-8 The southern distribution of Caiman yacare , found in the río Paraguay and lower río Paraná systems..........................................................................................33 2-9 Localities for Caiman yacare recorded during surveys of the southern distribution...............................................................................................................35

PAGE 15

xv 2-10 Distribution of Caiman crocodilus : C. c. fuscus in the north, and C. c. crocodilus in the south (the limited C. c. apaporiensis distribution is not shown). – localities for samples used in this study. – localities for published sequences included in analyses for this study (see Farias et al . 2004)......................................41 2-11 Simplified map of Hon duras with localities of Caiman crocodilus fuscus samples used in this study. Samples collected during a CITES surveys of Atlantic drainage rivers and coastal lagoons (see King et al . 1990) and used in this study..................................................................................................................42 2-12 Distribution of the broad-snouted caiman, Caiman latirostris (blue) in South America, showing areas sympatric with C. yacare (green). locality for samples used in this study........................................................................................43 2-13 Distribution of the black caiman, Melanosuchus niger (blue) in South America, showing areas sympatric with C. yacare (green). locality for samples used in this study. – localities for published sequences included in analyses for this study (see Farias et al . 2004)....................................................................................45 2-14 Distribution of the Dwarf caiman, Paleosuchus palpebrosus (blue), showing areas sympatric with Caiman yacare (green). locality for samples used in this study..................................................................................................................46 2-15 Distribution of the Schneid er’s smooth-fronted caiman, Paleosuchus trigonatus (blue), showing areas sympatric with Caiman yacare (green).................................47 3-1 Adult male Caiman yacare from the río Mamoré basi n (left). Note strongly contrasting mandibular markings characte ristic for the species. Adult male Caiman c. crocodilus from río Apure basin in ce ntral Venezuela (middle). Mandibles lack any markings, color is uni form. A similar appearance is seen in C. c. fuscu s in Mexico (right)...................................................................................53 3-2 Peak dry season concentration of adult Caiman yacare in the Bolivian Pantanal. Basking at mid-day (left) and eyeshine re flected at night (rig ht). Over 700 large caiman were counted in this artificial cat tle tank that measured approximately 85 m x 105 m.................................................................................................................54 3-3 A group of yacares fishing in flowi ng water. Individuals face the water flow, mouths open, and snap at passing fish (left, ©C. Yamashita). Caiman feed on concentrated fish in the dry season water bodies. An adult caiman eating an armored catfish ( Hoplosternum ) is shown (right)....................................................61 3-4 Freshwater apple snail ( Pomacea lineata ) from central Bolivia, a common invertebrate food item for C. yacare in the flooded savannas (left). Caiman may also feed on carrion. An adult Caiman crocodilus in central Venezuela has seized a capybara carcass, Hydrochoeris hydrochaeris , (right)...............................62

PAGE 16

xvi 4-1 Juvenile yacare captured by hand in eas tern Bolivia. Jaws were secured with electrical tape (l eft). Capture of Caiman yacare in arroyo Negro, southern Bolivia, using a Ketch-All™ capture pole (right)....................................................77 4-2 Extraction of fresh blood sample from Caiman yacare in the field (left). Precipitation of DNA after isolation from a blood sample. Cloudy material in the center of the tube is purified genomic DNA. (right)................................................78 4-3 Unusual sample sources: trophy skull of black caiman ( Melanosuchus niger ) from central Bolivia. Skull measured 57 cm, representing an adult of about 4.7 m total length (left). Chair and swing fa shioned from a salted and dried black caiman belly skin hunted from a nearby caño (right)...............................................80 4-4 A laminar flow hood is essential for work with bone to avoid crosscontamination of samples while drilling with the Dremel tool (left). Various drill bits used, and an osteoderm sample from Paraguay (center). Example of a femur used for sample drilling (right).................................................................................82 4-5. Positions of primers and PCR amplifie d Fragments A and B in reference to the Caiman crocodilus 16S rDNA mitochondrial gene.................................................92 4-6 Figure 4-5. Positions of primers and PCR amplified Cyb Fragments 1, 1A, 1B, 2 and 2B in reference to the Caiman crocodilus cytochrome b mitochondrial gene..95 4-7 Polymorphic loci N10 (left), and R8 (right) PCR products on 1.2% agarose (w/v) gels with EtBr staining and UV luminescence.............................................104 5-1 Diagram of phylogenetic nomenclature fo r extant crocodylians. Arrows indicate stem-based group names (groups including a species and any other more closely related to it than to another species) , and black circles denote node-based group names (last common ancestor of two or more species and all of its descendents) (simplified from Brochu 2003)..............................................................................107 5-2 Maximum parsimony phylogram representing base changes for 2200 bp concatenated partial sequences of mito chondrial 16S and cytochrome b genes. Tree length is 2057, CI=0.761, RI= 0.565, RC=0.430 and HI=0.239. Species abbreviations are: Cc ( Caiman crocodilus ), Cy ( C. yacare ), Cl ( C. latirostris ), Mn ( Melanosuchus niger ), Pp ( Paleosuchus palpebrosus ), Pt ( P. trigonatus ) Am ( Alligator mississippiensis ), As ( A. sinensis ) C. mydas ( Chelonia myda s), Gallus ( Gallus gallus )........................................................................................................118 5-3 Minimum evolution distance phylogra m with outgroup rooting (score 0.79236). Note deep basal branch length for Paleosuchus . Species abbreviations are the same as in Figure 5-2.............................................................................................119 5-4 Bootstrap values for support of Alligatoroidea phylogeny....................................120

PAGE 17

xvii 5-5 Minimum evolution distance phylogra m (uncorrected ‘p’) for cytochrome b sequence comparison rooted with Melanosuchus (score 0.22141, branch length above; 1000 bootstrap replications with values below branch: top-ML, middleME, bottom-MP). Alternate trees and boot strap values on th e right. Species abbreviations are: Cc ( Caiman c. crocodilus ), Ccf ( Caiman c. fuscus), Cy ( C. yacare ), Cl ( C. latirostris ), Mn ( Melanosuchus niger )..........................................121 5-6 Minimum evolution distance tree for cytochrome b data set using GTR+ evolutionary model; score=0.15676. Bootstra p branch support for Ccf not strong and the relationship remains unclear. Boot strap values (1000 replications): topML, middle-ME, bottom-MP. Species abbreviations are: Cc ( Caiman c. crocodilus ), Ccf (( Caiman c. fuscus), Cy ( C. yacare ), Cl ( C. latirostris ). A = río Amazon, P = río Paraná, O = río Orinoco.............................................................123 5-7 Possible configuration at the end of the Cretaceous (ca. 65 mybp) when the caiman lineage may have migrated to S outh America. Major Central American blocks have sutured allowing fauna l movement (Me=Mexican, Ma=Mayan, Ch=Chortis). Volcanic arc (Aves Ridge) along the trailing western edge of the Caribbean Plate may have provided ‘ste pping stones’ for intercontinental faunal exchange. Arc ultimately moved east, forming the Lesser Antilles. Arrows indicate plate movement (after Donnelly 1988).....................................................128 5-8 Paleogeographic reconstruction of north ern South America and Central America before uplift of Isthmus of Panama and cl osing of two ocean basins (c. 10 mya). Arrows indicate marine paleocurrents a nd continental drainage and sediment transport. M = proto-lake Maracaibo out flow area (after Albert and Crampton 2005).......................................................................................................................130 5-9 Neogene palaeooceanographic and palae ogeographic evolution of the central American region. Hypothetical surface fl ow resulting from the disruption of warm Caribbean flow and intensificati on of the cool California current. A – Middle Miocene (15.1 to 12.9 mybp). B – Late Middle Miocene (11.8 to 7.0 mybp). C – Late Miocene to Early Pliocene (7.0 to 3.7 mybp). D – Early Pliocene (3.7 to 3.1 mybp). See text fo r explanation. (Simplified from DuqueCaro 1990)..............................................................................................................133 6-1 River drainage divisi ons used for segregating samples from the Amazon (northern) distribution of Caiman yacare . B = río Beni, M = río Mamoré, SM = río San Miguel, P = río Paraguá.............................................................................153 6-2 River drainage division used for segregating samples from the río Paraguay-Paraná (southern) distribution of Caiman yacare . Py = río Paraguay, Pa = río Paraná. Note length and location of western tributary río Pilcomayo....................154 6-3 Maximum parsimony (MP) tree for 25 cytochrome b haplotypes of Caiman yacare rooted with Caiman crocodilus ..................................................................156

PAGE 18

xviii 6-4 Genealogy network for 25cytochrome b haplotypes in Caiman yacare . Geographic division (blue line) occurs between river basins: A = Amazon, P = Paraguay-Paraná. Branch length between haplotypes and nodes equal one base change. Red symbols indicate the most common haplotypes in each distribution. Haplotype Cy-03 (rectangle) may represent the ancestral line..............................158 6-5 Allele frequency distributions of five microsatellite loci for Caiman yacare in the two main distribution areas: Northern (Amazon – light blue) and Southern (Paraguay-Paraná – dark grey)Norther n distribution include s populations from rios Beni, Mamoré, San Miguel and Paraguá . Southern distribution includes the rios Paraguay and Paraná.......................................................................................167 6-6 Allele frequency distributions of six microsatellite loci for Caiman yacare in the two main distribution areas: Northern (Amazon – light blue) and Southern (Paraguay-Paraná – dark grey). Northe rn distribution includes populations from ríos Beni, Mamoré, San Miguel and Paraguá . Southern distribution includes the ríos Paraguay and Paraná.......................................................................................168

PAGE 19

xix Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy PHYLOGEOGRAPHY AND CONSERVATION GENETICS OF THE YACARE CAIMAN ( Caiman yacare ) OF SOUTH AMERICA By Robert E. Godshalk May 2006 Chair: F. Wayne King Major Department: Wildlife Ecology and Conservation Crocodylians have a fossil record dating back to Archosaurian ancestors over 200 million years ago. One evolutionary line produc ed the Caimaninae, a largely Neotropical alligatorid group. The focus of this study, Caiman yacare , is found in wetland habitats of central South America. The yacare caiman is one of the worldÂ’s most abundant crocodylian species and represents a high pe rcentage of the world hide trade. Caiman yacare is often considered a subspeci es. Analyses were performed to determine the genetic distance from related species. DNA was extracted from all extant Alligatoroidea species: Alligator mississippiensis , A, sinensis , Paleosuchus palpebrosus , P. trigonatus , Melanosuchus niger , Caiman crocodilus , C. latirostris and C. yacare , and PCR amplified. Concatenated partial sequen ces of mitochondrial 16S and cytochrome b genes were compared. Divergence of Alligator from the monophyletic caiman group is estimated at about 69 million years. Divergence of Caiman yacare and C. crocodilus is estimated at about 7.5 my.

PAGE 20

xx Caiman yacare is a distinct Evolutionary Significant Unit that merits full species status. Analyses of C. c. fuscus from Honduras are presente d and is also distinct, significantly different from both C. c. crocodilus and C. yacare in all tests performed. Caiman yacare distribution includes portions of two large river systems. DNA was extracted from 211 samples representing th e northern (Amazon, n=143) and southern (Paraguay-Paraná, n=68) populations and 25 cytochrome b haplotypes are described. Each haplotype occurs uniquely to only one basin (río Amazon, n=16; río ParaguayParaná, n=9). The northern distribution is divide d into 4 drainages with shared (2, 5, 5, 1) and unique (1, 1, 7, 2) haplotypes. The southe rn distribution, a singl e river drainage, is divided into upper and lower s ections with shared (3) and unique (5, 1) haplotypes. The two most common haplotypes accounted for 76% of the samples in each distribution.. Analyses of molecular variance (AM OVA) demonstrated a correlation of Fst values and distance with 72.4% of the differe nce between the distributions. Microsatellite analyses were also under taken. Seven primer sets were developed and combined with four primers from a previous study of Caiman latirostris (Zucoloto et al . 2002). All loci were present in all populati ons and the average he terozygosity for 11 loci was 0.81 with a standard deviation of 0.068. As a result of high numbers of alleles (12 to 40) per locus, and relatively small samp le sizes, fine-scale structuring could not be determined. Results from Rst AMOVA tests show only 19.8% of the variation between the north and south distributions, with 74.4% of the variation with in the populations. These data may be valuable for wildlif e management during sustainable harvest planning. Mitochondrial haplotype discrimination also has futu re forensic possibilities for source locality determination for both meat and skins.

PAGE 21

1 CHAPTER 1 INTRODUCTION Crocodilians and humans have undoubtedly interacted since the dawn of humankind. The pantropical distribution of Crocodylia has resulted in a continued relationship as human populat ions expanded into new tropi cal areas. Crocodilians have contributed to many cultures around the gl obe in various ways. They have been incorporated into local diet s, folklore, mythology, literatur e, music, art and religion. “They have been revered a nd reviled, deified and vilifie d, protected and slaughtered” (Trompf 1989). Yet, in spite of the collectiv e knowledge amassed on most of the large, conspicuous vertebrate groups over the past few centuries, crocodilian biology remained largely unknown until relatively recently. Economic forces exerted strong negative e ffects long before these creatures were understood. During the beginning decades of the last century, evolving fashion trends in the international market required increasing crocodilian hide production. The largest and most accessible species were targeted first. As technologies advanced aiding both the field hunter and international shipping, even the most remote places and populations were affected. Beginning in the 1940s and 50s, the market turned to the numerous, but less valuable, caiman species (Fuchs et al . 1989). These stocks were al so drastically affected in a short time. At the same time, global concern was deve loping in regard to widespread wildlife population reductions, endangered species, a nd environmental degradation. This was especially accentuated with the expansion of poor, rural human populations in many

PAGE 22

2 developing countries of the tr opics. Slash and burn farmers, along with cattle, timber and industrial agriculture in terests, continuously reduced trop ical wildlands – a process that goes on today. In response to these threat s, focused conservation efforts have begun, increasing numbers of protected areas have been declared and programs for wildlife management have been developed throughout the tropics. Conservation of tropical wildlife has undergone tremendous development in the past 30 years. Results of research by nume rous dedicated biologists have increased the collective understand ing of many species and system s throughout the tropics. As new techniques develop, they are frequently in corporated into the toolboxes of research scientists. Prominent among new developments of the last decade are a suite of molecular techniques that provides useful information to many disciplines. In this dissertation, I have attempted to address a number of questi ons that can be resolved most effectively through understanding at a molecular level. Background Nearly all of the world's crocodilian species have suffe red widespread population reductions within the last fifty years (Fitzgerald 1988; King 1989; Ross 1998). These species are important ecologica l components in their wetland habitats, usually occupying a position as top predator. They are often c onsidered as keystone species within their respective habitats. Over-hunting of certain crocodilian species for their hides has reduced some populations to the point of ecological extinction (Chiriví-Gallego 1973; Alcala and Dy-Liacco 1989). Other species ha ve been severely reduced, limiting their functional roles within wetland ecosystems. The collateral effects ar e decreased stability of the ecosystems, loss of biodiversity, and decreased ec onomic potential (Ross 1998).

PAGE 23

3 Extensive global scale predation by humans for skins has recently threatened the ecological role of many crocodi lians. Currently, 16 of the 23 sp ecies of crocodilians have been placed on the endangered Appendix I of the Convention on International Trade of Endangered Species of Wild Flora and Fauna (CITES) (Ross 1998). The present situation has evolved since the early 1900s, mostly due to the demand for hides by the international leather industry. In addition to the direct eff ects of the commercial hide hunters, habitat alteration and encroaching human activities have also contributed to reduced crocodilian populations (Alcala and Dy-Liacco 1989). In a few species, most notably the Amer ican alligator and two South American caiman, sufficient wild populations are main tained to permit sustainable harvest under regulated programs. One of these, the yacare caiman ( Caiman yacare ), from central South America, is the focus of this dissertati on (see Figure 1-1). It is arguably one of the most abundant of the world’s cr ocodilian species and supplies a large percentage of skins in the global hide trade. Ye t, until about 15 years ago, vi rtually no detailed biological studies had been undertaken on the species. Sin ce that time, the majority of studies on Caiman yacare have been undertaken in the Pantanal of Brazil (e.g. Schaller and Crawshaw. 1982; Crawshaw 1987; Cintra 1988; Crawshaw 1991; Campos 1993; Mourão et al . 1994; Campos and Magnusson 1995; C outinho and Campos 1996; Mourão et al . 1996; Santos et al. 1996; Coutinho 2000, Mourão et al . 2000; Coutinho et al . 2001, Coutinho and Campos 2002). Most of the rese arch conducted in Argentina, Bolivia and Paraguay, the other range states, has resulted in population surveys that related to harvest programs. Brazil currently permits commercial exploitation of wildlife only with captive animals. There is a movement for creati ng some sustainable exceptions including a

PAGE 24

4 yacare caiman program. Studies in other parts of the yacare distribution have not been done and much of the basic biology and natural history remain unknown. Figure 1-1. A robust adult male yacare caiman ( Caiman yacare ) from the rio Itonamas drainage in central Bolivia Evolving Conservation Various forms of wildlife management ha ve existed for centuries, but in many Western cultures it was manifested in royal privilege. As the urban populations grew, a commercial market for wild meat arose, as we ll as fashion markets for hides and feathers of wild species. With deadly efficienc y, the market hunters be gan reducing wildlife populations. Only during the twentieth century did legislation have any real impact in conserving wild populations, and this usually occurred after great damage had taken place. In 1900, the Lacey Act was passed in the Un ited States, a major wildlife consumer nation, and it still has impact today. The main intent was to control the destructive effects

PAGE 25

5 of commercial hunting by controlling the market (King 1978). While market hunting within the United States was largely brought und er control, the situa tion deteriorated for many species worldwide. Concern for dwi ndling wildlife populat ions led to the Endangered Species Act (ESA) of 1973, whic h along with the Lacey Act, helped to control imports to the United States, an im portant market for international wildlife products (Jenkins 1987; King 1978, 1989). A worldwide protectionist movement de veloped in the 1960s and 1970s and many regulations, often modeled after ESA, were wr itten by nations attempting to protect their wildlife. The example of consumer market c ontrol had been set by the United States, and was adopted later by many European nations. In addition to strict import laws, strong consumer sentiment against wildlife produc t use evolved in many developed nations. Campaigns focusing on high-profile species, such as spotted cats and ivory, resonated with the general public and elicited widespread sympathy (Fitzgerald 1989; Shaw 1991). While preservation of endangered species is occasionally attained, the majority of the world's biota remains outside effective c ontrol. Cultures of most developing nations have long traditions of wildlife consumption, appropriately identified as a "subsidy from nature" (Hecht et al . 1988), and continue to rely on these benefits. A significant segment of the rural human population c ontinues to be maintained by subsistence hunting for food and market products, particularly in the tr opics (Robinson and Bennett 2000). While this assistance is seldom calculated into rural development projects and its value rarely accounted for, many endeavors could not c ontinue without it. Recognition of the legitimate use of wildlife by ru ral people in the tropics give s the conservation of those resources new focus and goals (Robinson and Redford 1991).

PAGE 26

6 Conservation strategies evol ved with new and distinctiv e characteristics during the 1980s. The World Conservation Strategy called for "management of human use of the biosphere so that it may yield the greatest sustainable benefi t to present generations while maintaining its potential to meet the needs and aspirations of futu re generations" (IUCN 1980:2). This focus led many programs to expl ore management of resource use based on economic incentives. The rationale is that allocation of right s to the resource benefits increases the advantages and importance of those resources for the users (Magnusson 1984; King 1989; Shaw 1991). Use rights and re sponsibilities are clearly defined and management goals are designed to minimize effects of the "tragedy of the commons" (Hardin 1968; Robinson and Redford 1991). These circumstances led to many of the current crocodilian conservation programs. Although several endangered species require co mplete protection for their recovery, other species appear to have benefited from sust ained-utilization management. One particular program, the wild harvest of the common caiman, Caiman c. crocodilus , in Venezuela has been very successful. It was used as a model for the plan for Bolivia. The Venezuelan harvest was initiated in 1982, and involves using large, privately owned cattle ranches as conser vation units. Cattle ranchers supplement ranch production through conservation and subsequent harvest of caiman that occu r naturally on their property. These ranches encompass the necessa ry habitat and caiman populations for this type of harvest, and, due to their economic interest in maintaining yield, the ranchers control poaching on their property (Gorzu la 1987; King 1989). Ambivalent attitudes toward caiman have often been transf ormed into protectionist sentiments.

PAGE 27

7 The Venezuelan government has scant reso urces for widespread enforcement of hunting controls and often such attempts are resisted or ignored. By allocating the rights of the harvest benefits to the landholder, a pers onal interest in prot ection of the resource is created. For this reason, land tenure is a key factor in the functioning of this program (Thorbjarnarson 1991b). Before initiating the plan, the Venezuelan government was involved in collecting information on the species. I worked for thei r Department of Wildlife gathering baseline biological data for caiman during this pe riod (Godshalk 1976; Mane ss 1976; Marcellini 1979). The government was thorough in the pl anning stages, and in coordinating the necessary elements of the private sector. Afte r observing the careful development of this program, and its subsequent success, I believed that a modified plan might be a useful and productive conservati on tool elsewhere. Bolivia was chosen for possible intr oduction of a caiman conservation and management plan based on the Venezuelan m odel. From my extens ive experience in Venezuela, and later travels in Bolivia, I fe lt that the transfer of technology would be appropriate. Adaptation to local conditions would be made easier by key similarities between Bolivia and Venezuela. Bolivia has extensive lowland hab itat, much of which resembles the Venezuelan Llanos savanna in climate, flora and fa una. A large portion of the lowlands is held in large, privately ow ned cattle ranches that are extensively managed with minimal impacts to the environment. The area is populated by Caiman yacare , closely related to the common caiman harv ested in Venezuela (Medem 1983; King and Videz 1989; Ergueta and Pacheco 1990).

PAGE 28

8 I met with IUCN Crocodile Specialist Gr oup consultants, government officials, civic and Indigenous leaders and helped fo rm a workable program. After a limited preliminary season harvest of 10,000 caiman, the program opened with a 30,000 caiman quota that was conducted entirely on Indige nous lands. The program continues, albeit with imperfections, and the mechanisms for improvement appear to be working. Bolivia has a history of intensive, un sustainable wildlife exploitation. Beginning with trade in caiman hides in the 1940s, th e commerce quickly expanded to all species that were profitable (Medem 1983; Claure 1986). This included other reptile skins, furs, feathers, and live birds, primates and tropical fish. In response to unregulated exports and widespread corruption, all commercial wild life use was made unlawful by the Bolivian government in 1989. Hopefully, the current caiman program will depart from the previous cycle and begin a new era of responsible management. Current Study This study seeks to learn about Caiman y acare from the most fundamental unit of heredity at the molecular level – the DNA. Few molecular studies have been undertaken on crocodilians and even fewer that include caiman. C. yacare is often included with C. crocodilus. There is growing interest for su stainable use programs by the yacare range states (Ross and Godshalk 2003) and molecu lar information gained here may have important management implications. Listed are some of the general questions to be addressed at a molecula r level in this study: What is the molecular p hylogeny of the Caimaninae? What is the phylogenetic position of C. yacare in relation to C. crocodilus ? Is C. yacare an identifiable evolutionary li neage and are there any subunits or cryptic populations?

PAGE 29

9 Is there evidence for diminished genetic di versity, inbreeding or past bottlenecks? Is there a geographic signature in the haplotype distributions? Do the results have poten tial forensic applications? This dissertation is structured in th e following manner. Chapter 2 outlines the geographic limits of distribution of Caiman yacare and the important landscape details such as temperature, precipitation, and biotic factors. Special atte ntion is given to the river systems and known localitie s for yacares. The distributi on can be divided into two distinct geographic areas. Th e northern distribution encompasses most of lowland Bolivia where the rivers drain northward and form part of the Amazon Basin. The southern distribution lies within the Pa raguay-Paraná drainage, whic h runs south to the south Atlantic. Chapter 3 summarizes pert inent literature on the biology of Caiman yacare . Many facets of yacare natural history are unknown or poorly understood. For certain details, information is drawn from studies of the closely related Caiman crocodilus . A synopsis of the taxonomy, current usage and unresolved issues within Caimaninae is given. Chapter 4 explains the various material s and methods. I collected a significant portion of the samples, largely whole blood, fr om yacares within the major drainages in the Bolivian lowlands. I obtained permissi on to use Paraguayan yacare osteological material in the collection at the Florida Museum of Natural History (FLMNH). After some disappointing preliminary trials, a successful protocol was developed and my geographic and molecular sample pool wa s increased. Twenty blood samples were obtained from Argentina, further expandi ng my investigation. The Honduran and Venezuelan samples from common caiman were donated through the FLMNH. Samples

PAGE 30

10 from additional caiman species were obtaine d from the St. Augustine Alligator Farm, St. Augustine, Florida, that helped complete my survey of the group. Chapter 5 and 6 are results, analyses and discussions of the molecular data. Chapter 5 contains details of the molecular p hylogeny of Caimaninae based on mitochondrial sequence data from the 16S rRNA and cytochrome b genes. New data are presented for all the species, as well as estimated divergen ce times and an analysis of the position of C. yacare within the clade. An analysis of the Caiman crocodilus fuscus subspecies relationship is presented.. Chapter 6 is a combined analys is of mitochondrial cytochrome b sequence data and microsatellite fragments. A phylogeographic analysis for C. yacare is described and recommendations for conservation are given. Hopefully, these analyses will provide useful information about the evolutionary history of caiman such as patterns of gene flow and relationships among species. Further, by understanding the phylogeography of C. yacare , and the genetic patterns and dynamics of populations between drainages, more effective management can be achieved. These data will be added to the growing body of knowledge for caiman in general, and for Caiman yacare in particular, to better understand and manage this remarkable reptile.

PAGE 31

11 CHAPTER 2 YACARE CAIMAN DISTRIBUTION – RIVER SYSTEMS, HABITATS AND DETAILS ON THE BOLIVIAN LOWLANDS Introduction The yacare caiman, Caiman yacare , inhabits most wetland areas within its distribution in central South America (see Fig. 2-1). It occu rs in portions of two major river systems, the Amazon and the Paraguay-Pa raná (see Fig. 2-2). The Andes, or its effects, form most of the western limits a nd the Brazilian Shield describes most of the northern and eastern edge. Understanding th e complexities of the geographic setting gives insight to potential immi gration and gene flow pathwa ys that could affect the species. The geologic history of South America is not well documented, in part due to the intense weathering of great expa nses east of the Andes. Vari ous drainage histories have been suggested. Salt-water intrusions have occurred far into the continent creating large inland seaways (Räsänen et al . 1995; Webb 1995). These, as we ll as uplifted salt-bearing marine sediments, resulted in large scale salt depositions such as found in the immense Uyuni salt flats of southwestern Bolivia and in large areas of the xeric Gran Chaco of western Paraguay. The Andean orogenesis has a long and complicated history that began in the south and central re gions about 45 mybp and progres sed northward. By 10 mybp the east and west drainages e ssentially had been divided by a barrier at least 1000 m high and nearly 8000 km long (Webb 1991).

PAGE 32

12 Figure 2-1. Map of the distri bution of the Yacare caiman, Caiman yacare in South America, with political boundaries shown. Distributi on includes portions of Bolivia, Brazil, Paraguay and Argentina. The history of the Amazon basin is in completely understood. An inland Lago Amazonas has been proposed with potentia l portals north, east or south (Fraily et al . 1988). Present day topography divides the Ori noco, Amazon and Paraguay-Paraná basins by rises of only 400 m. Miocene marine fossils are present in the Andean foreland basin of Argentina and Paraguay correlating w ith high sea levels about 11 mybp (Haq et al .1987). Absence of similar fossils in Amazoní a may be an artifact of weathering and acidic dissolution (Webb 1995). Räsänen et al. (1995) also conclude that sedimentary and fossil evidence indicates a semi-marine embaymen t or seaway existed in the west-central

PAGE 33

13 Amazon at that time. Large-scale lacustrine deposits point to possi ble basin creation by plate subduction or downwarping (Fraily et al . 1988). Recent works indicate that until the Middle Miocene (16 mybp), the western Am azon region drained north with a delta forming at present day Lake Maracaibo (Lundberg et al . 1998, Colinvaux and De Oliveira 2001; Vonhof et al . 2003). Figure 2-2. South America with details of the two main river systems within the Caiman yacare distribution, the Amazon (north) a nd the Paraguy-Paraná (south). Most tributaries outside the distribution ar e not shown for map simplification. Red arrows indicate approximate latitudinal range limits. At that time the continent was separate d from Central America by more than 200 km of open ocean (Coates and Obando 1996). A complete terrestrial connection is estimated to have occurred in the Pliocene (c a. 3 mybp) which led to the Great American

PAGE 34

14 Biotic Interchange (GABI) and consequent radiations of te rrestrial vertebrates (Coney 1982; Webb 1978, 1991). The radiation of the ic hthyofauna has a very different history. The distributions of the curre nt ichthyofauna within the Orinoco, Amazon and ParaguayParaná basins display a high de gree of overlap indicating pr evious connections. Diversity of extant teleost fish suggests that diversit y and distributions of Amazonian fish genera were in place by the Middle Miocene (Albert et al . 2005). Three river turtles, Podocnemis expansa , P. unifilis , and Chelus fimbriatus, are found in both the Orinoco and Amazon drainages. Phrynops geoffroanus has a similar distribution and also occurs in Paraguay-Paraná system (Pritchard and Trebbeau 1984). The Brazilian Shield consists mostly of crystalline granites and hard metamorphic basalts of Precambrian origin. Exposed porti ons have resisted erosion and it strongly influences the drainage patterns over much of central South America (Pareja L. et al . 1978). Numerous rivers have their origin in Shield highlands, in cluding many southern Amazon tributaries. A western Shield extension in southeastern Bolivia results in a weak divide for headwaters draining north to th e Amazon or south toward the río Paraguay. The northern distribution of C. yacare , the Amazon portion, is almost entirely contained within Bolivian borders. The sout hern distribution, comp rising a portion of the Paraguay-Paraná basin, is shared within th e borders of Bolivia, Brazil, Paraguay and Argentina. The Paraguay basin watershed drai ns south off the Brazilian Shield forming the Pantanal and has been described as the world’s largest wetland (Mittermeier et al . 1990). This river continues sout hward, dividing Paraguay in tw o, where it unites with the Paraná. The distribution con tinues south on the río Paraná, also of Shield origin, for about 150 km before reaching what a ppears to be a climatic barrier.

PAGE 35

15 Amazon River Basin Hydrology The northern distribution of C. yacare is contained in the río Amazon basin, all affluents of the río Madera (Madeira in Brazil, see Fig. 2-3). The area measures approximately 1100 km on a NW-SE axis and is about 450 km wide. The major rivers of lowland Bolivia meander northward, coalescing to form the río Madeira at the northern extreme of the country. This river begins at the junction of tw o important tributary systems, the Beni and the Mamoré, and flow s about 100 km before entering Brazil. The río Madeira basin produces a pproximately one trillion m 3 of water each year, which is about one-fifth of the water discharged by the Amazon (Goulding 1981). In South America, a river’s name may change at th e junction of a major tributary or country boundary. This can lead to confusion as many rivers may have multiple names along its course. For example, the río San Julián has at least seven names along its course in central Bolivia. An upland ridge of the Brazilian Shield, th e Serra dos Parecis, presents a barrier just to the north of Bolivia and forms the basin edge for part of the drainage. These uplands frame the northern li mits of distribution of Caiman yacare as well, confining the río Abuná drainage west of the Madeira and th e río Iténez (or Guaporé in Brazil) to the east. Affluents drain south fr om the highlands, with only short distances of good habitat before reaching these two rivers. The río Abuná cuts eastward across a hard rock bed and does not exhibit the large meanders and repeated course changes seen in the rivers to the south which cross softer alluvial soils. The río Iténez, however, is not as severely affected by the Shield material and the meanders mo re closely resemble the southern rivers.

PAGE 36

16 Figure 2-3. The northern distribution of Caiman yacare , found within tributaries of the upper río Madeira system of the Am azon basin. Dashed line marks the southern limit of the Amazon drainage, dotted line is 500m elevation contour. The río Madeira has cut northeast through this resistant material to empty the basin, comprising most of lowland Bolivia. It enters a constricted channel as it crosses this substrate enroute to the Amazon further dow nstream. This stretches from the BrazilBolivia border over 200 km downs tream to the town of Pôrto Velho. The central 125 km

PAGE 37

17 of numerous rapids, grouped under the name ‘Caldera do Inferno’ (Hell’s Caldron), appears to be the northernmost ex tent of the species. Brazaitis et al . (1990, 1993, 1996) list this as an intergradation zone between C. yacare and C. crocodilus based on general morphology and jaw coloration patterns. Vanz olini and Gomes (1979) list a locality record for C. crocodilus from Cachoeira San Antonio, the final rapid on the Madeira near the town of Pôrto Velho. Brazaitis (1 993) lists Humaitá, about 250 km further downstream, as an intergrade locality. He also includes intergrades at Borba on a distribution map, an additiona l 400 km downstream (1990: 113). While most rivers in the Mamoré drainage are confined wholly within Bolivia, a few northern rivers flow northeast from Peru, across the Bolivian Department (=state) of Pando, to unite with the Mamoré (see Fig. 23). These include the rios Abuná, Orthon, and Madre de Dios. This area is generally covered with Amazon hylea and riverbanks support dense gallery forest. These rivers, like most in lowland Bolivia, exhibit immense meanders and oxbow lakes except as mentione d above. Caiman do not typically inhabit the large, forested rivers and, while they ar e encountered there, th ey mainly inhabit the associated complexes of wetlands. These rivers coalesce en route toward the Madeira, draining a vast area. The constricted channe l in Brazil impedes drainage during peak runoff and water backs up the swollen rive rs. This can cause flooding over immense areas in eastern Pando. King and Videz ( 1989) surveyed there and found yacares common on the Beni, Yata and Mamoré rivers near their confluence. The extent of distribution of C. yacare in this northernmost region is poorly documented and little is known about the western limits. Caiman crocodilus (‘ caimán blanco ’ in Peru) occupies habitats in these Pe ruvian affluents further upriver and it

PAGE 38

18 appears to inhibit C. yacare expansion into those areas. Brazaitis (1990) listed the río Negro, a Brazilian tributary of the Abuná ar ea near northwestern Pando as another zone of intergradation with C. crocodilus . This was based on specimens captured near the Brazilian town of Brasileia. Nothing is know n about interspecific interactions in the zones of sympatry and th is needs investigation. No localities have been published for C. yacare in Peru although records exist nearby in the lower río Madre de Dios in Bolivia. Videz (pers. comm.) confirmed the presence of C. yacare on the Madre de Dios in Peru, upriv er of the border town of Puerto Heath. He also noted that upon reaching Pu erto Maldonado, about 100 km upriver, all caiman observed were Caiman crocodilus , ‘ caimán blanco ’. Medem (1983) also reported C. crocodilus from Puerto Maldonado. Medem (1983) reported that yacare caiman were common in the Bolivian sections of the rios Orthon, Madre de Dios and Manuripi during his 1973 survey. King and Videz (1989) also reported localitie s along the Manuripi further west, nearly to the Peruvian bord er. The Pando area was never a zone of high production during the years of in tense hunting for hides, partly due to lower densities of caiman, and in part due to undeveloped infras tructure and difficulties of transporting skins. The current caiman distribution here should thus generally reflect the natural historic pattern. Underlying Brazilian Shield material aff ects the drainage of several rivers in eastern Pando where numerous rapids are fo rmed. These are serious enough to impede river vessel traffic and possibly add to the barr ier effect of the ‘Cal dera do Inferno’ found a relatively short distance downstream. The rapids include the fo rmidable ‘Cachuela

PAGE 39

19 Esperanza’ on the Beni, ‘Cachuela Nueva España’ on the Madre de Dios, ‘Cachuela Banamera’ on the Mamoré, and the ‘Cachuela Madera’ on the Madera among others. The major rivers flowing from the south meander across vast alluvial plains of Andean origin. This large lowland mass of the northern, central and eastern Bolivia occupies nearly 60% of the national te rritory. The Bolivian lowlands comprise approximately 660,000 km 2 and the yacare caiman is found in most wetland habitats there. Few records for Caiman exist above 500 m above sea level (msl) and this approximate contour was chosen for th e included maps where appropriate. The central Bolivian lowlands exhibit very little relief . In general, they slope downward from south to north along a very ge ntle gradient, reaching approximately 90 msl at Manoa in the northern tip of Boliv ia (Zeitum Lopez 1991, see Fig. 2-3). The vast plains area has an average elevation of 200 m, punctuated with a micro relief of almost imperceptible depressions and rises. The elevations are locally known as islas , islands of trees that frequently esca pe inundation during the wet season. The depressions, or bajios , are subject to widespread flooding often formi ng large, shallow lakes over the savannas. These plains are frequently underlain with a clay hardpan that impedes percolation and sheet flow runoff drains much of the fl at landscapes. During th e seasonal drought the receding water eventually collects in isolat ed bodies or in mosaics of articulating waterways. The low-gradient rivers frequently change course, often cutting off large bends to form oxbow lakes, curiches (lakes with floating islands of vegetation), and other bodies of water isolated from the main current (F ig.2-4). These are comm only occurring features

PAGE 40

20 in the landscape and provide good habitat for caiman. The region has several thousand shallow, permanent lakes (Denevan 1966). Figure 2-4. Rio Mamoré, central Beni Departme nt. Repeated change of course creates many isolated bodies of water off the main channel, which provide excellent habitat for Caiman yacare (left). Smaller caños meander across open savannas and, along with associated water bodies, al so provide excellen t habitat (right). Large areas are subject to flooding during th e heavy rains that occur during the wet season. Precipitation typically ex ceeds the capacity of the land scape to absorb or drain it and rivers overflow their banks annually. Th e waters recede as the dry season progresses, although some bodies of water in low-lyi ng areas persist permanently. Flooding often lags up to two months behind the maximum precipitation as the runoff accumulates in downstream areas and backs up the swollen rivers (Denevan 1966). The western lowlands are drained by th e río Beni, and the mighty río Mamoré. They cross the expansive savannas of the Lla nos of Moxos in the central Beni Plains. These rivers frequently flood surroundi ng areas in the wet season, inundating the

PAGE 41

21 numerous interfluvial bajios . In addition, this area ha s a large concentration of floodplains lakes, some of immense size. Ma ny of these retain wa ter all year, except during exceptional drought. The actual limits to local distribution in the southwest appear to be due to a combination of altitude and av ailable habitat. The Andean foothills often rise quickly from the flat foreplains and appropriate ups tream wetland habitats may be restricted well below the theo retical 500 msl altitude limit. The eastern region drains north off the re sistant granite upla nds of the Brazilian Shield. Rivers such as the Itonamas, Blan co and Paraguá feed into the río Iténez (Guaporé in Brazil). The latter river drainage forms part of the distribution boundary in Brazil, as mentioned above, along the north eastern edge. Abundant caiman locality records throughout these lowland regions illustrate the plasticity to ut ilize the varied and locally variable habitats (see Fig. 2-5). The southeastern limit of the Amazon dist ribution is defined by local geography. A weak drainage division is formed by the Ch iquitana Highlands in eastern Santa Cruz. This formation overlays the Brazilian Shield and reaches heights up to 1,250m (Sanabria F. 1990). The land to the west of the highla nds is part of the Amazon drainage. The highlands themselves give rise to seasonal affluents of the río Paraguay that run eastward. Most of the rivers in the Bolivian Amazon have their origin in the Andes. Rapid physical weathering occurs there and the wide variety of substrate materials provides a sediment load rich in minerals (Sioli 1968). Th ese rivers are typically very turbid, with water the color of cafe-au-lait, a nd are termed whitewater rivers (‘ agua blanca’ ). Light penetration is negligible. I have measured S ecchi disc transparency at less than four cm on occasion.

PAGE 42

22 Figure 2-5. Localities for Caiman yacare recorded during surveys of the northern distribution ( -Medem 1983; King and Videz 1989; Brazaitis et al . 1990; Godshalk and King 2002 sample localities for this study). Rivers that flow off the Brazi lian Shield, such as most of the tributaries of the Río Iténez, have reduced sediment load a nd corresponding low nutrient and conductivity levels as well. They are often very clea r, and are known as clear water rivers (‘ agua

PAGE 43

23 clara’ ). The eroded crystalline bedrock sometimes forms sandy beaches along these rivers. Blackwater rivers occur frequently in th e central Brazilian Amazon. Sediment load is low and the water is clear as viewed thr ough a glass container. The water is tea colored from tannins believed to be derived from decay ing forest materials. Light penetration is reduced and the rivers appear very dark or black when seen from above. These rivers are generally acidic and nutrient poor, resulting in a depaupera te fauna (Junk 1975). This is not a common river type in Bolivia. Environment The northern (Amazon) distribution ranges from approximately 9°30' S to about 19°15' S latitudes and therefore the climate is largely tropica l. Due to the absence of relief over most of the distribution lands cape, air masses move unimpeded and form relatively smooth gradients for change s in temperature and precipitation. Temperature The mean annual temperatures vary from about 26° C in the north, to about 24° C in the central region, rising to about 26° C sout heast in the arid Chaco (see Fig 2-6). The warmest months are during the southern hemi sphere summer, generally in November and December. The minimum temperatures are regi stered during June and July when cold polar air masses may sweep north across the pl ains. These fronts bring cold, damp air, which precipitates in a chilling drizzle and ma y remain stationary for a week or more. Near freezing temperatures occasi onally reach as far north as 15 û S latitude.

PAGE 44

24 Figure 2-6. Mean annual temperatures for the area of distribution of Caiman yacare . Isotherms in û C . Precipitation Precipitation is hyperseasonal across all of the lowlands except piedmont areas, with wide variation on length, timing and se verity of the dry season. The majority of precipitation falls in a short period during th e wet season and is usually heaviest from December to February. Several months in the dry season may produce little or no rain at all depending on locality. Most of the preci pitation originates from humid air masses moving south and west from the Amazon basi n. For this reason, the mean annual rainfall averages are generally higher in the no rth, at about 1900-2000 mm, decreasing towards the south and east to about 600 mm (see Fig 2-7). Orographic lifti ng at the Andean foothills increases rainfall locally in those areas.

PAGE 45

25 Figure 2-7. Mean annual pr ecipitation for the area of distribution of Caiman yacare . Isohyets in mm. Great variation in annual pr ecipitation can occur. For ex ample, 1992 proved to be a strong El Niño year. It rained nearly every day for thirty days in May. Trinidad, the capital of Beni Department in Bolivia, has a normal annual precipitation of about 1800 mm, with 260 mm falling in February (41 y ear average). In 1992, almost 800 mm were registered in February alone (SEMENA da ta, 1992), and many meteorological stations had received average annual precipitation totals by May. Vast savanna areas were flooded and thousands of persons affected. Widespread flooding also occurred at this time in the southern distribution of yacare. In southern Brazil and northern Argentina, over 130,000 persons were displaced from their homes by flooding of the Paraguay, Paraná and Iguazú rivers. Pr ecipitation continued sporadic ally through the normally dry months as well causing the inundation s to recede even more slowly.

PAGE 46

26 Soils The vast central lowland soils are relativel y recent, derived from the eroded parent material in the Andes. Rich alluvial soils , deposited during annual flooding, predominate along most watercourses of the lowland plains . However, these soils generally show the effects of extremes in rainfall and temperatur e. The high precipitation of the lowlands has leached away much of the so luble nutrients. Red or pale yellow latosols are found on much of the interfluvial areas (Morales 1990). The elevated temperatures throughout the year facilitate the breakdown of organic soil matter and lead to low cation exchange capacity. These nutrient poor soils are deficient in calcium and magnesium and exhibit low base saturation (Sánchez 1976). They are generally acidic, a nd are suitable for sustained agriculture only with costly fertilizer inputs. Scarce nutrien ts are quickly exhausted and excess aluminum, iron or manganese produ ce toxicity. Fixation of soil phosphorus renders it unavailable and frequently this is a limiting factor (Peducassé 1982). Clearing native vegetation exposes most of the scarce nutri ents in the thin topsoil to quick erosion (Sánchez 1976). As a result, extensive cattle ranching is the dominant land use over the majority of yacare distribution. This has re sulted in minimum negative impact to caiman and, in some areas, increased dry season ha bitat due to construction of cattle tanks. Biota Flora The lowlands of both distribution areas support diverse vegetation resulting from the unique geographical position. Plant spec ies from the Amazon, Chaco and Cerrado regions combine to form mixed assembla ges (Beck 1984). The lowland flora can be broadly divided into four main categorie s: Amazon rainforest, seasonally inundated

PAGE 47

27 savannas, cerrado, and chaco (Solomon 1989) . The wetland habitats utilized by Caiman yacare are predominantly in savanna areas , and forest to a lesser degree. In northern Bolivia, the Amazon hylea covers most of the Department of Pando and part of northern Beni Department. This ar ea has low human densities and has not yet suffered much degradation. The dense forest, po or soils and lack of infastructure have discouraged large-scale settleme nt. Further south, large expa nses of savanna are present in the Beni Plains ( Pampas de Mojos ). Gallery forest generally follows the watercourses, and tree islands ( islas ), are interspersed across the ope n grasslands. The gallery forests along the río Mamoré and tributaries west a nd southwest are extens ions of the hylea, although much of the Beni forests and islas are semi-evergreen (Jones 1980). The origin of savannas is a subject of debate by various authors (Harris 1980, Sarmiento 1984). Evidence indicates that the region has been occupied by indigenous tribes for thousands of years and anthr opogenic fire has been proposed, (Denevan 1966, Hammond 1980, Beck 1984). Once grasslands have been established, fire can maintain savanna conditions and may also enlarge th e area. The effects of climate, seasonal flooding and the nutrient poor, impermeable soils may also be sufficient to prevent forest encroachment (Sarmiento 1984). Alternate theories have proposed more recently citing perturbances by catastrophic floods, tectonic movements or succession from sea or lake beds (Räsänen et al . 1987; Fraily et al . 1988; Räsänen et al . 1995; Colinvaux, and De Oliveira 2001) The large variety of plant associations in the savannas is determined by micro-relief and length of flooding. The savanna plant sp ecies are numerous but large areas are dominated by Paspalum acuminatum , Panicum spp., Rhynchospora spp., Thalia

PAGE 48

28 geniculata , Eleocharis spp., Andropogon sp., and Paratheria prostrata (Haase 1989, Haase and Beck 1989). Dense monospecific palm stands occur in some areas subject to long periods of flooding, usually composed of Copernicia alba or Mauritia flexuosa . The islas do not have uniform vegetation and ha ve not been thoroughly studied, but may represent remnants of gallery forest isolated after rive r course changes (Beck 1984). They are typically dominated by fire resistant, and drought resistant specie s. There is a wide diversity of plant species f ound in these plains regions, important areas for caiman populations. The precipitation generally decreases toward s the southeast, lead ing to an area of semi-evergreen forest in Santa Cruz Depart ment. This area supports the largest human populations in the Bolivian lowlands, and mu ch land has been converted to commercial agriculture for production of cotton, sugar, ri ce, and soybeans. This has occurred in areas less likely to contain good caiman habitat. Fauna The lowland fauna of Bolivia is also di verse, having representatives from the surrounding realms. Approximately 2300 verteb rate species have been described for Bolivia (Ergueta 1990). The Bolivian ichthyof auna is the most poorly known of the vertebrates. The majority of the species found in lowland Bolivia are common to the central Amazon basin and few endemic species have been described. Only a few limited surveys have been undertaken. Terrazas ( 1970) estimated that 340 species are found in Bolivia. This appears to be quite low as a study of only larger fish of the río Mamoré found 280 species present (Lauzanne and L oubens 1985). Ergueta and Salazar (1991) reported about 400 known species and estimated 200 additional species exist in Bolivia.

PAGE 49

29 The diverse terrestrial vertebrate fauna of the Bolivian lowlands is largely Amazonian. Many species have been important historically as sour ces of food, hides and other products. This continues today although generally with diminished intensity and economic importance. Subsistence hunting is st ill a very important as pect of life for many lowland indigenous groups (Townsend 1995, 1996). Mestizo colonists often supplement their supplies with wild game meat, but prefer only a subs et of the wide range of species hunted by the Indians (Redford and Robinson 1987). Large areas of the Bolivian lowland habitat remain relatively intact. In much of these areas, the wildlife component s are also nearly intact. Bu t recent demographics show a trend for rapid, large-scale human expans ion into the lowlands and conservation measures must be taken to reduce the potential impact on wildlife populations. The species of reptiles and amphibians th at inhabit the lowlan ds are not well known to biologists or local inhabitants. Ergueta (1987) reported 252 species of reptiles and 110 amphibians, perhaps only half the number actually present (Ergue ta and Salazar 1991). Few studies have been undertaken beyond making checklists. Turtles are common in diets locally and are the reptiles most widely eaten by people. The large river turtle, Podocnemis expansa , is sought throughout the Amazon basin for meat, eggs and oil. Over-exploita tion has led to a serious decline in numbers (Ernst and Barbour 1989). The smaller, yellow-spotted Amazon River turtle, P. unifilis , is also widely used for meat but remains re latively abundant in ma ny areas in the yacare distribution. Tegu lizards, ( Tupinambis spp.), a nest predator and the large-bodied anaconda ( Eunectes murinus ), a predator of all yacare ag e classes, are commonly seen.

PAGE 50

30 During the past century, crocodilians have b een the most important reptiles in the commercial rural economy. Three of the five species found in Bolivia have been actively hunted with little control until very recently. Sm all-scale poaching still occurs but not at previous levels. Over-hunting caused serious population declines and CITES took steps to persuade the Bolivian government to curtail th e illicit trade. Intern ational controls have proven successful in reducing the negative impact of trade on caiman in Bolivia. The avifauna is also diverse due to the wide range in habitats. Kempff M. (1985) listed 1,512 species and subspecies, about one third of described Neotropical species. Most lowland species are common to the Amaz on basin, but some Sout hern Cone species like the conspicuous rhea ( Rhea americana ) , are widely seen in th e savanna habitats. The Psittacidae is very diverse, represented by 44 species. The lowland landscapes are typically inhabited by a remarkable abundan ce of birdlife. The wetlands are always attended by large storks, he rons and egrets such as Jabiru mycteria , Mycteria americana , Casmerodius albus , as well as other predatory species. Over 316 mammal species have been record ed in Bolivia and the fauna is also predominantly Amazonian (Groombridge and Jenkins 1994). The capybara, Hydrochaeris hydrochaeris , is abundant in many lowland areas. The otters ( Pteronura brasiliensis and Lutra longicaudis ), and several spotted cats, such as jaguars ( Panthera onca ), and ocelots ( Felis pardalis ), also are commonly present. Human Dimension Humans have inhabited the Bolivian lowl ands for more than 10,000 years (Ibarra G. and Querejazu L. 1986). Previous cultures of the central Beni Plains used extensive raised-bed agriculture, similar to the chinampas of Mexico. These organized societies produced hundreds of kilometers of elevat ed pathways and dikes. Denevan (1966, 1970)

PAGE 51

31 estimates the pre-contact population ma y have exceeded 350,000. Contact with Europeans resulted in genera l decline of indige nous populations from disease, slave trade, displacement, and general so cial disruption (Denevan 1991). The high diversity of lowland tribes resu lts from influences of the surrounding geographic areas as seen above in flora a nd faunal groups. Twenty-eight tribes are currently recognized, with perh aps twice that number existing in the sixteenth century (Denevan 1966). Rough estimates in the mi d 1980s calculate about 100,000 indigenous people inhabit the lowlands (Montes de Oca 1989). The expansion into the lowlands by non-i ndigenous people developed very slowly although the pace has greatly accelerated in the last twenty years. Major areas currently remain sparsely inhabited in terms of individu al per square kilometer as most growth has occurred in urban areas. Th e Beni Department (213,564 km 2 ) grew from 32,200 (1900) to an estimated 280,000 in 1990 (Montes de Oca 1989). Due to poor soil conditions and other factors outlined above, extensive cattle ranching has developed as the principal land use. By 1988, about 60% of the population was c oncentrated in urban centers with more than 2,000 people (Muñoz Reyes 1991). People con tinue to migrate from the countryside and become more urbanized. In Santa Cruz Department (370,621 km 2 ), large-scale popula tion increases began in the 1950s after the comp letion of the Cochabamba-Santa Cruz highway. This coincided with the Agrarian Reform of 1952 and resulted in a tremendous expansion of commercial agriculture. Alt hough the population is more than 70% urban within the Department, mechanized agri culture has converted substa ntial areas of the natural landscape in the western and central porti ons of Bolivia's largest department.

PAGE 52

32 Paraguay-Paraná River Basin The southern distribution of Caiman yacare is found within the drainage of the ríos Paraguay and Paraná. This di stribution area is approximate ly 950 km long on the northsouth axis, with a width of about 250 km. Distribution is bounded on the north and east by Brazilian Shield uplands, on the west by th e Andes and the xeric Gran Chaco and in the south by climatic limits for the species (see Fig. 2-8). Hydrology The upper río Paraguay basin presents some of the best habitat for the species and results in astounding densities. The río Paraguay drainage lies just to the east of the headwaters for the Amazonian río Iténez, orig inating in the same Shield formation known as the Serra dos Parecis. An arc of steep upl ands drains quickly s outh and west, forming about 140,0002 of wetlands, the Pantanal. The flat landscape inundates to form a complex hydrographic matrix peculiar to this area. The Pantanal is considered the world's largest wetland where diverse wildlife occurs in abundance (Alho et al . 1988). A series of large lakes and marshes straddle the border with Bolivia, and the area frequently floods in the wet season (Mittermeier et al . 1990, Crawshaw 1991). The río Paraguay flows along the western side of the area, with most of the Pa ntanal in Brazil. As in much of the yacare distribution, the hyperseasonal rains cause periodic flooding followed by severe drought. The Chiquitano Highlands, mentioned above, is an isolated formation found in southeastern Santa Cruz Department of Bolivia . It is composed of Cambrian Era material and overlays the Pre-Cambrian strata of th e Brazilian Shield (Montes de Oca 1989.) The resulting hills rise above th e surrounding plains and form the western boundary for the upper Paraguay drainage.

PAGE 53

33 Figure 2-8. The southern distribution of Caiman yacare , found in the río Paraguay and lower río Paraná systems. As the río Paraguay flows south, a parallel ridge of Shield uplands limits wetland habitats east of the river. Most of the tr ibutaries are short and flow through uplifted landscapes that confine surface water to rivers and streams (Scott et al . 1990). The principal yacare habitats occur in the central river corridor as it meanders southward. Low meadows occur on both east and west side s of the river, forming large marshes during the wet season.

PAGE 54

34 The xeric Gran Chaco is a flat alluvial pl ain west of the river. Situated in an Andean rain shadow, it becomes increasingly dry as it extends west towards Bolivia. Occasional rivers dissect it and caiman dist ribution there is limited by the presence of water in the dry season. Large deposits of sa line soils, especially in the Upper Chaco, sometimes render areas unsuitable as evaporating surface water produces extremely high salt concentrations. King et al . (1994) reported salinity of 73‰ at the Represa Mosquito in the central Chaco of Paraguay. Wet season pr ecipitation dissolves the layer of surface salts that are re-deposited dur ing the annual dry-down of wetla nd habitats. As a result of this process, Chaco ponds, streams and rivers hold water with salt c oncentrations varying locally from slightly brackish to very sa lty. Medem (1983) states that yacares avoid saltwater although no threshold has been establ ished. A large part of this drainage in western Paraguay, and penetrating into Bolivia, is occupied by thorny scrub which is dry most of the year and contains little or no ha bitat suitable for yacares. At least two ranches in this area have introduced yacares in artif icial cattle tanks that maintain water during the dry season (King et al . 1994). The río Pilcomayo is the largest river en tering from the west and localities have been recorded along its length, including si tes in Bolivia (Medem 1983; King and Videz 1989). Flow ceases at the height of the dry s eason and it becomes a series of disconnected pools throughout its length. Th e river bisects the largest marsh in Paraguay, Estero Patiño, which occupies a low area of about 1,500 km2. It occurs about 270 km above the river mouth and is an area of high caiman densities. During the wet season, the river flows along its course and a great marsh also forms at the union w ith the río Paraguay. This is also the case further downstream at the junction of the Paraguay and Paraná.

PAGE 55

35 Figure 2-9. Localities for Caiman yacare recorded during surveys of the southern distribution ( Medem 1983; Brazaitis et al . 1990; Scott et al . 1990; Waller and Micucci 1993; King and Videz 1989; King et al . 1994; Godshalk and King 2002, samples used in this study; – Localities for additional samples used in this study, see text for details). The río Paraná provides good habitat for about 200 km above that junction, to about the level of the town of Encarnación. The floodplain constricts further upriver limiting habitat possibilities. Th e rocky bed and shores plus the swift current provide

PAGE 56

36 little for caiman in this area. Medem (1983) reported localities for C. yacare in the upper river, above Iguazu Falls, based on a public ation by Krieg (1948). Krieg described the animals as spectacled caiman (‘ brillenkaimane ’ in German) with no species name and Medem assigned them to C. yacare as the most likely species. No one has since investigated this area and I consider the species assignment unlikely. The Estero Iberá wetland region just south of the Paraná is also an important habitat area. The río Salado, further downstream, appears to be the southe rn limit to the dist ribution according to Medem (1983). This appears to coincide with the 15 û C minimum average temperature tolerance frequently used for most crocodilians. The more cold resistant C. latirostris distribution extends slightly further south (see Fig. 2-12). Passive migration occurs as yacares are swept downstream, sometimes as far as Buenos Aires (Medem 1983, Waller 2003) but no evidence of breeding populations south of the río Salado has been reported. Environment The southern distribution (rios Paraguay -Paraná) of yacare caiman ranges from approximately 14°30' S to about 30°00' S latitu des with a tropical to temperate climate gradient. The Gran Chaco is a continuation of the alluvial plains found north in Bolivia, so there are many similarities with the norther n distribution area. As in the northern area, the lack of relief over most of this distri bution is responsible fo r the relatively smooth gradient changes in temperature and precipitation. Temperature Annual mean temperatures range from 26 û C in the northern areas to about 15 û C at the southern extreme (see Fig.2-6). The isothe rm shapes reflect the combined effects of altitude and latitude. Maximums may exceed 40 û C during the warm months of December to February. Winter fronts move through as described above for the northern distribution,

PAGE 57

37 but here the effect is more pronounced. Fr ost can be observed in all parts of the distribution south of the Pant anal (Fariña Sanchez 1973). Precipitation As seen in the temperature regime a bove, precipitation forms smooth gradients continuous with those observed in the Amaz on (northern) distribution area. However, local variation can be remarkable. Rainfall is hyperseasonal over the entire range with a dry season May to September. There is virtua lly no precipitation in the months of June, July and August. Most rainfall occurs from October to April, the austral summer. The upper río Paraguay basin drains Brazilian Sh ield uplands which receive >1200mm. The central floodplain rece ives 1100-1400 mm annual precipitat ion as it courses south (see Fig. 2-7). Rainfall decreases west across the xeric Gran Chaco to a minimum of 400 mm in the Upper Chaco of western Paraguay, Bolivia and Argentina. Even in there, wet season rainstorms may be torrential a nd cause temporary widespread flooding. Soils The soils over the southern distributi on can be divided into two main groups according to origin. West of the río Paraguay lies the great alluvial plain of the Gran Chaco. It is a continuation of the Bolivian pl ains of the northern y acare distribution and has similar soil characteristics. Poor drainage occurs over most of this large, flat area due to lack of relief and limited percolation. Much of the area was covered by marine intrusions for a period afte r the initial Andean uplift (Räsänen 1995). Resulting salt deposits affect soils and make much of the su rface water brackish. So ils are nutrient poor but qualities generally improve as one moves fr om west to east toward the río Paraguay. Typically the sand content decr eases and organic matter increa ses, although alkaline soils are also increasingly more common. In the ce ntral river corridor, persistent flooding in

PAGE 58

38 many areas has produced nutrient poor hydromor phic soils that are se verely leached. Due to the difficult soil problems and severe dry seasons, agriculture has limited possibilities there. As seen in the northern yacare dist ribution, extensive cattle ranching is the dominant land use over most of the southern distribution of C yacare . The soils and topography east of the río Paraguay have a completely different history. The flat floodplain forms a narrow belt that parallels the river, characterized by poor drainage and large flooded marshes. Th ese hydromorphic soils have been leached by the intense precipitation, a nd dominant land use here is al so extensive cattle ranching. The Brazilian Shield forms the undulati ng topography further east and underlies Permian-Triassic deposits. More recent overlay formations are seen on these uplands as one moves south. Erosion of the crystalline and sandstone upstream Shield layers has produced sandy, nutrient poor soils down slope. Biota Flora The Pantanal wetland of the upper río Pa raguay is a complex of wooded savanna with elements from surrounding areas: Am azon, Cerrado, Atlantic Forest and Chaco. Gallery forest or semi-deciduous forests ar e found along watercourses. This pattern is also seen along the yacare distribution sout h in the lower Chaco. Inundated palm savannas ( Copernicia alba ) occur as described for Boliv ia. The woody vegetation reaches 15-25 m height and the canopy is fairly open. The forests are often associated with Panicum grasslands. Common speci es are in the genera Schinopsis , Prosopis , Tabebuia , and Aspidosperma . Most are found on very slight ri ses above the surrounding savanna landscape that provide slightly drier conditions. Vegetati on height varies with average annual precipitation, decreasing westward towa rd the rain shadow in the upper Chaco.

PAGE 59

39 The xerophytic species possess in creased leaf thickness, spin es and drought tolerance. Low, dense thorn scrub forests with halophytic shrubs and cacti are found to the west in the Upper Chaco. Large grass savannas occur eas t of the río Paraguay and great open marshes form in the low areas. Forests line watercourses and the vegetation becomes higher and more dense further east with increasing elevation a nd precipitation. There is a relatively rapid transition from suitable lowland caiman habitat to the Cerrado environments of the hilly uplands. Fauna The vertebrate fauna has a large Amazonian influence. The ichthyofauna contains many species common to the Amazon basin, pos sible remnants of a connection with Lago Pebas (Lundberg et al . 1998). In the Pantanal, 262 species of fish have been reported (Coutinho 2000). A total of 645 species of birds have been reported in Paraguay by Hayes (1995) and over 130 mammal sp ecies are known (Coutinho 2000). These include all of the major mammalian and avian predators of yacare caiman seen in Bolivia. In general terms, the two distribution areas are very similar in their faunal compositions. Human Dimension Indigenous groups occupy much of the northern yacare distribution and have inhabited this area for thousands of years. In the southern distribution area AmerIndians continue to inhabit large areas of the Paraguayan Chaco. Post-Conquest diseases and displacement by settlers have taken a severe toll in Braz il, Argentina and in eastern Paraguay. Cattle ranching has occupied much of the better hunting areas and forced tribal groups into the drier, more inhospitable ar eas of the Upper Chaco where neither largescale agriculture nor ranching is feasible. Well perforation has pushed these development

PAGE 60

40 limitations westward but subsurface saline aquife rs have frequently frustrated endeavors in much of the arid west. In general, land use over much of the s outhern distribution has minimal negative impact on Caiman yacare populations. Some activities, such as water impoundment and raised bed road construction, actually increase dry season habitat for caiman. The largest negative impact to the sp ecies in both north a nd south distribution areas comes from commercial hide hunting (see Chapter 3). The Caiman Species Complex The Neotropical crocodylians of the Allig atoroidea are collectively referred to as ‘caiman’. The group is generally regarded as monophyletic in origin, although the interspecific relationships are not clear. I obtained samples of all the caiman species in order to gain better understanding of the group at a molecular level. Caiman crocodilus The yacare caiman appears most closely related to the common caiman, Caiman crocodilus . The yacare is listed as a subspecies, Caiman c. yacare , by many authors. The common caiman occurs from southern Mexi co and Guatemala through Central America and into most lowland habitats of northern South America (see Fi g. 2-10). With such a large geographic distribution, some regiona l variation is expected and has been documented. Currently, three or four subs pecies are commonly considered by most authors (Medem 1981, 1983; Webb et al . 1987; Gorzula and Seijas 1989; King and Burke 1989; Ross 1989; Ross 1998). Medem (1981) conduc ted a thorough literature review and supported four subspecies. However, he argued for use of the name Caiman sclerops (Schneider 1801). This has not been adopted by subsequent authors.

PAGE 61

41 Figure 2-10. Distribution of Caiman crocodilus : C. c. fuscus in the north, and C. c. crocodilus in the south (the limited C. c. apaporiensis distribution is not shown). – localities for samples used in this study. – localities for published sequences included in analyses for this study (see Farias et al . 2004) The Central American population is sometimes considered as C. c. chiapasius (Bocourt 1876). Populations in the Pacific drainages on the narrow strip of coastal lowlands of Colombia and Ecuador ar e also assigned to this subspecies. C. c. fuscus (Cope 1868) is often given to populations di stributed in the Atlantic drainages of Colombia and western Venezuela. No defi nitive systematic work has been done to separate the two subspecies or to establish firm geogra phic boundaries. Some authors collapse these two into one subspecies , as I have done for simplicity, with fuscus taking historic precedence. Samples obtained fr om some of the northernmost populations, located in Honduras, were used in this st udy as comparative mate rial (see Fig 2-11).

PAGE 62

42 Figure 2-11. Simplified map of Honduras with localities of Caiman crocodilus fuscus samples used in this study. Samples collected during a CITES surveys of Atlantic drainage rivers and coastal lagoons (see King et al . 1990) and used in this study. Caiman c. crocodilus (Linnaeus 1758) occupies mo st of the Orinoco and Amazon basins. It is also reported from Atlantic drainages of eastern Venezuela, Guyana, Suriname and French Guiana. No determinati on has been made where the western coastal limit of C. c. crocodilus meets the eastern coastal limit for C. c. fuscus . The slender snouted C. c. apaporiensis was described by Medem (1955) from an ‘isolated’ population above waterfalls on the río Apaporis. Subsequent information shows this may be part of a larger cline of longi rostrine caiman reaching from central Colombia through the ‘ llanos ’ savannasof Venezuela (Ayara güena 1983, Gorzula 1994). The validity of this subspecies may be in doubt. Continuous clinal variation may also explain differences between the Central American and northern South American forms. The relationships between the various subspecies remain unclear.

PAGE 63

43 Caiman latirostris The remaining four species of caiman are sympatric with C. yacare . The broad snouted caiman. C. latirostris ( overo or choco in Bolivia,’ jacaré do papo amarelo ’ in Brazil, ‘ jacaré mariposo ’ in Paraguay, ‘ jacaré overo ’ in Argentina ), has the largest distribution overlap of the caiman species group (see Fig. 2-12). The two species are separated on a finer scale by habitat pr eferences (Medem 1983, Waller and Micucci 1993). Waller (1987) states that C. latirostris is more successful at colonizing ephemeral water bodies. Figure 2-12. Distribution of the broad-snouted caiman, Caiman latirostris (blue) in South America, showing areas sympatric with C. yacare (green). locality for samples used in this study. It can attain up to 3.5 m in length and th e finished tanned belly hide is smoother than the other members of the genus (Ro ss and Magnusson 1989). H unters report that C. latirostris was never abundant in the northern distribution of C. yacare (King and Videz 1989, Godshalk, unpublished data). It remains very scarce throughout Bolivia, the only

PAGE 64

44 portion of its range in Amazonía. It is relativ ely common in some areas of its distribution and shows a great plasticity of habitat requireme nts. It also is more cold tolerant which explains its distributi on further south than C. yacare . Melanosuchus niger The black caiman, Melanosuchus niger , is the largest of the caiman species ( ‘caimán negro’ in Bolivia, ‘ jacaré açu ’ in Brazil). The maximum length is widely accepted as 5 m and there are reports of up to 7 m (Medem 1983; King and Videz 1989; Thorbjarnarson 1998). The wide distribution includes most of the Amazon basin and some Atlantic drainages of the Guianas (see Fig 12-13). Black caiman are sympatric throughout the northern distribution of C. yacare and the two species are often found together in the same lagoons. Due to the large adult size, the black caiman is considered the most valuable of the caiman species for the hide trade. The smoot h ventral osteoderms permit glazing of the finished hide. The pitted osteoderms of yacare and common caiman prevent this glossy finish and therefore have lower value. Inte nse hunting pressure from 1940 to about 1975 reduced M. niger populations throughout the range. Recovery has been very sporadic and today it still remains very rare throughout much of its distribution. Depletion of the M. niger and Caiman latirostris stocks led hunters to turn to the more abundant but lower value C. crocodilus and C. yacare . Sympatric Caiman species often utilize the ‘vacant’ habitat after removal of dominant Melanosuchus and this may inhibit recovery of populations.

PAGE 65

45 Figure 2-13. Distribution of the black caiman, Melanosuchus niger (blue) in South America, showing areas sympatric with C. yacare (green). locality for samples used in this study. – localities for publishe d sequences included in analyses for this study (see Farias et al . 2004) Paleosuchus The genus Paleosuchus , the smooth-fronted caimans, is comprised of two South American species that are sympatric with C. yacare (‘ cocodrilo’ in Bolivia, ‘ jacaré pagua ‘ and ‘ jacaré coroa ’ in Brazil, ‘ jacaré itá ’ in Paraguay). Both are small, very similar in appearance and are poorly known by biologists a nd local mestizos (Medem 1983; King and Videz 1989; King and Godshalk 1997). These diminutive caiman frequently inhabit small whitewater and bl ackwater streams and wetlands in marginal areas (Godshalk 1982,1998a, 1998b; Magnusson 1989). They are the most terrestrial of New World crocodylians, w ith the possible exception of the Cuban crocodile, Crocodylus rhombifer . The well developed osteoderms throughout the hide, strongly keeled on the dorsum, negate any commercial value. They are taken preferentially by some Indigenous groups as a food item

PAGE 66

46 The dwarf caiman, P. palpebrosus , also known as CuvierÂ’s smooth-fronted caiman, has a wide distribution in north and central South America (see Fig 12-13). It attains a maximum length of 1.8 m in Surinam (Ouboter 1996). SchneiderÂ’s smooth-fronted caiman, P. trigonatus , also has a distribution that overlaps part of the area for C. yacare (see Fig 12-14). This species is slightly larger and can attain 2.3 m in total length (Medem 1981). Little is known about the basic biology of either species. They occupy a wide variety of habitats and are often f ound in the vicinity of other caiman species. Paleosuchus appears to be largely a terrestrial pred ator. Details of interspecific resource partitioning are poorly understood (Godshalk 1998a,1998b). Fig. 2-14. Distribution of the Dwarf caiman, Paleosuchus palpebrosus (blue), showing areas sympatric with Caiman yacare (green). locality for samples used in this study.

PAGE 67

47 Fig. 2-15. Distribution of the Schne iderÂ’s smooth-fronted caiman, Paleosuchus trigonatus (blue), showing areas sympatric with Caiman yacare (green).

PAGE 68

48 CHAPTER 3 NATURAL HISTORY AND BASIC BIOL OGY OF THE YACARE CAIMAN, Caiman yacare Introduction The evolutionary history of crocodili ans stretches over more than 200 million years, from the late Triassic (Buffetaut 1979, 1989; Brochu 2003). The basic crocodilian form, so successful in wetland habitats, evol ved early in their hi story. It has been conservatively retained with relatively little change through time leaving extant crocodilians to appear as "living fossils " (Meyer 1984). This may be misleading as crocodilians have evolved continually over ti me and the modern forms generally have only very recent fossil records (Brochu 2003). The continued persistence of crocodilians over long geologic periods, incl uding their survival through some mass extinction events, is a tribute to the superb adaptations of Crocodylia. Three major lines evolved leading to the extant forms: true crocodiles (Crocodyloidea), gavialoids (Gavialoid ea), and the alligators and caimans (Alligatoroidea)[per Brochu 2003]. These lineag es were distinct by the end of the Mesozoic (Brochu 2003). Alligatorids have a long fossil history dating from the late Cretaceous. The Tertiary radiations were La urasian, with a North American center and additional records from Europe (Taplin and Grigg 1989). It is belie ved that a precursor form gave rise to the sister taxa Alligatorinae and Caimaninae. The first specimen for the genus Alligator appears in the fossil record abou t 35 mybp, during late Eocene (Brochu 1999, 2003). Large gaps exist in the fossil hi story on the development of the caiman

PAGE 69

49 lineage. Brochu (2003) describes the knowledge of the caiman evolutionary history as the least complete of all crocodilians. At some point in time, the lineage leading to the caiman group is believed to have dispersed to South America. The first alligatorid fossils in South America appear in formations dating to the Paleocene. Broc hu (2004a,b) estimates the alligator-caiman divergence at about 64 mybp in the lower Paleocene. Eocaiman cavernosus first appears in strata in southern Argentina from th e early Eocene, about 50 mybp and may be the precursor to the modern caiman species (L angston 1965; Brochu 2003). The fossil record provides no clear information about the relations hips or recent past of the modern caiman genera. The group appears to have exclusive South American radia tion, with a recent invasion of Central America by Caiman crocodilus . Langston (1965) described a close relationship of Melanosuchus and Caiman genera and likely derivation from Eocaiman. Some researchers believe this group is monophyletic and the genus Melanosuchus is not justified. No fossil record has been discovered for the genus Paleosuchus . While definitely a member of the Caimaninae, the two Paleosuchus species are distinct from the rest of the group in many features and are generally considered more primitive. Both biochemical and morphologi cal analyses place the Paleosuchus species as most distant from the rest of the group (White and Densmore 2001; Brochu 2003). A complex combination of climate, hab itat, and zoogeography has resulted in a diversity of caiman in central South America as described in Chapter 2. Five sympatric alligatorid taxa occur within the yacare dist ribution and interaction with a sixth species, Caiman crocodilus , occurs in the northern range limits (Medem 1983; King and Videz 1989; Brazaitis et al . 1990; King, pers com.).

PAGE 70

50 The world stocks of the more valuable, so ft-bellied ‘true’crocodile species declined from unsustainable hide hunting beginning in the 1920s. These species are large, slowmaturing crocodiles that experienced severe population declines wit hout recovery. By the 1950s, tanners could not get enough of the high qu ality, ‘classic’ grade skins to maintain year-round operations. Caiman hunting increa sed to fill the void and stocks of Melanosuchus niger and Caiman latirostris were quickly depleted. Shortly following this, populations of Caiman yacare and C. crocodilus were affected throughout their distributions. As a result , tanning technology changed to utilize the rough belly osteoderms. While traditional glazing was not possible, caiman hide s were given a shine with the application of polyurethane. The ma rket also shifted to predominantly use the softer, but smaller, flank pieces. The trade of endangered crocodilians was curtailed thr ough international agreements and trade controls. The US Enda ngered Species Act included crocodilians in 1968, and CITES (Convention on International Trade in Endangered Species of Wild Fauna and Flora) began to operate in 1975. In most crocodilian species, populations have recovered to some degree but are still banned from the trade. The smaller, less valuable but more abundant crocodilian species have been substituted in the hi de trade since then. The Yacare Caiman Taxonomy of the caiman complex has not been firmly established among investigators. Félix de Azara first describe d the yacare caiman from his travels along the río Paraguay in 1801. Daudin named it as the species Crocodilus yacare in a publication in 1802. He also compared and describe d the broad-snouted caiman, naming it Crocodilus fissipes and recognizing the diffe rences between it and Crocodilus yacare . Later both species were assigned to the genus Caiman Spix 1825 by Müller and Hellmich

PAGE 71

51 (1936). Since that time, yacare has been described as a subspecies of Caiman crocodilus by various authors (Schaller and Crawsh aw 1982; Brazaitis 1986; Groombridge 1987; Gorzula and Seijas 1989; King and Videz 1989; Ross and Magnusson 1989), and it also has been considered a full species by ma ny others (Medem 1960, 1983; Thorbjarnarson 1991a; King and Burke 1989, 1997; Ross 1998; G odshalk and King 2002). Adding to the confusion, two subspecies within the yacar e distribution were proposed from material based solely on commercial sk ins without known localities, C. c. mattogrossiensis and C. c. paraguayensis (Fuchs 1971, 1974). Even though these subspecies were included in the CITES Identification Manual (1983), they ha ve been criticized on both biological and taxonomic bases (Frair and Behler 1983; Medem 1983; King and Burke 1989, 1997) and are largely ignored by researchers. Scalation patterns and gross morphological characteristics do not discretely separate Caiman crocodilus and Caiman yacare , and large overlaps exist. Busack and Pandya (2001) used a Principal Component Analysis to evaluate morphological measurements from C. c. crocodilus , C. c. fuscus , C. c. chiapasius and C. yacare . Using a combination of tests they found that yacares were correctly identif ied 100% of the time, but the subspecific identities of the co mmon caiman were much less certain. Recent biochemical and molecular information s upports these findings (Densmore and Owen 1989, White and Densmore 2001). But in a comparison of skull morphometrics and ontogenetic growth, Monteiro et al . (1997) could not support a di stinction. In any case, under the rules of nomenclature, DaudinÂ’s or iginal description (1802) holds precedence until a formal synonymy or a species or s ubspecies designation has been published.

PAGE 72

52 Male yacares can reach nearly 3.0 m total length (TL), while females rarely grow more than about 1.8 m TL (Medem 1983; Craw shaw 1987). There is great variation in color and the animals generally darken with age. Mature yaca re caiman typically are dark brown, gray or black dorsall y, lightening to medium gray laterally, and light gray to yellowish ventrally. Older, la rger animals may have a unifo rmly dark gray appearance with a lighter ventral surface and are called ‘ lagartos negros ’ in Bolivia (see Fig. 1-1; Medem 1983). The species common name in Argentina is ‘ jacaré negro ’ (=black caiman), ‘ jacaré tinga ’ in Brazil, and jakaré hû in Paraguay. A distinguishing feature of C. yacare is the presence of dark mandibular and maxillary blotches (Fig. 3-1). This coloration is also commonly found in Melanosuchus and C. latirostris as well but rarely found in adults of the Caiman crocodilus subspecies. The pattern is variable among i ndividuals and is often different in number on each side of the head. This blotching was used as a diagnostic character by Brazaitis et al . (1990, 1993) when assessing the zone of intergradation between C. crocodilus and C. yacare near the Bolivia-Brazil border. Young emerge with strongly contrasting tr ansverse bands on the tail. These often darken with age and are lost, although some a dults retain this pattern. Some individuals remain lighter overall and exhibit yellow, brow n or red tones, or may be strongly spotted laterally. These lighter colored animals are sometimes called lagartos blancos by hunters. Local belief holds that the blancos live in moving water, while the negros inhabit quiet water and lakes (Medem 1983).

PAGE 73

53 Figure 3-1. Adult male Caiman yacare from the río Mamoré basi n (left). Note strongly contrasting mandibular markings characte ristic for the species. Adult male Caiman c. crocodilus from río Apure basin in ce ntral Venezuela (middle). Mandibles lack any markings, color is uniform. A similar appearance is seen in C. c. fuscu s in Mexico (right). Movement Seasonal movements are dictated by the fluc tuations of the a nnual weather cycles. The hyperseasonal precipitation and resulting flooding affects habita ts throughout all of the distribution as mentioned above. During high water, caiman disperse into all available habitats. As the dry season pr ogresses caiman typically move with the receding water, often returning to the same dry season locat ion. In large savanna areas of the Mojos Plains in Bolivia, the Pantanal, and cent ral río Paraguay corri dor, the lakes, ponds wetland areas are severely reduced in the dry season drought, leading to great concentrations of caimans (see Fig 3-2). This is also very similar to the situation seen with C. c. crocodilus in the Venezuelan Llanos. With the onset of the wet season, caiman begin to disperse and reclaim the broade r landscape and the cycle begins again.

PAGE 74

54 Figure 3-2. Peak dry season concentration of adult Caiman yacare in the Bolivian Pantanal. Basking at mid-day (left) and eyeshine reflected at night (right). Over 700 large caiman were counted in this artificial cattle tank that measured approximately 85 m x 105 m. An interesting survival strategy for copi ng with exceptionally hot or dry periods has evolved in some crocodilian species. Caiman are reported to estivate in various parts of its range. The behavior is apparently wi despread but exhibited by a relatively small number of individuals in a given dry pe riod. Estivation has been described for the common caiman in Brazil (Bates 1892), in Colombia (Chirivi-Gallego 1973; Medem 1981), and in Venezuela (Sta ton and Dixon 1975). I observed C. crocodilus almost completely hidden under the cracked mud of dried pond beds in Venezuela. This behavior allows some caiman to avoid lim itations of available dry season habitat. Estivation was noted by Eder (1791) and Cardús (1886) for yacares in Bolivia but has not been reported in modern literature. Torpid caiman are easily killed by hunters and it is possible that there was a reduction of animals that displayed this behavior. Thorbjarnarson (1991a) found that buried caiman were mostly small males, perhaps

PAGE 75

55 subordinate individuals escaping agonistic encounters. Estivati on behavior in the common caiman is well known to the llanero cowboys of Venezuela where populations were not severely reduced. In contrast, most Bolivians th at I interviewed, primarily younger hunters and lowland inhabitants, had ne ver witnessed this fo r yacare. Perhaps the loss of the estivating subset of yacare populati ons explains the current situation. A similar process may have occurred with Orinoco crocodiles ( Crocodylus intermedius ) in Venezuela. These large and once numerous cr ocodiles estivated commonly in historic times but appear to do so rarely at present (Godshalk 1978). Similar seasonal movements occur in bot h the yacare and common caiman for the wet season dispersals and dry season concentrations. Movements of yacares in the Pantanal were first investigated by Schall er and Crawshaw (1982) for the dry season. Most animals moved less than 4 km, with 9.4 km the greatest distance. Generally, research indicates less movement and smaller home ranges than seen in larger crocodilian species. Caiman crocodilus displays similar seasonal move ment but likewise on a small scale (Gorzula 1978; Ayarza güena 1983; Ouboter and Nanhoe 1989; Thorbjarnarson 1991a). Radio tracked animals tended to return to dry season sites or remain in the general area. Intraspecific interactions may also initiate long-range or overland migration. In Venezuela I witnessed Caiman crocodilus during nocturnal migrations over 1 km from the nearest water in the dry season. Young may accompany females during these movements, possibly seeking areas safe from cannibalism by dominant males. Thorbjarnarson (1991a) reported that young of several size classes were led overland by an adult female. Ruiz (1988) states that yacare caiman commonly undertake active

PAGE 76

56 overland forays. This overland capacity means that they may pass from one drainage to another under certain conditions (Sch aller and Crawshaw 1982; Medem 1983). There are ontogenetic shifts in habitat requi rements and use that result in a suite of movements. It begins with hatching and moveme nt to protective vege tation at the water’s edge. Hatchlings are often found together in mixed crèches at times with multiple protective females. Later, as the caiman grow they change location with the seasons. The caiman segregate by body size and wetland char acteristics. Finally, subadult males may face forced migration to evade mature terr itorial, or cannibalistic, males. Of the 700 caiman in the tank shown in Fig. 2-2, near ly all were >1.5 m and none were <1 m. Ontogenetic food shifts also occur with ca iman growth. This entails changing habitat preferences as new prey items enter the diet. A large shift toward vertebrates occurs at about four years old (Aquino-Otriz 1988). Interspecific conflict may also initia te immigration. Anecdotal information indicates that the larger Melanosuchus dominates habitat selection and interactions with yacare , often excluding the smaller yacare fr om waterbodies. However, once hunting removed black caiman from an area, yacares quickly invaded the vacated habitat. The resulting dense adult yacare population whic h will prey on young caiman moving into the area, may pose a barrier for re-establishment of black caiman at a later date (Magnusson 1982; King and Videz 1989). Similar inform ation has been reported for depleted C. latirostris populations in Argentina (Muniz Saaved ra 1983). This process has contributed to the slow recovery of th e endangered Orinoco crocodile, Crocodylus intermedius , (Thorbjarnarson and Hernandez 1990) and the American crocodile, Crocodylus acutus , (Seijas 1986) in Venezuela.

PAGE 77

57 Movement by passive migration must not be overlooked. Caiman are present in all the larger rivers though yacare is not primarily a riverine species. The fluvial network ultimately interconnects most of the habitats and, over the course of time, individuals move downstream with the flow. For the Amazonian populations this represents a northward flow and intergradation with Caiman crocodilus along a portion of the río Madeira in Brazil. For the southern di stribution, individual s moving past the 30 û S latitude find environmental st resses beyond the species limits. Reproduction The reproductive cycle for Caiman yacare was investigated in detail by Coutinho (2000) in the Pantanal. The speci es follows the general pattern of elaborate courtship and mating behaviors outlined for Caiman crocodilus by Garrick and Lang (1977) and Staton and Dixon (1977). Initiation of re production varies w ith locality and appears selected for timing near the beginning of the seasonal ra infall. King and Videz (1989) reported that yacares in eastern Bolivia court and mate in September to November, then nest and lay eggs in November-December. Some indivi duals may oviposit as late as February. Lovisek (1977) also found nesting in November and December in the northern distribution. Crawshaw (1987) and Coutinho (2 000) found that courtship began in August and peaked in November, with egg laying from late December to February in the Pantanal. Caiman nest earlier in northern Boliv ia, from August to November (King and Videz 1989), while Ruiz (1988) found yacares in southern Beni Department lay their eggs in December and January. Medem (1983), c iting a second source, stated that nesting in Bolivia extended from the end of August to November and hatching occurred from December through January. Nesting coincides with high water levels with several

PAGE 78

58 possible benefits: good food availability, post-dry season dispersal which provides maximum nest site distribution, higher humidity and ambient temperatures for egg development, lower chances for nest predat ion by terrestrial vertebrates in flooded landscape, and reduced intraspecific compe tition affecting hatchlings (Crawshaw and Schaller 1980; Thorbjarna rson 1991a; Coutinho 2000). Yacares use the surrounding vegetation to build a mound nest that measures about 40 cm high and approximately 90-120 cm in diameter (Ergueta and Pacheco 1990). It may be constructed on floating mats of ve getation as well as dry ground. Medem (1983) stated that 23 to 41 eggs are laid by Boliv ian yacares, with a mean of 33.6 eggs per clutch. Coutinho (2000) found 15 to 41 eggs pe r clutch. He reported a 21.5% clutch to body mass ratio, the highest recorded for any crocodilian. The duration of incubation is about 65 days (Crawshaw 1987). The female yacare typically attends the nest during incubation, as also found in the other Caiman species. I witnessed this repeatedly during my investigations in Venezuela and Bolivia. Brazaitis (1989) and Crawshaw ( 1991) have shown that human disturbance of the nesting female can lead to decreased attendance and an increase in nest predation. Continuous poaching or even poorly timed ha rvest can affect the overall reproductive success of surviving females. At the end of incubation, unhatched young vo calize from within the egg and trigger a nest opening response in adult Caiman (Staton and Dixon 1977). Both sexes have been observed to assist the hatching process in C. c. fuscus in captivity (Alvarez del Toro 1974), although this has not been observed fo r yacares. However, Cr awshaw and Schaller (1980) reported deliberate nest opening by female yacare caiman.

PAGE 79

59 The young remain in pods for weeks to mont hs after hatching and may be actively defended by an adult caiman (Staton and Dixon 1977; Ayarzagüena 1983; Gorzula and Seijas 1989). A hatchling warning call may e voke a defensive response from adults of either sex. A pod may remain cohesive fo r over one year (Ouboter and Nanhoe 1984), although high mortality generally occurs during that first year. Growth rates are not well documented but appear closely linked to food availability, temperature and dry season stress (Gorzula and Seijas 1989; Thorbjar narson 1991a). Coutinho (2000) found high initial growth in Pantanal hatchlings that doubled their length to about 25 cm SVL in the first year. The sexes contrasted in growth only over 70 cm SVL, with female growth slowing. Members of the Caiman crocodilus / yacare complex mature at an earlier age and smaller size than most other crocodilians. Th ese are important factors in the population response following harvesting. These factors also have long-term survival benefits for the species. I examined a gravid female Caiman c. crocodilus in Venezuela with 50 cm SVL. Females may reach sexual maturity between 4-6 years. Maturity is ha rder to determine in males, but they are generally slightly larger when they become sexually active (Thorbjarnarson 1991a). Adult females do not necessarily nest each year. Coutinho et al . (2001) detailed sexual developm ent for yacares in the Pantanal Although some females intiated vitellogenesis at 55 cm SVL, only females >70 cm SVL, estimated at 7-years old, were found with ma ture ovaries with follicles, and advanced vitellogenesis. They estimated that most fema les are 10-years old before entering into full reproductive life. They also found that sperma togenesis occurred in caiman >60 cm SVL

PAGE 80

60 but reached a peak maturity in male yacares >90 cm SVL. This represents animals about 9-years old. Prey Caiman are euryphagous opportunistic feed ers with a gradual ontogenetic dietary shift. Though they are often described typi cally as sit-and-wait ambush hunters, the relative uniformity of prey over the large distribution of Caiman indicates dietary preferences. Young animals eat large quantities of invertebra tes. They tend to include more vertebrate prey, predominantly fis h, as they grow larg er (Aquino-Ortiz 1988; Fitzgerald 1988; Thorbjarna rson 1993; Coutinho 2000). A nu mber of specialized, active feeding behaviors have been observed. Pantan al yacares showed a shift toward calcium and phosphorus rich fish prey at about 40 cm SVL, a break point in the size specific growth rates (Coutinho 2000). Investigations of stomach contents have b een made to determine the importance of prey species in the diet but differential di gestion rates of various items and secondary ingestion can confuse findings (Jackson et al . 1974; Garnett 1984). Durable residual items such as coleopteran elytra, crab carapa ces, or mollusk opercula may be retained for long periods after digestion. The importance of prey species ultimately depends on the habitat use and presence of prey species (Magnusson et al . 1987; Fitzgerald 1988). Caiman employ foraging strategi es that select for the most easily caught prey. This may differ from absolute abundance of prey sp ecies present (Fitzgerald 1988). Caiman will also feed on carrion (Eder 1791; Stat on and Dixon 1975; Thorbjarnarson 1991a; Godshalk, unpubl. data). Fish are an important component of si ze >40 cm SVL caiman but especially for larger caiman (Ayarzagüena 1983; AquinoOrtiz 1988; Fitzgerald 1988; Coutinho 2000).

PAGE 81

61 Highly specialized fishing behavior has b een described where caiman use their body to herd fish against the bank, sometimes cooperatively (Schalle r and Crawshaw 1982; Thorbjarnarson 1993; King et al . 1998). Sometimes groups of yacares form shoulder to shoulder in narrow channels, facing upstream w ith their mouth open, snapping at fish that pass through their jaws. Indivi dual catch per unit effort is assumed to be higher in selection for this cooperative setting (see Fig. 33). This has been obser ved in at least four species and may represent a behavior that is more common but under-reported (King et al . 1998). Sometimes individuals place their body perpendicular to the flow and forcing much of the water past its head. The caiman se nse fish swept past th eir body in a form of ‘weir-fishing’ (Thorbjarnarson 1993). More ac tive pursuit may involve a head swipe, quick rush or even jumping, but many prey captures involv e minimal effort. Figure 3-3. A group of yacares fishing in flowi ng water. Individuals face the water flow, mouths open, and snap at passing fish (left, ©C. Yamashita). Caiman feed on concentrated fish in the dry season wa ter bodies. An adult caiman eating an armored catfish ( Hoplosternum ) is shown (right).

PAGE 82

62 Figure 3-4. Freshwater apple snail ( Pomacea lineata ) from central Bolivia, a common invertebrate food item for C. yacare in the flooded savannas (left). Caiman may also feed on carrion. An adult Caiman crocodilus in central Venezuela has seized a capybara carcass, Hydrochoeris hydrochaeris , (right). Invertebrates remain an im portant component of caiman diet throughout their life in many areas. This can be appreciated by the relative size and abundance in the environment of the two most common inve rtebrate food items, freshwater crabs ( Dilocarcinus ) and snails ( Pomacea , see Fig 3-4). Aquino-Ortiz (1988) conclude d that body size determines the feeding ecology that is expressed in different foraging modes. Di fferent size caiman also select appropriate habitats and size of prey sp ecies. Relative abundance of prey species fluctuates widely both seasonally and regionally and limited sampling can lead to biased information. Coutinho (2000) found that yacares shifted their diet to invertebrates as water peaked and receded in the Pantanal, switching back to ve rtebrates at low and increasing water levels. A number of factors may interplay as food av ailability does not account for all the data.

PAGE 83

63 Body Condition Body condition is a measure of the relative fat content and represents how well the individual is coping with the environment. Santos et al . (1994) reported a relationship between caiman dry season condition factor and habitats and available prey in a preliminary study. Seasonal changes in body co ndition have been investigated in the Pantanal and provide a model for caiman th rough out its range. Seasonal changes in temperature and water level were shown to be correlated with variation in feeding success and resulting body condi tion (Coutinho 2000; Coutinho et al . 2001; Bampi and Coutinho 2003). These two variables interact to produce four distinct periods in the annual cycle. With high water and temperatur es (January-March) caiman have greater feeding success and take high qua lity vertebrate prey. This results in lipid reserves and good body condition. Although the water remains hi gh for most of the following period (April-June), temperatures diminish, slowing digestive efficiency and there is a shift towards lower quality invertebrate pre y. Feeding success remains high however, offsetting the lower nutritional value. Caiman suffer most stress in the period of low water and temperatures of the Austral winter (July-September). Feeding su ccess is low and body condition reflects the stresses present. Temperatures rise preceding the wet season and there is a shift back to vertebrate prey (October-December). Due to low water levels, feeding success continues to be low. Body condition in this period is in direct relation to the intensity of the previous wet season and resulting dry season water depth (Coutinho 2000; Coutinho et al . 2001; Bampi and Coutinho 2003). Reproductive effort has a high energy cost to an animal and is directly related to body condition. It has been shown in the Pantanal that water levels in the río Paraguay

PAGE 84

64 can be used to predict flood levels over the region. This, in turn, is a predictor for the ratio of reproductive females nesting over the landscape. Future studies may find similar environmental correlations in other areas to aid in yacare management Predation of Caiman Predation of yacares involves many species (see Table 3-1) and includes all life stages. Nests are frequently predated and lo sses may increase in areas where disturbance of the females has suppressed defens ive behavior. The large tegu lizard, ( Tupinambis spp.) is reported as an im portant nest predator in Ve nezuela (Staton and Dixon 1977; Ayarzagüena 1983; Thorbjarnarson 1991a) and Brazil (Coutinho 2000). The same is true for the coati ( Nasua nasua ), perhaps the most important pr edator of yacare nests in the Pantanal (Crawshaw 1987; Cint ra 1989; Coutinho 2000). Othe r reported nest predators include raccoon ( Procyon spp.), crab-eating fox ( Cerdocyon thous ), jaguar ( Panthera onca ) and feral pig ( Sus scrofa ), but quantitative data ar e scarce (Rivero Blanco 1974; Gorzula 1978; Ayarzagüena 1983; Crawshaw 1987; Thorbjarnarson 1991a; Hoogesteijn and Mondolfi 1992; Coutinho 2000). Hatchling and young caiman are undoubtedly ta ken by the same species mentioned above as well as an array of other predators. An assortme nt of large predatory fish capable of consuming young caiman is f ound throughout yacare dist ribution, although few instances have been documented. Certai n species of fish commonly reach sizes capable of eating relatively larg e caiman. Some Pimelodid rive r catfish species, such as Brachyplatystoma (>100 kg) and Pseudoplatystoma (>60 kg), are quite capable of predation of even medium sized caiman. The behaviors of young caiman may, however, keep them from river areas normally inha bited by these fish. Young caiman remain primarily at the shallow margins of the water, often seeking shelter in floating or

PAGE 85

65 emergent vegetation. Alvarez del Toro (1974) reported that 14 gars ( Lepidosteus tropius) caught at a fish camp in Mexico contained hatchling fuscus caimans. Table 3-1. Vertebrate predators of Caiman spp. listed from reports throughout the Neotropics. Most are found within C. yacare distribution (Modified from Gorzula and Seijas 1989). Scientific Name Life Stage Source PISCES Hoplias macrophthalmus Hatchlings Medem 1983 Serrasalmus spp. Hatchlings Blohm 1973, Ayarzagüena 1983, Medem 1983 REPTILIA Drymarchon corais Hatchlings Alvarez del Toro 1974 Eunectes murinus to Adults Staton & Dixon 1975, Medem 1983, Thorbjarnarson 1991a Chelus fimbriatus Hatchlings Medem 1981 Phrynops geoffroanus Hatchlings Medem 1981 Melanosuchus niger to Adults Medem 1981 Tupinambis spp. Eggs Rivero Blanco 1974, Staton & Dixon 1977, Medem 1983, Crawshaw 1987, Thorbjarnarson 1991a, Coutinho 2000 AVES Ardea cocoi Juveniles Gorzula 1978 Nycticorax nycticorax Hatchlings Ayarzagüena 1983 Euxenura maguari Juveniles Staton & Di xon 1977, Ayarzagüena 1983 Mycteria americana Juveniles Gorzula 1978 Jabiru mycteria Juveniles Gorzula 1978 Buteogallus urubitinga Hatchlings Ayarzagüena 1983 Polyborus plancus Eggs Rivero Blanco 1974, Ayarzagüena 1983 MAMMALIA Cerdocyon thous Hatchlings Gorzula 1978, Ayar zagüena 1983, Crawshaw 1987, Thorbjarnarson 1991a, Coutinho 2000 Procyon spp. Eggs Alvarez del Toro 1974, Medem 1983, Nasua nasua Eggs Crawshaw & Schaller 1980, Thorbjarnarson 1991a, Coutinho 2000 Eira barbara Eggs Fitch & Nadeau 1980 Pteronura brasiliensis Hatchlings Cardús 1886, Emmons 1990 Felis pardalis Juveniles Alvarez del Toro 1974 Panthera onca Adults Medem 1983 Sus scrofa Eggs Crawshaw 1987, Thorbjarnarson 1991a, Coutinho 2000 Cebus apella Eggs Crawshaw & Schaller 1980

PAGE 86

66 Birds may account for sizable predation of young caiman though this is hard to quantify. Gorzula (1978) repor ted a white-necked heron ( Ardea cocoi ) ate a caiman that was approximately 80 cm in length. The common egret ( Casmerodius albus ) is common throughout the Neotropic wetlands and a very nimble predator. Large wading birds such as the wood stork ( Mycteria americana ), maguari stork ( Euxenura maguari ), and the jabiru stork ( Jabiru mycteria ) are of a size capable of f eeding upon caiman up to a half meter or more. Other fishing birds like the cormorants ( Phalacrocorax spp.), anhinga ( Anhinga anhinga ), or fishing hawk ( Busarellus nigricollis ) are also very common in caiman habitat and may occasionally take hatchlings as well. Subadult and adult caiman generally face lit tle threat from predation. Conspecific cannibalism and territorial fighting may be dynamic factors and sometimes cause reported mortality in other crocodilians (P ooley and Ross 1989). The larger black caiman ( Melanosuchus niger ) historically posed a threat to yacares but currently populations have been severely reduced or exterminated over much of its range. Besides man, the anaconda ( Eunectes spp .) and jaguar ( Panthera onca ) are the only other predators posing a threat to adult caiman (Staton and Dixon 1975; Medem 1981, 1983; Ayarzagüena 1983; Gorzula and Seijas 1989; Hoogesteijn and Mondolfi 1992). Population Parameters Information on structure and densities of caiman populations is crucial to a wellmanaged conservation program. Detailed data are not easily obtained. Applicable survey techniques have been developed and field te sted in various habitats (Magnusson et al. 1978, Messel et al. 1981). A uniform, repeatable survey system, which encompasses the

PAGE 87

67 variability of habitats for yacares should be adopted, incorporating sampling errors and the fraction of animals sighted. Surveys of caiman in a variety of neotropi cal areas have produced differing results (see Table 3-2). This may have to do with the wide variety of habita ts that were sampled, often without clear indication of the suitabi lity for caiman or taking into account the effects of environmental vari ability or human impacts. With strong seasonal migration patterns in some areas, compar isons are further confused. Table 3-2. Abundance of caiman (ind./km) in Neot ropical habitats from selected sources (Modified from Pacheco 1993). Species Abundance Country Source (ind/km) Caiman crocodilus 6.6-9.02 Peru Gorzula & Seijas 1989 8.51 Peru Verdi et al. 1980 0.7-19 Surinam Gorzula & Seijas 1989 19.8-86.5 Surinam Ouboter & Nanhoe 1989 2.52-23.4 Venezuela Gorzula & Seijas 1989 1.24-24.86 Venezuela Thorbjarnarson 1991a Caiman. yacare 0.01-70.0 Bolivia King & Videz 1989 0-1017 Paraguay King et al, 1994 3.0-5.6 Bolivia Ergueta & Pacheco 1991 0.15-8.4 Bolivia Pacheco 1993 5-1448 Bolivia Godshalk & King 2002 2.8-46.2 Argentina Waller 2003 Melanosuchus niger 0.25-15.8 Bolivia Pacheco 1993 0.28 Peru Verdi et al. 1980 Large portions of prime hab itat for caiman are composed of seasonally inundated savannas. Obviously, dry season surveys are logi stically more efficient as the caiman are concentrated. On a ranch in the southern Llanos of Venezuela up to 150 caiman per

PAGE 88

68 hectare of lagoon were recorded by Ayarzagüe na (1983). Extrapolating dry season values uniformly over the 78,000 ha ranch produced estimated wet season densities of 0.17 caiman/ha, including unsuitable dry land ar eas. Woodward and David (1985) estimated 1.11 caiman/ha on a large ranch in central Ve nezuela. Arial surveys in the Pantanal showed an average of >100 caiman/km2, w ith a total for the area reaching up to 30 million animals (Coutinho 2000) Determining the size/age class distribution is very problematical. Few detailed studies have been produced and populations vary between sites and species. Size approximations must be substituted for ag e. Caiman populations are generally broken into four size divisions. These size-classes utilize different habitats and surveys vary accordingly in difficulty and accuracy. Thorbjar narson (1991a) cautions that capture data may be biased toward the smaller classes and must be adjusted for correct interpretation of size distribution. Ecological Role Yacares are a significant element of th e complex wetland ecosystems they inhabit in lowland South America. Their important eco logical role as top car nivores has not been fully investigated. Fittkau (1970, 1973) hypothe sized that caimans accelerate the nutrient cycling in the electrolyte-poor Amazon wa ters. In his study areas, the ecosystems remained virtually intact except for the re moval of the caimans by hunters. Fishermen anticipated larger catches since they view ed caiman as competitors. The common view was stated by Roosevelt (1925:59) when he wr ote caiman "are always destructive to fish, and it is good to shoot them". However, in Fittkau's study areas, fish production decreased in successive year s after caiman removal.

PAGE 89

69 Fittkau (1973) examined experiment al nutrient release by caiman through defecation. He hypothesized that significan t amounts of nutrients originating from nutrient rich areas have been transferred via feces to nutrient poor river mouths. He concluded that the previous large populations of animals could have a positive influence on electrolyte flow in the ecosystems. Consid ering the impressive yacare densities found in recent surveys (King et al . 1994; Coutinho 2000, Godshalk and King 2002), primordial populations must have been astounding. No work on this intriguing idea has subsequently been undertaken. Detailed investigations on the effects of caiman removal have not been undertaken but supplementary and anecdotal information continues to grow. A commonly held belief in many countries is that piranhas multiply in the absence of caiman. Many studies confirm that caiman eat piranhas (Krieg 1928; Seijas and Ramos 1980; Vasques 1981; Medem 1981, 1983; Schaller and Crawshaw 1982; Ayarzagüena 1983; Coutinho 2000) but none have shown that caiman act as a contro l. It is ironic that only a few pages after the statement above by Roosevelt, he continues: "Undoubtedly the caymans were subsisting largely on these piranhas" (1925:88) . In many areas where caiman have been extirpated, local inhabitants claim that pira nha and other fish deemed "less desirable" have become more common (Godshalk 1978; Fitch and Nadeau 1980; Medem 1983). Cott (1961) studied stomach cont ents of Nile crocodiles, Crocodylus niloticus , and concluded that they fed prin cipally on non-commercial fish and did not compete with fisherman. However, he found that otters did compete with fisherman for preferred fish species. Several authors suggest that Alligator mississippiensis maintain year-round water-hole habitats that benefit many organi sms, including fish (Craighead 1968; Neill

PAGE 90

70 1971; Alcala and Dy-Lliacco 1989). Other servic es to the ecosystem will likely occur with a numerous, large predator. Future studies will undoubtedly demonstrate the important roles of caiman within their ecosystems. Use of Caiman The use of caiman dates back to pre-Eu ropean contact eras. Indigenous tribes throughout tropical South America have used caiman species for food, adornment, and medicinals, and some groups continue to do so today (Medem 1981, 1983; Godshalk pers. obs.). Both AmerIndians and Spaniard colonists were fond of yacare meat (Eder 1791; Azara 1801). Métraux (1942) listed several lowland trib es that regularly included caiman in their diet, especially the tail meat. Cardús (1886:387), who served as a missionary in eastern Bolivia, stated "the fl esh is craved by many." He described the Guarayo Indians (1886:393) stating that "a lthough they don't eat the caiman meat, being all white and foul smelling, ne vertheless eat the eggs, in sp ite of them having a very repugnant odor as well". Consumption of yacare today is not ve ry common. Bolivian mestizos have no cultural tradition for it and regard caiman meat as ‘I ndian food’. Caiman meat is sometimes available in various Argentine, Brazilian and Paraguayan restaurants as novelty game dishes but not on a widespread basis. Indigenous groups in Paraguay and Bolivia account for some meat and egg consumption but at such low levels as to have little effect on the wild populations. Crocodilian oil has long been extracted fo r medicinal use. This practice is seen commonly throughout the tropics where crocodi lians are found (Ross 1992). I visited six stores in Bolivia where folk and herbal re medies are sold and all carried ‘Aceite de Caimán’ (caiman oil). It was previo usly rendered from visceral fat of Melanosuchus

PAGE 91

71 niger but currently the more a bundant yacare is used. The oil is rubbed on a patient’s chest for clearing pulmonary congestion or ta ken with food for the same effect. Cardús (1886) mentions the topical value for h ealing deep wounds and general curative properties. Taken internally it also re solves constipation (Eder 1791, Cardús 1886). According to Oblitas (1971) the oil is used fo r anemia, tuberculosis, rheumatism, and as a health tonic for the weak or ill. Commercial Hide Trade Rural life in the yacare di stribution area remained virtua lly unchanged for centuries after European contact. The main transporta tion was by horse, oxcart, and river craft, a scenario that persists today in many areas. La rge steam vessels plied the major rivers and provided long-range transport for passe ngers and cargo. Many rural inhabitants supplemented their meager incomes by provi ding fuelwood for these ships at regular intervals along the routes (Carvalho U. 1983). The 1940s was a period of marked changes for the lowland campesino and had great impact on the wildlife. Small internal combustion engines were available for use on boats and quickly replaced the cumbersome steam engines. Many former fuelwood providers entered the booming market in wildlife skins. Caiman latirostris were first exploited in Ar gentina, starting around 1930, and stocks there were depleted quickly. H unters switched to the less valuable Caiman yacare in the 1940s and exhausted local populations rapidly (Medem 1983). Hunting began in Bolivia in 1942 with the harvest of the valuable black caiman, Melanosuchus niger . Argentine merchants taught Bolivian fisher man to hunt caiman and prepare hides but losses were high initially due to inexperi ence (Medem 1983). The hides were easily transported by river and were usually shipped, salted and dried, to Argentine dealers

PAGE 92

72 (Carvalho U. 1983). By the end of the 1950s, black caiman were commercially extinct and large scale hunting of Caiman yacare began in Bolivia, Brazil and Paraguay. This commerce had very little benefit to the production areas. Hides were purchased very inexpensively and, typically, ta nning and manufacture of finished articles occurred in Europe or Asia. Illegal trade was conducted over the rugged lowland borders and virtually no record remains of the thousands of hides that were transported. Most of the hides were destined for the European high fashion market (Medem 1983). Claure (1986) estimated that almost nine million ca iman (black and yacare) were exported from Bolivia during the most productive y ears between 1940 and 1955. An estimated 12 million black caiman hides were produced throughout the Amazon basin in 1950 (S. Fitzgerald 1989). During that period other importa nt changes also occurred that affected the trade. After World War II, surplus airc raft readily available in the neotropics and the nature of transport there was permanently transforme d. Airstrips were constructed throughout the lowlands and even the most remote areas were much more accessible. This accelerated resource depletion as hides were often flow n out of the hinterlands. Medem claimed that missionaries from the International Language In stitute subsidized ai rstrip costs through the sale of hides (King pers. com.). Additi onally, outboard motors became more widely available which allowed greater freedom of movement and wider spheres of human influence. Previously remote areas were more easily reached by hunters. With the recognition of the peril of cr ocodilian species worldwide, most range countries took conservation measures. A series of legal instruments were enacted during the 1960s and 1970s governing minimum caiman size limits, hunting seasons and

PAGE 93

73 mandating the incountry tanning of hides. These laws were ignored and hunting undersized animals continued undiminished th roughout the year. The legal procedures were easily avoided and many skins were expor ted across the porous frontiers of Bolivia, Brazil, Paraguay, and Argentina, often with assistance of government officials (Medem 1983; King and Videz 1989). Claure (1986) estimated that 669,000 yacares were taken from Bolivia alone during the period from 1956 to 1973. Medem (1983) stated that in addition to the 120,000 skin legal quota in 1980, 350,000 illegal skins passed to Paraguay for export. Today, the crocodilian hide trade involves approximately 1-1.5 million skins annually (Thorbjarnarson and Velasc o 1998; MacGregor 2002). Hides from Caiman species account for approximately 75% of th e world trade in crocodilian skins. This involves programs of captive production and su stainable use of wild populations. Yacare caiman previously made up a large percenta ge of that number (S. Fitzgerald 1989; Brazaitis 1989). The four yacare range states have sustai nable use programs in development (Ross and Godshalk 2003). Bolivia has begun an act ive monitoring and harv est program with a 30,000 skin quota for 2002, raised to 50,000 in 2003. They have experienced administrative problems but appear to be on the general path to a good program. Brazil has strict laws against harvest of wildlife da ting back to 1967. It has only recently been enforced effectively. Due to the low value of caiman skins, the high cost of captive production generally cannot be justified. Colo mbia is the one exception where guerrilla activity precludes a wild harvest. Recent Pa ntanal population survey s reveal incredible numbers of wild caiman and wildlife manage rs are trying to amend national wildlife law

PAGE 94

74 to accommodate this situation. Paraguay ha s a long history of wildlife exploitation accompanied by institutional corruption. Govern ment officials were unable to conduct their use program as planned and widespr ead over-harvesting o ccurred. Faced with international pressure and pot ential CITES sanctions, Para guay suspended all wildlife exports in 2004 for an indefinite period. Cu rrently, all yacare programs are being closely monitored by CITES and the IUCN Crocod ile Specialist Group fo r compliance with international regulations for legal trade and to insure that wild stocks are not jeopardized (Ross and Godshalk 2003).

PAGE 95

75 CHAPTER 4 MATERIALS AND METHODS Introduction A wide variety of materials and methods are described here. Most samples were and preserved in storage buffers in the fiel d. Some samples were obtained from museum specimens (see Appendix 1 for complete samp le data). DNA was then extracted, purified and quantified. These samples were then amplified using both published primers and primers designed for this study. Some anal yses used mitchondrial DNA. A library of nuclear microsatellites was constructed and a suite of primers were developed. Analyses were conducted using these and other publishe d primers. Molecular procedures were conducted in the Interdisciplinary Center fo r Biotechnology Research (ICBR) facilities at the University of Florida (UF) in Gaines ville, FL. The Genetic Analysis Laboratory (GAL), directed by Scientific Research Ma nager AnnMarie Clark, provides services within the ICBR Biotechnologies for the Ecological, Evolutionary and Conservation Sciences (BEECS). Many of the necessary techniques were acq uired in workshops I took through the ICBR Education Core. Prominent among these was the “Tools for Developing Molecular Markers” course. Other ICBR facilities were used for sequencing fragments and are mentioned below.

PAGE 96

76 Sample Collection Fresh Samples Most of the samples used in this study consist of fresh blood, although fresh muscle tissue was sometimes taken. Reptilian red blood cells are nucleated so blood presents an easily extracted, DNA-rich sample that poses minimum intrusion to the live animal. Most captures occurred from watercraft that in cluded dugout canoes, aluminum and wooden skiffs and the CITES survey project’s 3 m A von inflatable skiff. Caiman were located with the use of a hand-held 12v one million can dlepower halogen spotlight. Crocodilians have a tapetum that, in caiman, reflects bright orange lightwhen illuminated. Caiman up to 1 meter total length we re generally captured by hand, swiftly grabbing the animal by the neck and body. Larger specimens were usually captured with a pole and rope or cable noose system. In these situations, I often used a commercially available Ketch-All™ animal restraining pole. This pole has a plastic coated, locking cable that avoids injury to the animal and rubber grips for a secure handling of larger specimens. During a typical capture, a caiman would be spotted and approached slowly by watercraft. An assistant in the rear of the craft would maintain illumination on the caiman’s eyes as he guided the craft forward. I would extend myself over the bow of the craft, wearing a headlamp to leave my hands free for capture or to handle the pole and noose. Some nocturnal captures were made while walking along lakeshores or river edges. Caiman were often encountered in the shallows at the waters edge or, at times, out on dry land. It appears that shining the beam of a s potlight or headlamp into the eyes of the caiman reduces the reptile’s ability to detect peripheral movement and makes nocturnal approach and capture possible.

PAGE 97

77 Figure 4-1. Juvenile yacare cap tured by hand in eastern Bolivia. Jaws were secured with electrical tape (l eft). Capture of Caiman yacare in arroyo Negro, southern Bolivia, using a Ketch-All™ capture pole (right). After capture, the caiman’s jaws were secured with a 15 mm wide rubber band or plastic electrical tape. Jaw musculature al lows the cvaiman to close its jaws with incredible bite force, but they possess lit tle comparable strength for opening the jaws. Therefore, securing the jaws shut requires minimum force and materi al. All alligatoroids, including caiman, have an overbite so the teeth of the upper jaw remain exposed when the mouth is closed, and a ‘secured’ animal can still inflict serious injury if care is not taken. Blood samples were taken from the cervica l sinus using sterile 2 cc syringes and needles (Olson et al . 1975). The needle was introduced through the nucha l skin about 5– 30 mm posterior to the cranial table form ed by the parietal and squamosal bones, depending on the size of the animal (see Fig. 42). The needle was angled slightly caudad

PAGE 98

78 and a negative pressure was created in the syringe by lifting the pl unger slightly during insertion. In this manner, perf oration of the cervical sinus was immediately detected and a sample of about 1 ml of blood was drawn and introduced into a vial containing blood lysis buffer. Figure 4-2. Extraction of fresh blood sample from Caiman yacare in the field (left). Precipitation of DNA after isolation fr om a blood sample. Cloudy material in the center of the tube is purified genomic DNA. (right). Blood lysis buffer, in a protocol modi fied from White and Densmore (1992), contains 100 mM Tris-HCl (pH of 8.0), 100 mM EDTA (p H of 8.0), 10 mM NaCl and 1.0% SDS (sodium dodecyl sulfate) by weight per volume (w/v). Nine ml of this nontoxic buffer, with no anticoagulant, was placed in 10 ml screw cap tubes for field use. Upon extraction, a blood sample was immedi ately added to a labeled tube in an approximate 1:10 blood to buffer ratio. The tube was then inverted gently several times to mix, the saline solution thoroughly lysing th e cells and preserving the mixture.

PAGE 99

79 The resulting viscosity depends on the ratio of blood to buffer. Once the animal is in hand, one is inclined to take more bl ood than necessary. This only makes potential difficulties later in the laboratory when tr ying to separate a small amount from the gelatinous sample mass for DNA extraction (see below). Samples were maintained in the field at ambient temperatures, away from sunlight or excessive heat. Animals were measured, marked by scute removal to avoid any inadvertent recaptures, and the animal was released at th e capture site. Captur e site latitude and longitude were identified with a hand-held Garmin GPS, models 38, 40 or 12XL. Total work-up time from capture to release was mi nimal, generally requiring no more than 1520 minutes per animal. Large caiman require d more time for securing the animal to ensure a safe procedure. With practice, the blood extraction techni que was swift and efficient. In a few instances, usually with hatchlings, the cervi cal sinus could not be located. Occasionally with large, adult males, the needle length was insufficient and the sinus could not be reached through the thick cervical musculat ure. When blood samples could not be obtained, other tissues were coll ected. Sterile scissors were used to remove one or two caudal scutes of the terminal vertical scal e row. For a few Bolivian samples, freshly harvested and salted skins of known origin were sampled. Some of the Venezuelan C. crocodilus samples were taken from large salted, boneless meat slabs ( salon in Venezuela) representing single animals. Cu ts were made in the tissue to ensure penetration of the preserving buffer. These samples, about 1 g, were placed in a container with SED (saline, EDTA, DMSO) tissue storage buffer, modified from Amos and Hoelzel (1991) and Proebstel et

PAGE 100

80 al. (1993)(saturated NaCl; 25mM EDTA pH 7.5; 20% DMSO). Long-term storage (>5 years) at ambient temperature is possible in these buffers with no a pparent degradation of the sample. Dried Samples Only a few samples consisted of dried fragments of muscle or ligament tissue picked from skeletal preparati ons. Several trophy black caiman ( Melanosuchus niger ) skulls in Bolivia were sampled this way. One sample came from a dried black caiman skin that had been fashioned into a swing (see Figure 4-3). The animal had been taken from the nearby río Ipurupuru, about 1 y ear prior to my visit to the ranch. Figure 4-3. Unusual sample source s: trophy skull of black caiman ( Melanosuchus niger ) from central Bolivia. Skull measured 57 cm, representing an adult of about 4.7 m total length (left). Chair and swing fa shioned from a salted and dried black caiman belly skin hunted from a nearby caño (right).

PAGE 101

81 Bone Samples All of the Paraguayan samples originated from the osteological collections in the Florida Museum of Natural History (FLMNH) at the University of Florida (see Appendix 1). The Museum has a substantial collection of crocodilian skeletal material and 21 Paraguayan specimens of C. yacare were selected for DNA extraction. For specimens with post-cranial material, the femur was selected for the ease of drilling, relative diameter of the marrow channel and minimu m impact to the specimen. Some specimens are represented only by a skull. On these, th e quadrate/quadratojuga l bones were chosen for drilling. A perforation was made just late ral to the mandiblular hinging condyle of the quadrate and on an angle toward s a point under the orbit. Pe netration was usually about 1-2 cm, depending on the relative size of the specimen. All bone sample preparations were perf ormed in a NuAire laminar flow hood to reduce the chance of cross-contamination. Be nch surfaces were cleaned initially with 10% bleach solution, followed by 95% ethano l and at least 10 minutes of UV light sterilization. All equipment and surfaces we re cleaned with bleach and alcohol between samples. All tools such as scalpel, forceps, and dental picks were soaked in 10% bleach, then 95% alcohol and flame sterilized. Glove s were always changed before beginning new samples. A Dremel MultiPro™ variable speed rotary tool was used to perforate the bone. A variety of drill bit types and sizes were c hosen according to sample bone size but a Dremel No. 196 cutting bit was preferred most often. Drill bits were first inspected for bone or marrow residue with a 10x loupe and cleaned with a fine wire brush until the steel appeared clean under magnification. In a few instances, bits were cleaned of sticky

PAGE 102

82 residue by drilling into coarse hardwood. The cleaning procedure then continued with the wire brush, bleach, alcohol , and flame sterilization. The exterior bone surface selected for gr inding was first cleaned with 95% ethanol and a Kimwipe™. A pilot hole was started a nd the grindings from the outermost layer discarded. A piece of folded weigh paper was us ed to catch the grindings and they were then distributed into 1.5 ml microtubes for subsequent de-calcification and DNA extraction. Replicate sample tubes were ma de for each specimen. The samples varied greatly in color and appearance from fine, dr y, pure white powder to course, dark red, nearly black, masses that were very sticky a nd difficult to extract from the marrow cavity. Figure 4-4. A laminar flow hood is essentia l for work with bone to avoid crosscontamination of samples while drilling with the Dremel tool (left). Various drill bits used, and an osteoderm sample from Paraguay (center). Example of a femur used for sample drilling (right) All samples were collected with prop er documentation. The Bolivian Wildlife authorities issued CITES permit No. 00470 for export of specimens. Entry into the US was under CITES import permit No. 816827 issued by the U.S. Fish and Wildlife Service Office of the Management Authority. Arge ntine samples were exported by Proyecto

PAGE 103

83 Yacare, S.A. with CITES permit No. 023752 and imported under US CITES permit No. 03US714329/9. Honduran tissue samples were obtained by the FLMNH during a CITES funded population survey mission. Venezuelan material was obtained by FLMNH from the caiman survey team of ProFauna, the Venezuelan G overnment Wildlife Service. Extraction of DNA Phenol-Chloroform-Isoamyl Alcohol Extraction All DNA isolations from fresh tissue we re performed with a phenol-chloroform (PCl) isolation protocol modified from Hillis et al. (1996) at the ICBR-GAL, UF. This produces high concentrations of very pure DNA that can be stored for long periods without significant dete rioration. The majority of the sa mples used in this study were blood in lysis buffer, the easiest sample type to process for DNA isolation. Using a plastic transfer pipette, approximately 300 µl of sample was placed in a 1.5 ml microtube. The sample-buffer mix forms a viscous complex, sometimes presenting problems for removal of the desired amount. With experien ce, most samples could be transferred by manipulating the pipette carefu lly. With some of the more viscous samples, it was necessary to cut the sample mass near the pipette tip with sterile scissors. Most blood samples required no further preparation, 500 µl of STE (0.1 NaCl; 10mM Tris-Cl, pH 8.0; 1 mM EDTA, pH 8.0; Sambrook et al. 1989) was added, it was vortexed and the extraction would begin. The most viscous samples produced poor isolation results initially. La ter, by using the digestion protocol for tissue on these dense samples, excellent results were produce d. Improved results were obtained using a reduced initial quantity of these samples, however, so extraction began with only 200 µl of the sample in a 1.5 ml microtube. To this was added 500 µl of STE, 50 µl of 20% SDS and 20 µl of proteinase K (20 mg/ml). The tube was vortexed for 15 seconds for

PAGE 104

84 thorough mixing and emulsification of the visc ous sample matrix. The tube was then incubated in a 55 û C water bath for 24 hours with re peated vortexing. This produced a uniform and less viscous solution for the DNA extraction. Most of the non-blood tissue samples used were excised tail scutes that were preserved fresh in tissue buffe r. A 100 mg portion of each ti ssue sample was transferred to a sterile 1.5 ml tube abd combined with 500 µl of STE, 50 µl of 20% SDS and 20 µl of proteinase K (20 mg/ml) enzyme. The tube was then vortexed and incubated in a 55 û C water bath for 24 hours, with periodic vor texing. Usually the tissue was completely digested after 1 day. If tissue remained intact, 10 µl of pr oteinase K (20 mg/ml) enzyme was added and the sample digested for additional time. Some samples were dry tissue from bone surfaces of skulls or skeletons. The tissues were first washed by adding 1 ml of sterile water, vorte xing and removing the supernatant. This step was repeated. The samp les were then rehydrated by adding 1 ml of phosphate buffered saline (PBS, pH 7.4, Sambrook et al. 1989), vortexing, and placed in a 55 û C water bath for 10 minutes. The sample was spun in an Eppendorf-Minispin Plus™ at 14,500 RPM for 1 minute and the supernatant was removed. Soaking in PBS helped remove surface contaminants and rehydrated the cellular tissue. The procedure was repeated 4 times. Then 500 µl of STE, 50 µl of 20% SDS and 20 µl of proteinase K (20 mg/ml) enzyme were added and the tube wa s vortexed. This was then placed in a 55 û C water bath and digested for th e appropriate amount of time, up to 7 days in some cases, with repeated vortexing. Additional proteinase K (20 mg/ml) was usually added in 10 µl increments as needed for complete digestion.

PAGE 105

85 Extractions from Bone Bone samples proved the most problematic , requiring the most time invested and generally resulting in low amounts and poor qua lity of the DNA extracted. These samples were subjected to extensive decalcification with 0.5 M disodium ethylenediaminetetraacetate (EDTA, pH 8.0, Sambrook et al. 1989) before the PCl isolation. Calcium (Ca++) ions inhibit PCR (polymerase chain reacti on) amplifications so bone samples must undergo special preparation before isolati on (Tuross 1994). Bone samples were washed with sterile water and rehydrated with PBS, as described above for dry tissue samples. To decalcify the samples, 1 ml of EDTA (0.5 M, pH 7.5) was added to the sample and it was vortexed to mix. It was then placed on a Thermolyne™ Vari-Mix rocking mixer at ambient temperature for 24 hours. The rocki ng kept the sample continually washed by EDTA. The sample was centrifuged at 12,000 g for 1 minute and the supernatant removed with a pipette and discarded. This wa s repeated at least 4 times. The supernatant taken from the sample on the fifth day was pl aced in a sterile 1.5 ml tube containing 300 µl of 4% ammonium oxylate (Labchem, Inc., Pittsburgh, PA) to check for the presence of Ca++ ions. Calcium, if present, forms a precipitate, indicating that more EDTA washes were required. When no precip itate formed, the sample was considered decalcified. At that point, the sample was centrifuged and the EDTA supernatant was removed. One ml of sterile water was added, the sample wa s vortexed and centrifuged 1 minute at 12,000 g. The supernatant was removed and discarded. The sterile water wash was repeated four additional times for trace EDTA remova l and the PCl protocol was started. The samples taken from bone varied from extremes of dry, pure white powder to sticky, viscous masses consisting mostly of fa tty marrow. As these decalcified in the EDTA, cloudy substances, probably lipids, appe ared floating on the aqueous surface after

PAGE 106

86 centrifugation. These were removed using 1000 µl pipettes with cut tips to allow suction of these viscous masses. Care must be taken that none of this cloudy material is removed with the supernatant when checking for Ca++ ions as it can easily be confused with the precipitant. Over the days of EDTA treatments, the sample material reduces in volume and changes texture to re semble very fine silt. After the last H2O wash, the remaining sample mate rial was combined with 500 µl of STE, 50 µl of 20% SDS and 20 µl of prot einase K (20 mg/ml) enzyme and the tube was vortexed. It was then incubated in a 55 û C water bath, with repeated vortexing, until digestion was complete. To begin the DNA PCl protocol, approxi mately 700 µl of buffered phenol was added to the sample material, it was then vor texed and placed on a rocker for 5 minutes at ambient temperature. The samples were centrifuged for 5 minutes at 12,000 g. The aqueous layer was very carefully removed usi ng a glass Pasteur pipette and transferred to a sterile 1.5 ml tube. This first step is critical with blood samp les and care must be taken not to transfer the cellular debris that is suspended delicatel y on the phenol/STE interface. This debris is easily disturbed and, in order to leave the majority of the cellular material behind, some of the aqueous layer had to be left behind as well. The discarde d aqueous quantity is greatest in this first step and decreases as the extraction progresses and debris is reduced. Some aqueous layer must be left in every step however, to ensure that no phenol or chloroform is carried over with the aqueous layer. This phenol process was repeated for all samples.

PAGE 107

87 Approximately 700 µl of phenol-chlorofor m isoamyl alcohol (25:24:1) was added to the sample, it was shaken to mix and placed on a rocker for 5 minutes. The samples were then centrifuged for 5 minutes at 12,000 g. The aqueous layer was then extracted and placed in a sterile 1.5 ml microtube. Th is phenol-chloroform process was repeated for all samples. Approximately 700 µl of chloroform-isoamyl alcohol (24:1) was then added to the sample, it was shaken to mix and placed on a rocker for 5 minutes. The samples were centrifuged for 5 minutes at 12,000 g. The aqueou s layer was then extracted and placed in a new 1.5 ml microtube. The aqueous phase was re-extracted with the chloroformisoamyl alcohol process for all samples produc ing a final tube with approximately 500 µl of aqueous sample solution remaining. Care wa s taken not to carry over any chloroformisoamyl alcohol in the final step. About 50 µl of 3 M sodium acetate (or 1/10 the aqueous volume) was added, the tube topped off with 95% ethano l and inverted several times to mix well. At this point, most of the DNA precipitated and was frequen tly visible in the tube as a small cloudy substance (see Fig 4-2). The samp le was then placed at -20 û C for 24 hours to complete the DNA precipitation. The next day the sample was centrifuged for at least 5 minutes at 12,000 g. This yielded a small, whitish colo red pellet of DNA at or near the bottom of the tube. The ethanol was carefully removed w ith a pipette and the pellet was gently washed with 1 ml 70% ethanol to remove the precipitating sa lts and centrifuged for 5 minutes at 12,000 g. Most of the ethanol was then removed with a pipette and the open tube s were placed in a 37 û C incubator to evaporate the remaining al cohol. Once the tube s were thoroughly dry,

PAGE 108

88 the pellets were re-suspended in approximate ly 100 µl of warm TE (10 mM Tris-Cl, pH 7.4; 1 mM EDTA, pH 8.0; Sambrook et al. 1989). The quantity of TE was varied according to the relative size of the pellet. Subsequent PCR amplification of the isol ations revealed probl ems with most PCl DNA extractions from bone samples. Apparently , PCR inhibitors had been co-isolated in the final product. A protocol modified from Ye et al. (2004) proved successful in overcoming this problem. Equal amounts of bone sample and CTAB lysis buffer ( 2% cetyltrimethylammonium bromide,100 mmo l/L Tris HCl pH 8.0, 20 mmol/L EDTA, 1.4 mmol/L NaCl and 0.2% 2-mercaptoethanol ), 300-400 µl each, were combined in a sterile, ceramic mortar. Using a pestle, the bone was ground as finely as possible and the mixture was reduced to a thin paste. When necessary, additional CTAB was added to maintain the proper working consistency and the sample was transferred to a 1.5 ml microtube using a sterile steel spatula. Add itional CTAB was added as needed to produce a consistency that could be vortexed. The sample was incubated on the bench at room temperature for 12 hours with periodic vortexing. As mentioned above, great care was take n to minimize the possibility of crosscontamination of bone samples. Between ex tractions, the ceramic mortars and pestles were thoroughly washed and then soaked in a bleach solution. They were then immersed in sterile water for 4 hours to remove all tr aces of bleach, rinsed with 95% ethanol and carefully wrapped in aluminum foil for autoclaving. The sample tubes were incubated in a 65º C water bath for 2 hours with vortexing every 15 minutes. The tubes were then centrifuged at 1000 g for 25 minutes. The resulting supernatant was caref ully removed and placed in new 1.5 ml tubes and 500 µl

PAGE 109

89 of chloroform-isoamyl alcohol (24:1) was a dded to each. The samples were gently mixed on a rocker for 5 minutes and then centrifuged at 1000 g for 25 minutes. The upper aqueous phase was removed and tr ansferred to a new 1.5 ml tube. These isolation samples were further purified of PCR inhibitors by using a Qiaquick™ DNA Purification Kit (Qiagen, In c., Chatsworth). Five volumes of proprietary Qiaquick™ PB buffer was added to each sample and mixed. This mixture was then loaded into a Qiaquick™ spin column that contains a silica-based membrane for binding DNA. The column was centrifuged at 12,000 g for 1 minute, and the flowthrough was discarded. The process was repeat ed until all the mixture had passed through the spin column. Next, 750 µl of proprieta ry Qiaquick™ PE buffer was added to each column and centrifuged at 12,000 g for 2 minut es for thorough reagent removal. The flow-through was discarded and the column was transferred to a new 1.5 ml tube. DNA was eluted with addition 50 µl of proprietary Qiaquick™ AE buffer to the column and a 5 minute incubation on the bench. The tube wa s centrifuged at 12,000 g for 1.5 minutes. To increase yield, the eluted product was re-intr oduced to the column, incubated another five minutes followed by centrifugation at 12,000 g for 1.5 minutes. Subsequent PCR amplification verified th e success of these procedures with the bone samples. Bone samples that had previous ly been isolated using the PCl extraction protocol described above were also processed with the Qi aquick™ PCR Purification Kit with very good results. The PCR inhibitors were successfully removed with the kit. Quantity and Quality All DNA isolations were quantified using an Eppendorf spectrophotometer, BioPhotometer™ model, to verify the extraction success and to produce working dilutions of 5-15 ng/µl. Isolations from bl ood samples ranged from 25 to 2,556 ng/µl with

PAGE 110

90 an average of 628 ng/µl. Tissue sample isolat ions produced less DNA, ranging from 26 to 588 ng/µl with an average of 190 ng/µl. B one samples produced the least amount of DNA and generally presented problems in PCR amplification. Bone isolations produced highly variable DNA concentrations ranging from 20 to 1080 ng/µl. The total DNA obtained from the phenol-chlor oform extraction is a critical factor, but the quality of DNA is also important. The ratio of Å260/Å280 light wavelength absorbance values was recorded during sp ectrophometer measurements. A ratio of 1.802.0 indicates pure DNA. Lower values indicate protein or pigment contamination. Most sample measurements ranged from 1.65-1.90. Additional Isolations After testing all the samples in PCR r eactions, a few samples proved difficult to amplify. Occasionally with th ese samples, a comparison DNA extraction was performed using a Qiagen DNeasy™ isolation kit, follo wing the manufacturer’s protocols. The sample is digested using pr oteinase (10 mg/ml) K enzyme and proprietary Qiagen ATL buffer in a 55°C water bath. A second proprie tary Qiagen buffer, AL and ethanol, are added to the digestion and then transferre d to a special QIAGEN Mini Spin column (QIAGEN, Inc.) where the DNA binds to a proprietary resin ma trix. A series of proprietary buffers are then used to wash th e salts and contaminants from the matrix by centrifugation. The DNA is eluted from the matrix with the final QIAGEN AE reagent. The results using this quick, expensive kit were fine, but no impr ovement was observed in the quality of the isolation when co mpared to phenol-chloroform extractions. Ultimately, although very labor-intensive, all sa mples were extracted with the excellent, and least costly, PCl protocol.

PAGE 111

91 Molecular Investigations Once DNA isolations of the samples were completed and they were prepared in working dilutions, molecular analysis began. As the main focus for study, I had the core 172 Caiman yacare samples from Bolivia, 21 Paraguayan samples from the FLMNH herpetological collection and 20 blood samples from Argentina. To contrast these, I obtained 12 Caiman c. crocodilus samples from Venezuela, and 18 samples of Caiman c. fuscus from Honduras. For additional comparison, I also included specimens from other species of Caimaninae: 8 samples from sympatric Caiman latirostris , 7 samples from sympatric Melanosuchus niger , 2 samples from sympatric Paleosuchus trigonatus and 3 samples from sympatric Paleosuchus palpebrosus . I also incorporated 2 samples of Alligator mississippiensis from Florida as the outgroup (see Appendix 1 for details). Fortunately, the entire mitochondrial (mt) genome for both a Louisiana Alligator missisippiensis (Janke and Arnarson 1997; Ge nBank accession number NC_001922.1) and a Venezuelan Caiman crocodilus (Janke et al. 2001; GenBank accession number NC_002744.2) have been published. This faci litated direct comp arison with my PCR amplification product sequences and in the design of new primers. The Caiman mtDNA genome is 17,900 bp long compared to the Alligator mtDNA genome of 16,646 nucleotides. All further position references are relative to the Caiman mitochondrial genome of Janke et al. (2001) unless otherwise noted. 16S rRNA Gene The 16S rRNA gene was chosen for analys is of interspecific relationships. The Caiman 16S rRNA gene is 1593 bp long, from position 1058 to 2650. I first used universal primers from Kessing et al. (1989) for the mitochondrial 16S gene to amplify Fragment A (see Fig. 4-5, Tabl e 4-1). This is a highly conserved gene and amplification

PAGE 112

92 for a diverse range of taxa can be achieved w ith these or other primers. For this reason, steps were taken to avoid contamination from other sources and the PCR reactions were set up in the NuAire laminar flow hood. The resulting product was 614 bp long 16S Fragment A. After sequencing the PCR produc ts, alignment of the fragments with the published Caiman crocodilus 16S sequence (Janke et al. 2001) was performed using Sequencher 4.1 (Gene Codes Corp., 2004). The fr agment ends correspond to positions 1881 and 2493 of the Caiman sequence (Janke et al. 2001). Figure 4-5. Positions of primers and PCR amp lified Fragments A and B in reference to the Caiman crocodilus 16S rDNA mitochondrial gene. I designed a new set of 16S primers to amplify a 581 bp complementary 16S Fragment B (see Table 4-1), with an 84 bp ove rlap for a final sequence reconstruction of 1,108 bp. Using the published Caiman crocodilus mtDNA genome as a template (Janke et al. 2001a), I created a primer set using Prim er3, an online primer construction engine based at the Massachusetts Institute of Technology (Rosen and Slatesky 2000, http://frodo.wi.mit.edu/cgi-bin/primer3/pr imer3_www.cgi). I conducted preliminary PCR amplifications using C. yacare samples with successful re sults. This could not be repeated with the Alligator templates, however. Using Sequencher 4.1 to align my primers with the published Alligato r genome, a number of nucleotide differences in the reverse primer (Cya-16SaR) were discovered. I used the Primer3 engine to create a new reverse primer (Ami-16SaR) th at was based on the published Alligator genome template

PAGE 113

93 and was compatible with the existing Cya-16S aF forward primer. This led to successful amplification of Fragment B for the Alligator templates (see Fig. 4-5). Most primers were designed for this study and all primers used in this research were optimized for the most efficient am plification. A Biometra® T-Gradient model thermal cycler was used to determ ine the optimal temperatures. An 11 û C thermal gradient provided 1 û C steps when using 0.2 ml 12 tube strips for accurate temperature assessment. Sometimes the MgCl2 concentration was manipulated to induce priming with certain samples. Table 4-1. Oligonucleotide primers for PCR amplification of mitochondrial 16S rRNA gene Fragments A and B for sequence analysis. Position refers to 5’ nucleotide of forward primers (A or F) a nd 3’ nucleotide of reverse primers (B or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001) Fragment Name Sequence Position Source A 16S A CGCCTGTTTATCAAAAACAT 1881 Kessing et al. 1989 A 16S B CTCCGGTTTGAACTCAGATC 2493 Kessing et al. 1989 B Cya-16SaF AAAAGCCTACCGAACCCAGT 1385 This study B Cya-16SaR CGGCCATTCAACCTGTAGTC 1963 This study B Ami-16SaR TTGTATATACCGC GGCCATT 1975 This study Most amplification of PCR reactions for 16S were 25 µl volumes. For these I used 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 2.0 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the forward and reverse primers and 2 µl ( 10-30 ng) of each sample. Difficult samples were successfully amplified by increasing MgCl2, up to 3.5 mM. All PCR amplifications were performed on Biometra thermocyclers, models UNOThermoblock, UNO-II, or TGradient. The PCR conditions for Fragment A began with a denaturing step for 5 minutes at 94 û C. This was followed by 35 cycles of 94 û C for 1 minute, 52 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Owing to the generic nature

PAGE 114

94 of these primers and nucleotide mismatches w ith the target templates, some samples did not amplify well. In addition to increased MgCl2 as mentioned above, I also employed a modified touchdown PCR program that produced excellent results. The PCR began with a denaturing step for 5 minutes at 94 û C. This was followed by 8 cycles of 94 û C for 1 minute, 52 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A second, less stringent series followed of 30 cycles of 94 û C for 1 minute, 50 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final ex tension step was performed at 72 û C for 12 minutes before cooling to 4 û C. Fragment B PCR conditions we re as follows: a denaturing step for 5 minutes at 94 û C, followed by 35 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Because the Fragment B primers were specifically designed for Caiman and Alligator , no touchdown PCR was necessary. Most PCR amplifications of Fragment A were run on Applied Biosystems, Inc. (ABI) 377 or 3700 automated sequencers at the ICBR Custom Sequencing Core, UF. A few of Fragment A amplifica tions and most of Fragment B PCR products were sent the ICBR-GSSL (Genome Sequencing Service Laborat ory) at UF, a facility using Amersham MegaBACE™ 1000 96 capillary sequencers. The Cytochrome b gene The cytochrome (cyt) b gene was chosen for analys is of both interspecific and intraspecific relationships. The Caiman cyt b gene is 1150 bp long, from position 14,461 to 15,610. Universal primers for the mitochondrial cytochrome b gene from Kocher et al. (1989) were first used (see Ta ble 4-2). This is a highly vari able region and of great value for phylogenetic research (Avise 2000, 2004). Am plification for a diverse range of taxa can be achieved with this prel iminary primer set. The resu lting product cyb Fragment 1A

PAGE 115

95 was 374 bp long (see Fig. 4-6). After sequenc ing the PCR products, alignment of the fragments was performed using Se quencher 4.1 with the published Caiman crocodilus cyt b sequence (Janke et al. 2001). Cyb Fragment 1A corresponds from position 14,526 to 14,899 of the Caiman crocodilus sequence. Further inve stigation was indicated following the preliminary results. Figure 4-6. Figure 4-5. Positions of primers and PCR amplified Cyb Fragments 1, 1A, 1B, 2 and 2B in reference to the Caiman crocodilus cytochrome b mitochondrial gene. Amplifications using primers modified from Glenn et al . (1998) successfully produced Cyb Fragment 1. This fragment is 668 bp long, and contains the shorter Fragment 1A within it. It corresponds from positions 14,461 to 15,128. Using the Oligo 6.7 primer software package (Rychlik and Rychlik 2001), a new set of cyt b primers were designed to amplify an overlapping fragme nt, Cyb Fragment 2. This complementary fragment is 649 bp long, with a 117 bp ove rlap for sequence concatenation, and corresponds from positions 15,012 to 15,660. The reconstructed fragment was 1200 bp long, corresponding from positions 14,461 to 15,660 and represents the entire cyt b gene. Although most C. yacare blood samples were successfully amplified with these original primer sets, the C. yacare bone samples, Alligator and related Caimaninae

PAGE 116

96 samples proved more difficult. Cyb Fragment 1 was successfully amplified for all Alligatordae using the modified Glenn et al. (1998) primers. More efficient primer sets for Cyb Fragment 2 were then designed us ing the online Primer3 engine (Rosen and Slatesky 2000), producing very similar length fragments (see Table 4-2). Cyb Fragment 2B is 683 bp long, with a 120 bp overlap and corresponds to positions 15,009 and 15,691. The concatenated fragment is 1231 bp long, corresponding from positions 14,461 to 15,691 representing the entire cyt b gene. Using Primer3, other primers were designed as needed to complete the amplifications fo r intransigent samples (see Table 4-2). Table 4-2. Oligonucleotide primers for PCR am plification of mito chondrial cytochrome b gene fragments for sequence analysis. Position refers to 5’ nucleotide of forward primers (H or F) and 3’ nucleoti de of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001) Frag. Name Sequence Position Source 1A H14841 AAAAAGCTTCCATCCAACAT CTCAGCATGATGAAA 14526 Kocher et al. 1989 1A L15149 AAACTGCAGCCCCTCAGAAT GATATTTGTCCTCA 14899 Kocher et al. 1989 1 Cya-14461F ATGACCCACCAACTACGAAAAT 14461 Glenn et al. 1998* 1 Cya-15128R GGGTGAAATGGGATTTTGTC 15128 Glenn et al. 1998** 2 Cya-15012F CCTCCTCCCATTCATCATCCTAGC 15012 This study 2 Cya-15660R CCGTTGAATGGGAATGATTGGTTG 15660 This study 1B Cya-14425F CCCCCGTTGTTCTTCAACTA 14425 This study 1B Cya-15078R GGATCCTCGTTCGTGTAGGA 15078 This study 2B Cya-15009F CTTCCTCCTCCCATTCATCA 15009 This study 2B Cya-15600R TGGTTTGGTTGTTTTATTTTCTAGGG 15600 This study 2B Cya-15692R TACGAGGGTGCTGGGTTAAG 15692 This study 2B Cya-15814R TTAGAAYGTCGGCTTTGG 15814 This study 2B Ami-14763F ACGCTTCACTGCCCTACACT 14763 This study 2B Ami-15454 GGTTCCGTCCACTTCTGTCTTACAA Glenn et al. 1998*** 2B Ami-15688R CAAGGCCAGCGCTTTTCCTT This study 2B Mni-14731F CATCGTACCATGAATCTGAG This study * Alligator primer number L14254 ** Alligator primer number H14881, modified by 1 nucleotide *** Alligator primer number H15454 Most PCR amplification reactions for cytochrome b were 25 µl. For these 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 1.5-3.0 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the forward and reverse prim ers and 2 µl (10-30 ng) of each sample.

PAGE 117

97 All PCR amplifications were performed on Biometra thermocyclers, models UNOThermoblock, UNO-II, or TGradient. The P CR conditions for cyb Fragment 1A began with a denaturing step for 5 minutes at 94 û C. This was followed by 35 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute. A final extension step was performed at 72 û C for 5 minutes before cooling to 4 û C. The PCR conditions for both Cyb Fragment 1, 1B and Cyb Fragment 2 and 2B began with a denaturing step for 5 minutes at 94 û C. This was followed by 37 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. The preliminary PCR amplifications of c yb Fragment 1, 1-B, 2 and 2-B were run on an Applied Biosystems, Inc. (ABI) 377 and 3700 automated sequencers at the ICBR Custom Sequencing Core, UF. The resulting sequences were aligned with Sequencher 4.1 and screened for polymorphic s ites. Reactions were subsequently set up for the remaining samples in 96-well plates. Most of the cyt b PCR amplification pr oducts were sent the ICBR-GSSL at UF, a facility using Amersham MegaBACE™ 1000 96 capillary sequencers. Sequences were evaluated and concatenat ed using Sequencher 4.2. File were then imported into PAUP 4.0b (Swofford 1998) fo r phylogenetic analyses and tree generation. Arlequin3 (Excoffier et al . 2005) was used to perform statistical analyses. MEGA3 (Kumar et al. 2004) was also used for comparison of results. Microsatellite Loci Construction of an Enrich ed Microsatellite Library For analyses of fine scale population structure of Caiman yacare , greater resolution was required than mitochondrial DNA could provide, therefore I de cided to use microsatellite markers.

PAGE 118

98 Most eukaryote genomes contain thousands of loci comprised of short nucleotide sequence couplets such as (G T)n or (CA)n that are re peated in tandem (Tautz et al. 1986). The repeated sequence fragments may be 2 to 6 nucleotides, but the CA dinucleotide is most common. Hamada et al. (1982) described poly (CA) microsatellites with a frequency of 5-10x104 individual sites in the mammalian genome. I first explored the use of primers designed for Alligator (Glenn et al. 1998, Davis et al. 2001a). A survey of the published Alligator microsatellite primers resulted in some that failed to amplify with C. yacare. Amplification proved problematic for most loci and some were not polymorphic. After testi ng microsatellite pr imers described for Caiman latirostris (Zucoloto et al. 2002), six amplified well and were found to be polymorphic in C. yacare and they were tested fo r this study (see Table 4-3). I began work to design primers specific for C. yacare . Finding and describing microsatellite loci for a species is a time cons uming process. In this case, I constructed an enriched genomic library, ligated fragments in to plasmid vectors and inserted them into bacteria for replication, screen ed the resulting clones for tandem repeats, then sequenced the fragments and developed PCR primers to amplify the tandem repeat. The hybridization enrichment me thod, modified from Kandpal et al. (1994) and by Kijas et al. (1994), was selected for simplicity (Moraga et al . 2001). Genomic DNA from two samples with a large geographic separation was combined: one from the río Beni in the Amaz on basin of western Bolivia, and the other from the río Paraná basin of eastern Boliv ia. This was to provide a wide range of potential microsatellite allele s. About 5 µg of pooled genomic DNA was combined with 1 unit Sau 3AI enzyme and the appropriate buffer and the mixture was digested overnight

PAGE 119

99 at 37 û C. The enzyme fractionates DNA at sp ecific recognition sites throughout the genome. The following day, the digested DNA was lo aded into a 1.2% (w/v) agarose gel, along with a size standard, for size selection. The gel was stained with ethidium bromide (EtBr), visualized with UV, and the area corresponding to 400-1500 bp was cut from the gel using a clean scalpel. DNA was removed from the gel slice with a QIAquick® Gel Extraction Kit as per manufacturer’s protocols. This is a pr oprietary process using a DNA binding resin, washes and buffers to remove the gel a nd purify the DNA. DNA was eluted in 10 mM Tris-HCL (pH 8.5) and quantified using a Ho efer fluorometer Dyna Quant 200 (Hoefer Pharmacia Biotech, Inc., San Francisco, CA). The DNA concentration was adjusted to 10 ng/µl. Linkers, complimentary to the sticky ends generated by the Sau3 AI digest, were ligated to the digested and fractionate d DNA. The linkers facilitate later PCR amplification using the complementary, non-ligated Sau3 AI strand as the primer. Linkers were ligated to the fractionate d genomic DNA digest with T4 DNA ligase (400 U/µl, New England Biolabs, Inc. Beverly, MA). Exce ss linkers were removed and the product was PCR amplified following protocols of Moraga et al. (2001). The amplified, linker-ligated fragment libr ary was hybridized to a biotinylated CA repeat probe [5’-(CA)15TATAAGATA-Biotin]. The library was enriched for microsatellites by using VECTREX® Avidin D (Vector Laboratory, Burlingame, CA) using modified Vector protocols (Moraga et al. 2001). The VECTREX® Avidin matrix has a high affinity to biotin and efficiently captures the hybridized DNA fragments. The

PAGE 120

100 matrix was then subjected to a series of stringent washes to remove nonspecifically bound DNA fragments. Bound fragment s were eluted in water. The captured, enriched fragments were P CR amplified and the products analyzed for successful amplification in a 1.2% agar ose gel. A dot blot nylon membrane, using serial dilutions of the final PCR produc t and unprocessed genomic DNA, and Quick Light™ chemiluminescent CA repeat probe (L ifecodes Corp., Stamford, CT) were used to determine if enrichment was attained. Th e membrane was exposed to X-ray film that was developed and analyzed. Positive dot blots displayed a signa l across the dilution range. Having successfully produced a microsat ellite-enriched DNA fragment library, the next step was direct ligation of these fragments to plasmid vectors for subsequent cloning in E. coli . The Invitrogen TA Cloning Kit (Invi trogen, San Diego, CA) with pCR2.1® plasmid vectors was used with the recommende d protocol. The colonies on the resulting plates were screened for the CA repeats. The recombinant DNA molecule was transformed into competent E. coli strain INV F’ (Invitrogen) using the manufacturer’s protocol. At the end of incubation, concentrated cells were spread on an LB-ampicillin agar plate. Only the transformed cells which have ampicillin resistance conferred by the pCR2.1® plasmid gene will grow. This preliminary reduction was important as less than 1% of the E. coli bacteria acquire the vector. In addition, segregation of the transforme d bacteria was required. The agar plate was also initially prepared with 20 µl of X-gal (5-Bromo-4chloro-3-indolyl-ß-Dgalactopyranose). X-gal is a chromogenic substrate for the product of the E. coli lac Z

PAGE 121

101 gene, ß-galactosidase. For transformed colonies without the fragment insert, the functional lac Z gene produces ß-galactosidase and the colo nies appear blue in the presence of X-gal. Transformed E. coli with plasmid vectors containing the target fragment insert have a disrupted lac Z gene and appear white. Microsatellite Library Screening – The initial step involved manufacturer’s protocols in screening the plas mid library for the desired (C A)n repeats. Colonies were lifted from the agar plates using Magn Graph nylon hybridization membranes (.045 micron, Osmonics, Inc), cross-linked with UV, and hybridized with the Quick Light™ system and chemiluminescent (CA)n probe. It was then exposed to X-ray film that was developed and analyzed. The microsatellite positive clones were id entified, picked and incubated overnight. Plasmids were isolated and captured using a QIAprep™ (Qiagen) Spin Miniprep kit and manufacturer’s protocol. A dot blot test membrane was prepar ed with serial dilutions of each clone using the Quick Light™ system and hybridized with a chemiluminescent (CA)n probe. This was exposed to X-ray f ilm that was developed and analyzed. To confirm the appropriate inse rt size >300 bp, 2 µl purified, positive clone samples were then digested with Ecor 1 enzyme and run on a 1.2% (w/v ) agarose gel. Selected clones were sent to the ICBR Custom Se quencing Core, UF, for processing. In all, 18 agar plates were screened, and 191 clones were picked, incubated, and screened with dot blot tests results. Of the 191 probed clon es, 92 (48%) appeared positive from the dot blot analyses. A total of 72 cl ones were sequenced. Of these, 45 (63%) clone sequences had microsatellite repeats. Only 4 sequences (5%) of these lacked sufficient flanking regions for primer design and 27 se quences (37%) lacked microsatellites.

PAGE 122

102 Sequences were aligned using Sequencher 4.1 and analyzed for duplication. Only one pair of clones duplicated the same microsatellite insert. Microsatellite Primers Once the clone sequences were reviewed individually in Sequencher 4.1, primer design proceeded with the online Primer3 software package (Rosen and Slatesky 2000). The default crit eria were not changed and included an optimum primer size of 20 bp (minimum 18 bp, maximum 27 bp) and primer Tm optimum of 60 û C (minimum 57 û C, maximum 63 û C, see table 4-3 for details). Primer pairs generally had a Tm difference <1 û C, with a maximum of 2.49 û C. Primer G-C content was generally 50-55%, ranging from 35 to 69%, and all primers were reviewed for complementarity and low 3’ stability. Complementarity scores included selfcomplementarity, 3’ complementarity, and pr imer-dimer formation. Product length was usually selected for 150-350 bp, to maintain separation in pool-plexing of primer pair PCR product combinations and subsequent analyses (see Table 4-3). Most PCR amplifications for microsatellite primer screening were 25 µl volumes. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM of each dNTPs, 0.26 µM each of the forward and revers e primers and 1 µl (10-30 ng) of sample. The PCR conditions for microsatellite prim ers began with a denaturing step for 5 minutes at 94 û C. This was followed by 37 cycles of 94 û C for 1 minute, 58 û C (for most primers, see Table 4-3) for 45 seconds, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Once optimal conditions of reagents and te mperature were established, the primers were then screened for allelic polymorphism s. Samples from different geographic areas

PAGE 123

103 and different cyt b haplotypes were used in the anal yses. Ten to fifteen µl of PCR products were run at 135 V, up to 2.5 hours, for the full length of 150 mm 1.2% agarose (w/v) gels to separate heteroz ygous alleles (see Fig 4-7). Gels were stained with EtBr and visualized with UV. Table 4-3. Oligonucleotide primers and parame ters for PCR amplification of genomic microsatellite loci in Caiman yacare . Locus Sequence Repeat Tm Source Claµ-1F CATAAACCTTGGGGCTGGT A (AC)15 59.82 Zucoloto et al. 2002 Claµ-1R ACACCATGCTGAAAGAAGCC 60.26 Zucoloto et al. 2002 Claµ-4F CCATGAGTGCTTGAACAG (AC)15 53.23 Zucoloto et al. 2002 Claµ-4R CAGTCTCTACCCAAGATGTG 58.26 Zucoloto et al. 2002 Claµ-5F GCGTAGACAGATGCATGGAA AC)16(AC)22 59.83 Zucoloto et al. 2002 Claµ-5R CAGTCTGAAGCTAGGGCAAA 58.26 Zucoloto et al. 2002 Claµ-6F GAAATATGGGACAGGGAGGA (AC)17 58.79 Zucoloto et al. 2002 Claµ-6R GGTTGGCTGCATGTGTA TGT 59.45 Zucoloto et al. 2002 Claµ-7F CGGGGTCTTGGTGTTGACTA (G T)12, (GT)6 60.94 Zucoloto et al. 2002 Claµ-7R CGGGACCAGGAGCTGTATAA 60.09 Zucoloto et al. 2002 Claµ-10F TGGTCTTCTCTTCGTGTCCT (CA)12, (CT)19 57.42 Zucoloto et al. 2002 Claµ-10R ATGAGCCCCTCTATGTTCCT 57.67 Zucoloto et al. 2002 Cya-D2F ATATCATTGATGCCCCCAAA (GT)27 59.98 This study Cya-D2R ATGAGCCCCTCTATGTTCCT 60.21 This study Cya-E5F GCAGGGGAGGGAGAAATAAG (CA)18 60.03 This study Cya-E5R GCCAAAGACTTGGACTCTGC 60.00 This study Cya-E9F GCTACACAGCCCCATGAAGT (CA)28 60.14 This study Cya-E9R CCCCCTCTTCAACCTAGGAC 59.93 This study Cya-F2F CTGAGCAGCAAGGTCAACAA (GT)17(CG)1(GT)15 60.18 This study Cya-F2R GCTTTAAATGCCATGCCCTA 60.06 This study Cya-I14F CCCTCATCCCTCTTCAATCA (GT)24 60.00 This study Cya-I14R GGCCAAACCAAAGTAAAGCA 60.11 This study Cya-N6F ATGCCCCATTAAAAGGTGTG (CA)21 59.69 This study Cya-N6R ACCCTAGGTACCCCAAGCAC 60.00 This study Cya-N10F TGCTGACCATTTTACTTCTTT GA (GT)27 57.50 This study Cya-N10R CTTCCCCAGCAACCTGAATA 60.07 This study Cya-O22F TCACCTTGAGTGAGCTGCAT (CA)26 59.58 This study Cya-O22R GCCTGAATTTGGCTTGAC AT 60.08 This study Cya-R8F GCCCAAGTTGAAGGTGTGTT (CA)19 60.01 This study Cya-R8R AAGGGCAGAGTCCAGTTTCA 59.84 This study Cya-S4F TGTGCATGTA TGTGCATTTG (CA)26 58.93 This study Cya-S4R GCCTGAAAGTTTGCCAAGAG 59.99 This study Primers for polymorphic loci require fluorescent dye labelin g for fragment analyses on Amersham MegaBACE 1000 sequencers. At this point in most proj ects, one primer of each pair is re-made with one of thre e costly labels (HEX=green, FAM=blue, TAM=yellow). In order to economize this expensive process, a 3 primer system described by Schuelke (2000) was employed. With this protocol, the forward primer of

PAGE 124

104 each pair is re-made with an 18 bp M13(-21) (TGTAAAACGACGGCCAGT) universal sequence at the 5Â’ end. Figure 4-7. Polymorphic loci N10 (left), a nd R8 (right) PCR products on 1.2% agarose (w/v) gels with EtBr staining and UV luminescence. This system was discussed at length th rough messages on the Evolution Directory (EvolDir) web site and protocol modifications were suggested for improving efficiency, including an improved 19 bp M13(-21) (CACGACGTTGTAAAACGAC) universal sequence. The forward primer was re-ordered with this new sequence at the 5Â’ end. Three separate primers (CACGACGTTGTAAAACGAC) comprised of th is same improved M-13(-21) sequence were ordered, each with a HEX, FAM or TA M fluorescent label at the 5Â’ prime end. With this system, fluorescent labeling of th e PCR fragments occurs within the reaction. The labeled M13(-21) primer, and reverse pr imer are added to the PCR cocktail in equimolar amounts. The forward primer with th e 5Â’ M13(-21) tail is added at one tenth the amount (changed from one quarter as pe r Schuelke 2000 protocol) so when it is consumed in the reaction, the labeled M13(-21) primer takes over. The initial PCR annealing temperatures were set for forward primer conditions. After the initial 10 cycles at the appropriate temperature, the forward pr imer is theoretically exhausted and the last

PAGE 125

105 30 cycles are run at 48 û C, the most efficient annealing temperature of the M13(-21) primer. In this manner, labeled primer purch ases were kept at a minimum and the choice of fluorescent dye color was made at the time of amplification. PCR amplifications for microsatellite prim ers using the 3 primer system were 15 µl volumes in 96 well plates. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the labeled M13(21) and reverse primers, 0.026 µM of the forward primer and 0.6 µl (5-20 ng) of each sample. Five primers were prepared conventi onally with fluorescent labels. PCR amplifications for these primers were also 15 µl volumes in 96 well plates. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the reverse and labeled forward primer, and 0.6 µl (5-20 ng) of each sample. After analysis of test plates for succe ssful amplification a nd labeling, pool-plex 96 well plates were prepared with 3 loci comb inations and sent for processing to the ICBRGSSL using ABI Prism® 37000 96 capillary sequ encers. Fragments were analyzed using GeneMapper v 3.0 software (Applied Biosyste ms, Inc. 2002). Microsatellite analyses were conducted using Arlequin3 (Excoffier et al. 2005) and PopGene v. 1.32 (Yeh et al. 1999).

PAGE 126

106 CHAPTER 5 MOLECULAR PHYLOGENY OF EXTANT ALLIGATOROIDEA, WITH DETAILS ON THE CAIMANINAE Introduction Crocodylian origins are estimated to have arisen over 200 million years before the present (mybp) (Buffetaut 1989; Brochu 2001). Over this great span of time, there has been a remarkable radiation of species from their Archosaurian ancestor, including novel fresh water, marine and terrestrial forms. Described extinct taxa outnumber the extant species five to one (Brochu 2003). This rich evolutionary traject ory continues today although current diversity is low. Modern fo rms are traditionally divided into three lineages: Crocodiloidea (‘true’ crocodiles and relatives), Gavialoidea (gharials), and Alligatoroidea (alligators and caimans) all w ith fossil records dating from the Late Cretaceous (Brochu 2003; see Fig 5-1). Recent molecular data associates gharials as a sister group with Tomistoma within the Crocodyloidea (Harshman 2003; Janke et al . 2005). The extant species have developed rela tively recently and, due to poor fossil formation in their preferred habi tats, little evidence exists for most living genera. Twentythree species are currently recognized and their morphol ogical characters lend scant inference to phylogeny for the groups. Molecula r analyses were employed in this project to define the topology with greater confiden ce and to add information on their possible evolution.

PAGE 127

107 The Alligatoroidea ( Alligator mississippiensis and all crocodylians closer to it than Crocodylus niloticus or Gavialis gangeticus ) is composed of 2 Alligator species and a monophyletic clade of 6 caiman species. Th e Caimaninae is a subgroup composed of Caiman crocodilus and all crocodylians closer to it than to Alligator mississippiensis . This includes Caiman yacare , C. latirostris , Melanosuchus niger and the more primitive Paleosuchus palpebrosus and P. trigonatus (see Fig. 5-1). Figure 5-1. Diagram of phylogenetic nomencl ature for extant crocodylians. Arrows indicate stem-based group names (groups including a species and any other more closely related to it than to an other species), and black circles denote node-based group names (last common ances tor of two or mo re species and all of its descendents) (simplified from Brochu 2003). The early fossil records place alligatorid de velopment in Laurasia due to finds in Europe and North America from the late Cretaceous (Taplin and Grigg 1989). An undiscovered precursor is believed to have led to the alligator and caiman lineages. . Alligator appears to have originated in North America with fossils from the late Eocene, about 35 mybp (Brochu 1999). A subseque nt radiation to Asia led to A. sinensis in China. Few fossils exist for Caimaninae and Broc hu (2003) describes the knowledge of the caiman evolutionary history as the least co mplete of all crocodilians. Brochu (2004a,b) estimates the alligator-caiman divergence o ccurred by the lower Paleocene, about 65

PAGE 128

108 mybp. Caiman fossils from North America are rare but some from the Eocene have been located. An ancestral lineage is believed to have dispersed to South America where subsequent radiation resulted in the extant species. The Caimaninae appears to have an exclusive South American ra diation with a re cent invasion of Central America by Caiman crocodilus . Eocaiman cavernous remains have been recovered in southern Argentina from Eocene strata. The fossils are dated at about 50 mybp and Eocaiman is believed to have given rise to the modern ge nera (Langston 1965; Brochu 2003). Melanosuchus has the best fossil record for the group with speci mens dating to the Pliocene (Medina 1976). Fossil remains attributed to a precursor of Melanosuchus and Caiman latirostris date from the Miocene (Langston 1965). The fossil record provides no clea r information about the relationships of the m odern caiman genera. No fossils has been reported for Paleosuchus . The two Paleosuchus species are distinct from the rest of the Caimaninae in many features and are generally considered more primitive. Biochemical and morphological analyses place Paleosuchus basal to the rest of the group (White and Densmore 2001; Brochu 2003). Materials and Methods All of the samples prepared for this study c onsist of fresh blood or fresh tissue (see Table 5-1). Caiman up to 1 meter total leng th were generally cap tured by hand, swiftly grabbing the animal by the neck and body. Larg er specimens were captured with a cablenoose system. Upon capture, the caiman’s ja ws were secured with a 15 mm wide rubber band or plastic electrical tape. Blood samples were taken from the cervical sinus using sterile 2 cc syringes and needles (Olson et al . 1975). The needle was introduced into the nuchal skin about 5–30 mm posterior to the cr anial table formed by the parietal and

PAGE 129

109 squamosal bones, depending on the size of the animal, and a sample of about 1 ml was drawn and introduced into a 10 ml vial c ontaining 9 ml of blood lysis buffer. Blood lysis buffer, in a protocol modi fied from White and Densmore (1992), contains 100 mM Tris-HCl (pH of 8.0), 100 mM EDTA (p H of 8.0), 10 mM NaCl and 1.0% SDS (sodium dodecyl sulfate) by weight per volume (w/v). Tissue samples, about 1 g, were placed in a container with SED (s aline, EDTA, DMSO) tissue storage buffer, modified from Amos and Ho elzel (1991) and Proebstel et al. (1993)(saturated NaCl; 25mM EDTA pH 7.5; 20% DMSO). Samples were maintained in the field at ambient temperatures, away from sunlight or excessi ve heat. Long-term storage (>5 years) at ambient temperature is possible in these bu ffers with no apparent degradation of the sample. All samples were obtained with proper documentation. The Bolivian Wildlife authorities issued CITES permit No. 00470 for export of specimens. Entry into the US was under CITES import permit No. 816827. Argentine samples were exported by Proyecto Yacare, S.A. with CITES permit No. 023752 and imported under US CITES permit No. 03US714329/9. Honduran tissue sa mples were obtained by the FLMNH during a CITES funded population survey missi on. Venezuelan material was obtained by FLMNH from the caiman survey team of ProFauna, the Venezuelan Government Wildlife Service. Alligator samples were provided by the Florida Fish and Wildlife Conservation Commission. Paleosuchus trigonatus samples from the St Augustine Alligator Farm, St. Augustine, Florida were kindly provided by Dr. Kent Vliet. All DNA isolations were performed with a phenol-chloroform isoamyl alcohol (PCl) isolation protocol modified from Hillis et al. (1996) with 95% ethanol precipitation

PAGE 130

110 at the ICBR-GAL, UF. Concentrated DNA was resuspended in TE buffer (10 mM TrisCl, pH 7.4; 1 mM EDTA, pH 8.0; Sambrook et al. 1989). All DNA isolations were quantified using an Eppendorf spectrophotom eter, BioPhotometer™ model to produce working dilutions of 5-15 ng/µl. Table 5-1. Data on species and material used for analyses of Alligatoroidea. Species Material Origin Source Alligator mississippiensis literature Louisiana, USA Janke and Arnason 1997 Alligator mississippiensis blood Florida, USA This study Alligator sinensis literature China Wu et al . 2003 Caiman c. crocodilus literature Venezuela Janke et a l. 2001 Caiman c. crocodilus literature Brazil Farias et al . 2004 Caiman c. crocodilus tissue Venezuela This study Caiman c. fuscus tissue Honduras This study Caiman yacare blood Bolivia This study Caiman latirostris blood Argentina This study Melanosuchus niger literature Brazil Farias et al . 2004 Melanosuchus niger blood Bolivia This study Paleosuchus palpebrosus blood Bolivia This study Paleosuchus trigonatus blood S. America (unknown) This study Gallus gallus literature Canada Desjardins and Morais 1990 Molecular Investigations Once DNA isolations of the samples were completed and they were prepared in working dilutions, molecular analysis began. I collected, or had access to, samples resulting in unique ha plotypes (in brackets) Caiman c. crocodilus (n=12[4]), Caiman c. fuscus (n=20[1]) , Caiman yacare (n=213[25]), Caiman latirostris (n=5[1]), Melanosuchus niger (n=5[2]), Paleosuchus palpebrosus (n=3[3]) , and Paleosuchus trigonatus (n=2[2]). The entire mitochondrial (mt) genome for Caiman crocodilus (Janke et al. 2001, GenBank accession number NC_002744.2) had been published and served as the base reference sequence. The Caiman mtDNA genome is 17,900 bp long and all further position references described below are relative to the Caiman mitochondrial genome of Janke et al. (2001) unless otherwise noted. Published sequences for Alligator

PAGE 131

111 mississippiensis (Janke and Arnason 1997; Ge nBank accession number NC_001922.1), and A. sinensis (Wu et al . 2003, GenBank accession number NC_004448.1), were incorporated in analyses. Additional sequences for Caiman crocodilus and Melanosuchus niger from Farias et al . 2004 were also used. Birds have Archosaurian ancestry and are c onsidered the closest living relatives to crocodylians (Sereno 1999; Brochu 2001; Va rgas and Fallon 2005). Sequences for the domestic chicken, Gallus gallus (Desjardins and Morais 1990, GenBank accession number NC_001323.1) were used for comparis on as an outgroup in some analyses. Recent molecular studies have determined th at Testudines are likely a sister clade to crocodylians (Zardoya and Meyer 1998; Hedges and Poling 1999; Kumazawa and Nishida 1999; Iwabe et al . 2005). The mtDNA genome for the green sea turtle was published and sequences were also used as an outgroup in some analyses, GenBank accession number NC_000886.1). 16S rRNA Sequences The highly conserved 16S rRNA gene was chosen for phylogenetic analyses. The Caiman 16S rRNA gene is 1593 bp long, from position 1058 to 2650. I first used universal primers from Kessing et al. (1989) for the mitochondrial 16S gene to amplify Fragment A (see Table 5-2). Th is is a highly conserved ge ne and amplification for a diverse range of taxa can be achieved with these primers. The resulting product was 614 bp long 16S Fragment A. Alignment of the PCR amplified fragments with the published Caiman crocodilus 16S sequence (Janke et al. 2001) was performed using Sequencher 4.1 (Gene Codes Corp., 2004). The fragment ends correspond to positions 1881 and 2493 of the Caiman sequence (Janke et al. 2001).

PAGE 132

112 A new set of 16S primers were designed to amplify a 581 bp partially overlapping 16S Fragment B (see Table 5-2), with an 84 bp overlap for a final sequence reconstruction of 1,108 bp. Using the published Caiman crocodilus mtDNA genome as a template (Janke et al. 2001), I created a primer set using Primer3, an online primer construction engine based at the Massachus etts Institute of Technology (Rosen and Slatesky 2000, http://frodo.wi. mit.edu/cgi-bin/primer3/primer3_www.cgi). I conducted preliminary PCR amplifications using C. yacare samples with successful results. This could not be repeated with the Alligator , Melanosuchus , or Paleosuchus templates, however. Using Sequencher 4.1 to ali gn my primers with the published Alligato r genome, a number of nucleotide differences in the reve rse primer (Cya-16SaR) were discovered. I used the Primer3 engine to create a new re verse primer (Ami-16SaR) that was based on the published Alligator genome template and was compatible with the existing Cya16SaF forward primer. This led to successf ul amplification of Fragment B for the Alligator templates. Table 5-2. Oligonucleotide primers for PCR amplification of mitochondrial 16S rRNA gene Fragments A and B used for sequence analysis. Position refers to 5’ nucleotide of forward primers (A or F) a nd 3’ nucleotide of reverse primers (B or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001) Fragment Name Sequence Position Source A 16S A CGCCTGTTTATCAAAAACAT 1881 Kessing et al. 1989 A 16S B CTCCGGTTTGAACTCAGATC 2493 Kessing et al. 1989 B Cya-16SaF AAAAGCCTACCGAACCCAGT 1385 This study B Cya-16SaR CGGCCATTCAACCTGTAGTC 1963 This study B Ami-16SaR TTGTATATACCGC GGCCATT 1975 This study Most PCR amplification reactions for 16S were 25 µl volumes. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 2.0 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the

PAGE 133

113 forward and reverse primers and 2 µl (1030 ng) of each sample was used. Difficult samples were successfully amplified by increasing MgCl2, up to 3.5 mM. All PCR amplifications were performed on Biometra thermocyclers, models UNOThermoblock, UNO-II, or TGradient. The PCR conditions for Fragment A began with a denaturing step for 5 minutes at 94 û C. This was followed by 35 cycles of 94 û C for 1 minute, 52 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Owing to the generic nature of these primers and nucleotide mismatches w ith the target templates, some samples did not amplify well. In addition to increased MgCl2 as mentioned above, I also employed a modified touchdown PCR program that produced excellent results. The PCR began with a denaturing step for 5 minutes at 94 û C. This was followed by 8 cycles of 94 û C for 1 minute, 52 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A second, less stringent series followed of 30 cycles of 94 û C for 1 minute, 50 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final ex tension step was performed at 72 û C for 12 minutes before cooling to 4 û C. Fragment B PCR conditions we re as follows: a denaturing step for 5 minutes at 94 û C, followed by 35 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Because the Fragment B primers were specifically designed for Caiman and Alligator , no touchdown PCR was necessary. Most PCR amplifications of Fragment A were run on Applied Biosystems, Inc. (ABI) 377 or 3700 automated sequencers at the ICBR Custom Sequencing Core, UF. A few of Fragment A amplifica tions and most of Fragment B PCR products were sent the

PAGE 134

114 ICBR-GSSL (Genome Sequencing Service Laborat ory) at UF, a facility using Amersham MegaBACE™ 1000 96 capillary sequencers. Cytochrome b Sequences The highly variable cytochrome (cyt) b gene was chosen due to great value for phylogenetic research and widespread use in intraspecific relations hip analyses (Avise 2000, 2004). The Caiman cyt b gene is 1150 bp long, from position 14,461 to 15,610. After sequencing the PCR products, alignment of the fragments was performed using Sequencher 4.1 Amplifications using primers modified from Glenn et al . (1998) successfully produced Cyb Fragment 1. This fragment is 668 bp long, and contains the shorter Fragment 1A within it. It corresponds from positions 14,461 to 15,128 with the published Caiman crocodilus sequence. Using the Oligo 6.7 primer software package (Rychlik and Rychlik 2001), a new set of cyt b primers were designed to amplify an overlapping fragment, Cyb Fragment 2 (see Table 5-3). This complementary fragment is 649 bp long, with a 117 bp overlap for sequence concaten ation, and corresponds from positions 15,012 to 15,660. The reconstructed fragment was 1200 bp long, corresponding from positions 14,461 to 15,660 and represents the entire cyt b gene. Cyb Fragment 1 was successfully amplified for all Alligatoroidea using the modified Glenn et al. (1998) primers. Alligator , Melanosuchus , and Paleosuchus samples proved more difficult for amplification of Frag ment 2. More efficient primer sets were designed using the online Primer3 engine (R osen and Slatesky 2000), producing very similar length fragments (see Table 5-3). Cyb Fragment 2B is 683 bp long, with a 120 bp overlap and corresponds to positions 15,009 and 15,691. The concatenated fragment is 1231 bp long, corresponding from positions 14,461 to 15,691 representing the entire cyt b

PAGE 135

115 gene. Using Primer3, other primers were designed as needed to complete the amplifications for intransigent samples (see Table 5-3). Table 5-3. Oligonucleotide primers for PCR am plification of mito chondrial cytochrome b gene fragments for sequence analysis. Position refers to 5’ nucleotide of forward primers (H or F) and 3’ nucleoti de of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001) Frag, Name Sequence Position Source 1 Cya-14461F ATGACCCACCAACTACGAAAAT 14461 Glenn et al. 1998* 1 Cya-15128R GGGTGAAATGGGATTTTGTC 15128 Glenn et al. 1998** 2 Cya-15012F CCTCCTCCCATTCATCATCCTAGC 15012 This study 2 Cya-15660R CCGTTGAATGGGAATGATTGGTTG 15660 This study 1B Cya-14425F CCCCCGTTGTTCTTCAACTA 14425 This study 1B Cya-15078R GGATCCTCGTTCGTGTAGGA 15078 This study 2B Cya-15009F CTTCCTCCTCCCATTCATCA 15009 This study 2B Cya-15600R TGGTTTGGTTGTTTTATTTTCTAGGG 15600 This study 2B Cya-15692R TACGAGGGTGCTGGGTTAAG 15692 This study 2B Cya-15814R TTAGAAYGTCGGCTTTGG 15814 This study 2B Ami-14763F ACGCTTCACTGCCCTACACT 14763 This study 2B Ami-15454 GGTTCCGTCCACTTCTGTCTTACAA Glenn et al. 1998*** 2B Ami-15688R CAAGGCCAGCGCTTTTCCTT This study 2B Mni-14731F CATCGTACCATGAATCTGAG This study *Alligator primer number L14254 ** Alligator primer number H14881, modified by 1 nucleotide *** Alligator primer number H15454 Most PCR amplification reactions for cytochrome b were 25 µl. For these 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 1.5-3.0 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the forward and reverse prim ers and 2 µl (10-30 ng) of each sample. The PCR conditions for Cyb Fragment 1, 1B and Cyb Fragment 2 and 2B began with a denaturing step for 5 minutes at 94 û C. This was followed by 37 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. The preliminary PCR amplifications of Cyb Fragments 1, 1B, 2 and 2B were run on an Applied Biosystems, Inc. (ABI) 377 and 3700 automated sequencers at the ICBR Custom Sequencing Core, UF. The resulting sequences were aligned with Sequencher 4.1 and screened for polymorphic s ites. Reactions were subsequently set up for the remaining

PAGE 136

116 samples in 96-well plates. Most of the cyt b PCR amplification pr oducts were sent the ICBR-GSSL at UF, a facility using Amersham MegaBACE™ 1000 96 capillary sequencers. Data Analysis Sequences for 16S and cytochrome b were separately evaluated, and aligned with Sequencher 4.2. The resulting 16S contigs were th en trimmed of missing data resulting in consensus sequences of 1077 bases depending on insertion or deletions. The same process was performed for the cytochrome b data, resulting in sequences of 1145 bases depending on indels. The sequences were th en concatenated into one data set of approximately 2200 bases per species. Both Gallus gallus and Chelonia mydas served as out groups for initial analyses of Alligato roidea as mentioned above. Files were then imported into PAUP 4.0b (Swofford 1998) for phylogenetic analyses and tree generation. The data set was subjected to the hierarchi cal likelihood ratio test (LRT, Huelsenbeck and Rannala 1997), by means of Modletest v.3.06 (Posada and Crandall 1998) to assign the most appropriate of 56 models. A general tim e reversal plus rate heterogeneity among sites (GTR+G+ ) was selected. Maximum likelihood (ML), minimum evolution (ME) distance and maximum parsimony (MP) trees we re constructed by means of branch and bound or tree-bisection-reconnection algorithms. MEGA3 (Kumar et al. 2004) was also used for comparison of result parameters. Analyses were performed using 16S and cytochrome b separately and combined. Additiona l statistical support was provided by Bayesian inference using MrBayes v. 3.1.1 (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003). Data sets were progre ssively decreased in genera and increased in sample size, see discussion below for details

PAGE 137

117 Results and Discussion The concatenated sequence data set desc ribed above, composed of about 65% of the 16S rRNA gene and nearly the entire cytochrome b gene, was employed for a preliminary analysis of Alligatoroidea. One constructed sequence for each of the following species was used for the analyses: Alligator sinensis (As), A, mississippiensis (Am), Caiman latirostris (Cl), C. yacare (Cy), C. crocodilus (Cc), Paleosuchus palpebrosus (Pp), P. trigonatus (Pt), Melanosuchus niger (Mn), as well as Gallus gallus (Gg) and Chelonia mydas (Cm) as outgroups. Alligatoroidea Figure 5-2 shows the relationships betw een the species of Alligatoroidea under criteria for maximum parsimony. The genera l topology follows that given in recent publications using DNA (White and Densmo re 2001), and a combination of DNA and morphological characters (Brochu 2003). There were a total of 2328 unordered characters of equal weight, 1167 constant characte rs, 517 parsimony-uninformative variable characters and 644 parsimony-informative vari able characters. Gaps were treated as missing data. Starting trees were obtaine d via simple stepwise addition using C. crocodilus as the reference taxon. Tree-bisection -reconnection (TBR) branch swapping algorithm was employed. Bootstrapping was performed with branch-and-bound search and a 50% majority-rule consensus tree wa s enforced. A phylogram was constructed under criteria for minimum e volution (see Figure 5-3). A maximum likelihood model of evolution (GTR+G) was established by Modeltest v3.06 using the Akaike Information Criteri on (AIC) (see Figure 5-4). The assumed nucleotide frequencies (A=0.31910, C=0.30880, G=0.14440, T=0.22770) were calculated from the data set. A substitution matrix was constructed from the data set by Modeltest

PAGE 138

118 and six substitution types were used. Starti ng branch lengths were obtained using the Rogers-Swofford approximation method and no molecular clock was enforced during the test. All but one branch had uniformly hi gh bootstrap agreement (see Figure 5-4). This branch refers to the association of Paleosuchus . The most parsimonious alternate tree forms a polytomy with Alligator basal to the caimans. While this does not represent the ‘true’ phylogeny, it underscores the distinct molecular composition for Paleosuchus and their evolutionary distance from the remainder of the Caimaninae. Figure 5-2. Maximum parsimony phylogram representing base changes for 2200 bp concatenated partial sequences of mito chondrial 16S and cytochrome b genes. Tree length is 2057, CI=0.761, RI= 0.565, RC=0.430 and HI=0.239. Species abbreviations are: Cc ( Caiman crocodilus ), Cy ( C. yacare ), Cl ( C. latirostris ), Mn ( Melanosuchus niger ), Pp ( Paleosuchus palpebrosus ), Pt ( P. trigonatus ) Am ( Alligator mississippiensis ), As ( A. sinensis ) C. mydas ( Chelonia myda s), Gallus ( Gallus gallus ). This analysis supports the historic assumption of m onophyly within Alligatoroidea and also within Caimaninae. Traditional syst ematic diagnoses for the clade have relied heavily on morphological cranial characters with certain assu mptions on derived states. The apparent weaker support for the Paleosuchus position mentioned above disappears

PAGE 139

119 with the removal of the distant out groups. When Alligator serves as the most distant group, the bootstrap percentage values for unrooted trees following the same criteria above equal 100 for MP (10,000 replicates), 100 for ME (10,000 replicates) and 100 for ML (1000 replicates) for the Paleosuchus branch in question. Figure 5-3. Minimum evolution distance phylogram with outgroup rooting (score 0.79236). Note deep basal branch length for Paleosuchus . Species abbreviations are the same as in Figure 5-2. Although some authors have proposed a close relationship between Melanosuchus and C. latirostris , (Poe 1966; Brochu 2001, 2003), the above trees do not support that assertion. Caiman latirostris appears to occupy a relatively intermediate position between Melanosuchus and Caiman crocodilus according to this data set. In the uncorrected “p” pairwise distance matrix, scores were Mn-Cl=0.10143 vs. Cl-Cc=0.09258. Branch length comparisons on the ML tree are Mn-Cl=0.142 vs. Cl-Cc=0.120 and for the ME tree: MnCl=0.106 vs. Cl-Cc=0.089. These scores indicate a slightly closer re lation ship between the Caiman species for this data set.

PAGE 140

120 Figure 5-4. Bootstrap values for support of Alligatoroidea phylogeny. Bootstraps values are maximum likelihood (ML-top), mini mum evolution distance (ME-middle) and maximum parsimony (MP-bottom) . Bootstrap values are for 1000 replicates. Bayesian inference in brackets (5x106 generations, 50,000 trees, 25% burnin). Species abbrev iations as in Figure 5-2. Caimaninae With the basic Caimaninae phylogeny firmly es tablished for this data set, the next step was to determine the relationships within Caiman . A new data set was constructed in which Gallus and Chelonia were eliminated. A concatenated sequence for Caiman c. fuscus from Honduras, near the northern extrem e for the species, was constructed and added to this new data set. The subspecies di agnosis is not clear as there is great overlap in characters with C. c. crocodilus, and the detailed distributi on has not been determined. Questions remain whether the taxon represents a distinct evolutionary lineage or simply clinal variation within Caiman crocodilus distribution. Results of analyses were inconclusive regardi ng the relationship of C. c. fuscus with C. c. crocodilus and C. yacare. A reduced concatenated data set was constructed which eliminated distant Paleosuchus also with inconclusive results.

PAGE 141

121 A new data set was constructed using only cytochrome b sequences. The 16S sequence was removed to see if there had b een a confounding effect from this slowly evolving gene on these closely related lin eages (see Figure 5-5). A 1092 base sequence was used for each taxa with the following nucleotide frequencies for the data set: A=0.28610, C=0.35320, G=0.11390, T=0.24680. A GTR+I+ model was selected using model test and a TBR branch swapping algorithm was employed. Figure 5-5. Minimum evolution distance phylog ram (uncorrected ‘p’) for cytochrome b sequence comparison rooted with Melanosuchus (score 0.22141, branch length above; 1000 bootstrap replications with values below branch: top-ML, middle-ME, bottom-MP). Alternate trees and bootstrap values on the right. Species abbreviations are: Cc ( Caiman c. crocodilus ), Ccf ( Caiman c. fuscus), Cy ( C. yacare ), Cl ( C. latirostris ), Mn ( Melanosuchus niger ). Tree generation under various criterion resulted bootstrap associations of C . fuscus with C. yacare over 50% of the time as seen by the bootstrap values in ML, ME and MP hierarchies, but sometimes was weakly associated with C. crocodilus (ML=29, ME=11, MP=20). Alternatively, C. yacare was associated with C. crocodilus (ML=11, ME=35, MP=20) (see Fig. 5-5). A new data set was constructed for Alligatoroidea from cytochrome b sequences, with Gallus as the outgroup. A likelihood test rati o was performed showing no significant difference for trees with and without molecu lar clock enforcement. An estimate of the

PAGE 142

122 evolutionary time for C. fuscus was subsequently made using rooting time estimates found in the literature. The ear ly origin of birds has been calculated at about 240 mya (Brochu and Norell 2000) and the split between alligator and caiman was estimated from 65 mya (Brochu 2003) to 75 mybp (Wu et al . 2003). With this data set, the alligator– caiman split was estimated at 69 mybp, cons istent with the above calculations. Based on ML distance values generated with enforcement of a molecular clock using both the concatenated and cytochrome b only data sets, the Cc-Cy/Ccf split occurred about 7-8 mybp, with the emergence of Caiman yacare and C. c. fuscus at ~7 mya. This predates the rise of the Isthmus of Panama and closure of the Panama Seaway between the Atlantic and Pacific basins. Migr ation scenarios to acc ount for this possible occurrence are discussed below. Further investigation was undertaken to clarify the Caiman crocodilus complex and a new data set was constructed with 1193 base sequences for cytochrome b and partial Control Region. As many authors consider th e lineages to be subspecies of the common caiman, analyses were constructed to repres ent one species with various populations. Sequences for two haplotypes from each of the major basins occupied by C. crocodilus (Orinoco and Amazon) and C. yacare (Amazon and Paraná) were selected for comparison with the C. fuscus haplotype. Nucleotide frequencies were A=0.30400, C=0.34510, G=0.10740, and T=0.24350. The GTR+G+ evolutionary model was selected using Modeltest v.3.06. C. yacare and C. crocodilus were consistently separated (see Figure 5-6). With this data set C. fuscus is often associated with C. yacare but with weak support and little additional resolution was established. Alternate trees associated

PAGE 143

123 C. fuscus with different arrangements of C. yacare haplotypes (ML=43, ME=35, MP=46) more often than with C. crocodilus (ML=32, ME=23, MP=31). Figure 5-6. Minimum evolution distance tree for cytochrome b data set using GTR+ evolutionary model; score=0.15676. Boot strap branch support for Ccf not strong and the relationship remain s unclear. Bootstrap values (1000 replications): top-ML, middle-ME, botto m-MP. Species abbreviations are: Cc ( Caiman c. crocodilus ), Ccf (( Caiman c. fuscus), Cy ( C. yacare ), Cl ( C. latirostris ). A = río Amazon, P = río Paraná, O = río Orinoco. A new data set was constructed with cytochrome b sequences for analysis with Arlequin v 2.0 (Schneider et al . 2000), and v 3 (Excoffier et al . 2005). This exercise was to not only to determine the relative position of Caiman c. crocodilus with C. c. fuscus , but also include comparisons with C. yacare, often regarded an a dditional subspecies. A total of 274 caiman representing 40 haplotypes were used for analysis (see Table 5-4). For the purpose of analysis, the six popul ations were associated into 4 groups related to the sample localites: Central Amer ica (Honduras), Orinoco (Venezuela), central Amazon (French Guiana, Brazil) and southern Amazon (Bolivia, Paraguay). The French Guiana (FG) sample is grouped with the Brazilian population due to the most

PAGE 144

124 parsimonious geographic connection for resulting gene flow. This logic is also applied to associating the C. yacare populations even though the Pa raguay haplotype technically lies outside the Amazon basin boundaries, as does the FG sample. An AMOVA pairwise difference test revealed 83.55% vari ation among groups, 12.63% variation among populations within groups, and 3.82% percent variation within popula tions, all significant at the p=0.05 level. Significant differences also were found for the population pairwise Fst test results at an alpha leve l of 0.05 (see Tables 5-5, 5-6) Table 5-4. Information for 1120 base cytochrome b sequence data set for 40 Caiman haplotypes used in analyses. Species Population Habitat Location Haplotypes N= Source Caiman c. fuscus Honduras Caribbean coast 1 20 This Study Caiman c. crocodilus Venezuela central Orinoco 5 14 This Study Caiman c. crocodilus French Guiana Atlantic coast 3 8 Farias et al . 2004 Caiman c. crocodilus Brazil central río Amazon 6 24 Farias et al . 2004 Caiman yacare Bolivia southern río Amazon 16 143 This Study Caiman yacare Paraguay northern río Paraná 9 65 This Study Table 5-5. Pairwise Fst test results, all significant at an level of 0.05. Honduras VenezuelaFr GuianaBrazilBolivia Paragua y 1) Honduras 0 2) Venezuela 0.9853 0 3) French Guiana 0.9972 0.9000 0 4) Brazil 0.9915 0.90730.6002 0 5) Bolivia 0.9559 0.95430.95480.9563 0 6) Paraguay 0.9713 0.96390.96640.96820.7581 0 Table 5-6. Population average pairwise di stances. Above diagonal: Average number of pairwise differences between populations ( xy). Diagonal elements: Average number of pairwise diffe rences within population ( x); Below diagonal: Corrected average pairwise difference ( xy-( x+ y)/2), all significant at an level of 0.05. Honduras VenezuelaFr GuianaBrazilBolivia Paragua y 1) Honduras 0.0000 46.428647.250046.375038.4615 39.8462 2) Venezuela 45.5934 1.670312.464311.589342.2028 42.2187 3) French Guiana 47.0000 11.3791 0.50001.625041.2780 41.1885 4) Brazil 46.0145 10.39361.0145 0.721040.4309 40.3135 5) Bolivia 37.4821 40.388240.048639.0910 1.9589 7.4128 6) Paraguay 39.0995 40.636940.191839.20635.6868 1.4933

PAGE 145

125 As seen in column 1 of the two distance tables above, the closest group association to Central America (Ho) is southern Amazon (B o, also row 1 in Table 5-6). As seen in by the Fst values (Table 5-5), there is no gene flow among C. fuscus and the other populations. The lowest value is still very high (0.6002) between Brazil and French Guiana. The results from this data set indicate that the C. fuscus samples used here represent a distinct evolutionary lineage. Further inves tigation will be required to determine if there is a clinal variation from H onduras to the distribution extrem es of southern Ecuador and western Venezuela. The caiman of the central Venezuelan coast have not been clearly identified, and the relationship to C. crocodilus in rest of the country is unclear. The C. fuscus populations west of the Ande s have been isolated from C. crocodilus populations for about 5 million years by vicariance events as outlined below. To investigate the relationship of distance to divergence, a Mantel distance matrix was constructed (see Table 5-7). Distances we re calculated using the most parsimonious migration routes along coasts and rivers gi ven current geographic features and hydrology. Distances were measured using a Run-Mate ™ digital chartmeter and detailed maps (1:5,148,000; National Geographic 1981). These distances are rough estimates due to map scale and respective detail and because precise tracing of the outlines was not possible. In any case, true distances will onl y increase values in the matrix. As most populations were represented by multiple samp le localities, a single central point was selected for determining distances. The distance from Honduras to easter n French Guiana followed the Caribbean coast to the Kaw Swamp (see Farias et al . 2004). The distance from Honduras to central

PAGE 146

126 Orinoco followed the coast, pass ing the Coast Range (Cordillera de la Costa) to just east of Caracas, then south along rivers to the Or inoco in Apure State. Point localities for other measurements include Brazil: río Amaz on (río Solimões) just upriver of Manaus (see Farias et al . 2004), Bolivia: central río San Mi guel in Departamento Beni and Paraguay: río Paraguay, just upstream of th e río Pilcomayo. Distances between these points were measured along main watercourses with consideration of current distribution, crossing basin divides in the areas of most narrow separati on. All results (1000 permutations) indicate significant divisi on between populations with respect to geographic and genetic distances. Table 5-7. Distances between populations (kilometers) of Caiman samples used in Mantel analysis. Genetic Group Population Honduras Venezuel a Fr GuianaBrazil Bolivia Paraguay Central America 1) Honduras 0 central Orinoco 2) Venezuela 4,000 0 central Amazon 3) French Guiana6,2002,500 0 central Amazon 4) Brazil 6,0002,0001,800 0 southern Amazon 5) Bolivia 7,9003,9003,7001,900 0 southern Amazon 6) Paraguay 10,0006,0005,8004,000 2,100 0 In the hypothetical reconstr uction of conditions during th e time frame leading up to and including the proposed migration events, three interrelated pro cesses must be taken into consideration: 1) genesis of norther n South America includi ng the influence of Andean tectonics and development of the Amazon and Orinoco basins, 2) dynamic conditions of changing paleocu rrents affecting the Panama Se away and environs, and 3) the progression of emerging land mass and ev entual closure of the Seaway by the isthmus.

PAGE 147

127 Northern South America and caiman migrations South American geography has been st rongly influenced by the Guyana and Brazilian shields. These pre-Cambrian massifs dominated drainage and erosion patterns in the pre-Andean landscape. Early continen tal drainages flowed off the shield in a generally western directi on to the Pacific (White et al . 1995). The influence of these highland regions continues today and they are the source for many affluents of the Orinoco, Amazon and Paraná drainages. The South American Plate has been impacted by the influence of several peripheral plates. Most notable has been the converge nce with the oceanic Nasca Plate along most of the western margin, leading to the rise of the Andes. This m ountain building process began at least 90 mybp and has not been uni form over the 7000 km length. The tectonic phases have undergone at least 6 periods of intense development (Sempere et al . 1997). The Andes north of Ecuador rose most rece ntly and have divided into 3 principal branches (cordilleras): Western, Centra l and Eastern. The genesis of the Andes fundamentally changed the drainage a nd weather patterns ove r the continent. As the mountains rose, an axial trough was formed by crustal downwarping in the foreland basin, east of the Andean spine. Prev ious drainage to the west was blocked, and drainages off the forming Andes flowed east into the trough. As a result of fluctuating sea levels, periodic intrusions penetrated the continent along these trough basins. This occurred from the Caribbean into the west ern Amazon to central Bolivia and the south Atlantic into the proto-Paraná area to southern Paraguay (Haq et al. 1987; Hoorn et al . 1995; Lundberg et al . 1998). By the late Cretaceous, the central Am azon appears to have had two major drainages: a minor one flowing east toward th e present delta, and the other draining west

PAGE 148

128 toward the alternately brackish or marine foreland basin. The early Paleocene hydrology was characterized by reduced marine intrusions and increased fresh water flow northward from a huge watershed. The caiman lineage is believed to have invaded South America form North America at about this time (see Fi gure 5-7). This draina ge system dominated much of South America from ~67 to ~8 mybp (Lundberg et al . 1998). A fresh water plume would have extended outward from the coastal margin and created brackish conditions for a great distance. Figure 5-7. Possible configurati on at the end of the Cretaceo us (ca. 65 mybp) when the caiman lineage may have migrated to South America. Major Central American blocks have sutured a llowing faunal movement (Me=Mexican, Ma=Mayan, Ch=Chortis). Volcanic arc (Aves Ridge) along the trailing western edge of the Caribbean Plate ma y have provided ‘stepping stones’ for intercontinental faunal exchange. Arc ultimately moved east, forming the Lesser Antilles. Arrows indicate plat e movement (after Donnelly 1988). Successful migration may also have depe nded on the oceanic currents to extend this brackish environment along the volcanic Av e arc. Fluctuating sea levels would have, at times, reduced inter-island distances and provided a ‘stepping stone’ archipelago enhancing faunal connections. There are numerous records of continental interchanges in

PAGE 149

129 both directions between the late Cretaceous and early Paleocene, including freshwater fish, birds, reptiles a nd mammals (Hallam 1994). The pattern in northern South America continued for the next 50 million years as the Andes grew in height and width. Drai nage water volume was increased due to orographic lifting and the cordillera imposed strong weather modifications not only on the continent but also within the hemisphere. A western Amazon inland sea, “Lago Pebas’, has been proposed due to pooling of fresh water and periodic marine intrusions (Hoorn 1993, 1994). The cyclic marine intrusions continued but, partly due to erosional infilling of the forebasin, they did not reach the previous interior extensions. Caiman evolution continued and several large, bizarre forms appeared such as the broad-snouted Mourasuchus (Brochu 2003). Flow increased northward into the proto-Or inoco basin from a system that extended from Bolivia. Outflow existed in a large area near present day Lake Maracaibo and Falcón further east (Hoorn et al . 1995). Collision with the Caribbean Plate to the north and Cocos Plate to the northwest produced profound changes. There were significant drainage disruptions as the eastern Cordillera began to rise (ca. 12 mybp) in Colombia and Venezuela. The Magdalena river valley was isolated and its large input to the Maracaibo basin ceased. From ~10 to ~8 m ybp there was an increased drainage via the Maracaibo and Falcón basins (see Fig. 5-8). De ep marine incursions ceased and to the south, the proto-Paraná drainage exte nded from southern Bolivia (Lundberg et al . 1998). In the intervening time period to the presen t, active tectonism in the northern Andes resulted in closing the Maraca ibo basin and forcing the río Orinoco eastward. The course of the Amazon also changed to the current Atlantic outflow. Major rearrangements

PAGE 150

130 occurred resulting in multiple interbasin exchanges of species (Albert et al . 2005). This accounts for the apparent incongruent distribu tions of many aquatic species and may be reflected to some degree in caiman distribut ion. A dynamic mosaic of marine, brackish and freshwater environments were evolving over a huge tropical area. Conditions were excellent at this time for caiman species development in northern South America. Figure 5-8. Paleogeographic re construction of northern South America and Central America before uplift of Isthmus of Pa nama and closing of two ocean basins (c. 10 mya). Arrows indicate marine pa leocurrents and continental drainage and sediment transport. M = proto-lake Maracaibo outflow area (after Albert and Crampton 2005). Measurements of the outflow of the río Amazon today indicate the magnitude of potential effects of an ancient river draining into the Maracaibo basin and adjacent Falcón basin. The present Amazon annual av erage discharge is about 175,000 m3/second (Sioli 1984) and the 250 km wide sediment fan reaches 680 km into the Atlantic (Putzer 1984). Depending on the season, the fresh water plum e is transported northwest 600-800 km to French Guiana (Goulding et al . 2003). The ancestral plume of fresh water flowing north into the Panama seaway would have affected the salinity of near s hore seawater for many hundreds of square kilometers.

PAGE 151

131 Marine ancestry is indicated by the lingual salt glands found in Crocodylus , Osteolaemus and Tomistoma that include primarily freshwater forms (Taplin 1988; Leslie and Taplin 2001). Cloacal tissues for cons ervation of water through urine modification also indicate adaptati ons for marine existence. The recently derived Crocodylus genus (~5-6 mybp) has a pantropical distribution th at has been credited to trans-oceanic migration events (Taplin et al . 1982; Taplin et al . 1985; Taplin and Grigg 1989; Leslie and Taplin 2001). A recent analysis of the New World Crocodylus species revealed a monophyletic clade with a probable colonizati on from Africa (Gratton 2003) In contrast, the Alligatoroidea do not possess lingual salt gl ands or display urine modification (Leslie and Taplin 2001). The lack of resistance to marine conditions has been attributed as a primary reason for geographic restriction of the group. Despite apparent physiological restric tions, caiman (and alligators) are not restricted solely to fresh water. Caiman populations resulting from natural migration events are found on islands off Colombia, Ven ezuela, Nicaragua and Panama, as well as Trinidad and Tobago. Occupied habitats with in caiman distributions include brackish lagoons, estuaries and mangrove swamps (M edem 1981, 1983; Gorzula and Seijas 1989; Grigg et al . 1998). These occurrences demonstrat e partial salt tolerance and the possibility for coastal marine migrations. In addition to effects fr om a river discharge plume and partial salt tolerance, proposed oceanic currents of the time period would favor transport of both caiman and brackish water mass toward the emerging peninsula (see Figure 5-8, 5-9). Evolution of the Isthmus In the Middle Miocene (15.1-12.9 ma), the shallow seaway was open allowing warm Atlantic water to enter the Pacific and continental vertebrate faunas were isolated.

PAGE 152

132 (see Fig 5-9A). By the late Middle Miocene (11.8-7.0 ma), there was partial emergence of the Panama Isthmus as sea levels vary, cau sing a disruption of wa rm, westerly Atlantic currents although shallow water connections persisted. The cool California current flowed south along the west margin of Cent ral America (see Figure 5-9B). Terrestrial faunal interchanges were first reco rded at ~9.3-~8.0 ma (Marshall et al . 1979, 1982; Webb 1985; Duque-Caro 1990). From the Late Miocene into the Early P liocene (7.0-3.7 ma), up lifting continued to diminish the shallow water connections but a later reduction of the California current reestablished the westerly flow of the warm Atlantic current (see Figure 5-9C). Sea levels fell initially but rose later in the period. Continued uplift for ced the Atlantic current north, initiating the proto-Gulf Stream. The Equato rial counter-current evolved bringing cool water to the region. As time progressed, sea levels fell and the isthmus emerged as continuous (or nearly) dry land (see Figure 5-9D; Duque-Caro 1990). The Great American Biotic Interchange (GABI) commenced in force (Stehli and Webb 1985; Webb 1991). Eustatic sea level changes may have pr oduced temporary seaway connections in the late Pliocene in southern Panama. (Cr onin and Dowsett 1996). Even today, areas of the Darien in Panama have a maximum height of only tens of meters above sea level. Addition physiological factors may have c ontributed to successful colonizations of Central America by caiman before a continuou s terrestrial landscape. Multiple paternity has been shown for Alligator mississippiensis (Davis et al . 2001b). There is also evidence that caiman ( Paleosuchus ) may employ sperm storage in reproduction (Davenport 1995). If these factors prove true also for Caiman , a single migration event of a gravid female could result in establishment of a new popula tion with a robust ge netic compliment. The

PAGE 153

133 apparent founder effect of low mtDNA haplot ype variability found in the Honduras series (1 haplotype, n=20) might be expl ained by such an occurrence. Figure 5-9. Neogene palaeooceanographic and pa laeogeographic evolu tion of the central American region. Hypothetical surface fl ow resulting from the disruption of warm Caribbean flow and intensificati on of the cool California current. A – Middle Miocene (15.1 to 12.9 mybp). B – Late Middle Miocene (11.8 to 7.0 mybp). C – Late Miocene to Early Pliocene (7.0 to 3.7 mybp). D – Early Pliocene (3.7 to 3.1 mybp). See text fo r explanation. (Simplified from DuqueCaro 1990). There is compelling evidence that the conditions outlined above were advantageous for caiman expansion north. Land mammals such as ground sloths migrated north from South America during the Late Miocene. Proc yonids from North America traversed the route south about the same time. In addi tion to vagile mammalian species, several freshwater fish taxa made successful migrati ons to Central America before a continuous land connection (Lundberg 1992; Bermingha m and Martin 1998; Martin and Bermingham 2000; Lovejoy and Collette 2001).

PAGE 154

134 The Central American freshwater icthyofauna is largely of South American origin (Bussing 1985).In Bermingham and MartinÂ’s (19 98) analysis, they described freshwater fish species the Central American Atlantic coas t slope drainages as mo re isolated than the Pacific slope drainages. They found the extant fish fauna to be the result of multiple colonizations across the isthmus beginning about ~7 mybp near the close of the Miocene. Phlygeographic signal due to s ea level rise and the re-estab lishment of the seaway (Haq et al. 1987) is observed in several studies (Bermingham and Martin 1998; Knowlton and Weigt 1998; Zeh et al . 2003). The biogeographic history of Caiman c. fusc us may have followed a pattern similar to these histories. Andean tectonism, st arting in northern Colombia and western Venezuela ~10 mybp (Hoorn et al.1995), gave rise to the Eastern Cordillera. This produced drainage division and vicariant se paration of the existing caiman populations. The area biogeography outlined above, produ ced conditions for caiman colonization of Central America. The current evolutionary lineage, Caiman c. fuscus, is considered a subspecies of Caiman crocodilus but today is reproductively isolated west of the Andes. The Honduras population, the most distant Atlantic drainage C. c. fuscus from the source C. c. crocodilus population, has a ~7 ma hist ory and has significant sequence divergence. Conclusions The following conclusions are given with the ca veat that this data set is limited to molecular analyses of mitchondrial DNA se quences from 16S and/or cytochrome b genes only. Given the immense distribution of th e Caimaninae, point locality sampling produces large intervening gaps in the data and transitional intermediate haplotypes may exist undetected.

PAGE 155

135 1) Alligatoroidea forms a monophyletic clade derived from a common Archosaurian ancestor as recognized hist orically by conventional, morphlogical systematics. 2) The alligator-caiman split occurred about ~69 mybp, similar to estimates by Brochu (2004) of 64 mybp and Wu (2003) of 75 mybp. 3) Caimaninae forms a monophyletic clade with Paleosuchus in the basal position and Caiman yacare as the most recently derived species. 4) The lineage leading to Paleosuchus split from the othe r caiman ~53 mybp (54 mybp; Brochu 2004). Caiman crocodilus dates from about 7.5 mybp. 5) The lineage leading to Melanosuchus split from the group at about 30 mybp, with a more recent splitting of the C. latirostris line at about 24 mybp. 6) Caiman crocodilus from the Orinoco and Amaz on basins are currently geographically (reproductively) isolated a nd have formed distinct genetic populations. 7) Caiman yacare and Caiman c. fuscus (Honduras population) are unique and separate evolutionary lineages, which arose independently from Caiman crocodilus ~6 mybp in a combination of vicariance and disp ersal events. They are also geographically, and hence reproductively, isol ated from populations of C. crocodilus of the Orinoco and Amazon basins.

PAGE 156

136 CHAPTER 6 PHYLOGEOGRAPHY AND CONSERVATION OF Caiman yacare Introduction The yacare caiman has been subjected to large-scale commercial hunting during most of the last 50 years. Much of the exploi tation in the range states of Bolivia, Brazil, Paraguay and Argentina occurred with governme nt authorization. Little effective control existed in spite of legislation specifying limits for export quotas, minimum size, and hunting seasons. Although Brazil made all comm ercial exploitation of wild populations illegal in 1967, porous frontiers in all of the range states allowed free movement of wildlife products. Government officials and military personnel were frequently involved in the exploitation, complicating control effo rts. CITES has imposed trade sanctions, and both Bolivia and Paraguay, the largest exporte rs of yacare hides, have independently suspended exports at different points in time. All four range states have sustainable use programs in development, although none have firm biological bases for determining th e harvest regions. In Bolivia, where the only functioning wild harvest curren tly occurs, size-class abundan ce data from limited surveys is the only factor employed. While adequa te as a minimum meas ure when sufficient surveys are undertaken, addi tional population data would provide for more robust management. The Venezuela program placed a pr iority on obtaining biological data for a decade before harvest, and a self-financing program developed incorporating user fees. The under-funded program in Bolivia continues to suffer from inadequate survey personnel, lack of control of areas harv ested and annual over-harvesting. Wildlife

PAGE 157

137 management programs are often non-existe nt of severely under-funded. Hides are reported to be collected clandestinely thr oughout the year (J. Aparicio, pers. comm.). The following study produces the first molecular data for Caiman yacare populations. Information on the phylogeny, haplot ype partitioning, gene flow and cryptic populations will be valuable fo r wildlife managers to incorporate into sustainable use planning. Sample Preparation The samples prepared for this study consis t of fresh blood or fresh muscle tissue. Caiman up to 1 meter total length were ge nerally captured by hand, and larger specimens were captured with a cablenoose system. The caiman’s jaws were secured and blood samples were taken from the cervical sinus using sterile 2 cc syri nges and needles (Olson et al . 1975). The needle was introduced into the nuchal skin about 5–30 mm posterior to the cranial table formed by the parietal a nd squamosal bones, depending on the size of the animal, and a sample of about 1 ml was drawn and introduced into a 10 ml vial containing 9 ml of blood lysis buffer. Blood lysis buffer (modified from White and Densmore 1992), contains 100 mM Tris-HCl (pH of 8.0), 100 mM EDTA (pH of 8.0), 10 mM NaCl and 1.0% SDS (sodium dodecyl sulfate) by weight per volume (w/v). Tissue samples, about 1 g, were placed in a container with SED (saline, EDTA, DMSO) tissue storage buffer, (saturated NaCl; 25mM EDTA pH 7.5; 20% DMSO; modified fr om Amos and Hoelzel 1991 and Proebstel et al. 1993). Samples were maintained in the fi eld at ambient temperatures, away from sunlight or excessive heat. Long-term stor age (>5 years) at ambient temperature is possible in these buffers with no appa rent degradation of the sample.

PAGE 158

138 All of the Paraguayan samples (n=21) orig inated from the osteological collections in the Florida Museum of Natural History (F LMNH) at the University of Florida. For specimens with post-cranial material, the femur was selected for the ease of drilling, relative diameter of the marrow channel a nd minimum impact to the specimen. Some specimens are represented only by a skull. On these, the quadra te/quadratojugal bones were chosen for drilling. A pe rforation was made just late ral to the mandiblular hinging condyle of the quadrate and on an angle toward s a point under the orbit. Penetration was usually about 1-2 cm, depending on th e relative size of the specimen. All bone sample preparations were perf ormed in a NuAire laminar flow hood to reduce the chance of cross-contamination. Be nch surfaces were cleaned initially with 10% bleach solution, followed by 95% ethano l and at least 10 minutes of UV light sterilization. All equipment and surfaces we re cleaned with bleach and alcohol between samples. All tools were soaked in 10% blea ch, then 95% alcohol and flame sterilized. Gloves were always changed be fore beginning new samples. A Dremel MultiPro™ variable speed rotary tool was used to perforate the bone using Dremel No. 196 cutting bit. Drill bits were first inspected for bone or marrow residue with a 10x loupe and cleaned with a fi ne wire brush until the steel appeared clean under magnification. The cleaning procedure th en continued with bleach, alcohol, and flame sterilization. The exterior bone surface selected for gr inding was first cleaned with 95% ethanol and a Kimwipe™. A pilot hole was started a nd the grindings from the outermost layer discarded. A piece of folded weigh paper was us ed to catch the grindings and they were

PAGE 159

139 then distributed into 1.5 ml microtubes for subsequent DNA extracti on. Replicate sample tubes were made for each specimen. All samples were obtained with proper documentation. The Bolivian authorities issued CITES export permit No. 00470 and US entry was under CITES import permit No. 816827. Argentine samples were exported with CITES permit No. 023752 and imported under US CITES permit No. 03US714 329/9. Venezuelan material was obtained by FLMNH from the Venezuelan Go vernment Wildlife Service. DNA isolations were usually performed w ith a phenol-chloroform isoamyl alcohol (PCl) isolation protocol modified from Hillis et al. (1996) with 95% ethanol precipitation at the ICBR-GAL, UF. Concentrated DNA was resuspended in TE buffer (10 mM TrisHCl, pH 7.4; 1 mM EDTA, pH 8.0; Sambrook et al. 1989). All DNA isolations were quantified using an Eppendorf spectrophotom eter, BioPhotometer™ model to produce working dilutions of 5-15 ng/µl. Bone samples proved the most problematic , requiring the most time invested and generally resulting in low amounts and poor qua lity of the DNA extracted. Bone samples require special preparation to eliminate PCR inhibitors that may be co-isolated during extraction, even with EDTA de-calcification st eps. A successful prot ocol, modified from Ye et al . (2004), overcame these problems. Equal amounts of CTAB lysis buffer (2% cetyltrimethylammonium bromide,100 mmol/L Tris-HCl pH 8.0, 20 mmol/L EDTA, 1.4 mmol/L NaCl and 0.2% 2mercaptoethanol ) and bone sample, 300-400 µl each, were combined in a sterile, ceramic mortar. The bone was crushed with a st erile pestle as finely as possible and the mixture was reduced to a thin paste. Add itional CTAB was added when necessary to

PAGE 160

140 maintain the proper working consistency. The sample was transferred to a 1.5 ml microtube using a sterile steel spatula and additional CTAB was added as needed for vortexing. The sample was incubated on the be nch at room temperature for 12 hours with periodic vortexing. Great care was taken to minimize the possi bility of cross-contamination of bone samples. Ceramic mortars and pestles were thoroughly washed and soaked in a bleach solution between extractions. They were then immersed in sterile water for 4 hours to remove all traces of bleach, rinsed with 95% ethanol and carefully wrapped in aluminum foil before autoclaving. The bone samples were incubated in a 65 ºC water bath for 2 hours with vortexing every 15 minutes. The tubes were then centrifuged at 1000 g for 25 minutes. The resulting supernatant was caref ully removed and placed in new 1.5 ml tubes and 500 µl of chloroform-isoamyl alcohol (24:1) was a dded to each. The samples were gently mixed on a rocker for 5 minutes and then centrifuged at 1000 g for 25 minutes. The upper aqueous phase was removed and tr ansferred to a new 1.5 ml tube. These isolation samples were further purified of PCR inhibitors by using a Qiaquick™ DNA Purification Kit (Qiage n, Inc., Chatsworth) following the manufacturer’s protocols. Five volumes of proprietary Qiaquick™ PB buffer was added to each sample, mixed and then loaded into a Qiaquick™ spin column that contains a silica-based DNA binding membrane. The co lumn was centrifuged at 12,000 g for 1 minute, the flow-through was discarded, and the process was repeated until all the mixture had passed through the spin colu mn. Next, the membrane is washed of contaminants by centrifugation with proprieta ry Qiaquick™ PE buffer. The column was

PAGE 161

141 transferred to a new 1.5 ml tube and th e DNA was eluted with addition 50 µl of proprietary Qiaquick™ AE buffer to the column. After a 5 minute incubation on the bench, the tube was centrifuged at 12,000 g for 1.5 minutes. To increase yield, the eluted product was re-introduced to the column, incubated another five minutes followed by centrifugation at 12,000 g for 1.5 minutes. Subsequent PCR amplification verified th e success of these procedures with the bone samples. Bone samples that had previous ly been isolated using the PCl extraction protocol described above were also processed with the Qi aquick™ PCR Purification Kit with very good results. The PCR inhibitors were successfully removed with the kit. All DNA isolations were quantified using an Eppendorf spectrophotometer, BioPhotometer™ model, to produce working di lutions of 5-15 ng/µl. Isolations from blood samples ranged from 25 to 2,556 ng/µl with an average of 628 ng/µl, tissue ranged from 26 to 588 ng/µl with an average of 190 ng/µl. Bone samples produced highly variable DNA concentrations ranging from 20 to 1080 ng/µl. The ratio of Å260/Å280 light wavelength absorbance value measurements in most samples ranged from 1.65-1.90. Sequence Data The cytochrome (cyt) b gene was selected for intras pecific relationship analyses. The Caiman cyt b gene is 1150 bp long, from position 14,461 to 15,610. This is a highly variable region and of great value fo r phylogenetic research (Avise 2000, 2004). The entire mitochondrial (mt) genome for a closely related Caiman crocodilus (Janke et al. 2001, GenBank accession number NC_002744.2) from Venezuelan specimen has been published. This f acilitated direct comparison with my PCR amplification product sequences and in the design of new primers. The Caiman mtDNA genome is

PAGE 162

142 17,900 bp long and all further position references are relative to the Caiman mitochondrial genome of Janke et al. (2001) unless otherwise noted. Amplifications using primers modified from Glenn et al . (1998) successfully produced 668 bp Cyb Fragment 1. It corresponds from positions 14,461 to 15,128. Using the Oligo 6.7 primer software package (R ychlik and Rychlik 2001), a new set of cyt b primers were designed to amplify a partiall y overlapping fragment, Cyb Fragment 2. This new fragment is 649 bp long, with a 117 bp overlap for sequence concatenation, and corresponds from positions 15,012 to 15,660. The reconstructed fragment was 1200 bp long, corresponding from positions 14,461 to 15,660 and represents the entire cyt b gene (see Table 6-1). Table 6-1. Oligonucleotide primers for PCR amplification of Caiman yacare mitochondrial cytochrome b gene fragments for sequence analysis. Position refers to 5Â’ nucleotide of forward prim ers (H or F) and 3Â’ nucleotide of reverse primers (L or R) relative to Caiman crocodilus mitochondrial genome (Janke et al. 2001). Frag. Name Sequence Position Source 1 Cya-14461F ATGACCCACCAACTACGAAAAT 14461 Glenn et al. 1998* 1 Cya-15128R GGGTGAAATGGGATTTTGTC 15128 Glenn et al. 1998** 2 Cya-15012F CCTCCTCCCATTCATCATCCTAGC 15012 This study 2 Cya-15660R CCGTTGAATGGGAATGATTGGTTG 15660 This study 1B Cya-14425F CCCCCGTTGTTCTTCAACTA 14425 This study 1B Cya-15078R GGATCCTCGTTCGTGTAGGA 15078 This study 2B Cya-15009F CTTCCTCCTCCCATTCATCA 15009 This study 2B Cya-15600R TGGTTTGGTTGTTTTATTTTCTAGGG 15600 This study 2B Cya-15692R TACGAGGGTGCTGGGTTAAG 15692 This study 2B Cya-15814R TTAGAAYGTCGGCTTTGG 15814 This study *Alligator primer number L14254 ** Alligator primer number H14881, modified by 1 nucleotide The Cyb Fragment 1 was successfully amplified using the modified Glenn et al. (1998) primers. Although most C. yacare blood samples were successfully amplified with these original primer sets, the C. yacare bone samples proved more difficult with Fragment 2 amplification. More efficient primer sets for Cyb Fragments 1 and 2 were designed using the online Primer3 engine (R osen and Slatesky 2000) producing nearly

PAGE 163

143 identical Fragments 1B and 2B (see Table 61). Cyb Fragment 2B is 683 bp long, with a 120 bp overlap and corresponds to positions 15,009 and 15,691. The concatenated fragment is 1231 bp long, corresponding to positions 14,461 to 15,691 representing the entire cyt b gene. Using Primer3, other primers were designed as needed to complete the amplifications for intransigent samples (see table 6-1). Resulting se quences varied from 1065 bp to 1225 bp with an average of 1143 bp. Most PCR amplification reactions for cytochrome b were 25 µl. For these 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 1.5-3.0 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the forward and reverse primer s and 2 µl (10-30 ng) of each sample. All PCR amplifications were performed on Biometra thermocyclers, models UNOThermoblock, UNO-II, or TGradient. The PCR conditions for all cyt b fragments began with a denaturing step for 5 minutes at 94 û C. This was followed by 37 cycles of 94 û C for 1 minute, 58 û C for 1 minute, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 12 minutes before cooling to 4 û C. The preliminary PCR amplifications of cyt b fragments were run on an Applied Biosystems, Inc. (ABI) 377 and 3700 automated sequencers at the ICBR Custom Sequencing Core, UF. The resulting sequen ces were aligned with Sequencher 4.5 and screened for polymorphic sites. Reactions we re subsequently set up for the remaining samples in 96-well plates. Most of the cyt b PCR amplification pr oducts were sent the ICBR-GSSL at UF, a facility using Amersham MegaBACE™ 1000 96 capillary sequencers. For haplotype ve rification, new PCR amplificat ion products were sent to

PAGE 164

144 ICBR Custom Sequencing Core, UF using Applied BioSystem, Inc. 3700 capillary sequencers. Sequences were evaluated and concatenat ed using Sequencher 4.5. Files were then imported into PAUP 4.0b10 (Swofford 1998) for phylogenetic analyses and tree generation. Likelihood model selection was performed using MODELTEST v.3.06 (Posada and Crandall 1998). Tree support through Bayesian inference was accomplished using MrBayes v.3.1.1 (Huelsenbeck and R onquist 2001; Ronquist and Huelsenbeck 2003). MEGA3 (Kumar et al. 2004) was also used for co mparison of results. TCS v. 1.21 software (Clement et al . 2000) was used to create haplotype genealogy networks by implementing the “statistical parsim ony” algorithm described by Templeton et al . (1992). Arlequin v.2.0 (Schneider et al . 2000) and Arlequin v. 3.0 (Excoffier et al . 2005) were used for AMOVA analyses, genetic distan ce measurements, and Mantel tests. Microsatellite Loci Construction of an Enrich ed Microsatellite Library For analyses of fine scale population structure of Caiman yacare , microsatellite markers were required for greater resolution than mitochondrial DNA could pr ovide. The inheritance pathway differs giving additional informati on about gene flow. Most eukaryote genomes contain thousands of loci comprised of short nucle otide sequence couplets such as (GT)n or (CA)n that are repeated in tandem (Tautz et al. 1986). The repeated sequence fragments may be 2 to 6 nucleotides, but the CA di-nucleotide is most common. Hamada et al. (1982) described poly(CA) microsatel lites with a frequency of 5-10x104 individual sites in the mammalian genome. I first explored the use of primers designed for Alligator (Glenn et al. 1998, Davis et al. 2001a). A survey of the published Alligator microsatellite primers had poor results

PAGE 165

145 with C. yacare . Amplification proved problematic fo r most loci and some were not polymorphic. Microsatellite primers described for Caiman latirostris (Zucoloto et al. 2002) were tested. Five amplified well and were found to be polymorphic in C. yacare and have been incorporated in this study (see Table 6-2). Work began to design primers specific for C. yacare . In this case, an enriched genomic library was constructed, fragments were ligated into plasmid vectors and inserted into bacteria for re plication, and the resulting cl ones were screened for tandem repeats. The selected fragments were se quenced and PCR primers were designed to amplify the tandem repeat. The hybridizat ion enrichment method, modified from Kandpal et al. (1994) and by Kijas et al. (1994), was selected for simplicity (Moraga et al . 2001). Genomic DNA was combined from two samp les with large geographic separation: one from the northern population (Amazon) of western Bolivia, and the other from the southern population (ParaguayParaná) of eastern Bolivia. Th is was to provide a wide range of potential microsate llite alleles. About 5 µg of pooled genomic DNA was combined with 1 unit Sau 3AI enzyme and the appropriate buffer, and the mixture was digested overnight at 37 û C. The enzyme fractionates DNA at specific recognition sites throughout the genome. The following day, the digested DNA was lo aded into a 1.2% (w/v) agarose gel, along with a size standard, for size selection. The gel was stained with ethidium bromide (EtBr), visualized with UV, and the area corresponding to 400-1500 bp was cut from the gel using a clean scalpel. DNA was removed from the gel slice with a QIAquick® Gel Extraction Kit as per manufacturer’s protocols. This is a propriet ary process using a DNA

PAGE 166

146 binding resin, washes and buffers to re move the gel and purify the DNA. DNA was eluted in 10 mM Tris-HCL (pH 8.5) and qua ntified using a Hoefer fluorometer Dyna Quant 200 (Hoefer Pharmacia Biotech, Inc., Sa n Francisco, CA). The DNA concentration was adjusted to 10 ng/µl. Linkers, complimentary to the sticky ends generated by the Sau3 AI digest, were ligated to the digested and fractionate d DNA. The linkers facilitate later PCR amplification using the complementary, non-ligated Sau3 AI strand as the primer. Linkers were ligated to the fractionate d genomic DNA digest with T4 DNA ligase (400 U/µl, New England Biolabs, Inc. Beverly, MA). Exce ss linkers were removed and the product was PCR amplified following protocols of Moraga et al. (2001). The amplified, linker-ligated fragment libr ary was hybridized to a biotinylated CA repeat probe [5’-(CA)15TATAAGATA-Biotin]. The library was enriched for microsatellites by using VECTREX® Avidin D (Vector Laboratory, Burlingame, CA) using modified Vector protocols (Moraga et al. 2001). The VECTREX® Avidin matrix has a high affinity to biotin and efficiently captures the hybridized DNA fragments. The matrix was then subjected to a series of stringent washes to remove nonspecifically bound DNA fragments. Bound fragment s were eluted in water. The captured, enriched fragments were P CR amplified and the products analyzed for successful amplification in a 1.2% agar ose gel. A dot blot nylon membrane, using serial dilutions of the final PCR produc t and unprocessed genomic DNA, and Quick Light™ chemiluminescent CA repeat probe (Lif ecodes Corp., Stamford, CT) was used to determine if enrichment was attained. The me mbrane was exposed to X-ray film that was developed and analyzed. Positive dot blots displayed a signal across the dilution range.

PAGE 167

147 Having successfully produced a microsat ellite-enriched DNA fragment library, the next step was direct ligation of these fragments to plasmid vectors for subsequent cloning in E. coli . The Invitrogen TA Cloning Kit (Invi trogen, San Diego, CA) with pCR2.1® plasmid vectors was used with the recommende d protocol. The colonies on the resulting plates were screened for the CA repeats. The recombinant DNA molecule was transformed into competent E. coli strain INV F’ (Invitrogen) using the manufacturer’s protocol. At the end of incubation, concentrated cells were spread on an LB-ampicillin agar plate. Only the transformed cells which have ampicillin resistance conferred by the pCR2.1® plasmid gene will grow. This preliminary reduction was important as less than 1% of the E. coli bacteria acquire the vector. In addition, segregation of the transforme d bacteria was required. The agar plate was also initially prepared with 20 µl of X-gal (5-Bromo-4chloro-3-indolyl-ß-Dgalactopyranose). X-gal is a chromogenic substrate for the product of the E. coli lac Z gene, ß-galactosidase. For transformed colonies without the fragment insert, the functional lac Z gene produces ß-galactosidase and the colo nies appear blue in the presence of X-gal. Transformed E. coli with plasmid vectors containing the target fragment insert have a disrupted lac Z gene and appear white. Microsatellite Library Screening – The initial step involved manufacturer’s protocols in screening the plas mid library for the desired (C A)n repeats. Colonies were lifted from the agar plates using Ma gniGraph nylon hybridization membranes (.045 micron, Osmonics, Inc), cross-linked with UV, and hybridized with the Quick Light™

PAGE 168

148 system and chemiluminescent (CA)n probe. Memb ranes were then exposed to X-ray film that was developed and analyzed. The microsatellite positive clones were id entified, picked and incubated overnight. Plasmids were isolated and captured using a QIAprep™ (Qiagen) Spin Miniprep kit and manufacturer’s protocol. A dot blot test membrane was prepar ed with serial dilutions of each clone using the Quick Light™ system and hybridized with a chemiluminescent (CA)n probe. The membrane was exposed to X-ray film that was developed and analyzed. To confirm the appropriate inse rt size >300 bp, 2 µl purified, positive clone samples were then digested with Ecor 1 enzyme and run on a 1.2% (w/v) agarose gel. Selected clones were sent to the ICBR Custom Sequencing Core, UF, for processing. In all, 18 agar plates were screened, and 191 clones were picked, incubated, and screened with dot blot tests results. Of the 191 probed clon es, 92 (48%) appeared positive from the dot blot analyses. A total of 72 cl ones were sequenced. Of these, 45 (63%) clone sequences had microsatellite repeats. Only 4 sequences (5%) of these lacked sufficient flanking regions for primer design. Sequences were aligned using Sequencher 4.5 and analyzed for duplication. Only one pair of clones duplicated the same microsatellite insert. Microsatellite Primers Sequences were reviewed individually in Sequencher 4.1, and primers were designed with the onlin e Primer3 software package (Rosen and Slatesky 2000). The default criteria were not changed and included an optimum primer size of 20 bp (minimum 18 bp, maximum 27 bp) and primer Tm optimum of 60 û C (minimum 57 û C, maximum 63 û C, see table 4-3 for details). Primer pairs generally had a Tm difference <1 û C, with a maximum of 2.49 û C. Primer G-C content was generally 50-

PAGE 169

149 55%, ranging from 35 to 69%, and all primer s were reviewed for complementarity and low 3’ stability. Complementarity scor es included self-complementarity, 3’ complementarity, and primer-dimer formati on. Product length was usually selected for 150-350 bp, to maintain separation in pool -plexing of primer pair PCR product combinations and subsequent analyses (see Table 6-2). Table 6-2. Oligonucleotide primers, repeat t ype and Tm used for PCR amplification of genomic microsatellite loci in Caiman yacare . Locus Sequence Repeat Tm Source Claµ-4F CCATGAGTGCTTGAACAG (AC)15 53.23 Zucoloto et al. 2002 Claµ-4R CAGTCTCTACCCAAGATGTG 58.26 Zucoloto et al. 2002 Claµ-5F GCGTAGACAGATGCATGGAA AC)16(AC)22 59.83 Zucoloto et al. 2002 Claµ-5R CAGTCTGAAGCTAGGGCAAA 58.26 Zucoloto et al. 2002 Claµ-6F GAAATATGGGACAGGGAGGA (AC)17 58.79 Zucoloto et al. 2002 Claµ-6R GGTTGGCTGCATGTGTA TGT 59.45 Zucoloto et al. 2002 Claµ-7F CGGGGTCTTGGTGTTGACTA (G T)12, (GT)6 60.94 Zucoloto et al. 2002 Claµ-7R CGGGACCAGGAGCTGTATAA 60.09 Zucoloto et al. 2002 Claµ-10F TGGTCTTCTCTTCGTGTCCT (CA)12, (CT)19 57.42 Zucoloto et al. 2002 Claµ-10R ATGAGCCCCTCTATGTTCCT 57.67 Zucoloto et al. 2002 Cya-D2F ATATCATTGATGCCCCCAAA (GT)27 59.98 This study Cya-D2R ATGAGCCCCTCTATGTTCCT 60.21 This study Cya-F2F CTGAGCAGCAAGGTCAACAA (GT)17(CG)1(GT)15 60.18 This study Cya-F2R GCTTTAAATGCCATGCCCTA 60.06 This study Cya-I14F CCCTCATCCCTCTTCAATCA (GT)24 60.00 This study Cya-I14R GGCCAAACCAAAGTAAAGCA 60.11 This study Cya-N10F TGCTGACCATTTTACTTCTTT GA (GT)27 57.50 This study Cya-N10R CTTCCCCAGCAACCTGAATA 60.07 This study Cya-O22F TCACCTTGAGTGAGCTGCAT (CA)26 59.58 This study Cya-O22R GCCTGAATTTGGCTTGAC AT 60.08 This study Cya-R8F GCCCAAGTTGAAGGTGTGTT (CA)19 60.01 This study Cya-R8R AAGGGCAGAGTCCAGTTTCA 59.84 This study Cya-S4F TGTGCATGTA TGTGCATTTG (CA)26 58.93 This study Cya-S4R GCCTGAAAGTTTGCCAAGAG 59.99 This study Most PCR amplifications for microsatellite primer screening were 25 µl volumes. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma Polymerase Taq or Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM of each dNTPs, 0.26 µM each of the forward and revers e primers and 1 µl (10-30 ng) of sample. The PCR conditions for microsatellite prim ers began with a denaturing step for 5 minutes at 94 û C. This was followed by 37 cycles of 94 û C for 1 minute, 58 û C (for most

PAGE 170

150 primers, see Table 6-2) for 45 seconds, and 72 û C for 1 minute and 15 seconds. A final extension step was performed at 72 û C for 10 minutes before cooling to 4 û C. Once optimal conditions of reagents and temperature were established, samples from different geographic areas and different cyt b haplotypes were used to screen the primers for allelic polymorphism s. Ten to fifteen µl of P CR products were run at 135 v, up to 2.5 hours, for the full length of 150 mm 1.2% agarose (w/v) gels to separate heterozygous alleles. Gels were staine d with EtBr and visualized with UV. Fragment analyses of polymorphic loci re quire fluorescent dye la beling of primers. In order to economize, a 3 primer system m odified from Schuelke (2000) was employed. Efficiency improvements were suggested by various labs via the discussion list on the Evolution Directory (EvolDir) web site . With this protocol, each forward primer of is re-made with an improved 19 bp M-13(-21) (CACGACGTTGTAAAACGAC) universal sequence at the 5’ end. Three separate primers comprised of this same M-13(-21) sequence (CACGACGTTGTAAAACGAC) were ordere d, each with a HEX, FAM or TAM fluorescent label at the 5’ prime end. With th is protocol, fluorescen t labeling of the PCR fragments occurs within the reaction. The la beled M13(-21) primer, and reverse primer are added to the PCR cocktail in equimolar amounts. The forward primer with the 5’ M13(-21) tail is added at 0.10x that amount in the improved protocol (changed from 0.25x in Schuelke 2000). After the initial 10 cycl es (30 cycles in Schuelke 2000) at the appropriate forward primer te mperature, it is theoretically incorporated. The abundant, labeled M13(-21) primer proceeds as the fo rward primer in successive 30 cycles (10 cycles in Schuelke) which are run at 48 û C (52 û C in Schuelke 2000), a more efficient

PAGE 171

151 annealing temperature for the M13(-21) primer. In this manner, the choice of fluorescent dye color was made at the time of amplifica tion and labeled primer purchases were kept at a minimum. PCR amplifications with microsatellite pr imers using the 3 primer system were 15 µl volumes in 96 well plates. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the labeled M13(21) and reverse primers, 0.026 µM of the forward primer and 0.6 µl (5-20 ng) of each sample. Five primers were prepared conventi onally with fluorescent labels. PCR amplifications for these primers were also 15 µl volumes in 96 well plates. For these, 1x Sigma™ (Sigma Chemical Co., St Louis) PCR Buffer was used, 1 unit of Sigma JumpStart® Taq , 1.5 mM MgCl2, 200 µM each of dNTPs, 0.26 µM each of the reverse and labeled forward primer, and 0.6 µl (5-20 ng) of each sample. After analysis of test plates for succe ssful amplification a nd labeling, pool-plex 96 well plates were prepared with 2-4 loci combinations and se nt for processing to the Sick Children’s Hospital, Toronto, for analys is using ABI Prism® 3700 96 capillary sequencers. Fragments were analyzed using GeneMapper v 3.0 software (Applied Biosystems, Inc. 2002). Haplotype anal ysis was conducted using Arlequin v2.0 (Schneider et al . 2000) and v3.0 (Excoffier et al . 2005), and PopGene v. 1.31 (Yeh et al. 1999). Results Cytochrome b Two partially overlapping cytochrome b sequences for each of the 214 samples were evaluated, corrected by analysis of ch romatograms and use of reference sequences,

PAGE 172

152 and concatenated using Sequencher 4.5. Some samples were re-amplified as necessary for base verification and production of erro r-free final sequences. The concatenated sequences for 214 samples, averaging 1143 bp, were compared and analyzed for pairwise differences to determine haplotypes. New haplotypes were analyzed with maximum parsimony (MP), maximum likelihood (ML) a nd minimum evolution (ME) comparisons for associations with other haplotypes. A c onservative approach was taken in establishing haplotypes and samples were reamplified and sequenced repeatedly. The evolving group of haplotypes was cons istently re-analyzed in Sequencher 4.5, especially the 12 ‘singleton’ haplotypes represented by only 1 individual. Matrices of pairwise distances were calc ulated, haplotype trees were constructed and one step associations were carefully analyzed to ve rify differences. Haplot ype samples were reamplified and re-sequenced to confirm any questionable base calls. Sample DNA extractions do not amplify with equal success and some templates consistently presented problems with background noise. Upon fina l haplotype assignment, the group was assembled into a contig in Sequenche r 4.5 and imported to Paup v.4.0b10 for further analyses. Additional analyses were perfor med with Arlequin v.2000 and v.3.0, as well as DNAsp v.4.10.4 (Rozas et al . 2003). As the haplotype gr oup was assembled, the sample sequence data set was repeatedly compar ed and assigned as needed. Twenty-five cytochrome b haplotypes were finally determined. Analysis began by assessing the haplotypes as one group. The samples, and also the haplotypes as seen below, were split in to two groups representing the two major distribution areas (north a nd south) (see Fig. 6-1).

PAGE 173

153 Figure 6-1. River drainage di visions used for segregating samples from the Amazon (northern) distribution of Caiman yacare . B = río Beni, M = río Mamoré, SM = río San Miguel, P = río Paraguá. These were further divided into ‘populations’, taki ng into consideration the sampling localities and hydrogeographic topog raphy. Although the yacare caiman is not principally a riverine crocody lian, they are always present even in the largest rivers. River systems are assumed to be an importa nt conduit for migration and resulting gene flow and are used here to help defi ne boundaries for the analyses. The Amazon distribution (northern) was divi ded into 4 main drainages: the río Beni, río Mamoré, río San Miguel and río Paraguá.

PAGE 174

154 The Paraguay-Paraná (souther n) distribution represents a very different, linear topology and was divided into two portions (See Fig. 6-2). It has been noted in molecular studies of fish that genetic diversity gene rally increases downstream, especially when taking into account additive functions of major tributaries. In the s outhern distribution of Caiman yacare , the río Pilcomayo is a major tribut ary with very large populations along its considerable length (~700 km). With this in mind, the arbitrar y division was placed just upriver of the union the río Pilcomayo and the río Paraguay. Figure 6-2. River drainage divi sion used for segregating samp les from the río ParaguayParaná (southern) distribution of Caiman yacare . Py = río Paraguay, Pa = río Paraná. Note length and location of western tributary río Pilcomayo. After all sample haplotypes were determined, the samples were assigned and a drainage distribution table was constructed (see Table 63). Of the 214 samples, 16

PAGE 175

155 haplotypes are distributed in the northern di stribution and 9 in the southern distribution. There is a complete division of haplotypes with none shar ed in both distributions. A comparison of all haplotypes (1197 bp) reveal ed 23 polymorphic sites with 10 singleton variable sites and 12 parsimony informative site s. A total of 1063 invariable sites were detected and, as a result of different primer sets and inconsistent success in amplification, 112 sites had missing data. Table 6-3. Distribution of 25 cytochrome b haplotypes for Caiman yacare . No haplotypes detected are shared among basins. Localit y for original haplotype specimens: Bo=Bolivia, Pa=Paraguay. Haplotype: Beni Mamore San MiguelParagua Paraguay Parana Amazon N=20 N=29 N=74 N=20 N=44 N=22 Cy01Bo 2 0 0 0 0 0 Cy03Bo 16 18 39 0 0 0 Cy05Bo 2 7 7 0 0 0 Cy02Bo 0 1 0 0 0 0 Cy04Bo 0 1 15 18 0 0 Cy29Bo 0 1 1 0 0 0 Cy30Bo 0 1 2 0 0 0 Cy07Bo 0 0 1 0 0 0 Cy12Bo 0 0 1 0 0 0 Cy14Bo 0 0 1 0 0 0 Cy21Bo 0 0 2 0 0 0 Cy22Bo 0 0 1 0 0 0 Cy23Bo 0 0 1 0 0 0 Cy28Bo 0 0 3 0 0 0 Cy08Bo 0 0 0 1 0 0 Cy27Bo 0 0 0 1 0 0 Paraguay-Paraná Cy15Bo 0 0 0 0 22 1 Cy16Bo 0 0 0 0 3 3 Cy17Bo 0 0 0 0 1 0 Cy18Bo 0 0 0 0 0 1 Cy19Bo 0 0 0 0 10 17 Cy20Bo 0 0 0 0 1 0 Cy24Bo 0 0 0 0 3 0 Cy25Pa 0 0 0 0 1 0 Cy26Pa 0 0 0 0 2 0 The haplotypes trees were constructe d using PAUP 4.10b with MP, ME and ML evolutionary models without resolving all haplotype asso ciations. Haplotypes always

PAGE 176

156 segregated between the two distribution ba sins with none shared, but the internal topologies varied and many branches had poor bootstrap support (see Figure 6-3). Figure 6-3. Maximum parsimony (MP) tree for 25 cytochrome b haplotypes of Caiman yacare rooted with Caiman crocodilus . Number above branch refers to mutational steps, number below refers to percentage consensus of 98 equally parsimonious trees. (length=158, CI=0.918). Bracket A = Amazon basin, northern distribution, P = Paraguay-Pa raná basin, southern distribution. Haplotypes are unique to each dist ribution. Abbreviations: Cc_REF( Caiman crocodilus reference sequence), Cy_ ( C. yacare plus haplotype number). Maximum parsimony (MP) analyses used heur istic searches starting with stepwise addition trees and replicated 100 times. Br anch swapping was performed by the treebisection-reconnection (TBR) method. While 98 MP trees were constructed with the

PAGE 177

157 identical scores, they had high concordance on the main branch arrangements and differed mainly on tip associations, especially within the southern distribution (see Figure 6-3). This becomes evident when a comparison is made between the 2 distributions with this 1161 base data set. The northern distribution (16 cyt b haplotypes, mean 1155 base length) has 7 variable parsimony uninformative sites with 7 parsimony informative sites. The mean frequency base compositions are 30.3%A, 25.4%T, 34.2%C, and 10.1%G. The southern distribution (9 cyt b haplotypes, mean 1151 base length) has 9 variable parsimony uninformative sites with only 2 parsimony informative sites. The mean frequency base compositions are 30.2%A, 25.2%T, 34.3%C, and 10.3%G. These data show an under-representation of guanine as described for the mitochondrial genome by Zhang and Hewitt (1996). The Tamura and Nei model (TrN+I) of e volution was selected using Modeltest 3.06 for ML analyses. Excess transitions, unequal nucleotide frequencies and a variation of substitution rate among different sites are taken into account in this model. In addition, the proportion of invariable sites (+I) is used in the calculation. Employing the likelihood test ratio (LTR), a molecular clock mode l was compared and accepted and a phylogram tree, rooted at the C. latirostris split at 24 mybp, was constructed and measured. Results show that the haplotype dive rgence time between basins is relatively recent and is estimated at ~1.5 mybp an subsequent within basin radiation of populations occurred in the last 500,000 years. This follows a similar pattern seen for Caiman crocodilus cytochrome b sequences, with the divergence betw een the Amazon and Orinoco basins occurring at ~2.75 mybp and within basin ra diation occurring during the last million years (see chapter 5).

PAGE 178

158 As many haplotypes differ with single muta tional steps, haplot ypes trees could not be easily resolved using bootstrap values. A haplotype network was constructed using TCS v.2.1.1 to better understand the phyloge ny. A 15 single mutational step difference occurs between the 2 most distant haplotype s in the network. Within both haplotype networks, a 7 mutational step difference occurs between the most distant haplotypes (see Fig. 6-4). While some haplot ype reticulations remain unres olved, the overall pattern and distinction between basins are clear. The separation between basin clades is 2 or 3 mutational steps depend ing on the reticulation. Figure 6-4. Genealogy network for 25cytochrome b haplotypes in Caiman yacare . Geographic division (blue line) occurs between river basins: A = Amazon, P = Paraguay-Paraná. Branch length between haplotypes and nodes equal one base change. Red symbols indicate the most common haplotypes in each distribution. Haplotype Cy03 (rectangle) may repres ent the ancestral line.

PAGE 179

159 Results from an Analysis of Molecular Variance (AMOVA) (Excoffier et al., 1992) implemented in Arlequin v3.0 (Excoffier et al . 2005) added information on the distribution and separation of haplotypes: 72.4% of the variat ion was attributed between the 2 basins, 7.7% of variation was attributed between populations w ithin the basins, and 19.9% of the variation was attributed to within population differences. The pairwise distance method Fst value between basins was 0.7576, si gnificant at the p <0.001 level. Within the northern distribution, two co mmon haplotypes account for 76.2% of the 143 samples from the basin (Cy_03=51%, Cy_04=25.2%). Private haplotypes (i.e. occurring in only one population) were de tected in all 4 populations of the Amazon distribution. A similar situation exists in the southern distribution. Two common haplotypes account for 75.7% of th e 66 samples from the basin (Cy_19=40.9%, Cy_15=34.8%). Private haplotypes we re also detected in both populations of the southern distribution. An AMOVA analysis on the drainage populations provides additional information on within and between basin comparisons (see Table 6-4). Table 6-4. Analysis of Fst values for populations of Caiman yacare . Top set: Population pairwise distance Fst values, Bottom set: Matrix of significant Fst values, p <0.05. Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni 0 Mamoré -0.005760 San Miguel 0.040440.007610 Paraguá 0.766670.643280.462040 South Paraguay 0.787870.766210.743530.813770 Paraná 0.894190.85349 0.797580.950180.22362 0 Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni + + + Mamoré + + + San Miguel + + + Paraguá + + + + + South Paraguay + + + + + Paraná + + + + +

PAGE 180

160 Pairwise distance comparisons were made within and between all populations and a matrix was constructed using Arlequin 3.0 (E xcoffier et al. 2005). A corrected average pairwise difference was also computed (see Table 6-5). Table 6-5. Average pairwise distances for populations of Caiman yacare . Above the diagonal: average number of pairwise differences between populations ( XY). Diagonal elements: average number of pairwise differences within population ( X). Below diagonal: corrected average pairwise difference ( XY-( X + Y)/2). Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni 1.20000 1.41034 1.66757 3.00000 7.55581 7.80909 Mamoré -0.00493 1.63054 1.81081 2.82414 7.43865 7.70219 San Miguel 0.08978 0.01775 1.95557 2.47838 7.41169 7.65479 Paraguá 2.30000 1.90887 1.40059 0.20000 6.77442 7.00909 South Paraguay 6.04662 5.71419 5.52471 5.76523 1.81838 1.56237 Paraná 6.96667 6.64450 6.43458 6.66667 0.41075 0.48485 Additional information was computed usi ng standard molecular diversity indices (see Table 6-6). Mitochondrial cyt b haplotype diversity ( h ) (Nei, 1987), and the number of transitions and transversions we re estimated using DNAsp v.4.10.4 (Rozas et al . 2003). The nucleotide diversity ( ) (Nei, 1987) was measured using Arlequin v. 3.0 (Excoffier et al . 2005). With this program, I also estim ated neutral selec tion effects based on the Tajima D test statistic that employs an infinite site model (Tajima 1989). Table 6-6. Measures of genetic diversity wi thin populations of the río Amazon (north) and río Paraguay-Paraná (south) distributions. Population No. of Haps. transition/ transversion Nei's haplotype diversity ( h ) Tajima's D Fu's Fs Nucleotide diversity ( ) North (143) 16 15 / 2 0.573 +/0.033 -0.956 -4.638 0.00169 +/-0.00106 Beni (20) 3 4 / 0 0.358 +/0.127 0.186 1.672 0.00104 +/-0.00077 Mamoré (29) 6 8 / 2 0.571 +/0.087 -0.254 -0.261 0.00140 +/-0.00095 San Miguel (74) 12 11 / 0 0.677 +/0.050 -0.364 -2.870 0.00170 +/-0.00108 Paraguá (20) 3 2 / 0 0.195 +/0.115 -1.513 -1.863 0.00017 +/-0.00025 South (65) 9 11 / 1 0.778 +/0.031 -1.089 -1.714 0.00137 +/-0.00092 Paraguay (43) 8 11 / 1 0.687 +/0.061 -1.046 -2.678 0.00161 +/-0.00105 Paraná (22) 4 2 / 0 0.398 +/0.122 -0.263 -1.546 0.00043 +/-0.00043 In order to assess potential change in popul ation size, I employed the test devised by Fu (1997) (FS) implemented in Arlequin 3.0 (Excoffier et al . 2005). This tests the

PAGE 181

161 probability of having no fewer than the obs erved number of haplotypes in a random sample. The test assumes a neutral infinite-s ites model and constant population size and significance is determined through simulations estimating the probability of obtaining FS values equal to or lower than the observed (Hudson, 1990). A Mantel test was implemented by Arlequi n v.3.0 to test the correlation between genetic pairwise distance Fst values and riverine dist ances between populations. A positive correlation was determined (r=0.6962) and was established through a permutation test (2000 permut ations) and was significant at the p<0.01 level. Spatial autocorrelations are significant following the Fst values given in Table 6-4. using the distance matrix constructed for the test (see Table 6-7) Microsatellites PCR amplifications were performed us ing primers for 12 polymorphic loci and processed as described above. In spite of ex cellent success in cyt b amplifications with the Paraguayan bone samples, microsatellite amplifications proved more difficult. The less abundant nuclear DNA is more prone to degradation over time and some loci failed to amplify in certain samples. Due to missing allelic data, some samples from the río Paraguay population of southern distributi on had to be removed from the analyses. Fragment data were analyzed and genotyped using GeneMapper v 3.0 (Applied Biosystems, Inc. 2002). Fragment lengths were assigned, tabulated into data sets and alleles determined for the 12 loci. Genotype analyses were performed using Arlequin v.2 (Schneider et al . 2000) and v3.0 (Excoffier et al . 2005), and PopGene v. 1.31 (Yeh et al. 1999). After the alleles were determined, one of the first tasks wa s to test allele frequencies for Hardy-Weinberg Equilibrium (HWE) using the G2 goodness of fit metric in PopGene v1.3.1 (see Table 6-7). For the locus Claµ-7, allele frequencies were not in

PAGE 182

162 HWE in 5 of 6 populations. Although 17 alle les were detected fort this locus, examination of the individual genotypes re vealed very high frequencies of a small number of alleles. The most common a llele accounted for 43.5%, and the 3 most common alleles totaled 83.6%. This locus was removed in subsequent analyses due to potential linkage or selecti on that could bias analyses. Table 6-7. Microsatellite loci, al lele number and frag ment size data used for analyses in Caiman yacare. Test of allele frequencies for Hardy-Weinberg Equilibrium (HWE). (+ = frequencies in HWE, = frequencies out of HWE). Locus No. (sizes) Beni Mamoré Sn MiguelParaguá North Paraguay Paraná South Claµ-5 40 (173-271) + + + + + + + Claµ-4 12 (275-297 + + + + + + + Cy-R8 21 (235-277) + + + + + + + Cy-022 19 (179-221) + + + + + + + + Cy-I14 17 (235-273 + + + + + + + Cy-F2 26 (203-263 + + + + + + + + Cy-S4 27 (183-267) + + + + + + + + Cy-N10 31 (293-399) + + + + + + + + Claµ-10 32 (213-253 + + + + + + + + Cy-D2 32 (253-317 + + + + + + + + Claµ-6 22 (169-269 + + + + + + + + Claµ-7 17 (181-229) + Further inquiry revealed that the southern Paraná populati on allele frequencies were not in HWE in 4 of the remaining 11 loci. Combining the 2 populations, all 11 loci had allele frequencies in HWE (see Table 6-7). Subsequent analyses were performed with these populations in both combined a nd uncombined states, as noted below. Several statistical analyses were perf ormed to detect population differentiation. Pair-wise locus comparisons fo r all populations were performe d, and allele frequencies, observed heterozygosity (HO) and expected heterozygosity (HE) were calculated using PopGene 1.31 (see Table 6-8 and 6-9). The pa irwise comparisons show structuring

PAGE 183

163 similar to that seen in the mitochondrial haplotypes, with overall differences generally greater between populations of the north and south distribution than among them. Table 6-8. Average pairwise distances for populations of Caiman yacare . Above the diagonal: average number of pairwise differences between populations ( XY). Diagonal elements: average number of pairwise differences within population ( X). Below diagonal: corrected average pairwise difference ( XY-( X + Y)/2). Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni 9.11282 9.44009 9.58125 9.24554 9.37292 9.62554 Mamoré 0.31804 9.13128 9.57232 9.27124 9.42601 9.70915 San Miguel 0.23973 0.22157 9.57022 9.34710 9.38209 9.62998 Paraguá 0.72776 0.74424 0.60062 7.92273 8.17560 9.16925 South Paraguay 0.92289 0.96675 0.70336 0.32061 7.78723 9.18342 Paraná 0.66092 0.73529 0.43666 0.79968 0.88159 8.81643 The levels of observed heterozygosity (HO) of the six populations are within a relatively narrow range with the río Mamoré having the lowest value (0.736) and the río Paraguá having the highest (0.836). The two di stributions (north a nd south) displayed identical HO values of 0.811 (see Table 6-9). The mean allele number of the various populations sampled ranged from 9.6 (río Beni ) to 18.2 (río San Miguel) over the eleven loci used in the final analyses. The mean al lele number for the northern distribution was 22.2 and 18.5 for the southern distribution. Many population studies that u tilize microsatellite data derive additional resolution through the use of private al leles. These unique alleles occur only within a given population or region. Due to the hyper-variable na ture of a number of loci (3 loci with more than 20 alleles), and limited samples fr om several populations, private alleles could not be determined with any certainty. Sampli ng more individuals in all populations would give a more accurate measure for privat e alleles. Given the relatively low Fst values and apparent gene flow in C. yacare , private alleles may not be common.

PAGE 184

164Table 6-9. Summary statistics for all loci across all sampled populations for Caiman yacare . North = Beni, Mamoré, San Miguel and Paraguá, South = Paraguay and Paraná. N=number of individuals sampled, AN = number of alleles per locus, HO = observed heterozygosity, and HE = expected heterozygosity. Beni Mamoré San Miguel Paraguá Locus N AN HO HE Locus N AN HO HE Locus N AN HO HE Locus N AN HO HE Claµ-5 20 10 0.850 0.891 Claµ-5 29 19 0.793 0.888 Claµ-5 73 30 0.822 0.942 Claµ-5 23 21 0.870 0.950 Claµ-4 18 7 0.778 0.632 Claµ-4 27 9 0.630 0.846 Cl aµ-4 70 10 0.700 0.723 Claµ-4 18 5 0.556 0.549 Cy-R8 20 12 0.950 0.936 Cy-R8 29 11 0.828 0.864 Cy-R8 73 18 0.863 0.913 Cy-R8 23 11 0.870 0.852 Cy-022 20 10 0.800 0.877 Cy-022 29 13 0.931 0.912 Cy-022 73 14 0.904 0.899 Cy-022 23 10 0.870 0.836 Cy-I14 19 7 0.632 0.764 Cy-I14 26 8 0.539 0.732 Cy -I14 72 13 0.681 0.880 Cy-I14 21 8 0.857 0.681 Cy-F2 20 9 0.950 0.859 Cy-F2 29 11 0.793 0.891 Cy-F2 73 16 0.863 0.896 Cy-F2 23 14 0.826 0.873 Cy-S4 20 7 0.950 0.856 Cy-S4 29 13 0.828 0.895 Cy-S4 73 18 0.863 0.898 Cy-S4 23 15 0.913 0.905 Cy-N10 20 10 0.950 0.882 Cy-N10 29 16 0.931 0.914 Cy-N10 73 26 0.904 0.942 Cy-N10 23 15 0.957 0.905 Claµ-10 20 8 0.900 0.835 Claµ-10 29 11 0.690 0.829 Cl aµ-10 73 23 0.740 0.891 Claµ-10 23 7 0.565 0.781 Cy-D2 20 15 0.800 0.942 Cy-D2 27 13 0.704 0.918 Cy-D2 72 26 0.792 0.940 Cy-D2 23 18 0.913 0.951 Claµ-6 20 10 0.650 0.827 Claµ-6 29 11 0.862 0.819 Claµ-6 73 14 0.836 0.758 Claµ-6 23 13 0.913 0.836 Mean 20 9.6 0.784 0.825 Mean 26.5 11.9 0.736 0.848 M ean 72.5 18.2 0.784 0.867 M ean 22.5 12.2 0.799 0.822 Paraguay Paraná North South Locus N AN HO HE Locus N AN HO HE Locus N AN HO HE Locus N AN HO HE Claµ-5 27 21 0.778 0.945 Claµ-5 21 8 0.476 0. 803 Claµ-5 145 35 0.828 0.945 Claµ-5 48 23 0.646 0.922 Claµ-4 19 7 0.632 0.683 Claµ-4 19 3 0.368 0. 551 Claµ-4 133 12 0.677 0.757 Claµ-4 38 7 0.500 0.617 Cy-R8 27 13 0.815 0.811 Cy-R8 22 7 0.455 0. 748 Cy-R8 145 20 0.869 0.914 Cy-R8 49 13 0.653 0.849 Cy-022 29 13 0.828 0.896 Cy-022 22 10 0.682 0.803 Cy-022 145 17 0.890 0.911 Cy-022 51 14 0.765 0.874 Cy-I14 24 10 0.792 0.852 Cy-I14 16 8 0.750 0. 883 Cy-I14 138 13 0.674 0.840 Cy-I14 40 12 0.775 0.867 Cy-F2 29 7 0.379 0.711 Cy-F2 22 6 0.227 0. 632 Cy-F2 145 20 0.855 0.909 Cy-F2 50 10 0.620 0.548 Cy-S4 29 18 0.862 0.935 Cy-S4 22 16 0.909 0.874 Cy-S4 145 21 0.876 0.910 Cy-S4 51 21 0.882 0.928 Cy-N10 28 17 0.929 0.934 Cy-N10 21 13 1.000 0.891 Cy-N10 144 28 0.924 0.933 Cy-N10 49 19 0.959 0.924 Claµ-10 29 11 0.690 0.855 Claµ-10 22 11 0.318 0.864 Claµ-10 145 28 0.724 0.884 Claµ-10 51 14 0.529 0.862 Cy-D2 23 18 0.870 0.911 Cy-D2 18 12 0.889 0.886 Cy-D2 142 30 0.796 0.947 Cy-D2 41 22 0.878 0.942 Claµ-6 29 8 0.931 0.844 Claµ-6 22 7 0.955 0. 831 Claµ-6 145 17 0.828 0.825 Claµ-6 51 8 0.941 0.837 Mean 26.6 13 0.773 0.853 mean 21 8.9 0.640 0.797 mean 143 21 0.813 0.889 mean 43 14 0.713 0.846

PAGE 185

165 Fst values of population subdivision were calculated using Arlequin3 (Excoffier et al . 2005) according to Weir and Cockerham ( 1984). These are frequently compared to Rst values when using mi crosatellite data. Rst, values (Saltkin 1995) are derived from the sum of squares allele size differences over loci using Arlequin3 (Excoffier et al . 2005). Analyses of molecular variances (AMOVA) we re calculated using both statistics for comparison. Populations were analyzed separately and grouped according to the Northern (río Amazon) and Southern (río Paraguay-Par aná) distributions. Bo th metrics exhibited population structuring similar to the results found with the mtDNA sequences. Comparison of the results can be seen in Tables 6-10 and 6-11. Table 6-10. Analyses of Molecula r Variance (AMOVA) of sampled Caiman yacare populations for Fst (upper panel) and Rst (lower panel) values. Results in panels are presented for populations an alyzed separately and grouped into Northern (río Amazon) and Southern (río Paraguay-Paraná) distributions. Source of variation-Fst d.f. sum of squares Variance components Percentage of variation among populations 5 106.212 0.2732 Va 5.77 within populations 386 1721.457 4.4597 Vb 94.23 total 391 1827.668 4.7329 Source of variation-Fst d.f. sum of squares Variance components Percentage of variation among groups 1 45.474 0.1944 Va 4.02 among populations 4 60.737 0.1765 Vb 3.65 within populations 386 1721.457 4.4597 Vc 92.32 total 391 1827.668 4.831 Source of variation-Rst d.f. sum of squares Variance components Percentage of variation among populations 5 101604.829 307.03474 Va 17.39 within populations 386 563176.811 1459.0073 Vb 82.61 total 391 664781.640 1766.042 Source of variation-Rst d.f. sum of squares Variance components Percentage of variation among groups 1 67947.680 387.121 Va 19.75 among populations 4 33657.149 114.439 Vb 5.84 within populations 386 563176.811 1459.007 Vc 74.42 total 391 664781.640 1960.567

PAGE 186

166 From these data, it is clear that the ma jority of the variation detected is found within the populations. Vari ation detected among groups and among populations is minimal due to the high overlap of shared alleles. Table 6-11. Analysis of Fst and Rst values for microsatellite alleles in populations of Caiman yacare . Top set: Population allele Rst values (sum of square differences) above diagonal; pairwise distance (number of alleles) Fst values below diagonal, Bottom set: Matrix of significant values: Rst, p 0.05 above diagonal; Fst, p <0.05 below diagonal. Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni 0.0036 0.0456 0.1741 0.3012 0.3550 Mamoré 0.0337 0.0211 0.1410 0.2410 0.2822 San Miguel 0.0243 0.0227 0.1061 0.2368 0.2677 Paraguá 0.0688 0.0755 0.0439 0.2366 0.1934 South Paraguay 0.0800 0.0810 0.0610 0.0878 0.0773 Paraná 0.0994 0. 1016 0.0704 0.09622 0.0392 Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni + + + + Mamoré + + + + + San Miguel + + + + + Paraguá + + + + + South Paraguay + + + + Paraná + + + + + A Mantel test was implemented by Arlequi n v.3.0 to test the spatial correlation between population Rst values and riverine distances. A positive correlation was determined (r=0.8532) and was establis hed through a permutation test (2000 permutations) and was significant at the p< 0.01 level. Spatial autocorrelations are significant following the Rst values using the distance matrix constructed for the test (see Table 6-12). Table 6-12. Matrix of Rst and geographic distances used fo r Mantel test. Upper diagonal contains riverine distances (kilometers) among the sampled Caiman yacare populations. Lower diagonal contains the Rst values (Slatkin 1995). Beni Mamoré San Migue l Paraguá Paraguay Paraná North Beni 900 1100 1350 2000 2750 Mamoré 0.0036 750 1000 1500 2250 San Miguel 0.0456 0.0211 750 1500 2250 Paraguá 0.1741 0.1410 0.1061 1000 1750 South Paraguay 0.3012 0.2410 0.2368 0.2366 750 Paraná 0.3550 0. 2822 0.2677 0.1934 0.0773

PAGE 187

167 Claµ_50.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14 0.16 0.18173 181 191 1 9 5 2 0 1 2 0 5 2 0 9 213 217 221 2 25 2 29 2 3 3 2 3 7 2 4 1 245 249 2 53 2 67allelesfrequency Claµ_40.00 0.10 0.20 0.30 0.40 0.50 0.60275277279281283285287289291293295297allelesfrequency Cy_R80.00 0.05 0.10 0.15 0.20 0.25 0.30235239243247251255259263267273277allelesfrequency Cy_O220.00 0.05 0.10 0.15 0.20 0.25 0.30179189193197201205209213217221allelesfrequency Cy_I140.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35237243249253257261265271allelesfrequency Figure 6-5. Allele frequency distributi ons of five microsatellite loci for Caiman yacare in the two main distribution areas: Northern (Amazon – light blue) and Southern (Paraguay-Paraná – dark grey). Nort hern distribution includes populations from rios Beni, Mamoré, San Miguel and Paraguá. Southern distribution includes the rios Paraguay and Paraná. All tested microsatellite loci were shared over the six populations of Caiman yacare as designated for this study. Due to high allele numbers for several loci, population sampling was likely insufficient to accurately determine presence, or frequency, for many alleles within several popu lations. In general, possibly due to the lower sample numbers in the Southern group, some alleles were not detected. This

PAGE 188

168 absence of alleles could be due to sampling error or may re flect genetic drift. Allele frequencies were pooled for populations of th e Northern and Southe rn distributions for comparison (see Figures 6-5, 6-6). All loci s how differences in fr equency distributions between the Northern and Sout hern populations. Some distri butions are more informative as seen with locus Cy_F2 (see Figure 6-6). Cy_F20.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70203211217225229233237241245249253257261allelesfrequency Cy_S40.00 0.05 0.10 0.15 0.20183187203211215219223227231235239243247allelesfrequency Cy_N100.00 0.05 0.10 0.15 0.20293 2 9 7 3 0 1 30 7 311 3 1 5 31 9 323 3 2 7 3 3 1 335 339 3 4 3 3 4 7 357allelesfrequency Claµ_100.00 0.05 0.10 0.15 0.20 0.25 0.30213 219 2 2 3 2 2 9 235 2 5 1 2 5 5 259 263 2 6 7 271 275 2 7 9 2 8 5 289 2 9 5allelesfequency Cy_D20.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14251 259 263 267 271 275 279 283 287 291 295 299 303 307 31 1 315allelesfrequency Claµ_60.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35169171173175177179181183185187189191193197201allelesfrequency Figure 6-6. Allele frequency distributi ons of six microsatellite loci for Caiman yacare in the two main distribution areas: Northern (Amazon – light blue) and Southern (Paraguay-Paraná – dark grey). Nort hern distribution includes populations from ríos Beni, Mamoré, San Miguel and Paraguá. Southern distribution includes the ríos Paraguay and Paraná.

PAGE 189

169 Discussion Populations of Caiman yacare are subject to strong, le gal harvest pressures. Thousands of adult animals are culled annually from wild populations in Bolivia. The harvest program has been temporarily suspe nded in Paraguay but preparations are being made to resume. Yet surprisingly, little information has been accumulated about the biology of the yacare caiman. Much of what is known comes from the Brazilian Pantanal where conditions are unique and wild harv est is not permitted. However, wildlife managers must incorporate all relevant information in order to make informed decisions regarding sustainable use of this species. Data generated from this study undersco re the potential and the limitations of molecular information. These data complimen t knowledge on the biology of the species. From the mtDNA sequence data, derived from ma trilineal inheritance, a distinct split is seen between the North and South haplotype distributions (see Tabl e 6-3, Figures 6-3). Results from the Analysis of Molecular Va riance (AMOVA) attributes 72.4% of the variation between the two distributions. Considering the sample localities and overall sample size, shared haplotypes would likely have been detected. Of the 214 anim als sampled for this study, 22 caiman were from the Bolivian Pantanal. This is clos e to the Amazon basin boundary, an area of potential mixing between the North and South distributions. Additiona lly, 9 samples were from northern Paraguay along the main río Para guay corridor, relatively close to another potential mixing area (see Figure 2-9). These data suggest that th e geographic separation of the río Amazon basin (Northern distributi on) and the ríos Para guay-Paraná basin may be a significant barrier for migration of females.

PAGE 190

170 Large adult male caiman establish breeding areas which are defended against competing males. These dominant male patrol their area and aggressively exclude not only reproductive adult males, but also sub-adults (Thorbjarnarson 1991a, 1993; Coutinho 2000). Studies on seasonal movements show that while caiman are displaced by the rigors of the intense wet and dry season re gime, their net movement is generally not great. Adult females, however, show less move ment and the adult males respond to this by establishing territories relatively close to the dry season refugia. In this manner, sub adult males are forced further away from the dry season habitat (Thorjarnarson 1991a, 1993). The genealogy network graphically displa ys the division betw een haplotype groups of the two major distributions (see Figure 6-4). A minimum of 5 evolutionary steps separate the two haplotype groups. Most re lationships of within group haplotypes are separated by single evolutiona ry step. Although some relati onships between haplotypes cannot be resolved by transition-transversion analyses and lead to multiple pathways, these reticulations do not di srupt the overall structure of the 2 group networks. The effect of the large basin haplotype di vision is emphasized by the relative within group uniformity. The Fst values for cytochrome b haplotypes indicate that while some geographic structuring is evident, the rí o Beni, the río Mamoré and río San Miguel populations are not significantly different fr om each other (see Table 6-4). This is understandable given the topography and e nvironments resulting from periodic wet season flooding. Annual flooding events often ne gate drainage boundaries and large-scale pooling, particularly acute approaching rive r junctions, frequen tly extend hundreds of km2. This region is covered with isolated lake s and ponds of various sizes that may serve

PAGE 191

171 as ‘stepping stones’ for gene flow between river drainages that provide much shorter distances than linear riveri ne mileage would imply. Ephemeral streams crossing the open savanna s may also serve as genetic conduits. These conditions of seasonally coalescing wa terbodies, permanent lakes and intermittent streams are common but found to a lesser degree in the eastern region of río San Miguel. For this reason, Fst values are slightly hi gher than Bemi or Mamoré values but still not significant at the p <0.05 level. I coll ected 74 samples from the río San Miguel population and similarities detected in sample s from both the río Beni (N=20) and río Mamoré (N=29) populations reflect the fact that they are largel y subsets of the San Miguel population. This is not appear to be the case wi th the río Paraguá population (N=20). The geography of the area has isolat ed the river to a greater de gree than the other three populations mentioned above. The Fst values indicate the distinction although this is primarily due to the presence of 2 unique haplotypes. In a comparison of Fst values and average pairwise distances, it can be observed that the greatest similarity for populations w ithin the Northern dist ribution is with the nearest geographic neighbor (see Tables 6-4 and 6-5). For the Southern distributi on, the río Paraguay population is significantly different from the Paraná population at p <0.05 level (s ee Tables 6-4 and 6-5) . In the linear hydrogeographic setting of this distri bution, there are no consistent short-cut conduits for gene flow outside of the main river corridors as seen on the flooded sa vannas of Bolivia (see Chapter 2). In this sense, riverine and st raightline distances are essentially the same.

PAGE 192

172 The genetic diversity is su mmarized in Table 6-6. The NeiÂ’s haplotype diversity, h , is defined as the probability that two randomly chosen haplotypes in a sample are different (Nei 1987). Few mol ecular studies have been undertaken for Alligatoroidea. Data from Glenn et al (2002) with the Am erican alligator, Alligator mississippiensis showed surprisingly low mtDNA diversity (n=25, 0.153, 1317 bp). This has been attributed to an historic bottleneck. In Caimaninae , Farias et al .(2004) described values for common caiman, Caiman crocodilus (n=32, 0.692, 1192 bp) and black caiman, Melanosuchus niger (n=47, 0.715, 1192 bp) that demonstr ate much higher diversity. A recent study over a wider geographic range with C. crocodilus (Vasconcelos et al .2006) reported a higher value (n=125, 0.733, 1085 bp). Results from this study show a high overall genetic diversity (n=208, 0.824, 1175 bp) for the combined North and South distributions. Genetic equilibrium of the mtDNA haplot ypes was examined using the TajimaÂ’s D test (Tajima 1989) and FuÂ’s Fs test (Fu 1997). These tests were designed to examine selection but a significant de viation from genetic equilibrium with mtDNA haplotypes is likely a result from recent bottlenecks and/or population expansions where no selective advantage exists (Hartl a nd Clark 1997). Test were performed on the 6 individual populations, the North and South di stributions, and for the total of all samples. Results of these tests were non-significant at p <0.05 leve l. Populations of yacare caiman have been depressed in some areas of Bolivia (King a nd Videz 1989) due to unregulated harvest but these test results do not i ndicate any consequential popula tion expansion. Although local populations may have been reduced in the recent past, it appear s that the population genetic diversity has not been comp romised at any tested scale.

PAGE 193

173 Results of a Mantel test also confirms structuring according to river geography. There is a positive correlation of isolation by distance. These data indicate an important processes of haplotype distribution among the populations. Movements by female caiman are sufficient to distribute common mtDNA hapl otypes across large area s, but restricted enough for the evolution and establishment of geographically isolated haplotypes. Microsatellite analyses present both sim ilar and conflicting results. These loci are bi-parentally inherited an d spread throughout the nuclear genome. The loci are theoretically under neut ral selection pressure although th ey may hitch-hike if closely associated with a beneficial allele under selection pressure. It also appears that microsatellite mutation rate may increase over time with increas ed repeat number (Schlötter 2000). Alternatively, extremely long microsatellites are uncommon and there may be some selection for maintaining micros atellites within certain size limits (Nauta and Weissing 1996). This may help explain th e conservation of loci among species. Loci from a related caiman, Caiman latirostris , were successfully amplified for Caiman yacare in this study (Zucoloto et al .2002). Primers developed for loci from Alligator mississippiensis also amplified microsatellites for C. yacare, though not used in this study, despite a separation of a pproximately 70 million years. Microsatellite heteozygosity in the sampled populations of Caiman yacare are higher than, but comparable to, values repor ted in the few studies that have been conducted with crocodylians. The overall he terozygosity was 0.82 with the lowest individual population Ho value found in the río Paraná at 0.64. Davis et al . (2002) reported an overall heterozygosity of 0.69 w ith the Marsh Island, LA population showing a high of Ho of 0.79, and low of 0.55 in the Santee, SC population. Gratten (2003) tested

PAGE 194

174 several Crocodylus species w ith values from 0.24-0.46 (see Table 6-13). A study of 48 non-piscine vertebrate speci es found heterozygosities of 0.58 ± 0.25 (Johns and Avise 1998; DeWoody and Avise 2000). Table 6-13. Comparison of microsatellite hetrozygosities for crocodilian species. Ho Loci N Source Caiman yacare 0.8111192this study Alligator mississippiensis 0.691131.2Davis et al . 2002 Crocodylus moreletti 0.49933.3Dever et al . 2002 Crocodylus porosus 0.467148Gratten 2003 Crocodylus siamensis 0.36944Gratten 2003 Crocodylus novaeguineae 0.24676Gratten 2003 Results from the AMOVA with Rst values show low varia tion between distributions (19.5%) and high variation found within the popu lations (74.4%). As a result of the high numbers of alleles per locus (12–40), and th e relatively small sample size for some populations, fine-scale structuring could not be demonstrated. Results from a Mantel test show a correlation of Rst valu es and distance that is similar to that seen with the cytochrome b sequence data. Future efforts will be undertaken to verify the allele sizes and determine if there were amplificati on errors. Efforts are underway to obtain additional samples that may help determine the extent of structuring through use of microsatellites. It has been shown that phyl ogeographic patterns that are shared by sympatric or codistributed species may shed light on a co mmon historic biogeogr aphy (Avise 2000). In this manner, comparative molecular analyses for the sympatric crocodylian species may help the understanding of the development of the Amazon. Numerous vertebrate species are common to both the Amazon and Paraguay river basins but no comparative studies

PAGE 195

175 have been undertaken. Furthe r investigation into the mol ecular patterns exhibited by these species may assist in reconstructing the geographic histories that until now have proved elusive. Conservation Genetics Genetic information is serving an incr easing role in cons ervation biology. New techniques have been developed that produce more data, more accurately in less time and for less money than previously imagined. The utility of the resulting data depends on the unique situation of species in question. For example, genetic profiling is crucial for work with endangered species when considering captive breeding, re-introduction programs or in population viability modeling. This , however, is not the case with Caiman yacare . The yacare caiman, along with the common caiman, C. crocodilus , are perhaps the two most numerous crocodylian species. C onservation genetics have a much different role to play in the management of abundant populations. Da ta generated from the analyses in this study as documented above will have impact in the following areas: Taxonomic clarification – Conservation efforts can become focused when taxonomic uncertainties are resolved. Daugherty et al . (1990) observed that accurate taxonomies “are not irrelevant abstractions, bu t the essential foundations of conservation practice”. Caiman yacare is now regarded by some resear chers as a subspecies of the common caiman. This unwarranted convention has been adopted by some authors since Werner (1933) first referred to the yacare caiman as Caiman crocodilus yacare . This Apparently, this is based only on superficial si milarity rather than biological objectivity and is counter to the rules of the Intern ational Commission of Z oological Nomenclature. The data presented above emphasizes the position of C. yacare as a species. This presents no legal ramifications as CITES, the US Fish and Wildlife Service, the European Union

PAGE 196

176 and all the range states (Argentina, Bolivia , Brazil, and Paraguay) have previously regarded C. yacare as a species in their laws and regulations. The unexpected results regarding the geneti c distance as descri bed above with the Caiman crocodilus fuscus haplotypes (see Chapter 5) mu st be resolved through further study. Review of the biogeogra phic history suggests compar able reproductive separation from the Orinoco basin populations of Caiman crocodilus . This may lead to a new species designation for the Central Am erican populations, now considered C. c. chiapasius . Resolving the identity for the Pacifi c drainage populati ons from Colombia, also considered C. c. chiapasius , is also necessary but very difficult given the political situation there. There remains the uncertain identity of populations from the Caribbean drainages of northern Colombia and western Ven ezuela. Genetic analyses of these C. c. fuscus ( sensu stricto ) populations would help complete the ge netic mosaic of this wide ranging group. This could be undertaken with the co llaboration of Colombian and Venezuelan colleagues. A taxonomic change for thes e latter populations might entail legal repercussions, as Colombian hide exports fr om farming operations in the region are very large (>500,000/yr, Velasco and de Sola 2005) an d the regulations have been in place for over 15 years. Data presented here also show significan t genetic distance between the Orinoco and Amazon populations of C. crocodilus . Reproductive isolation is complete owing to geographic separation and the processes toward separate speciation continue. Further investigations are required to reveal the extent of sepa ration and clarify the taxonomic status.

PAGE 197

177 Genetic diversity – Wildlife species are typicall y characterized according to the genetic diversity detected within and betw een the various subpopulations. Appropriate species management will seek to quantify and maintain a broad level of natural hetrozygosity. Relative heterozygosity levels shou ld be maintained to avoid inbreeding or bottlenecks. The results of this st udy show that the populations of C. yacare show a relatively high degree of genetic diversity co mpared to non-piscine vertebrates. Large population numbers remaining in the range st ates have maintained high numbers of haplotypes in the cytochrome b gene and alleles for the micr osatellite sites surveyed. No indications of inbreeding or reduction in heterozygosity due to population bottlenecks were detected as measured by Fis values (data not shown). As this species is subject to commercial exploitation that selectively eliminates large adult males, wildlife managers will ne ed to periodically monitor the genetic diversity, drift or reduction in heterozygosity. Controversy exists over this type of harvest strategy, especially in a sp ecies where dominant males represent a disproportionate segment of the breeding populat ion. Continual removal of the largest members, and potentially a large segment of the breedi ng alpha male population, may ultimately cause an unfavorable genetic drift. Characterizing movements – Effective wildlife mana gement requires knowledge of species movements. Biological investig ations can yield direct information on movements at a certain scale. Genetic data can give insight to movement on a much different scales. As descri bed above, the cytochrome b data reveal that female yacare caiman do not appear to cross the barrier be tween the Amazon and Paraguay-Paraná river basins with enough regularity to establish haplotypes that are shared between basins. This

PAGE 198

178 result was unanticipated, although movements by females are generally less than males. This barrier may not impede migration of males, however, and result in the lack of distinct segregation of allele s. This movement would be very difficult to quantify by other methods. Further studies will be necessa ry if wildlife managers want more fine scale movement data. While traditional F-statistic s and other genetic metrics produce migration indices, often expressed as Nm values, wildlife ma nagers must be aware of the inherent limitations. These data are directly related to, and limited by, the quality and quantity of the source data. Information on the actu al migration numbers per generation, and implications to the study populations, are us ually not clear. For example, the genetic contribution of 10 migrants per generation from the río Paraguá to the Pantanal area, with local effective populations (Ne) in the millions, would likely be insignificant. Introduced alleles and haplotypes would be subject to lineage sorting and genetic swamping by the sheer magnitude of numbers. The situation is very different for an endangered species where the Ne is low and the relative genetic c ontribution of each migrant is higher. Reserve design – Molecular data can be incorpor ated in future reserve design and modification of existing protected areas. Th ese wetlands must be interspersed over a large area to form buffer zones for general r ecuperation in the event of over harvesting. Especially important to keep in mind are areas such as the río Paraguá where unique haplotypes and alleles are found. The pres ent study should be considered only preliminary as important areas, such as the western río Madre de Díos, southern Mamoré, eastern Iténez, eastern Pantanal and P ilcomayo were not surveyed at all.

PAGE 199

179 Natural corridors must be maintained for continued gene flow. This does not pose a problem in the near future for Bolivia and Paraguay where the land use in the primary habitat is for extensive cattle ranching. Ra pid development in Brazilian Pantanal and northern Argentina gives cause for concern. C onversion to large-scal e agriculture is a growing problem with reduction of habitat and contamination with herbicides, fertilizers and pesticides. There is also a growing potential for large-sc ale habitat disruption if the proposed waterway (Hidrovía) from the Brazil to the Atlantic is approved and constructed. A good example of a positive improvement to reserve design comes from Bolivia. Noel Kempff Mercado National Park is located on the Brazilian Shield along the río Paraguá. It includes important catchment basi ns and ecosystems on the east side of the river but the previous park boundary basically para lleled the río Paraguá was but set back several kilometers. Recent legislation increase d the Park holdings to include the main river corridor and protect the connectivity of affluents to the main river. This improvement may imply potenti al protection for unique C. yacare populations described in this study from the río Paraguá drainage. Evaluating translocations – Given the large numbers in wild Caiman yacare populations, translocations do not appear ne cessary in the near future. However, unforeseen events leading to local extirpa tion through poaching or contamination might necessitate re-introductions in selected areas. Very low popula tion level may also lead to inbreeding depression. Genetic profiling of the translocation stoc ks would be advised for maintenance of prior diversity. Fortunately, yacare hides are currently of relatively low value and hunting

PAGE 200

180 ceases at a certain “catch per unit effort” thre shold. Typically, when this occurs a large population of sub-adults and juveniles remai n. Due to young age and small size at sexual maturity, Caiman yacare and C. crocodilus populations rebound relatively quickly without outside assistance. Some are reproductive at about 1 m total length. With many of the larger crocodylian species , both size and time to maturity work against them. In those species, many individu als attain a valuable size and are killed before they reach the reproductive ag e/size class. Quite often sympatric Caiman species move into the habitat previ ously occupied by the extirpat ed species, making additional hurdles to recovery. These are typically the species requiri ng translocations. Forensics – The field of wildlife forensics has moved forward quickly with the advances in molecular techniques. Data from this study give author ities the ability to identify raw Caiman yacare hides from those of common caiman with certainty. This segregation by visual means requires comparison of whole flanks and is imprecise at best. If future harvest zones are to be monitored, haplotype data could be used to give an indication for the area of or igin. The border between Paraguay and Bolivia has been notoriously porous with wildlif e products passing in both di rections. The distinction of Bolivian Amazon versus the Paraguayan haplot ypes is clear as shown in this study. The only zone of ambiguity would come from cai man originating in the Bolivian Pantanal. The next step will be to develop molecula r protocols for working with tanned hides and manufactured products. Un fortunately, the normal procedures for fresh tissue are rendered useless once the hide has been subjec ted to the chemical treatments of tanning. As many hides are processed to the chrome tan or “crust” stage before international shipment, sensitive molecular inspection is circumvented. Once these technical problems

PAGE 201

181 are overcome, many of the current difficulties of hide identificati on and product origin can be resolved. CITES regulations have greatly reduced the current volume of illegal crocodylian hides on the world market, but stri cter controls resulting from new tools can further reduce that portion. Genetic informa tion from studies such as this will be collected and implemented as new techniques are developed. Conclusions 1) Caiman yacare is a distinct Evolutionarily Significant Unit (ESU). It was originally described as a full species and has a distribution with nearly complete reproduction separation from the closely related common caiman, Caiman crocodilus . Comparison with Amazon C. crocodilus sequences shows genetic separation with no apparent influence from C yacare migation downstream. 2) Caiman yacare can be identified from every ot her crocodylian species, including Caiman crocodilus and subspecies, by comparison of cytochrome b or 16S sequences. 3) The Brazilian Shield, which forms most of the northern boundary and causes the río Madera constriction, is an effec tive barrier to upstr eam migration of C. crocodilus from the north. While hybridization is possi ble at the very limited areas of contact between the two species, there has been no de tectable genetic introgression. Analyses from this study show distinct genetic sepa ration and significant genetic distance between them. 4) The 25 Caiman yacare cytochrome b haplotypes are divided between the two main distribution areas, río Amazon and río Paraguay-Paraná, with no haplotypes shared between the two. This indicates that there is no effective migrati on of females between these two populations.

PAGE 202

182 5) Caiman yacare populations are weakly structured in relation to the pattern of river drainages within their distribution. This is shown in analyses for both mtDNA sequences and microsatellite allele frequencie s. While microsatellite allele frequencies differ according to geography, all loci are shared by all populations.

PAGE 203

APPENDIX A LOCALITY INFORMATION FOR SA MPLES USED IN THIS STUDY

PAGE 204

184 Lab No. Date Basin Drainage Property Site Latitude Longitude Cya001 21-Jul-96 1 río Beni Riberalta near town S 11°00'19.7" W 66°05'19.7" Cya002 1 río Beni Riberalta near town S 11°00'19.7" W 66°05'19.7" Cya006 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo Los Caimanes S 12°24'59.3" W 65°27'51.0" Cya007 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo Los Caimanes S 12°25'08.2" W 65°27'54.2" Cya008 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo Los Caimanes S 12°25'13.0" W 65°27'34.7" Cya009 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo Los Caimanes S 12°25'08.2" W 65°27'54.2" Cya010 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo Los Caimanes S 12°25'08.2" W 65°27'54.2" Cya011 26-Jun-96 2 río Mamoré Est. Ponderosa pond at entrance S 14°50'13.5" W 64°47'17.5" Cya012 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya013 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya014 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya015 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya016 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya017 26-Jun-96 2 río Mamoré Est. Ponderosa cañada S 14°50'06.3" W 64°47'18.3" Cya020 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya021 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya022 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya023 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya024 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya025 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya026 1-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'17.2" W 67°17'01.6" Cya027 2-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'23.0" W 67°16'35.7" Cya028 2-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'19.1" W 67°17'16.2" Cya029 2-Ago-96 1 río Beni Rurrenebaque 5 km E of Reyes S 14°17'19.1" W 67°17'16.2" Cya030 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya031 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya032 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya033 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6"

PAGE 205

185Cya034 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya035 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya036 10-Sep-96 4 río Paraguá (Itenez) Est. Capar uch Bahia San Roque S 14°55'32.5" W 61°05'38.6" Cya037 11-Sep-96 4 río Paraguá (Itenez) Est. Capar u Terraplen Caparuchito S 14°55'15.6" W 61°06'13.8" Cya038 11-Sep-96 4 río Paraguá (Itenez) Est. Capar u Terraplen Caparuchito S 14°55'15.6" W 61°06'13.8" Cya039 11-Sep-96 4 río Paraguá (Itenez) Est. Capar u Terraplen Caparuchito S 14°55'15.6" W 61°06'13.8" Cya040 11-Sep-96 4 río Paraguá (Itenez) Est. Capar u Terraplen Caparuchito S 14°55'15.6" W 61°06'13.8" Cya041 11-Sep-96 4 río Paraguá (Itenez) Est. Capar u Terraplen Caparuchito S 14°55'15.6" W 61°06'13.8" Cya042 12-Sep-96 4 río Paraguá (Itenez) Est. La Union Bahía Venecia S 15°03'58.7" W 61°04'04.5" Cya043 12-Sep-96 4 río Paraguá (Itenez) Est. La Union Bahía Venecia S 15°03'58.7" W 61°04'04.5" Cya044 12-Sep-96 4 río Paraguá (Itenez) Est. La Union Bahía Venecia S 15°03'58.7" W 61°04'04.5" Cya045 12-Sep-96 4 río Paraguá (Itenez) Est. La Union Bahía Venecia S 15°03'58.7" W 61°04'04.5" Cya 050 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre pozo 3 S 12°18'12.1" W 65°32'37.6" Cya 051 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre pozo 3 S 12°18'12.1" W 65°32'37.6" Cya 052 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre pozo 3 S 12°18'12.1" W 65°32'37.6" Cya 053 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre pozo4A S 12°17'10.1" W 65°33'02.7" Cya 054 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre pozo4A S 12°17'10.1" W 65°33'02.7" Cya 055 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre 2 Mamá S 12°18'07.7" W 65°32'33.6" Cya 056 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre 2 Mamá S 12°18'07.7" W 65°32'33.6" Cya 057 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre 2 Mamá S 12°18'12.1" W 65°32'37.6" Cya 058 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre 2 Mamá S 12°18'12.1" W 65°32'37.6" Cya 059 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre 2 Mamá S 12°18'12.1" W 65°32'37.6" Cya070 1-Jun-96 2 río Grande (Mamoré) Santa Cruz within city limits S 17°44'49.4" W 63°10'13.8" Cya075 10-Ago-96 2 río Grande (Mamoré) Loma de Arena repressa S 17°55'05.3" W 63°09'42.1" Cya085 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya086 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya087 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya088 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya089 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya090 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7"

PAGE 206

186Cya091 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya092 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya093 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 1 S 16°47'39.0" W 58°47'49.7" Cya094 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 4, Campo 1 S 16°48'23.8" W 58°48'38.3" Cya095 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis S 16°46'14.0" W 58°44'21.0" Cya096 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 5 S 16°48'14.9" W 58°48'51.5" Cya097 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 5 S 16°48'14.9" W 58°48'51.5" Cya098 7-Sep-96 5 río Paraguay (Paraná) Est. San Luis pozo 1, Campo 5 S 16°48'14.9" W 58°48'51.5" Cya099 6-Set-96 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya100 6-Set-96 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya101 6-Set-96 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya110 8-Sep-96 5 río Paraguay (Paraná) Est. Tel Aviv Cotoca S 16°52'53.3" W 58°41'46.9" Cya111 8-Sep-96 5 río Paraguay (Paraná) Est. Tel Aviv Cotoca S 16°52'53.3" W 58°41'46.9" Cya112 8-Sep-96 5 río Paraguay (Paraná) Est. Tel Aviv Cotoca S 16°52'53.3" W 58°41'46.9" Cya113 8-Sep-96 5 río Paraguay (Paraná) Est. Tel Aviv Cotoca S 16°52'53.3" W 58°41'46.9" Cya114 8-Sep-96 5 río Paraguay (Paraná) Est. Tel Aviv Cotoca S 16°52'53.3" W 58°41'46.9" Cya130 13-Ago-96 3 río San Miguel Est. El To co pozo Lucero S14°50'59.0" W 64°40'08.5" Cya131 13-Ago-96 3 río San Miguel Est. El Toco pozo 2 S 14°53'07.3" W 64°39'02.1" Cya132 13-Ago-96 3 río San Miguel Est. El Toco pozo 3 S 14°53'09.8" W 64°38'57.2" Cya133 13-Ago-96 3 río San Miguel Est. El Toco Lucero 4 (casa) S 14°53'94.4" W 64°38'56.6" Cya134 13-Ago-96 3 río San Miguel Est. El Toco Atajado de Primavera S 14°51'31.7" W 64°39'12.0" Cya135 13-Ago-96 3 río San Miguel Est. El Toco Atajado de Primavera S 14°51'31.7" W 64°39'12.0" Cya136 13-Ago-96 3 río San Miguel Est. El Toco Atajado de Primavera S 14°51'31.7" W 64°39'12.0" Cya137 13-Ago-96 3 río San Miguel Est. El Toco pozo1 S 14°53'06.6" W 64°39'03.3" Cya138 13-Ago-96 3 río San Miguel Est. El Toco pozo1 S 14°53'06.6" W 64°39'03.3" Cya140 22-Ago-96 3 río Blanco Est. Esperanz a laguna Tapada S 13°08'33.0" W 63°47'39.1" Cya141 22-Ago-96 3 río Blanco Est. Esperanz a laguna Tapada S 13°08'18.6" W 63°47'46.6" Cya142 22-Ago-96 3 río Blanco Est. Esperanz a laguna Tapada S 13°08'21.2" W 63°47'44.1" Cya143 22-Ago-96 3 río Blanco Est. Esperanz a laguna Tapada S 13°08'29.9" W 63°47'37.5" Cya144 22-Ago-96 3 río Blanco Est. Esperanz a laguna Tapada S 13°08'30.5" W 63°47'43.5"

PAGE 207

187Cya145 22-Ago-96 3 río Blanco Est. Esperanz a laguna Orope S 13°09'35.6" W 63°47'54.0" Cya146 22-Ago-96 3 río Blanco Est. Esperanz a laguna Orope S 13°09'46.8" W 64°47'46.8" Cya147 22-Ago-96 3 río Blanco Est. Esperanz a laguna Orope S 13°09'46.8" W 64°47'46.8" Cya150 25-Ago-96 3 río Blanco Est. Nazaret laguna Masaví S 13°11'30.9" W 63°45'46.7" Cya151 25-Ago-96 3 río Blanco Est. Nazaret laguna Masaví S 13°11'30.9" W 63°45'46.7" Cya152 25-Ago-96 3 río Blanco Est. Nazaret laguna Masaví S 13°11'30.9" W 63°45'46.7" Cya153 24-Ago-96 3 río Blanco Est. Nazaret laguna Puerto Feliz S 13°08'13.7" W 63°44'33.8" Cya154 24-Ago-96 3 río Blanco Est. Nazaret laguna Puerto Feliz S 13°07'50.5" W 63°44'09.5" Cya155 23-Ago-96 3 río Blanco Est. Nazaret laguna El Pio S 13°08'02.6" W 63°47'03.4" Cya156 23-Ago-96 3 río Blanco Est. Nazaret laguna El Pio S 13°08'02.6" W 63°47'03.4" Cya157 23-Ago-96 3 río Blanco Est. Nazaret laguna El Pio S 13°07'52.8" W 63°47'08.1" Cya158 23-Ago-96 3 río Blanco Est. Nazaret laguna Vivisca S 13°06'55.5" W 63°46'43.9" Cya170 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya171 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya172 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya173 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya174 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya175 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya176 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya177 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya178 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya179 24-Jun-96 3 río Itonamas Est. laguna Azul laguna Azul S 14°59'12,2" W 64°48'50.5" Cya200 12-Dec-97 1 río Beni Rurrenbaque north of Reyes S 14°17'19.3" W 67°17'21.8" Cya201 12-Dec-97 1 río Beni Rurrenbaque north of Reyes S 14°17'19.3" W 67°17'21.8" Cya202 13-Dec-97 1 río Beni Rurrenbaque north of Reyes S 14°17'19.3" W 67°17'21.8" Cya203 13-Dec-97 1 río Beni Rurrenbaque north of Reyes S 14°17'23.6" W 67°17'47.9" Cya204 3-Aug-96 1 río Beni Rurrenbaque north of Reyes S 14°13'50.9" W 67°09'43.3" Cya205 3-Aug-96 1 río Beni Rurrenbaque north of Reyes S 14°13'50.9" W 67°09'43.3" Cya206 3-Aug-96 1 río Beni Rurrenbaque north of Reyes S 14°13'50.9" W 67°09'43.3" Cya207 3-Aug-96 1 río Beni Rurrenbaque north of Reyes S 14°13'50.9" W 67°09'43.3"

PAGE 208

188Cya210 4-Jul-96 3 río Itonamas Est. Esperanza río Ipurupuru S 14°04'27.4" W 64°52'57.9" Cya211 4-Jul-96 3 río Itonamas Est. Esperanza río Ipurupuru S 14°05'19.0" W 64°52'49.2" Cya212 4-Jul-96 3 río Itonamas Est. Cutal (s alted hides) S 14°06'41.0" W 64°55'43.5" Cya213 4-Jul-96 3 río Itonamas Est. Cutal (s alted hides) S 14°06'41.0" W 64°55'43.5" Cya214 4-Jul-96 3 río Itonamas Est. Cutal (s alted hides) S 14°06'41.0" W 64°55'43.5" Cya215 4-Jul-96 3 río Itonamas Est. Cutal (s alted hides) S 14°06'41.0" W 64°55'43.5" Cya216 4-Jul-96 3 río Itonamas Est. Cutal (s alted hides) S 14°06'41.0" W 64°55'43.5" Cya217 6-Jul-96 3 río Itonamas Est. Esperanza C anada del Corral S 14°00'14.6" W 64°48'44.2" Cya218 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'08.0" W 64°49'56.1" Cya219 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'14.7" W 64°50'04.0" Cya220 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'14.0" W 64°50'08.1" Cya221 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'07.8" W 64°50'11.5" Cya222 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'05.6" W 64°50'08.1" Cya223 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'05.6" W 64°50'08.1" Cya224 9-Jul-96 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°57'05.6" W 64°50'08.1" Cya225 13-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya226 14-Jul-96 3 río Itonamas Est. San Ignacio Curiche de la Pampa S 13°28'55.3" W 64°48'55.5" Cya227 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya228 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya229 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya230 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya231 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya232 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya233 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya234 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya235 14-Jul-96 3 río Itonamas Est. San Ignacio laguna Todos Santos S 13°29'18.9" W 64°49'38.3" Cya250 16-Set-95 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya251 16-Set-95 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya252 16-Set-95 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7" Cya253 16-Set-95 5 río Paraguay (Paraná) Est. San Luis pozo San Luis S 16°47'23.6" W 58°46'07.7"

PAGE 209

189Cya254 20-Set-95 4 río Paraguá (Itenez) El Refugio río Paraguá S 14°46'07.5" W 61°02'14.1" Cya255 21-Set-95 4 río Paraguá (Itenez) El Refugio río Paraguá S 14°45'45.2" W 61°02'51.0" Cya256 21-Set-95 4 río Paraguá (Itenez) El Refugio río Paraguá S 14°45'45.2" W 61°02'51.0" Cya257 21-Set-95 4 río Paraguá (Itenez) El Refugio río Paraguá S 14°45'45.2" W 61°02'51.0" Cya258 26-Set-95 4 río Paraguá (Itenez) Piso Firme Est. Achachairú, río Paraguá S 13°37'58.2" W 61°42'37.0" Cya259 26-Set-95 4 río Paraguá (Itenez) Piso Firme río Paraguá S 13°37'23.8" W 61°42'45.4" Cya260 26-Set-95 4 río Paraguá (Itenez) Piso Firme río Paraguá S 13°36'56.4" W 61°43'46.5" Cya261 26-Set-95 4 río Paraguá (Itenez) Piso Firme río Paraguá S 13°38'47.4" W 62°45'47.4" Cya262 5-Oct-95 3 río San Miguel Est. San Miguelito rSJ/Palca de Tuna S 17°07'10.5" W 61°45'46.4" Cya263 6-Oct-95 3 río San Miguel Est. San Miguelit o río San Julián S 17°07'23.4" W 61°44'52.5" Cya264 6-Oct-95 3 río San Miguel Est. San Miguelit o río San Julián S 17°07'27.1" W 61°44'59.5" Cya265 17-Oct-95 3 río Itonamas Est. La Ceiba arroyo Chuto S 14°23'40.8" W 64°03'16.3" Cya266 17-Oct-95 3 río Itonamas Est. La Ceiba arroyo Chuto S 14°23'40.8" W 64°03'16.3" Cya267 18-Oct-95 3 río Itonamas Est. Villa Christina ojo de agua S 14°25'18.9" W 63°59'18.2" Cya268 18-Oct-95 3 río Itonamas Est. Villa Christina ojo de agua S 14°25'18.9" W 63°59'18.2" Cya269 31-Oct-95 2 río Itonamas TIPNIS arroyo Negro S 15°54'02.0" W 65°13'58.6" Cya270 31-Oct-95 2 río Itonamas TIPNIS arroyo Negro S 15°54'02.0" W 65°13'58.6" Cya271 31-Oct-95 2 río Itonamas TIPNIS arroyo Negro S 15°54'02.0" W 65°13'58.6" Cya272 6-Nov-95 2 río Itonamas El Peru arroyo Barcelona S 14°29'59.0" W 65°38'38.2" Cya273 6-Nov-95 2 río Itonamas El Peru arroyo Barcelona S 14°29'59.0" W 65°38'38.2" Cya274 9-Nov-95 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°56'47.9" W 64°50'03.1" Cya275 9-Nov-95 3 río Itonamas Est. Esmeralda arroyo Masineca S 13°56'47.9" W 64°50'03.1" Cya276 21-Nov-95 3 río Itonamas Est. San Ignacio laguna San Ignacio S 13°26'31.2" W 64°49'57.8" Cya277 21-Nov-95 3 río Itonamas Est. San Ignacio laguna San Ignacio S 13°26'31.2" W 64°49'57.8" Cya400 5 río Paraguay Puerto Ramos 7 km S Bahia Negra Alto Paraguay Cya401 1-Dec-86 5 río Paraguay Est. General Diaz laguna Alto Paraguay Cya402 8-Nov-86 6 río Pilcomayo Est. La Golondrina potrero Doña Pres. Hayes Cya403 17-Nov-86 5 río Paraguay 2 km N Bahia Negra Alto Paraguay Cya404 5-Dec-86 5 río Paraguay Est. General Diaz laguna Alto Paraguay Cya405 5-Dec-86 5 río Paraguay Est. General Diaz laguna Alto Paraguay

PAGE 210

190Cya406 1-Nov-86 5 río Paraguay Est. Tinfunque r Montelindo, Retiro Pozo Arias Pres. Hayes Cya407 1-Nov-86 5 río Paraguay Est. Tinfunque 18 km S Fortín Avalos SánchezPres. Hayes Cya408 22-Nov-86 5 río Paraguay Est. Capitan Carmelo Peralta 15 km N Pto. Cabello Alto Paraguay Cya409 17-Nov-86 5 río Paraguay 2 km N Bahia Negra Alto Paraguay Cya410 1986 5 río Paraguay Puerto Ramos 7 km S Bahia Negra Alto Paraguay Cya411 22-Nov-86 5 río Paraguay 3 km E of Est. General Diaz Alto Paraguay Cya412 19-Nov-86 5 río Paraguay Est. Dona Julia 1 km SE Pto. Cabello Alto Paraguay Cya413 22-Nov-86 5 río Paraguay Est. Capitan Carmelo Peralta Alto Paraguay Cya414 4-Dec-86 5 río Paraguay 3 km E of Est. General Diaz Alto Paraguay Cya415 23-Jan-87 6 río Paraguay Est. 26, 8 km NE Tacuaro Ñeembucú Cya416 16-Nov-86 5 río Paraguay Est. Dona Julia 1 km SE Pto. Cabello Alto Paraguay Cya417 22-Nov-86 5 río Paraguay Est. Capitan Carmelo Peralta 15 km N Pto. Cabello Alto Paraguay Cya418 13-Nov-86 5 río Paraguay Pto. Ramos 7 km S Bahia Negra Alto Paraguay Cya419 15-Oct-86 5 río Paraguay Est. Juan de Zalazar Tuya Alto Paraguay Cya420 22-Nov-86 5 río Paraguay Est. Capitan Carmelo Peralta 15 km N Pto. Cabello Alto Paraguay Cya450 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo Provincia de Formosa S 24°25.449' W 60°20.575' Cya451 10-Dec-04 6 río Paraguay (Argentina) Riacho Mont e Lindo 50 km N of Formosa City S 24°25.449' W 60°20.575' Cya452 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya453 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya454 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya455 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya456 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya457 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya458 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya459 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya460 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya461 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya462 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya463 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya464 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo Provincia de Formosa S 24°25.449' W 60°20.575'

PAGE 211

191Cya465 10-Dec-04 6 río Paraguay (Argentina) Riacho Mont e Lindo 50 km N of Formosa City S 24°25.449' W 60°20.575' Cya466 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya467 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya468 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cya469 10-Dec-04 6 río Paraguay (Argentina) Riacho Monte Lindo " S 24°25.449' W 60°20.575' Cfu501 4-Feb-89 Honduras río Sico N 15°52'56.0" W 85°57'50.0" Cfu502 4-Feb-89 Honduras río Palácios N 15°55'38.0" W 85°54'34.0" Cfu504 5-Feb-89 Honduras laguna Bacalar N 15°56'55.0" W 85°55'59.0" Cfu505 7-Feb-89 Honduras laguna Bacalar N 15°56'55.0" W 85°55'59.0" Cfu506 7-Feb-89 Honduras laguna Bacalar N 15°56'55.0" W 85°55'59.0" Cfu507 8-Feb-89 Honduras laguna Bacalar N 15°56'55.0" W 85°55'59.0" Cfu509 18-Abr-88 Honduras Crique Amatingni, Brus laguna N 15°40'28.0" W 85°38'24.0" Cfu510 17-Abr-89 Honduras río Plátano N 15°40'28.0" W 85°34'49.0" Cfu511 17-Abr-89 Honduras río Plátano/La Mosquitia N 15°40'28.0" W 85°34'49.0" Cfu512 16-Abr-89 Honduras río Plátano/La Mosquitia N 15°40'28.0" W 85°34'49.0" Cfu513 27-Abr-89 Honduras río Chapagua N 15°51'54.0" W 85°46'14.0" Cfu514 27-Abr-89 Honduras río La Criba, W Laguna Ibans N 15°54'37.0" W 85°50'30.0" Cfu515 2-Mar-89 Honduras río La Criba, W Laguna Ibans N 15°54'37.0" W 85°50'30.0" Cfu516 26-May-89 Honduras río Chapagua N 15°51'54.0" W 85°46'14.0" Cfu517 26-May-89 Honduras río Chapagua N 15°51'54.0" W 85°46'14.0" Cfu518 26-May-89 Honduras río Chapagua N 15°51'54.0" W 85°46'14.0" Cfu519 23-May-89 Honduras río Aguán N 15°49'22.0" W 85°47'04.0" Cfu520 23-May-89 Honduras río Aguán N 15°49'22.0" W 85°47'04.0" Ccc600 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc601 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc602 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc603 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc604 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc605 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc606 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9"

PAGE 212

192Ccc607 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc608 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc609 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc610 9-May-99 río Apure Venezuela, Edo. Apure Centro de Acopio N 07°55'09.6" W 67°27'17.9" Ccc611 22-May-99 río Apure Venezuela, Edo. Apure N 07°52'38.1" W 67°31'25.1" Ccc612 20-May-99 río Apure Venezuela, Edo. Apure N 07°54'40.5" W 67°24'20.1" Cla700 20-May-96 7 Estero Iberá Argentina S 27°42.675' W 57°11.641' Cla701 20-May-96 7 Estero Iberá Argentina S 27°42.675' W 57°11.641' Cla702 20-May-96 7 Estero Iberá Argentina S 27°42.675' W 57°11.641' Cla708 B. Sedgwick collection hatchling #8 Cla709 B. Sedgwick collection hatchling #9 Cla710 B. Sedgwick collection adult M, father Cla711 B. Sedgwick collection adult F, mother Cla712 B. Sedgwick collection same as 711, extra sample Cla713 28-Jun-95 B. Sedgwick collection adult M, Gator Jungle Mni800 24-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre S 12°26'25.7" W 65°31'27.1" Mni801 25-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre S 12°26'25.7" W 65°31'27.1" Mni802 4-Jul-96 3 río Itonamas Est. Esperanza (salted silla) río Ipurupuru S 14°04'20.6" W 64°52'58.7" Mni803 9-Jul-96 3 río Itonamas Est. Esperanza a rroyo Masineca S 13°56'56.6" W 64°49'51.6" Mni804 14-Jul-96 3 río Itonamas Est. San Ignacio río Yuatre S 13°28'31.4" W 64°47'26.3" Mni805 4-Aug-96 1 río Beni Rurrenebaque dried tissue on skull S 14°26'26.1" W 67°31'52.1" Mni806 4-Aug-96 1 río Beni Rurrenebaque dried tissue on skull S 14°26'26.1" W 67°31'52.1" Pp900 18-Jul-96 1 río Beni Riberalta within city limits S 11°00'34.6" W 66°05'07.1" Pp901 26-Jul-96 2 río Yata (Mamoré) Est. Rancho Alegre arroyo los Caimanes S 12°24'45.7" W 65°28'07.9" Ami004 26-Aug-99 Hendry Correctional Facility Hendry Co., FL Dennis Raskins/Nuisance gator Ami012 12-Aug-99 Rent Pond Duval Co., FL Michael Haley/Nuisance gator

PAGE 213

APPENDIX B MICROSATELLITE ALLELE FRE QUENCY COMPARISON OVER POPULATIONS

PAGE 214

194 Key to Population Names: ~~~~~~~~~~~~~~~~~~~~~~~~ Pop1 Beni Pop2 Mamore Pop3 San_Miguel Pop4 Paragua Pop5 Paraguay Pop6 Parana Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Claµ_5 1 173 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Claµ_5 2 179 0.0000 0.0000 0.0000 0.0000 0.0385 0.0000 0.0052 Paraguay Claµ_5 3 181 0.0000 0.0000 0.0000 0.0000 0.1346 0.0652 0.0259 Claµ_5 4 189 0.0500 0.0172 0.0139 0.0000 0.0192 0.0000 0.0155 Claµ_5 5 191 0.2750 0.2931 0.1319 0.0000 0.0962 0.0000 0.1347 Claµ_5 6 193 0.0000 0.0690 0.0139 0.0000 0.0577 0.0000 0.0233 Claµ_5 7 195 0.0000 0.0000 0.0139 0.0000 0.0000 0.0000 0.0052 San_Miguel Claµ_5 8 199 0.0000 0.0000 0.0208 0.0435 0.0000 0.0435 0.0181 Claµ_5 9 201 0.1000 0.0000 0.0347 0.0000 0.0385 0.0000 0.0285 Claµ_5 10 203 0.0000 0.0517 0.0208 0.1087 0.0385 0.0000 0.0337 Claµ_5 11 205 0.0500 0.0517 0.0208 0.0000 0.0385 0.0000 0.0259 Claµ_5 12 207 0.1000 0.0000 0.0000 0.0000 0.0192 0.0000 0.0130 Claµ_5 13 209 0.0000 0.1207 0.0625 0.0000 0.0000 0.0000 0.0415 Claµ_5 14 211 0.0750 0.0690 0.0486 0.0000 0.0769 0.0000 0.0466 Claµ_5 15 213 0.1000 0.0517 0.0903 0.0217 0.1154 0.0217 0.0725 Claµ_5 16 215 0.0000 0.0690 0.0833 0.0000 0.0577 0.0652 0.0570 Claµ_5 17 217 0.0250 0.0172 0.0486 0.0217 0.0769 0.2609 0.0674 Claµ_5 18 219 0.0500 0.0000 0.0694 0.0217 0.0385 0.1739 0.0596 Claµ_5 19 221 0.0000 0.0000 0.1042 0.0217 0.0000 0.3043 0.0777 Claµ_5 20 223 0.0000 0.0000 0.0069 0.0870 0.0192 0.0435 0.0207 Claµ_5 21 225 0.0000 0.0172 0.0278 0.0217 0.0000 0.0000 0.0155 Claµ_5 22 227 0.0000 0.0000 0.0278 0.0870 0.0192 0.0000 0.0233 Claµ_5 23 229 0.0000 0.0172 0.0208 0.1087 0.0000 0.0000 0.0233 Claµ_5 24 231 0.0250 0.0345 0.0139 0.0870 0.0192 0.0000 0.0259 Claµ_5 25 233 0.0000 0.0172 0.0069 0.0870 0.0000 0.0000 0.0155 Claµ_5 26 235 0.0000 0.0345 0.0208 0.0217 0.0192 0.0217 0.0207 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Claµ_5 27 237 0.1000 0.0000 0.0069 0.0217 0.0385 0.0000 0.0207

PAGE 215

195Claµ_5 28 239 0.0250 0.0000 0.0000 0.0435 0.0192 0.0000 0.0104 Claµ_5 29 241 0.0000 0.0000 0.0139 0.0217 0.0192 0.0000 0.0104 Claµ_5 30 243 0.0250 0.0172 0.0000 0.0435 0.0000 0.0000 0.0104 Claµ_5 31 245 0.0000 0.0172 0.0139 0.0217 0.0000 0.0000 0.0104 Claµ_5 32 247 0.0000 0.0172 0.0139 0.0000 0.0000 0.0000 0.0078 Claµ_5 33 249 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_5 34 251 0.0000 0.0000 0.0139 0.0000 0.0000 0.0000 0.0052 San_Miguel Claµ_5 35 253 0.0000 0.0172 0.0139 0.0000 0.0000 0.0000 0.0078 Claµ_5 36 257 0.0000 0.0000 0.0069 0.0435 0.0000 0.0000 0.0078 Claµ_5 37 267 0.0000 0.0000 0.0069 0.0217 0.0000 0.0000 0.0052 Claµ_5 38 271 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Claµ_5 # samples: 20 29 72 23 26 23 193 Claµ_4 1 275 0.0278 0.0000 0.0145 0.0833 0.0000 0.0000 0.0175 Claµ_4 2 277 0.0000 0.0000 0.0580 0.2222 0.0000 0.0000 0.0468 Claµ_4 3 279 0.0000 0.0741 0.0072 0.0000 0.0263 0.0000 0.0175 Claµ_4 4 281 0.0556 0.0370 0.0725 0.0000 0.0263 0.0000 0.0439 Claµ_4 5 283 0.0278 0.1111 0.0290 0.0278 0.0000 0.0000 0.0351 Claµ_4 6 285 0.5833 0.1296 0.4783 0.6389 0.5000 0.5750 0.4649 Claµ_4 7 287 0.1667 0.0926 0.1159 0.0278 0.2632 0.3750 0.1550 Claµ_4 8 289 0.1111 0.1296 0.0870 0.0000 0.1053 0.0000 0.0789 Claµ_4 9 291 0.0000 0.3148 0.1304 0.0000 0.0263 0.0000 0.1053 Claµ_4 10 293 0.0000 0.0926 0.0000 0.0000 0.0526 0.0500 0.0263 Claµ_4 11 295 0.0278 0.0000 0.0072 0.0000 0.0000 0.0000 0.0058 Claµ_4 12 297 0.0000 0.0185 0.0000 0.0000 0.0000 0.0000 0.0029 Mamore Claµ_4 # samples: 18 27 69 18 19 20 171 Cy_R8 1 235 0.0000 0.0172 0.0000 0.0000 0.0000 0.0000 0.0026 Mamore Cy_R8 2 237 0.0500 0.2069 0.1250 0.0652 0.0769 0.0208 0.1031 Cy_R8 3 239 0.1000 0.1207 0.1111 0.3043 0.0962 0.2500 0.1495 Cy_R8 4 241 0.0000 0.0000 0.0000 0.0000 0.0385 0.3750 0.0515 Cy_R8 5 243 0.0500 0.0000 0.0278 0.0000 0.0000 0.0000 0.0155 Cy_R8 6 245 0.1000 0.0690 0.0347 0.0000 0.0577 0.1042 0.0541 Cy_R8 7 247 0.0000 0.0000 0.0139 0.0435 0.0192 0.0625 0.0206 Cy_R8 8 249 0.0500 0.0345 0.1597 0.0000 0.0000 0.0000 0.0696 Cy_R8 9 251 0.1000 0.2414 0.1458 0.1522 0.0385 0.0208 0.1263 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Cy_R8 10 253 0.1000 0.0345 0.0556 0.0000 0.3846 0.1250 0.1031 Cy_R8 11 255 0.1250 0.0000 0.0139 0.1739 0.0769 0.0000 0.0490 Cy_R8 12 257 0.1000 0.0000 0.0000 0.0435 0.0962 0.0000 0.0284 Cy_R8 13 259 0.0500 0.0862 0.0556 0.0435 0.0385 0.0000 0.0490 Cy_R8 14 261 0.0750 0.0345 0.0208 0.0000 0.0192 0.0000 0.0232

PAGE 216

196Cy_R8 15 263 0.0750 0.1379 0.0347 0.0435 0.0385 0.0417 0.0567 Cy_R8 16 265 0.0250 0.0172 0.0694 0.0000 0.0000 0.0000 0.0309 Cy_R8 17 267 0.0000 0.0000 0.0208 0.0000 0.0000 0.0000 0.0077 San_Miguel Cy_R8 18 271 0.0000 0.0000 0.0556 0.0870 0.0000 0.0000 0.0309 Cy_R8 19 273 0.0000 0.0000 0.0139 0.0000 0.0192 0.0000 0.0077 Cy_R8 20 275 0.0000 0.0000 0.0069 0.0217 0.0000 0.0000 0.0052 Cy_R8 21 277 0.0000 0.0000 0.0347 0.0217 0.0000 0.0000 0.0155 Cy_R8 # samples: 20 29 72 23 26 24 194 Cy_022 1 179 0.0000 0.0000 0.0000 0.0000 0.0000 0.0417 0.0051 Parana Cy_022 2 183 0.2250 0.0172 0.0000 0.0000 0.0000 0.0000 0.0255 Cy_022 3 189 0.0000 0.0000 0.0000 0.0000 0.0179 0.0000 0.0026 Paraguay Cy_022 4 191 0.0000 0.0000 0.0278 0.0000 0.0179 0.0208 0.0153 Cy_022 5 193 0.2000 0.0862 0.0347 0.0217 0.0893 0.1667 0.0816 Cy_022 6 195 0.0750 0.1552 0.1319 0.0000 0.1250 0.0417 0.1020 Cy_022 7 197 0.0500 0.1552 0.1736 0.0217 0.1786 0.3750 0.1658 Cy_022 8 199 0.1750 0.0862 0.1181 0.3261 0.0000 0.0000 0.1122 Cy_022 9 201 0.0000 0.0172 0.0000 0.0870 0.0357 0.0000 0.0179 Cy_022 10 203 0.0250 0.1034 0.0208 0.0000 0.0357 0.0625 0.0383 Cy_022 11 205 0.0000 0.0517 0.1319 0.1739 0.0000 0.0000 0.0765 Cy_022 12 207 0.0250 0.1034 0.0833 0.1304 0.1071 0.1250 0.0944 Cy_022 13 209 0.0250 0.0172 0.1042 0.1087 0.1429 0.0000 0.0765 Cy_022 14 211 0.0250 0.0517 0.0347 0.0217 0.1607 0.0417 0.0536 Cy_022 15 213 0.1000 0.0517 0.0486 0.0217 0.0179 0.0417 0.0459 Cy_022 16 215 0.0500 0.1034 0.0694 0.0870 0.0714 0.0625 0.0740 Cy_022 17 217 0.0000 0.0000 0.0139 0.0000 0.0000 0.0000 0.0051 San_Miguel Cy_022 18 219 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Cy_022 19 221 0.0250 0.0000 0.0000 0.0000 0.0000 0.0208 0.0051 Cy_022 # samples: 20 29 72 23 28 24 196 Cy_I14 1 237 0.0263 0.0577 0.0704 0.0000 0.0000 0.0000 0.0393 Cy_I14 2 241 0.2632 0.2115 0.1338 0.0000 0.0000 0.0000 0.1124 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Cy_I14 3 243 0.0000 0.0000 0.0000 0.0000 0.0417 0.0000 0.0056 Paraguay Cy_I14 4 247 0.0000 0.0000 0.0423 0.0238 0.0000 0.0000 0.0197 Cy_I14 5 249 0.0000 0.0000 0.0070 0.0000 0.0000 0.0000 0.0028 San_Miguel Cy_I14 6 251 0.0263 0.0577 0.1338 0.0238 0.1875 0.2059 0.1124 Cy_I14 7 253 0.0526 0.1154 0.0986 0.0476 0.0417 0.0000 0.0730 Cy_I14 8 255 0.3684 0.4615 0.2042 0.3810 0.2500 0.1471 0.2809 Cy_I14 9 256 0.0263 0.0000 0.0000 0.0000 0.0000 0.0000 0.0028 Beni Cy_I14 10 257 0.0263 0.0192 0.1268 0.4286 0.2083 0.0882 0.1433 Cy_I14 11 259 0.0000 0.0000 0.0211 0.0000 0.0000 0.1765 0.0253

PAGE 217

197Cy_I14 12 261 0.2105 0.0577 0.1338 0.0238 0.1042 0.1176 0.1124 Cy_I14 13 263 0.0000 0.0192 0.0141 0.0476 0.0833 0.1471 0.0393 Cy_I14 14 265 0.0000 0.0000 0.0070 0.0238 0.0000 0.0588 0.0112 Cy_I14 15 267 0.0000 0.0000 0.0070 0.0000 0.0208 0.0000 0.0056 Cy_I14 16 271 0.0000 0.0000 0.0000 0.0000 0.0208 0.0588 0.0084 Cy_I14 17 273 0.0000 0.0000 0.0000 0.0000 0.0417 0.0000 0.0056 Paraguay Cy_I14 # samples: 19 26 71 21 24 17 178 Cy_F2 1 203 0.0000 0.0000 0.0069 0.0000 0.0000 0.0417 0.0077 Cy_F2 2 209 0.0000 0.0000 0.0000 0.0000 0.0185 0.0000 0.0026 Paraguay Cy_F2 3 211 0.0000 0.0000 0.0000 0.0000 0.0370 0.0208 0.0077 Cy_F2 4 213 0.0000 0.0000 0.0208 0.0000 0.7222 0.5417 0.1744 Cy_F2 5 217 0.0000 0.0000 0.0069 0.2174 0.0741 0.2500 0.0692 Cy_F2 6 221 0.0000 0.0000 0.0000 0.0000 0.0926 0.0833 0.0231 Cy_F2 7 225 0.0000 0.0000 0.0000 0.0000 0.0000 0.0208 0.0026 Parana Cy_F2 8 227 0.0000 0.0000 0.0000 0.0000 0.0185 0.0000 0.0026 Paraguay Cy_F2 9 229 0.0000 0.0000 0.0000 0.0435 0.0000 0.0000 0.0051 Paragua Cy_F2 10 231 0.0000 0.0000 0.0208 0.0000 0.0000 0.0000 0.0077 San_Miguel Cy_F2 11 233 0.0000 0.0345 0.0069 0.1739 0.0185 0.0000 0.0308 Cy_F2 12 235 0.0250 0.0172 0.0486 0.0000 0.0000 0.0000 0.0231 Cy_F2 13 237 0.0000 0.1034 0.0556 0.0217 0.0185 0.0000 0.0410 Cy_F2 14 239 0.0500 0.2241 0.0486 0.0217 0.0000 0.0208 0.0615 Cy_F2 15 241 0.1000 0.0690 0.0694 0.1957 0.0000 0.0208 0.0718 Cy_F2 16 243 0.1000 0.0862 0.2014 0.1522 0.0000 0.0000 0.1154 Cy_F2 17 245 0.2750 0.1379 0.1597 0.0217 0.0000 0.0000 0.1103 Cy_F2 18 247 0.0750 0.0690 0.0833 0.0435 0.0000 0.0000 0.0538 Cy_F2 19 249 0.1250 0.1379 0.0972 0.0217 0.0000 0.0000 0.0718 Cy_F2 20 251 0.2000 0.0517 0.0556 0.0000 0.0000 0.0000 0.0487 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Cy_F2 21 253 0.0500 0.0345 0.1111 0.0217 0.0000 0.0000 0.0538 Cy_F2 22 255 0.0000 0.0345 0.0000 0.0000 0.0000 0.0000 0.0051 Mamore Cy_F2 23 257 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Cy_F2 24 259 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Cy_F2 25 261 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Cy_F2 26 263 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Cy_F2 # samples: 20 29 72 23 27 24 195 Cy_S4 1 183 0.0000 0.0000 0.0000 0.0000 0.0357 0.0417 0.0102 Cy_S4 2 185 0.0000 0.0000 0.0000 0.0217 0.0357 0.0208 0.0102 Cy_S4 3 187 0.0000 0.0345 0.0000 0.0000 0.0357 0.0417 0.0153 Cy_S4 4 195 0.0000 0.0000 0.0000 0.0000 0.0000 0.0208 0.0026 Parana Cy_S4 5 203 0.0000 0.0000 0.0000 0.0000 0.0000 0.0208 0.0026 Parana

PAGE 218

198Cy_S4 6 209 0.0000 0.0000 0.0069 0.0217 0.0536 0.1667 0.0332 Cy_S4 7 211 0.0000 0.0000 0.0139 0.1522 0.0179 0.0417 0.0306 Cy_S4 8 213 0.1500 0.1207 0.1667 0.1087 0.1786 0.0417 0.1378 Cy_S4 9 215 0.0500 0.0000 0.0069 0.0870 0.0714 0.0000 0.0281 Cy_S4 10 217 0.0750 0.0690 0.0833 0.0870 0.0000 0.0417 0.0638 Cy_S4 11 219 0.0000 0.0000 0.0000 0.0217 0.0893 0.0000 0.0153 Cy_S4 12 221 0.1500 0.0517 0.0278 0.0870 0.0357 0.0000 0.0485 Cy_S4 13 223 0.0000 0.1034 0.0833 0.0000 0.0714 0.0833 0.0663 Cy_S4 14 225 0.0000 0.0345 0.0764 0.0000 0.0714 0.2708 0.0765 Cy_S4 15 227 0.2250 0.0172 0.0833 0.0000 0.0179 0.0000 0.0587 Cy_S4 16 229 0.1000 0.0862 0.0278 0.0652 0.0000 0.0000 0.0408 Cy_S4 17 231 0.2250 0.2414 0.2083 0.0217 0.0714 0.0833 0.1582 Cy_S4 18 233 0.0250 0.0862 0.0556 0.0217 0.0179 0.0208 0.0434 Cy_S4 19 235 0.0000 0.1034 0.0556 0.0000 0.0714 0.0208 0.0485 Cy_S4 20 237 0.0000 0.0172 0.0278 0.1087 0.0536 0.0208 0.0357 Cy_S4 21 239 0.0000 0.0172 0.0347 0.1087 0.0179 0.0000 0.0306 Cy_S4 22 241 0.0000 0.0000 0.0139 0.0217 0.0179 0.0208 0.0128 Cy_S4 23 243 0.0000 0.0172 0.0069 0.0217 0.0357 0.0208 0.0153 Cy_S4 24 245 0.0000 0.0000 0.0208 0.0435 0.0000 0.0000 0.0128 Cy_S4 25 247 0.0000 0.0000 0.0000 0.0000 0.0000 0.0208 0.0026 Parana Cy_S4 # samples: 20 29 72 23 28 24 196 Cy_N10 1 293 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Cy_N10 2 295 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Cy_N10 3 297 0.0000 0.0000 0.0208 0.0000 0.0000 0.0000 0.0077 San_Miguel Cy_N10 4 299 0.0500 0.0000 0.0000 0.0000 0.0000 0.0000 0.0052 Beni Cy_N10 5 301 0.0250 0.0172 0.0000 0.0000 0.0000 0.0000 0.0052 Cy_N10 6 305 0.0000 0.0690 0.0000 0.0000 0.0185 0.0435 0.0180 Cy_N10 7 307 0.0000 0.0172 0.0208 0.0000 0.0185 0.0870 0.0232 Cy_N10 8 309 0.2250 0.2069 0.0833 0.2174 0.0741 0.0217 0.1237 Cy_N10 9 311 0.0500 0.1379 0.0764 0.0652 0.0000 0.0000 0.0619 Cy_N10 10 313 0.1750 0.0345 0.1319 0.0217 0.0000 0.0435 0.0799 Cy_N10 11 315 0.1500 0.0345 0.0694 0.0652 0.0000 0.0435 0.0593 Cy_N10 12 317 0.0250 0.0517 0.0486 0.0870 0.0556 0.1087 0.0593 Cy_N10 13 319 0.0750 0.0517 0.0764 0.0435 0.0370 0.1739 0.0747 Cy_N10 14 321 0.0000 0.0517 0.0417 0.0000 0.1481 0.1957 0.0670 Cy_N10 15 323 0.0000 0.0345 0.0208 0.0652 0.1111 0.1304 0.0515 Cy_N10 16 325 0.0250 0.0000 0.0486 0.0870 0.0556 0.0217 0.0412 Cy_N10 17 327 0.0000 0.0000 0.0208 0.0217 0.0000 0.0217 0.0129 Cy_N10 18 329 0.0000 0.0000 0.0208 0.0217 0.0926 0.0000 0.0232

PAGE 219

199Cy_N10 19 331 0.0250 0.1034 0.0278 0.0217 0.0741 0.0652 0.0490 Cy_N10 20 333 0.0250 0.0690 0.0556 0.0000 0.0370 0.0000 0.0387 Cy_N10 21 335 0.1250 0.0690 0.0486 0.0000 0.0370 0.0217 0.0490 Cy_N10 22 337 0.0250 0.0172 0.0069 0.0000 0.0556 0.0000 0.0155 Cy_N10 23 339 0.0000 0.0000 0.0833 0.1739 0.0926 0.0000 0.0644 Cy_N10 24 341 0.0000 0.0172 0.0278 0.0217 0.0370 0.0000 0.0206 Cy_N10 25 343 0.0000 0.0000 0.0208 0.0652 0.0370 0.0217 0.0232 Cy_N10 26 345 0.0000 0.0000 0.0139 0.0217 0.0000 0.0000 0.0077 Cy_N10 27 347 0.0000 0.0172 0.0069 0.0000 0.0000 0.0000 0.0052 Cy_N10 28 349 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Cy_N10 29 355 0.0000 0.0000 0.0000 0.0000 0.0185 0.0000 0.0026 Paraguay Cy_N10 30 357 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Cy_N10 # samples: 20 29 72 23 27 23 194 Claµ_10 1 213 0.3000 0.1552 0.1319 0.0000 0.0357 0.0000 0.1071 Claµ_10 2 217 0.1000 0.3276 0.2222 0.1957 0.0893 0.1042 0.1888 Claµ_10 3 219 0.0000 0.0517 0.0833 0.3043 0.0357 0.1667 0.0995 Claµ_10 4 221 0.2000 0.1897 0.1389 0.0652 0.1429 0.1667 0.1480 Claµ_10 5 223 0.2000 0.0172 0.0417 0.3043 0.1607 0.1250 0.1122 Claµ_10 6 227 0.0000 0.0172 0.0000 0.0000 0.0179 0.0000 0.0051 Claµ_10 7 229 0.0250 0.0517 0.0833 0.0217 0.0000 0.0208 0.0459 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Claµ_10 8 231 0.0000 0.0000 0.0069 0.0000 0.0000 0.0417 0.0077 Claµ_10 9 235 0.0000 0.0000 0.0000 0.0000 0.2857 0.2292 0.0689 Claµ_10 10 237 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 11 251 0.0250 0.0000 0.0000 0.0000 0.0000 0.0000 0.0026 Beni Claµ_10 12 253 0.0250 0.0000 0.0069 0.0000 0.0179 0.0000 0.0077 Claµ_10 13 255 0.0250 0.0000 0.0000 0.0000 0.0000 0.0000 0.0026 Beni Claµ_10 14 257 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 15 259 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 16 261 0.0000 0.0862 0.0903 0.0000 0.0000 0.0000 0.0459 Claµ_10 17 263 0.0250 0.0000 0.0764 0.0000 0.0000 0.0000 0.0306 Claµ_10 18 265 0.0000 0.0172 0.0278 0.0000 0.0000 0.0000 0.0128 Claµ_10 19 267 0.0500 0.0000 0.0069 0.0000 0.0000 0.0208 0.0102 Claµ_10 20 269 0.0000 0.0000 0.0000 0.0870 0.0000 0.0000 0.0102 Paragua Claµ_10 21 271 0.0000 0.0000 0.0069 0.0217 0.0179 0.0208 0.0102 Claµ_10 22 273 0.0000 0.0000 0.0000 0.0000 0.0179 0.0208 0.0051 Claµ_10 23 275 0.0000 0.0000 0.0000 0.0000 0.0893 0.0208 0.0153 Claµ_10 24 277 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 25 279 0.0000 0.0000 0.0000 0.0000 0.0000 0.0208 0.0026 Parana Claµ_10 26 281 0.0000 0.0000 0.0069 0.0000 0.0893 0.0208 0.0179

PAGE 220

200Claµ_10 27 285 0.0000 0.0000 0.0069 0.0000 0.0000 0.0208 0.0051 Claµ_10 28 287 0.0250 0.0517 0.0000 0.0000 0.0000 0.0000 0.0102 Claµ_10 29 289 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 30 293 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 31 295 0.0000 0.0345 0.0139 0.0000 0.0000 0.0000 0.0102 Claµ_10 32 303 0.0000 0.0000 0.0069 0.0000 0.0000 0.0000 0.0026 San_Miguel Claµ_10 # samples: 20 29 72 23 28 24 196 Cy_D2 1 253 0.0500 0.0000 0.0070 0.0000 0.0000 0.0000 0.0082 Cy_D2 2 257 0.0000 0.0000 0.0141 0.0000 0.0000 0.0000 0.0055 San_Miguel Cy_D2 3 259 0.0000 0.0000 0.0141 0.0000 0.0000 0.0000 0.0055 San_Miguel Cy_D2 4 261 0.0000 0.0000 0.0493 0.0000 0.0217 0.0000 0.0219 Cy_D2 5 263 0.0500 0.1111 0.1268 0.0000 0.0000 0.0000 0.0710 Cy_D2 6 265 0.1000 0.0370 0.0423 0.0000 0.0000 0.0000 0.0328 Cy_D2 7 267 0.0500 0.0000 0.0211 0.0000 0.0217 0.0000 0.0164 Cy_D2 8 269 0.0250 0.0000 0.0000 0.0652 0.0217 0.0000 0.0137 Cy_D2 9 271 0.0000 0.0000 0.0211 0.0870 0.2174 0.0000 0.0464 Cy_D2 10 273 0.0000 0.0000 0.0141 0.0217 0.0435 0.0000 0.0137 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Cy_D2 11 275 0.1000 0.1481 0.0775 0.0217 0.1522 0.0000 0.0847 Cy_D2 12 277 0.1000 0.1296 0.0282 0.0000 0.0870 0.0263 0.0546 Cy_D2 13 279 0.0750 0.0000 0.0634 0.0435 0.0652 0.0000 0.0464 Cy_D2 14 281 0.0000 0.0000 0.0845 0.0870 0.1087 0.0000 0.0574 Cy_D2 15 283 0.0000 0.0741 0.0845 0.1087 0.0000 0.2105 0.0792 Cy_D2 16 285 0.1250 0.0926 0.0845 0.0435 0.0000 0.1316 0.0792 Cy_D2 17 287 0.0500 0.1296 0.0563 0.0435 0.0000 0.0000 0.0519 Cy_D2 18 289 0.0750 0.0741 0.0634 0.0435 0.0000 0.0526 0.0546 Cy_D2 19 291 0.1000 0.0556 0.0141 0.0652 0.0217 0.2368 0.0601 Cy_D2 20 293 0.0250 0.0370 0.0070 0.0000 0.0000 0.0000 0.0109 Cy_D2 21 295 0.0250 0.0185 0.0423 0.0435 0.0000 0.0526 0.0328 Cy_D2 22 297 0.0250 0.0556 0.0141 0.0000 0.0217 0.0000 0.0191 Cy_D2 23 299 0.0000 0.0185 0.0070 0.0217 0.0217 0.0000 0.0109 Cy_D2 24 301 0.0000 0.0000 0.0211 0.0435 0.0000 0.0526 0.0191 Cy_D2 25 303 0.0000 0.0000 0.0070 0.1087 0.0217 0.0789 0.0273 Cy_D2 26 305 0.0000 0.0000 0.0070 0.0870 0.0000 0.0000 0.0137 Cy_D2 27 307 0.0250 0.0000 0.0141 0.0000 0.0217 0.0000 0.0109 Cy_D2 28 309 0.0000 0.0000 0.0141 0.0000 0.0435 0.0263 0.0137 Cy_D2 29 311 0.0000 0.0185 0.0000 0.0000 0.0435 0.0789 0.0164 Cy_D2 30 313 0.0000 0.0000 0.0000 0.0000 0.0217 0.0263 0.0055 Cy_D2 31 315 0.0000 0.0000 0.0000 0.0217 0.0217 0.0263 0.0082 Cy_D2 32 317 0.0000 0.0000 0.0000 0.0435 0.0000 0.0000 0.0055 Paragua

PAGE 221

201Cy_D2 33 377 0.0000 0.0000 0.0000 0.0000 0.0217 0.0000 0.0027 Paraguay Cy_D2 # samples: 20 27 71 23 23 19 183 Claµ_6 1 169 0.0250 0.0172 0.0069 0.0000 0.0536 0.0000 0.0153 Claµ_6 2 171 0.0000 0.0517 0.1111 0.0217 0.0000 0.0000 0.0510 Claµ_6 3 173 0.0500 0.0345 0.0347 0.0000 0.1071 0.1667 0.0587 Claµ_6 4 175 0.1000 0.1207 0.3403 0.3261 0.0536 0.0625 0.2066 Claµ_6 5 177 0.0000 0.0517 0.0208 0.1739 0.0000 0.0000 0.0357 Claµ_6 6 179 0.3750 0.3448 0.3403 0.0652 0.0893 0.0833 0.2449 Claµ_6 7 181 0.0500 0.0000 0.0069 0.1739 0.1429 0.2292 0.0765 Claµ_6 8 183 0.0750 0.0172 0.0417 0.0217 0.3036 0.1875 0.0944 Claµ_6 9 185 0.0500 0.2069 0.0139 0.0217 0.1607 0.2292 0.0944 Claµ_6 10 187 0.0250 0.1034 0.0208 0.0217 0.0893 0.0208 0.0434 Claµ_6 11 189 0.0250 0.0345 0.0000 0.0000 0.0000 0.0208 0.0102 Claµ_6 12 191 0.0000 0.0000 0.0069 0.0217 0.0000 0.0000 0.0051 Locus Allele# Size Pop1 Pop2 Pop3 Pop4 Pop5 Pop6 Overall Private? -----------------------------------------------------------Claµ_6 13 193 0.0750 0.0000 0.0000 0.0870 0.0000 0.0000 0.0179 Claµ_6 14 197 0.1500 0.0172 0.0208 0.0217 0.0000 0.0000 0.0281 Claµ_6 15 201 0.0000 0.0000 0.0347 0.0000 0.0000 0.0000 0.0128 San_Miguel Claµ_6 16 245 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Claµ_6 17 251 0.0000 0.0000 0.0000 0.0217 0.0000 0.0000 0.0026 Paragua Claµ_6 # samples: 20 29 72 23 28 24 196

PAGE 222

202 LITERATURE CITED Albert, J.S., and W.G.R. Crampton. 2005. Diversity and Phylogeny of Neotropical electric fishes (Gymnotiformes). In : H. Bullock, C.D. Hopkins, A.N. Popper, and R.R. Fay, eds. Electric Fishes (in press). Albert, J.S., N.R. Lovejoy, and W.G.R. Crampton. 2005. Miocene tectonism and the separation of cis-and trans-Andean basi ns: Evidence from Neotropical fishes. Journal of South American Earth Sciences (in press). Alcala, A.C., and M.T.S. Dy-Liacco. 1989. Habitats. Pp. 136-153. In : C.A. Ross, ed. Crocodiles and alligators. Facts On File, Inc., New York, N.Y. 240 p. Alho, C.J.R., T.E. Lacher, Jr., and H.C. Gon çalves. 1988. Environmental degradation in the Pantanal ecosystem. BioScience 38(3): 164-171. Alvarez del Toro, M. 1974. Los Crocodylia de México. Instituto Mexicano de Recursos Naturales Renovables, Distri to Federal, Mexico. 70 p. Amos, B., and A.R. Hoelzel. 1991. Long-te rm preservation of whale skin for DNA analysis. Reports for the International Whaling Commision. Special Issue 13: 99103. Aquino-Ortiz, A.L. 1988. Ontogenetic food sh ifts and their relation to morphological changes in the crocodilian Caiman yacare . Master’s thesis. Univ. New Mexico, Albuquerque, New Mexico. 77 p. Avise, J.C. 2000. Phylogeography – The history and formation of species. Harvard Univ. Press., Cambridge, Massechussetts. 447 p. Avise, J.C. 2004. Molecular marker s, natural history, and evolution (2nd ed,). Chapman and Hall, New York, N.Y. 684 p. Ayarzagüena, J. 1983. Ecología del caimán de anteojos o baba ( Caiman crocodilus L.) en los llanos de Apure (Venezuela). Doñana Acta Vertebrata 10: 1-136. Azara, F. de. 1802 (1978). Apuntamientos para la historia natural de los quadrúpedos del Paragüay y rio de la Plata. Tomo II. Im prenta de La Viuda de Ibarra, Madrid. (Reprint edition, Arno Press, NY. 328 p.

PAGE 223

203 Bampi, M.I., and M.E. Coutinho. 2003. Polic y for the conservation and management of jacaré ( Caiman yacare ) in Brazil. Pp. 35-48. In : J.P. Ross, and R. Godshalk, eds. International Workshop for Management and Trade of Caiman yacare . Crocodile Specialist Group publication. Gainesville, FL. 82 p. Bates, H.W. 1892 (1969). The naturalist on the river Amazons. J.M. Dant & Sons, London. 427 p. Beck, S.G. 1984. Comunidades vegetales de las sabanas inundazidas en el NE de Bolivia. Phytocoenologia 12(2/3): 321-350. Bermingham, E., and A.P. Martin. 1998. Comparative mtDNA phylogeography of Neotropical freshwater fish es: testing shared history to infer the evolutionary landscape of lower Central Ameri ca. Molecular Ecology 7: 499-517. Blohm, T. 1973. Conveniencia de cria r crocodilídeos en Venezuela con fines económicos y para prevenir su extinción. Unpubl. report presented at the Simposio Internacional Fauna Silvestr e, Pesca fluvial y lacustre Amazónica. Manaus, Brazil, Nov. 1973, 32 p. Bocourt, M.F. 1876. Note sur quelques rept iles de l’Istme de Tehuantepec (Mexique) donnés par M. Sumichrast au muséum. J ournal de Zoologie, Paris 5(5-6): 386-411. Brazaitis, P. 1986. Management, reproduction and growth of Caiman crocodilus yacare at the New York Zoological Park. Pp. 389-399. In : Crocodiles. Proceedings 7th Working Meeting Crocodile Specialists Group. IUCN Pub., Editorial Texto S.R.L., Caracas, Venezuela. 446 p. Brazaitis, P. 1989. The caiman of the Pantanal : Past, present and future. Pp. 119-124. In: Crocodiles. Proc. 8th Working Meeting Crocodile Specialists Group. IUCN Pub., Gland, Switzerland. 204 p. Brazaitis, P., G. Amato, G. Rebelo, C. Yama shita, and J. Gatesy. 1993. Report to CITES on the biochemical systematics study of yacare caiman, Caiman yacare , of central South America. Unpublished report to the CITES Secretariat, Geneva Switzerland. 43 p. Brazaitis, P., R. Madden, G. Amato, G. Re belo, and C. Yamashita. 1996. Revised systematics of the Caiman: Results of morphological, statis tical, and molecular genetic studies. Unpublis hed review draft. 36 p. Brazaitis, P., C. Yamashita, and G. Rebelo. 1990. A summary report of the CITES central South American caiman study: Phase I – Brazil. Pp. 100-115. In: Crocodiles. Vol. 1. Proc. 9th Working Meet ing Crocodile Specialists Group. IUCN Pub., Gland Switzerland. 399 p. Brochu, C.A. 1999. Phylogeny, systematics and hi storical biogeography of Alligatoridae. Society of Vertebrate Pa leontology Memoir 6: 9-100.

PAGE 224

204 Brochu, C.A. 2001. Progress and future direct ions in Archosaur phylogenetics. Journal of Paleontology 75(6): 1185-1201. Brochu, C.A. 2003. Phylogenetic approach es toward crocodylian history. Annual Review of Earth and Planetary Sciences 31: 357-397. Brochu, C.A. 2004a. Calibration age and quart et divergence date estimation. Evolution 58(6): 1375-1382. Brochu, C.A. 2004b. Patterns of calibration ag e sensitivity with quartet dating methods. Journal of Paleontology 78(1): 7-30. Brochu, C.A., and M.A. Norell. 2000. Tempor al congruence and the origin of birds. Journal of Vertebrate Paleontology 20: 197-200. Buffetaut, E. 1979. The evolution of crocodi lians. Scientific Amer ican 241: 130-144. Buffetaut, E. 1989. Evolution. Pp. 26-41. In : C.A. Ross, ed. Croc odiles and alligators. Facts on File, New York. 240 p. Busack, S.D., and S. Pandya. 2001 Geographic variation in Caiman crocodilus and Caiman yacare (Crocodylia: Alligatoridae): Systematic and legal implications. Herpetologica 57(3): 294-312. Bussing, W.A. 1985. Patterns of distribution of the Central American ichtiofauna. Pp. 157-175. In : F.G. Stehli, and S.D. Webb, eds. 1985. The Great American Biotic Interchange. Plenum Pre ss, New York, N.Y. 532 p. Campos, Z. 1993. Effect of habitat on survival of eggs and the sex ratio of hatchlings of Caiman crocodilus yacare in the Pantanal, Brazil. J ournal of Herpetology 27(2): 127-132. Campos, Z., and W. Magnusson. 1995. Relationshi p between rainfall, nesting habitat and fecundity of Caiman crocodilus yacare in the Pantanal, Brazil. Journal of Tropical Ecology 11(3): 51-58. Cardús. J. 1886. Las misiones Franciscanas en tre los Infieles de Bolivia. 1886. Librería de la Inmaculada Concepción, Barcelona, Spain. 425 p. Carvalho U., A. 1983. Beni Interpretación historica. Imprenta Renovación. La Paz, Bolivia. 100 p. Chiriví-Gallego, H. 1973. Contribución al cono cimiento de la babilla o yacaré-tinga, ( Caiman crocodilus ) con notas acerca de su manejo y de otras especies de Crocodylia neotropicales. INDERE NA Pubs, Bogotá, Colombia. 126 p.

PAGE 225

205 Cintra, R. 1989. A comparison of sizes of Caiman in hunted and non-hunted areas in the Brasilian Pantanal. Pp. 124-127. In : Crocodiles. Proc. 8th Working Meeting Crocodile Specialists Group. IU CN Pub., Gland, Switzerland. 204 p. CITES – Convention on International Trade of Endangered Species. 1983. Identification manual. Secretariat of the C onvention. IUCN, Gland, Switzerland. Claure E., D. 1986. Fauna silvestre en bus ca de otro "Arca de Noe". Concurso Tricentenario de la Fundación de Trinidad, Trinidad, Bolivia. 66 p. Clement, M.D., D. Posada, and K.A. Crandall. 2000. TCS: a computer program to estimate gene geneologies. Molecular Ecology 9(10): 1657-1660. Coates, A.G., and J.A. Obando. 1996. The geolog ical evolution of the Central American isthmus. Pp 21-56. In : J. Jackson, A.F. Budd, and A.G. Coates, eds. Evolution and environment in tropical America. Un iv. of Chicago Press, Chicago. 425 p. Colinvaux, P.A., and P.E. De Oliveira. 2001. Amazon plant diversity and climate through the Cenozoic. Paleogeography Paleoclimatology Paleoecology 166: 51-63. Coney, P.J. 1982. Plate tectonic constraint s on the biogeography of Middle America and the Caribbean region. Annals Misso uri Botanical Garden. 69: 432-443. Cope, E.D. 1868. An examination of the Re ptilia and Batrachia obtained by the Orton expedition to Equador [sic] and the u pper Amazon, with notes on other species. Proceedings of the Academy of Natura l Sciences of Philadelphia 20: 96-140. Cott, H.B. 1961. Scientific results of an inqui ry into the ecology and economic status of the Nile crocodile ( Crocodylus niloticus ) in Uganda and Northern Rhodesia. Transcripts of the Zoological Society of London 29: 211-358. Coutinho, M.E. 2000. Population ecology and the conservation and management of Caiman yacare in the Pantanal, Brazil. Doctoral Dissertation. Univ. of Queensland, Australia. 272 p. Coutinho, M.E., and Z. Campos. 1996. Effect of habitat and seasonal ity on the densities of caiman in southern Pantanal, Brazil. Journal of Tropical Ecology 12: 741-747. Coutinho, M.E., and Z. Campos. 2002. A utilização de populações naturais de jacaré ( Caiman crocodilus yacare ) como mecanismo de cons ervação do Panatanal. Pp. 47-59. In : L.M. Verdade, and A. Larriera, ed s. Conservação e manejo de jacarés e crocodilos da América Latina. Vol.II. C.N. Editora, Piracicaba. 184 p. Coutinho, M.E., Z. Campos, F. Cardoso, P. Massara, and A. Castro. 2001. Reproductive biology and its implication for the management of caiman, Caiman yacare , in the Pantanal wetland, Brazil. Pp. 229-243. In : G.C. Grigg, F. Seebacher, and C.E. Franklin, eds. Crocodilian biology and evolution. Sure y Beatty & Sons, Chipping Norton. 446 p.

PAGE 226

206 Craighead, F.C. 1968. The role of the alli gator in shaping plant communities and maintaining wildlife in the southern Ever glades. Florida Naturalist 41: 2-7, 69-74, 94. Crawshaw, Jr., P.G. 1987. Nesting ecology of the Paraguayan caiman ( Caiman yacare ) in the pantanal of Mato Groso, Brazil. Mast er’s thesis. Univ. Florida, Gainesville, Florida. 68 p. Crawshaw, Jr., P.G. 1991. Effects of hunti ng on the reproduction of the Paraguayan caiman ( Caiman yacare ) in the pantanal of Mato Grosso, Brazil. Pp. 145-153. In : J.G. Robinson, and K.H. Redford, eds. Ne otropical wildlife use and conservation. Univ. Chicago Press, Chicago. 520 p. Crawshaw, Jr., P.G., and G.B. Schaller. 1980. Nesting of the Paraguayan caiman. Papéis Avulsos de Zoologia 33: 283-292. Cronin, T.M., and H.J. Dowsett. 1996. Bi otic and oceanographic response to the Pliocene closing of the Central American Isthmus. Pp. 76104. In : J. Jackson, A.F. Budd, and A.G. Coates, eds. Evolution and environment in tropical America. Univ. of Chicago Press, Chicago. 425 p. Cuvier, G. 1807. Sur les differentes espèces de crocodiles vivans et sur leurs caractères distinctifs. Annales du Muséum d'Hi stoire Naturelle, París 10: 8-86. Daudin, F.M. 1801 [1802]. Historie naturelle, générale et particulière des reptiles; ouvrage faisant suit à l’Historie naturell e générale et particulière, composée per Leclerc de Buffon; et rédigee par C.S. Sonnini, member de plusieurs sociétés savants. F. Dufart, París. Tomo II. 432 p. Daugherty, C.H., A. Cree, M. Hay, and M. B. Thompson. 1990. Neglected taxonomy and continuing extinctions of tuatara ( Sphenodon ) Nature 347: 177-179. Davenport, M. 1995. Evidence for possi ble sperm storage in the caiman, Paleosuchus palpebrosus . Herpetological Review 26: 14-15. Davis, L.M., T.C. Glenn, R.M. Elsey, I.L. Br isbin, Jr., W.E. Rhodes, H.C. Dessauer, and R.H. Sawyer. 2001a. Genetic structure of six populations of American alligators: A microsatellite analysis. Pp. 38-50. In : G.C. Grigg, F. Seebacher, and C.E. Franklin, eds. Crocodilian biology and evolution. Sure y Beatty & Sons, Chipping Norton. 446 p. Davis, L.M., T.C. Glenn, R.M. Elsey, H.C. Dessauer, and R.H. Sawyer. 2001b. Multiple paternity and mating patterns in the American alligator, Alligator mississippiensis . Molecular Ecology 10: 1011-1024.

PAGE 227

207 Davis, L.M., T.C. Glenn, D.C. Strickland, L. J. Guillette, Jr., R.M. Elsey, W.E. Rhodes, H.C. Dessauer, and R.H. Sawyer. 2002. Microsatellite DNA analyses support an East-West phylogeographic split of Ameri can alligator populations. Journal of Experimental Zoology (Mol Dev Evol) 294: 352-372. Denevan, W.M. 1966. The aboriginal cultural geography of the Llanos de Mojos of Bolivia. Univ. California Press, Berkeley, California. 185 p. Denevan, W.M. 1970. Aboriginal drained-fiel d cultivation in the Americas. Science 169: 647-654. Denevan, W.M. 1991. Prehistoric roads and cause ways of lowland tropical America. Pp. 230-242. In : C.D. Trombold, ed. Ancient road networks and settlement hierarchies in the New World. Cambridge Univ. Pr ess, Cambridge, Great Britain. 277 p. Densmore, L.D. 1983. Biochemical and im munological systematics of the order Crcocodilia. Evolutionary Biology 16: 397-465. Densmore, L.D., and R.D. Owen. 1989. Molecu lar systematics of the Order Crocodilia. American Zoologist 29: 831-841. Densmore, L.D., and P.S. White. 1991. The syst ematics and evolution of the Crocodylia as suggested by restrictio n endonuclease analysis of mitochondrial and nuclear ribosomal DNA. Copeia 1991: 602-615. Desjardins, P., and R. Morais. 1990 Sequen ce and gene organization of the chicken mitochondrial genome. A novel gene order in higher vertebrates. Journal of Molecular Biology 212(4): 599-634. Dever, J.D., R.E. Strauss, T.R. Rainwater, S.T. McMurray, and L.D. Densmore III. 2002. Genetic diversity, population subdivision, a nd gene flow in Morelet’s crocodile ( Crocodylus moreletii ) from Belize, Central Amer ica. Copeia 2002(4): 1078-1091. DeWoody J.A., and J.C. Avise. 2000. Microsatel lite variation in ma rine, freshwater and anadromous fishes compared with other animals. J. Fish Biol. 56: 461–473. Donnelly, T.W. 1988. Geologic constraint s on Caribbean biogeography. Pp. 15-37. In : J.K. Liebherr, ed. Zoogeography of Caribbean insects. Cornell Univ. Press. Ithaca, N.Y. 285 p. Duque-Caro, H. 1990. Neogene stratigra phy, paleoceanography and paleobiogeography in northwest South America and the evolution of the Panama Seaway. Palaeography, Palaeoclimatol ogy, Palaeoecology 77: 203-234. Eder, F.X. 1791 (1985). Breve descripción de las reducciones de Mojos. (Translation to Spanish by J.M. Barnadas). Impresiones Poligraf, Cochabamba, Bolivia. 424 p.

PAGE 228

208 Emmons, L.H. 1990. Neotropical rainforest mammals – a field guide. Univ. of Chicago Press. Chicago, Illinois. 279 p. Ergueta. P. 1987. Panorama preliminar de la he rpetología en Bolivia. Centro de Datos para la Conservación, La Paz, Bolivia. 34 p. Ergueta. P. 1990. La situación de la fauna en Bolivia. Centro de Datos para la Concervación, La Paz, Bolivia. 9 p. Ergueta, P., and L.F. Pacheco. 1990. Los cr ocodilios (Orden Croc odylia) de Bolivia. Ecología en Bolivia 15: 69-81. Ergueta, P., and T.S.J. Salazar. 1991. Fauna silvestre de Bolivia. Memoria explicativa. Instituto Geográfico Milita r, La Paz, Bolivia. 97 p. Ernst, C.H., and R.W. Barbour. 1989. Turtles of the world. Smithsoni an Institute Press, Washington, D.C. 313 p. Excoffier, L., P.E. Smouse, and J.M. Qua ttro. 1992. Analysis of molecular variance inferred from metric distances among DNA haplotypes: Application to human mitochondrial DNA restriction data. Genetics 131: 479-491. Excoffier, L., G. Laval, and S. Schneider . 2005. Arlequin ver. 3.0: An integrated software package for population gene tics data analysis. Evolutionary Bioinformatics Online (submitted). Farias, I.P., R. Da Silveira, B. de Thois y, L.A. Monjoelo, J. Thorbjarnarson, and T. Hrbek. 2004. Genetic diversity and population structure of Amazonian crocodilians. Animal Conservation 7: 265-272. Fariña Sanchez, T. 1973. The climate of Paraguay. Pp. 33-38. In : J.R. Gorham, ed. Paraguay: Ecological essays. Acad. of Ar ts and Sciences. Miami, Florida. 296 p. Fitch, H., and M. Nadeau. 1980. Assessment of Caiman latirostris and Caiman crocodilus yacare in northern Argentina. Unpublished Final Report for Project 793162. New York Zoological Society, 14 p. Fittkau, E.J. 1973. Crocodiles and nutrient metabolism of Amazon waters. Amazoniana 4(1): 103-133. Fittkau, E.J. 1970. Role of caimans in the nutrient regime of mouth-lakes of Amazon affluents. Biotropica 2(2): 138-142. Fitzgerald, L.A. 1988. Dietary patterns of Caiman crocodilus in the Venezuelan Llanos. Master’s thesis. Univ. New Mexi co, Albequerque, New Mexico. 74 p. Fitzgerald, S. 1989. International wildlife tr ade: Whose business is it? World Wildlife Fund, Washington, D.C. 459 p.

PAGE 229

209 Fraily, C.D., E.L. Lavina, A. Rancy, and J.P. de Souza Filho. 1988. A proposed Pleistocene/Holocene lake in the Amazon Basin and its significance to Amazonian geology and biogeography. Acta Amazonica 18(3-4): 119-143. Frair, W., and J.L. Behler. 1983. Review of Liste der Resenten Amphibien und Reptilien, Testudines, Crocodylia, Rhyncocephalia . Herpetological Review 14: 23-25. Fu, Y.-X. 1997. Statistical tests of neutra lity of mutations against population growth, hitchhiking and background se lelction. Genetics 147: 915-925. Fuchs, K. 1971. Die sudamerianischen Reptilhaute. Das Leder 22: 3-19. Fuchs, K. 1974. Die Krokodilhaute. Eduard Rother Verlag. Darmstadt. 183 p. Fuchs, K., C.A. Ross, A.C. Pooley, and R. Whitaker. 1989. Crocod ile-skin products. Pp. 188-195. In: C.A. Ross, ed. Crocodiles and al ligators. Facts on File, New York. 240 p. Garnett, S.T. 1984. The consequences of slow chiton digestion on crocodilian diet analysis. Journal Herpetology 19(2): 303-304. Garrick, L.D., and J.W. Lang. 1977. Social sign als and behaviors of adult alligators and crocodiles. American Zoologist 17: 225-239. Glenn, T.C., H.C. Dessauer, and M. Braun. 1998. Characterization of microsatellite DNA loci in American alligators. Copeia 1998: 591-602. Glenn, T.C., J.L Staton, A.T. Vu, L.M. Davis, J.R. Alvarado Bremer, W.E. Rhodes, I.L. Brisbin, and R.H. Sawyer. 2002. Low mitochondrial DNA variation among American Alligators and a novel non-coding region in crocodilians. Journal of Experimental Zoology (Mol Dev Evol) 294: 312-324. Godshalk, R. 1976. Contribución al c onocimiento del ciclo de vida del Caiman c. crocodilus . Parte II: Epocha de lluvia. Resu men. II Seminario sobre Chigüires ( Hidrochoerus hidrochoerus ) y Babas ( Caiman crocodilus ). CONICT, Maracay, Venezuela. Unpublished ms. Godshalk, R. 1978. El caimán del Orinoco, Crocodylus intermedius , en los llanos occidentales venezolanos, con observacione s sobre su distribución en Venezuela y recomendaciones para su conservaci ón. FUDENA (World Wildlife Fund Venezuela), Caracas, Venezuela. 84 p. Godshalk, R. 1982. The hab itat and distribution of Paleosuchus in Venezuela. Pp. 3138. In : Crocodiles. Proc. 5th Working M eeting Crocodile Specialists Group. IUCN, Gland, Switzerland. 409 p.

PAGE 230

210 Godshalk, R. 1998a. Paleosuchus palpebrosus . Pp 29-30. In : J.P. Ross, ed. Crocodiles. Status survey and action plan (2nd ed). IUCN/SSC Crocodile Specialist Group. IUCN, Gland, Switzerland. 96 p. Godshalk, R. 1998b. Paleosuchus trigonatus . Pp 31-32. In : J.P. Ross, ed. Crocodiles. Status survey and action plan (2nd ed). IUCN/SSC Crocodile Specialist Group. IUCN, Gland, Switzerland. 96 p. Godshalk, R., and F.W. King. 2002. Status of Caiman yacare in Bolivia: Results of 1995 and 1996 CITES surveys. Pp. 269-314. In : Crocodiles. Proc. 16th Working Meeting Crocodile Specialists Gr oup. IUCN, Gland, Switzerland. 419 p. Gorzula, S.J. 1987. The management of crocodilians in Venezuela. Pp. 91-101. In : G.J.W. Webb, S.C. Manolis and P.J. Whitehead, eds. Wildlife Management: Crocodiles and alligators. Surrey Beaty and Sons, Chipping Norton, Australia. 552 p. Gorzula, S.J. 1994. A longirostrine Caiman crocodilus from central Venezuela. Crocodile Specialist Group Newsletter 13(3): 16. Gorzula, S.J. 1978. An ecological study of Caiman crocodilus crocodilus inhabiting the lagoons in the Venezuelan Guay ana. Oecologia 35: 21-34. Gorzula, S.J., and A.C. Seijas. 1989. The common caiman. Pp. 44-61. In : Crocodiles: Their ecology, management, and conser vation. IUCN, Gland, Switzerland. 308 p. Goulding, M. 1981. Man and fishes on an Am azon frontier. Dr. W. Junk, The Hague. 137 p. Goulding, M.J., R. Bartham, and E.G. Ferre ira. 2003. The Smithsonian atlas of the Amazon. Smithsonian Institute Press, Washington, D.C. 253 p. Gray, J.E. 1862. A synopsis of the species of alligators. Annals and magazine of Natural History. London. (3)10: 327-331. Gratton, J. 2003. The molecular systematics, phylogeography and popul ation genetics of Indo-Pacific Crocodylus . Doctoral dissertation, Un iversity of Queensland, Australia. 222 p. Grigg, G.C., L.A. Beard, T. Moulton, M.T. Queirol Melo, and L.E. Taplin. 1998. Osmoregulation by the broad-snouted caiman, Caiman latirostris , in estuarine habitat in southern Brazil. Journal of Comparative Physiology 168B: 445-452. Groombridge, B. 1982. The IUCN Amphibia-Re ptilia red data book. Part 1: Testudines, Crocodylia, Rynchocephalia. IU CN, Gland, Switzerland. 426 p.

PAGE 231

211 Groombridge, B. 1987. The distribution and stat us of the world crocodilians. Pp. 9-21. In : G.J.W. Webb, S.C. Manolis, and P.J. Whitehead, eds. Wildlife management: Crocodiles and alligators. Su rrey Beatty & Sons, Chipping Norton, Australia. 552 p. Groombridge, B., and R.W.G. Jenkins, ed s. 1994. Biodiversity Data Sourcebook. Compiled by the World Conservation M onitoring Centre. World Conservation Press, Cambridge, UK. Hammond, N. 1980. Prehistoric human utilization of the savanna environments of Middle and South America. Pp. 73-105. In : D.R. Harris, ed. Human ecology in savanna environments. Academic Press, London. 522 p. Haase, R. 1989. Plant communities of a sa vanna in northern Bolivia I. Seasonally flooded grassland and gallery fore st. Phytocoenologia 18(1): 55-81. Haase, R., and S.G. Beck. 1989. Structure and composition of savanna vegetation on northern Bolivia: A preliminary re port. Brittonia 41(1): 80-100. Hallam, A. 1994. An outline of Phanerozoic biogeography. Oxford University Press, Oxford, G.B. 246 p. Hamada H., M.G. Petrino and T. Kakunaga . 1982. A novel repeated element with ZDNA-forming potential is widely found in evolutionarily diverse eukaryotic genomes. Proceedings of the National Academy of Sciences 79 : 6465-6469. Haq, B.U., J. Hardenbol, and P.R. Vail. 1987. Chronology of fluctuati ng sea levels since the Triassic. Science 235(4793): 1156-1167. Hardin, G. 1968. The tragedy of the commons. Science 162: 1243-44 Harris, D.R. 1980. Tropical savanna environm ents: Definition, distri bution, diversity and development. Pp. 3-27. In : D.R. Harris, ed. Human ecology in savanna environments. Academic Press, London. 522 p. Harshman, J., C.J. Huddleston, J.P. Bollback, T.J. Parsons, and M.J. Braun. 2003. True and false gharials: a nuclear gene phylogeny of crocodylia. Systematic Biology 52: 386-402. Hartl, D.L., and A.G. Clark. 1997. Principles of population genetics. Sinauer Assoc., Inc. Sunderland, Massachusetts. 542 p. Hayes, F.E. 1995. Status, distribution a nd biogeography of the birds of Paraguay. American Birding Association, Colorado Springs, Colorado. 232 p. Hecht, S.B., A.B. Anderson, and P. May. 1988. The subsidy from nature: Shifting cultivation, successional palm forests, a nd rural development. Human Organization 47(1): 25-35.

PAGE 232

212 Hedges, H.B., and L.L. Poling. 1999. A molecular phylogeny. Science 83: 998-1001. Hillis, D.M., B.K. Mable, A. Larson, S.K. Davis, and E.A. Zimmer. 1996. Nucleic Acids IV: Sequencing and cloning. Pp. 321-381. In : D.M. Hillis, C. Moritz, and B.K. Mable, eds. Molecular systematics (2nd ed.). Sinauer Assoc., Inc., Sunderland, Massachusetts. 655 p. Hoogesteijn, R., and E. Mondolfi. 1992. The jaguar. Armitano Editores C.A. Caracas, Venezuela. 182 p. Hoorn, C. 1993. Marine incursions and th e influence of Andean tectonics on the Miocene depositional history of northwestern Amazonia: results of a palynnstratigraphic study. Palaeogra phy, Palaeoclimatology, Palaeoecology 105: 267-309. Hoorn, C. 1994. Fluvial palaeoenvironments in the Amazonas Basin (Early Mioceneearly Middle Miocene, Colombia). Palaeography, Palaeoclimatology, Palaeoecology 109: 1-54. Hoorn, C., J. Guerrero, G.A. Sarmiento, and M. A. Lorente. 1995. Andean tectonics as a cause for changing drainage patterns in Miocene northern South America. Geology 23(3): 237-240. Hudson, R.R. 1990. Gene geneologues and the coalescent process. Oxford Surveys in Evolutionary Biology 7:1-44. Huelsenbeck, J.P., and D.M. Hillis. 1996. Parametric bootstrapping in molecular phylogenetics: Applications and performance. Pp 19-45. In : J.D. Ferraris, and S.R. Palumbi, eds. Molecular Zoology: Advances, strategies and protocols. J. Wiley and Sons, Inc., New York. Huelsenbeck, J.P., and B. Rannala. 1997. Phyl ogenetic methods come of age: testing hypothesis in an evolutionary context. Science 276: 227-232. Huelsenbeck, J.P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17: 754-755. Ibarra G., D.E., and R. Querejazu L. 1986. 30,000 años de prehistoria en Bolivia. Editorial Amigos del Libro, La Paz, Bolivia. 365 p. IUCN. 1980. World conservation strategy. Livi ng resource conservation for sustainable development. IUCN-UNRP-WWF, Gland, Switzerland. 44 p. Iwabe, N., Y. Hara, Y. Kumazawa, K. Shib amoto, Y. Saito, T. Miyata, and K. Katoh. 2005. Sister group relationships of turtles to the bird-cro codilian clade revealed by nuclear DNA-coded proteins. Molecular Biology and Evolution 22(4): 810-813.

PAGE 233

213 Jackson, J.F., H.W. Campbell, and K.E. Campbell. 1974. The feeding habits of crocodilians: Validity of the evidence from stomach contents. Journal of Herpetology 8(4): 378-381. Janke, A., and U. Arnason. 1997. The complete mitochondrial genome of Alligator mississippiensis and the separation between rece nt Archosauria (birds and crocodiles). Molecular Biology a nd Evolution 14(12): 1266-1272. Janke, A., D. Erpenbeck, M. Nilsson, and U. Arnason. 2001. The mitochondrial genomes of the iguana ( Iguana iguana ) and the caiman ( Caiman crocodylus [sic]): implications for amniote phylogeny. Proceed ings of the Royal Society of London, B, Biologial Science 268 (1467): 623-631. Janke, A., A. Gullberg, S. Hughes, R.K. Aggarwal, and U. Arnason. 2005. Mitogenomic analyses place the gharial ( Gavialis gangenticus ) on the crocodile tree and provide pre-K/T divergence times for most crocodi lians. Journal of Molecular Ecology 61: 620-626. Jenkins, R.W.G. 1987. The World Conservati on Stategy and CITES: Principles for the management of crocodilians. Pp. 27-31. In : G.J.W. Webb, S.C. Manolis, and P.J. Whitehead, eds. Wildlife management: Crocod iles and alligators. Surrey Beatty & Sons, Chipping Norton, Australia. 552 p. Johns, G.C., and J.C. Avise. 1998. A compara tive summary of genetic distances in the vertebrates from the mitochondrial cytochrome b gene. Molecular Biology and Evolution 15(11): 1481-1490. Jones, J. 1980. Conflict between whites a nd Indians on the llanos de Moxos, Beni Department: A case study in development from the cattle regions of the Bolivian Oriente. Doctoral disser tation. Univ. Florida, Gainesville, Florida. 321 p. Junk, W.J. 1975. Aquatic wildlif e and fisheries. Pp. 109-125. In : The use of ecological guidelines for the development in the American humid tropics. IUCN, Surrey, England. 249 p. Kandpal R.P., G. Kandpal, and S.M. Weissman. 1994. Construction of libraries enriched for sequence repeats and jumping clones, and hybridization selection for regionspecific markers. Proceedings of the Na tional Academy of Sciences 91: 88-92. Kempff M., N. 1985. Aves de Bolivia. Edito rial Gisbert y Cia., La Paz, Bolivia. Kessing, B., H. Croom, A. Martin, C. Mc Intosh, W.O. McMillan, and S. Palumbi, (compilers). 1989. The simple fool’s guide to PCR – Version 1.0. Univ. of Hawaii, Honolulu, Hawaii. 44 p.

PAGE 234

214 Kijas J.M.H., J.C.S. Fowler, C.A. Garbett, and M.R. Thomas. 1994. Enrichment of microsatellites from the citrus geno me using biotinylated oligonucleotide sequences bound to streptavidin-coated ma gnetic particles. BioTechniques 16: 657660. King, F.W. 1978. The wildlife trade. Pp. 253-271. In : H.P. Brokaw, ed. Wildlife and America. Contributions to an unders tanding of American wildlife and its conservation. Council on Environm ental Quality, Washington, D.C. King, F.W. 1989. Conservation and management. Pp. 216-229. In : C.A. Ross, ed. Crocodiles and alligators. F acts on File, New York. 240 p. King, F.W., A.L. Aquino, N.J. Sc ott, and R. Palacios. 1994. St atus of the crocodilians of Paraguay. Pp. 162-198. In : Crocodiles. Proc. 12th Working Meeting Crocodile Specialists Group, Vol. 2. IUCN Pub., Gland, Switzerland. 340 p. King, F.W., and R.L. Burke. 1989. Crocodilian, tuatara and turt le species of the world. A taxonomic and geographic reference. A ssociation of Systematics Collections, Washington. 216 p. King, F.W., M. Espinal, and C.A. Cerrat o. 1990. Distribution and status of the crocodilians of Honduras. Pp. 313-354. In : Crocodiles. Proc. 10th Working Meeting Crocodile Specialists Group, Vol. 1. IUCN Pub., Gland, Switzerland. 354 p. King, F.W., and R. Godshalk. 1997. A program for the sustainable utilization and management of caimans: A report to the G overnment of Bolivia on the final results of the 1995 and 1996 field seasons. Unpublishe d report to the Bolivian Government and the CITES Secretariat. 96 p. King, F.W., and D.H. Videz. 1989. The caima ns of Bolivia: A preliminary report on a CITES and Centro de Dessarollo Forestal sponsored survey of species distribution and status. Pp. 128-155. In : Crocodiles. Proc. 8th Wo rking Meeting Crocodile Specialists Group. IUCN Pub., Gland, Switzerland. 204 p. King, F.W., J. Thorbjarnarson, and C. Yamashita. 1998. Cooperative feeding, a misinterprested and under-reported behavi or of crocodians. 9 p. Available at http://www.flmnh.ufl.edu/natsci/her petology/herpbiology/Bartram.htm. Knowlton, N., and L.A. Weigt. 1988. New date s and new rates for divergence across the Isthmus of Panama. Proceedings of the Royal Society of London B 265: 22572263. Kocher, T.D., W.K. Thomas, A. Meyer, S.V. Edwards, F.X. Villablanca, and A.C. Wilson. 1989. Dynamics of mitochondrial DNA evolution in animals: Amplification and sequencing with conserve d primers. Proceedings of the National Academy of Science USA 86: 6196-6200.

PAGE 235

215 Krieg, H. 1948. Zwischen Anden und Atlant ik. Ed. Carl Hanser, München, Germany. 492 p. Kumar, S., K. Tamura, and M. Nei. 2004. MEGA3: Integrated Software for Molecular Evolutionary Genetics Analysis and Sequence Alignment. Briefings in Bioinformatics 5:150-163. Kumizawa, Y., and M. Nishida. 1999. Co mplete mitochondrial DNA sequences of the green turtle and blue-tailed mole skink: Statistical evidence for Archosaurian affinity of turtles. Molecular Biology and Evolution 16(6): 784-792. Langston, Jr., W. 1965. Fossil crocodilians from Colombia and the Cenozoic history of the Crocodylia in South America. Un iv. California Pubs. Geology 52: 1-169. Lauzanne, L., and G. Loubens. 1985. Peces del Rio Mamoré. Éditions de l'ORSTOM, Paris. 115 p. Leslie, A.J., and L.E. Taplin. 2001. Recen t developments in osmoregulation of crocodilians. Pp. 265-279. In : G.C. Grigg, F. Seebacher, and C.E. Franklin, eds. Crocodilian biology and evol ution. Surey Beatty & Sons, Chipping Norton. 446 p. Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum Classes, Ordines, Genera, Species, cum characteribus, difere ntiis, synonymis, locis. Ed. 10, Tomus 1. L. Salvii, Stockholm. 824 p. Lovejoy, N.R., and B.B. Collette. 2001. Phylogenetic relationships of New World needlefishes (Teleostei: Belonidae) a nd the biogeography of transitions between marine and freshwater habitats. Copeia 2001: 324-338. Lovisek, J. 1977. To catch a caiman. Rotunda 10(3): 20-25. Lundberg, J.G., and B. Chernolt. 1992. A Miocene fossil of the Amazonian fish Arapaima (Teleostei, Arapamidae) from the Magdalena Rivier region of Colombia – Biogeographic and evolutionary im plications. Biotropica 24(1): 2-14. Lundberg, J.G., L.G. Marshall, J. Guerrer o, B. Horton, M.C.S.L. Malabarba, and F. Wesselingh. 1998. The stage for Neotropical fish diversification: A history of tropical South American rivers. Pp. 13-48. In : L.R. Malabara, R.E. Reis, C.A.S. de Lucena, and Z.M.S. de Lucena, eds. P hylogeny and classification of Neotropical fishes. Edipurcs, Porto Alegre, Brazil. 603 p. MacGregor, J. 2002. Review and analysis of the trade and industry dynamics for market-based conservation, Pp. 12-18. In : Crocodiles. Proc. 16th Working Meeting Crocodile Specialists Group. IUCN, Gland, Switzerland. 419 p. Magnusson, W.E. 1982. Techniques in surv eying for crocodilians. Pp. 389-403. In : Crocodiles. Proc. 5th Working Meeting Crocodile Specialists Group. IUCN Pub, Gland, Switzerland. 409 p.

PAGE 236

216 Magnusson, W.E. 1984. Economics, developing countries, and the captive propagation of crocodilians. Wildlife So ciety Bulletin 12: 194-197. Magnusson, W.E. 1989. Paleosuchus . Pp. 101-109. In : Crocodiles: Their ecology, management, and conservation. IUCN, Gland, Switzerland. 308 p. Magnusson, W.E., G.J. Caughley, and G.C. Grigg. 1978. A double-survey estimate of population size from incomplete counts. Journal Wildlife Management 42: 174176. Magnusson, W.E., E. Viera da Silva, and A.P. Lima. 1987. Diets of Amazonian crocodilians. Journal He rpetology 21(2): 61-66. Maness, S.J. 1976. Contribución al c onocimiento del ciclo de vida del Caiman crocodilus . Parte I: Epocha de sequía. Resu men. II Seminario sobre Chigüires ( Hidrochoerus hidrochoerus ) y Babas ( Caiman crocodilus ). CONICT, Maracay, Venezuela. Unpublished ms. Marcellini, D.L. 1979. Activity patterns a nd densities of Venezuelan caiman ( Caiman crocodilus ) and pond turtle ( Podocnemis vogli ). Pp. 262-271. In : J. Eisenberg, ed. Vertebrate ecology in the northern Neotropics. Smithsonian Institution Press, Washington, D.C. Marshall, L.G., R.F. Butler, R.E. Drak e, G.H. Curtis, and R.H. Telford. 1979. Calibration of the Great American Interchange. Science 204: 272-279. Marshall, L.G., S.D. Webb, J.J. Sepkoski, and D.M. Raup. 1982. Mammalian evolution and the Great American Inte rchange. Science 215: 1351-1357. Martin, A.P., and E. Bermingham. 2000. Regional endemism and cryptic species revealed by molecular and morphological an alysis of a widespread species of Neotropical catfish. Proceedings of the Royal Society of London B 267 : 11351141. Medem, F. 1955. A new subspecies of Caiman sclerops from Colombia. Fieldiana: Zoology 37: 339-344. Medem, F. 1960. Notes on the Paraguayan caiman, Caiman yacare (Daudin). Mitteilungen aus dem Zoologischen Museum im Berlin 36: 129-142. Medem, F. 1981. Los Crocodylia de Sur America.Vol. 1. Los Crocodylia de Colombia. Editorial Carrera Limitada , Bogotá, Colombia. 354 p. Medem, F. 1983. Los Crocodylia de Sur Amer ica. Volumen 2. Universidad Nacional de Colombia, Bogotá, Colombia. 270 p. Medina, C.J. 1976. Crocodilians from the Late Tertiary in northwestern Venezuela: Melanosuchus fisheri sp nov. Bereviora 438: 1-14.

PAGE 237

217 Messel, H., G.C. Vorlicek, A.G. Wells, and W.J. Green. 1981. Surveys of tidal river systems in the Northern Territory of Au stralia and their crocodile populations. Monograph No. 1. The BlythCadell river systems st udy and the status of Crocodylus porosus in tidal waterways of Northern Australia. Pergamon Press, New York, N.Y. 463 p. Métraux, A. 1942. The native tribes of easte rn Bolivia and western Matto Grosso. Smithsonian Institution Bureau of American Ethnology 54: 1-182. Meyer, E.R. 1984. Crocodilians as living fossils. Pp. 105-131. In : N. Eldridge and S. M. Stanley, eds. Living fossils. Springer-Verlag, New York. 291 p. Mittermeier, R.A., I. de Gusmão Câmara, M.T. Jorge Pádua, and J. Blanck. 1990. Conservation of the Pantanal of Brazil. Oryx 24(2): 103-112. Monteiro, L., M. Cavalcanti, and H. Sommer. 1997. Comparative ontogenetic shape changes in the skull of caiman species (Crocodylia, Alligatoridae). Journal of Morphology 231: 53-62. Montes de Oca, I. 1989. Geographía y recurs os naturales de Bolivia. Ministerio de Educación y Cultura, La Paz, Bolivia. 574 p. Moraga A., D., W. Farmarie, D. Brazeau, and A. Clark. 2001. Tools for developing molecular markers. ICBR, Univ. of Fl orida, Gainesville, Florida. 71 p. Morales, C.B. de. 1990. Bolivia. Medio ambi ente y ecología aplicada. Instituto de Ecología, La Paz, Bolivia. 318 p. Mourão, G., P. Bayliss, M.E. Coutinho, C. Ab ercrombie, and A. Aruda. 1994. Test of an aerial survey for caiman and other wildlife in the Pantanal, Brazil. Wildlife Society Bulletin 20: 50-56. Mourão, G., Z. Campo, M.E. Coutinho, and C. Abercrombie. 1996. Size structure of illegal harvested and surviving caiman ( Caiman crocodilus yacare ) in the Pantanal, Brazil. Biological C onservation 75: 261-265. Mourão, G., M.E. Coutinho, R. Mauri, Z. Campo, W. Tomás, and W. Magnusson. 2000. Aerial surveys of caiman, marsh deer, and pampas deer in the Pantanal Wetland of Brazil. Biological C onservation 92: 175-83. Müller, L., and W. Hellmich. 1936. Amphi bien und Reptilien. I. Teil. Amphibia, Chelonia, Loricata. Wissenschaft Ergeb Deutsch. Gran Chaco Expedition. Strecker & Schröder, Stuttgart. 120 p. Muniz Saavedra, J. M. 1983. El yaca re nato. Fauna Argentina 3: 1-32. Muñoz Reyes, J. 1991. Geographía de Bolivia . Empresa Editora "Urquizo S.A.", La Paz. 521 p.

PAGE 238

218 Nauta, M.J., and F.J. Weissing. 1996. Constr aints on allele size at microsatellite loci: implications for genetic diffe rentiation. Genetics 143: 1021-1032. Nei, M. 1987. Molecular evolutionary geneti cs. Columbia University Press, New York, N.Y. 512 p. Neill, W.T. 1974. The last of the ruling rept iles – alligators, crocodiles and their kin. Coumbia Univ. Press, New York, N.Y. 486 p. Norell, M.A. 1989. The higher level relationshi ps of the extant Crocodylia. Journal of Herpetology 23: 325-335. Oblitas, P.E. 1971. Magia, hechicería y medi cina popular boliviana. Ediciones Isla. La Paz, Bolivia. 602 p. Olson, G.A., J.R. Hessler, and R.E. Fait h. 1975. Technics for blood collection and intravascular infusion of reptiles. La boratory Animal Science 25(6): 783-786. Ouboter, P.E 1996. Ecological studies on croc odilians in Suriname. Niche segregation and competion in three predators. SPB Academic Publishing, Amsterdam, The Netherlands. 139 p. Ouboter, P.E., and L.M.R. Nanhoe. 1984. An ecological study of Caiman crocodilus crocodilus in northern Surinam. Unpublished Report. Catholic Univ. Nijmegen, Nijmegen, The Netherlands. 65 p. Ouboter, P.E., and L.M.R. Nanhoe. 1989. Notes on the dynamics of a population of Caiman crocodilus crocodilus in Northern Suriname and its implications for management. Biological C onservation 48: 243-264. Pacheco, L. 1993. Abundance, distribution, a nd habitat use of crocodilians in Beni, Bolivia. Master’s thesis. Univ. Fl orida, Gainesville, Florida. 116 p. Pareja L., J., C. Vargas F., R. Suárez S., R. Ballón A., R. Carrasco C., and C. Villarroel A. 1978. Mapa geológico de Bolivia: Memoria explicativa. Yacimientos Petrolíferos Fiscales Bolivi anos y Servicio Geológico de Bolivia, La Paz, Bolivia. 27 p. Peducassé C., A. 1982. Mineral status of grazi ng beef cattle in the tropics of Bolivia. Master’s thesis. Univ. Florida, Gainesville, Florida. 121 p. Poe, S. 1996. Data set incongruence and th e phylogeny of crocodilians. Systematic Biology 45: 393-414. Pooley, A.C., and C.A. Ross. 1989. Mortality and predators. Pp. 92-101. In : C.A. Ross, ed. Crocodiles and alligators. Facts on File, New York. 240 p.

PAGE 239

219 Posada, D., and K.A. Crandall. 1998. MODELTEST: testing the model of DNA substitution. Bioinformatics 14: 817-818. Pritchard, P.C.H., and P. Trebbau. 1984. The tu rtles of Venezuela. Society for the Study of Amphibians and Reptiles. Cushing-Ma lloy, Inc, Ann Arbor, Michigan. 403 p. Proebstel, D.S., R.P. Evans, D.K. Shiozaw a, and R.N. Williams. 1993. Preservation of nonfrozen tissue samples from a salmonine fish Brachymystax lenok (Pallas) for DNA analysis. Journal of Ichthyology 9: 9-17 Putzer, H. 1984. The geological evalua tion of the Amazon basin and its mineral resources. Pp. 15-46. In : H. Sioli, ed. The Amazon. Dr. W. Junk Publishers, Dordrecht, The Netherlands. 763 p Räsänen, M.E., J.S. Salo, and R.J. Kalliola 1987. Fluvial perturbance in the western Amazon Basin: Regulation by long-term sub-Andean Tectonics. Science 238(4832): 1398-1401. Räsänen, M.E., A.M. Linna, J.C.R. Santos, and F.R. Negri. 1995. Late Miocene tidal deposits in the Amazonian fore land basin. Science 269(5222):3 86-390. Redford, K.H., and J.G. Robinson. 1987. The ga me of choice: Patterns of Indian and colonist hunting in the neotropics. American Anthropologist 89: 650-667. Rivero Blanco, C. 1974. Hábitos reproductivos de la baba en los llanos de Venezuela. Natura 52: 24-29. Robinson, J.G., and E.L. Bennett, eds. 2000. Hunting for sustainability in tropical forests. Columbia Univ. Press, N.Y. 582 p. Robinson, J.G., and K.H. Redford. 1991. The us e and conservation of wildlife. Pp. 3-5. In : J.G. Robinson, and K.H. Redford, eds. Neotropical wildlife use and conservation. Univ. Chi cago Press, Chicago. 520 p. Ronquist, F., and J.P. Huelsenbeck. 2003. MRBAYES 3: Bayesian phylogenetic inference under mixed models . Bioinformatics 19: 1572-1574. Roosevelt, T. 1925. Through the Brazilian wilderness. Charles Scribner's Sons, New York. 401 p. Rosen, S., and H. Skaletsky. 2000. Primer 3 on the WWW for general users and for biologist programmers. Pp. 256-386. In : S. Krawetz, and S. Misener, eds. Bioinformatics: Methods and protocols: Methods in molecular biology. Humana Press, Totowa, New Jersey. Ross, C.A., ed. 1989. Crocodiles and alliga tors. Facts on File, New York, N.Y. 240 p.

PAGE 240

220 Ross, C.A., and W.E. Magnusson. 1989. Living crocodilians. Pp. 58-73 In : C.A. Ross, ed. Crocodiles and alligators. Fact s on File, New York, N.Y. 240 p. Ross, J.P. 1992. Medicinal use of crocodili ans, threat or opportunity? Crocodile Specialist Group Newletter 11(4): 2-3. Ross, J.P., ed. 1998. Crocodiles. Status survey and action plan (2nd ed.) IUCN/SSC Crocodile Specialist Group. IU CN, Gland, Switzerland. 96 p. Ross, J.P., and R. Godshalk, eds. 2003. International workshop for management and trade of Caiman yacare . October 3-5, 2002. Gainesville, Florida. IUCN/SSC Crocodile Specialist Group Pub. 82 p. Rozas, J., J.C. Sánches-DelBarrio, X. Me sseguer, and R. Rozas. 2003. DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496-2497. Ruiz, E. 1988. Ecología del lagarto ( Caiman crocodilus yacare ) en la Estación Biológica Beni (Bolivia). Informe prelim inar. Unpublished manuscript. Estación Biológica del Beni, La Paz, Bolivia. 52 p. Rychlik, W., and P. Rychlik. 2001. Oligo – Primer analysis software, v. 6.76. Molecular Biology Insights, Inc. Cascade, CO. Sambrook, J., E. Fritsch, and T. Maniat is. 1989. Molecular cloning – A laboratory manual. Cold Spring Harbor Laboratory Press, Plainview, N.Y. Sanabria Fernández, H. 1990. Geografía de Santa Cruz. Libreria Editorial Juventud, La Paz, Bolivia. 202 p. Sánchez, P.A. 1976. Properties and management of soils in the tr opics. John Wiley and Sons, New York, N.Y. 618 p. Santos, S.A., M.J.S. Nogueira, M.S. Pi nheiro, Z. Campos, W. Magnusson, and G. Mourão. 1996. Diets of Caiman crocodilus yacare from different habitats in the Brazilian Pantanal. Herpetol ogical Journal 6: 111-117. Santos, S.A., M.J.S. Nogueira, M.S. Pi nheiro, G.M. Mourão, and Z. Campos. 1994. Condition factor of Caiman crocodilus yacare in different habitats of Pantanal Mato-Grossense. Pp. 314-318. In : Crocodiles. Proc. 12th Working Meeting Crocodile Specialists Group. Vol. 2. IUCN Pub., Gland, Switzerland. 340 p. Sarmiento, G. 1984. The ecology of Neotr opical savannas. Harvard Univ. Press, Cambridge, Massachussetts. 235 p. Schaller, G.B., and P.G. Crawshaw, Jr. 1982. Fishing behavior of Paraguayan Caiman ( Caiman crocodilus ). Copeia 1982(1): 66-72.

PAGE 241

221 Schlötter, C. 2000. Evolutionary dynam ics of microsatellite DNA. Chromosoma 109:365-371. Schneider, J.G. 1801. Historia amphibiorum natu ralis et literariae. Fasciculis secundis: Continens Crocodilos, Scincos, Chamaesuras, Boas, Pseudoboas, Elapes, Angues, Amphisbaenas et Caecilias. Jena. 374 p. Schneider, S., D. Roessli, and L. Excoffier . 2000. Arlequin ver. 2.000: A software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland. Schuelke, M. 2000. An economic method for th e fluorescent labeling of PCR fragments. Nature Biotechnology 18:233-234. Scott, J.N.J., A.L. Aquino, and L.A. F itzgerald. 1990. Distribution, habitats and conservation of the caimans (Alligatoridae ) of Paraguay. Vida Silvestre Neotropical 2(2): 43-51. Seijas, A.E. 1986. Situación actual de las poblaciones de babas y babillas ( Caiman crocodilus ) en la región norte-costera de Venezuela. Pp. 8-22. In : Crocodiles. Proceedings 7th Working Meeting Croc odile Specialists Group. IUCN Pub., Editorial Texto S.R.L., Caracas, Venezuela. 446 p. Seijas, A.E., and S. Ramos. 1980. Caract erización de la dieta de la baba, Caiman crocodilus, durante la estación seca en las sa banas moduladas del Estado Apure, Venezuela. Acta Biológi ca Venezolana 10(4): 373-389. SEMENA. 1992. Meteorological data (unpublishe d). Servico para Mejoramiento de la Navegación Amazonica, Trinidad, Bolivia. 12 p. Sempere, T., R.F. Butler, D.R. Richards, L.G. Marshall, W. Sharp, and C.C. Swisher III. 1997 Stratigraphy and chronology of Upper Cr etaceous-lower Paleogene strata in Bolivia and northwest Argentina. Bulletin of the Geologic Societ y of America. 109: 709-727. Sereno, P.C. 1999. The evolution of dinosaurs. Science 284: 2137-2147. Shaw, J.H. 1991. The outlook for sustainable ha rvests of wildlife in Latin America. Pp. 24-34. In : J.G. Robinson, and K.H. Redford, eds. Neotropical wildlife use and conservation. Univ. Chicago Press, Chicago. 520 p. Sioli, H. 1968. Hydrochemistry and geology in the Brazilian Amazon region. Amazoniana 1(3): 267-277. Sioli, H. 1984. The Amazon and its main affluents: Hydrograph, morphology of the river courses, and river types. Pp. 127-165. In : H. Sioli, ed. The Amazon. Dr. W. Junk Publishers, Dordrecht, The Netherlands. 763 p.

PAGE 242

222 Slatkin, M. 1995. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139: 457-462. Solomon, J.G. 1989. Bolivia. Pp. 455-463. In : D.G. Campbell, and H. D. Hammond, eds. Floristic inventory of trop ical countries: The status of plant systematics, collections, and vegetation, plus recomm endations for the future. New York Botanical Gardens, Bronx, New York. 545 p. Spix, J.W. von. 1825. Animalia nova; sive, Spec ies novae Lacertarum, quas in itinere per Brasiliam, annis MDCCCXVII-MDCCCXX colle git et descripsit. Monachii typis franc. seraph. F.S. Hübschmann, München 4: 1-26. Staton, M.A., and J.R. Dixon. 1975. Studi es on the dry season biology of Caiman crocodilus crocodilus , from the Venenzuelan Llanos. Memorias Sociedad Ciencias Naturales La Salle 35: 237-265. Staton, M.A., and J.R. Dixon. 1977. Breedi ng biology of the spectacled caiman, Caiman crocodilus crocodilus in the Venezuelan Llanos. U.S. Fish and Wildlife Rpt. 5: 121. Stehli, F.G., and S.D. Webb (eds.). 1985. The Great American Biotic Interchange. Plenum Press, New York, N.Y. 532 p. Swofford, D.L. 1998. PAUP -. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4. Sinauer Associ ates, Sunderland, Massachussetts. Tajima, F. 1989. Statitistical method for testing the neutral mu tation hypothesis by DNA polymorphism. Genetics 105: 437-460. Taplin, L.E. 1988. Osmoregulation in crocodi lians. Biological Review of the Cambridge Philosophical Society 63: 333-377. Taplin, L.E., and G.C. Grigg. 1989. Hist orical zoogeography of the eusuchian crocodilians: A physiological perspectiv e. American Zoologist 29: 885-901. Taplin, L.E., G.C.Grigg, and L. Beard. 1985. Salt gland function in fresh water crocodiles: Evidence for a marine phase in eusuchian evolution? Pp: 403-410. In : G.C. Grigg, R. Shine, and H. Ehmann, ed s. Biology of australasian frogs and reptiles. Surrey Beatty and Sons, Sydney, Australia. Taplin, L.E., G.C. Grigg, P. Harlow, T.M. Ellis, and W.A. Du nson. 1982. Lingual salt glands in Crocodylus acutus and C. johnstoni (sic), and their absence from Alligator mississippiensis and Caiman crocodilus . Journal of Comparative Physiology 149: 43-47. Tautz D., M. Trick and G.A. Dover. 1986. Cryptic simplicity in DNA is a major source of genetic variation. Nature 322: 652-656.

PAGE 243

223 Terrazas, W. 1970. Lista de peces bolivianos. Academia Nacional de Ciencias de Bolivia, La Paz, Bolivia. 24: 1-65. Templeton, A.R., K.A. Crandall, and C.F. Sing. 1992. A cladistic analysis of phenotypic associations with haplotypes inferred from restriction endonuclease mapping and DNA sequence data. III. Cladogr am estimation. Genetics 132: 619-633. Thorbjarnarson, J.B. 1991a. Ecology and be havior of the spectacled caiman ( Caiman crocodilus ) in the central Venezuelan Llanos. Doctoral dissertation. Univ. Florida, Gainesville, Florida. 390 p. Thorbjarnarson, J.B. 1991b. An analysis of the spectacled caiman ( Caiman crocodilus ) harvest program in Venezuela. Pp. 217-235. In : J.G. Robinson, and K.H. Redford, eds. Neotropical wildlife use and cons ervation. Univ. Chicago Press, Chicago. 520 p. Thorbjarnarson, J.B. 1993. Diet of the spectacled caiman ( Caiman crocodilus ) in the central Venezuelan Llanos . Herpetologica 49(1): 108-117. Thorbjarnarson, J.B. 1998. Melanosuchus niger . Pp 23-28. In : J.P. Ross, ed. Crocodiles. Status survey and action plan. 2nd Edition. IUCN/SSC Crocodile Specialist Group. IUCN, Gland, Switzerland. 96 p. Thorbjarnarson, J.B., and G. Hernandez. 1990. Recent investigations into the status of Orinoco crocodiles in Venezuela. Pp. 308-329. In : Crocodiles. Proc. 9th Working Meeting Crocodile Specialist Group. Vo l. 2. IUCN, Gland, Switzerland. 379 p. Thorbjarnarson, J.B., and A. Velasco. 1998. Venezuela’s caiman harvest program – An historical perspective and an alysis of its conservation benefits. Working Paper No. 11. Wildlife Conservation Soci ety. New York, N.Y. 66 p. Townsend, W.R. 1995. Living on the edge: Si rionó hunting and fishing in lowland Bolivia. Doctoral Disserta tion. Univ. of Florida. Ga inesville, Florida. 169 p. Townsend, W.R. 1996. Nyao itõ: Caza y pesca de los Sirionó. Instituto de Ecología, Universidad Mayor de San Andrés, La Paz, Bolivia. 130 p. Trompf, G.W. 1989. Mythology, religi on, art and literature. Pp. 156-171. In : C.A. Ross, ed. Crocodiles and alligators. Facts on File, New York. 240 p. Tuross, N. 1994. The biochemistry of ancien t DNA in bone. Cellular and Molecular Life Sciences 50(6): 530-535. Vanzolini, P.E., and N. Gomes. 1979. Notes on the ecology and growth of Amazonian caimans (Crocodylia, Alligatoridae). Papé is Avulsos Zool. São Paolo 32(17): 205216.

PAGE 244

224 Vargas, A.O., and J.F Fallon. 2005. Birds have dinosaur wings: The molecular evidence. Journal of Experimental Zoology 304B: 86-90. Vasconcelos, W.R., T. Hrbek, R. da Silveira, B. de Thoisy, B. Marioni , and I.P. Farias. 2006. Population genetic analysis of Caiman crocodilus (Linnaeus, 1758) from South America. Genetics and Mo lecular Biology 29:(in press). Vasquez, P.G. 1981. Bases bioecológicas para el manejo de los Alligatoridae en Jenaro Herrea (Requna-Peru). Univ. Nacional Ag raria, La Molina, Lima, Peru. 205 p. Velasco, A., and R. De Sola. 2005. Exportacione s de pieles de coc odrilos desde Latino América y el Caribe. Proceedings de la Reunión Régional de América Latina y el Caríbe del Grupo de Especi alistas en Cocodrilos. Sa nta Fe, Argentina. 352 p. Verdi, L., L. Moya, and R. Pezo. 1980. Observ aciones preliminares sobre la bioecología del lagarto Caiman crocodilus (Linnaeus, 1758) (Alligatorid ae) en la cuenca del río Samiria, Loreto, Perú. Seminario sobre los proyectos de inve stigación ecológica para el manejo de los recursos natura les renovables del bos que tropical húmedo. Octubre 1980. COTESU ORDELORETO, Di rección Regional de Agricultura y Alimentación Dirección Forestal Fauna, Iquitos, Perú. 37 p. Vonhof, H.B., F.P. Wesselingh, R.J.G. Kaa ndorp, G.R. Davies, J.E. van Hinte, J. Guerrero, M. Räsänen, L. Romero-Pittman and A. Ranzi. 2003. Paleogeography of Miocene western Amazonia: Isotopic com position of molluscan shells constrains the influence of marine incursions. Geol ogic Society of America Bulletin 115: 983993. Waller, T. 1987. Registro de las localidades de distribución de las especies del genero Caiman (Crocodylia, Alligatoridae) en Argentina. Amphibia & Reptilia (Conservación) 1(3): 68-75. Waller, T. 2003. Current situation and perspe ctive on the use and th e conservation of the yacare, Caiman yacare , in the Argentine Republic. Pp. 7-16. In : J.P. Ross, and R. Godshalk, eds. International work shop for management and trade of Caiman yacare . Crocodile Specialist Group Pub., Gainesville, Florida. 82 p. Waller, T., and P.A. Micucci. 1993. Releva miento de la distribución, habitat y abundancia de los crocodilios de la Provi ncia de Corrientes, Argentina. Pp. 341385. In : Zoocria de los Crocodylia. Memorias de la 1a Reunion Regional del CSG. Santa Marta, Colombia. 399 p. Webb, G.J.W., S.C. Manolis, and P.J. Whitehead. 1987. Wildlife management: Crocodiles and alligators. Su rrey Beatty & Sons, Chipping Norton, Australia. 552 p. Webb, S.D. 1978. A history of savanna verteb rates in the New World. Part II: South America and the Great Interchange. Annual Review of Ecology and Systematics 9: 393-426.

PAGE 245

225 Webb, S.D. 1985. Late Cenozoic mammal disp ersals between the Americas. Pp. 357386. In : F.G. Stehli, and S.D. Webb, eds. Th e Great American Biotic Interchange. Plenum Press, New York, N.Y. 532 p. Webb, S.D. 1991. Ecogeography and the Grea t American Interchange. Paleobiology 17(3): 266-280. Webb, S.D. 1995. Implications of the middle Miocene Amazon seaway. Science 269(5222): 362-362. Weir, B.S., and C.C. Cockerham. 1984. Estim ating F-statistics for the analysis of population structure. Evolution 38: 1358-1370. Wermuth, H., and K. Fuchs. 1978. Bestim men von Krokodilen und Ihrer Haute. Gustav Fischer Verlag, Stuttgart. 100 p. Wermuth, H., and R. Mertens. 1977. Li ste der rezenten Amphibien und Reptilien Testudines, Crocodylia, Rhyncocepha lia. Das Tierreich. Berlin. 174 p. Werner, F. 1933. Loricata (Reptilia) . Das Tierreich, Leipzig. 62: 1-40. White, H.J., R.A. Skopec, F.A. Ramirez, J.A. Rodas, and G. Bonilla. 1995 Reservoir characterization of the Holin and Napa Formations, western Oriente Basin, Ecuador. Pp. 573-596. In : A.J. Tankard, R. Suarez, and H.J. Welsink, eds. Petroleum basins of South America. American Association of Petroleum Geologists Memoir 62. White, P.S., and L.D. Densmore. 1992. Mitochondrial DNA isolation. Pp. 29-58 In : A. R. Hoezel, ed. Molecular Genetic Analysis of Populations: a practical approach. Oxford University Press, New York. White, P.S., and L.D. Densmore. 2001. DNA sequence alignments and data analysis methods: Their effect on the recovery of crocodylian relationships. Pp. 29-37. In : G.C. Grigg, F. Seebacher, and C.E. Fr anklin, eds. Crocodilian biology and evolution. Surey Beatty & Sons, Chipping Norton. 446 p. Woodward, A., and D. David. 1985. A study of the baba ( Caiman crocodilus ) population on the Hato Piñero in Ve nezuela. Unpublished manuiscript. Wu, X., Y. Wang, K. Zhou, W. Zhu, J. Nie, C. Wang, and W. Xie. 2003. Complete mitochondrial genome sequence of Chinese al ligator, and phylogeny of crocodiles. Chinese Science Bulletin 48(19): 2050-2054. Ye, J., J. Anquan, J.P. Esteban, Z. Xiufen, J. Chengtao, Z. Xingchun, H. Lan, and T. Zheng. 2004. A simple and efficient me thod for extracting DNA from old and burned bone. Journal of Fore nsic Science 49(4): 1-6.

PAGE 246

226 Yeh, F.C., R.C. Yang, T.B. Boyle, Z.H. Ye, and J.X. Mao. 1999. POPGENE v1.32, the user-friendly software for population ge netic analysis. Mole cular Biology and Biotechnology Center, University of Alberta, Edmonton, Alberta, Canada. Zardoya, R., and A. Meyer. 2001. The evol utionary position of turtles revised. Naturwissenschaften 88: 193-200. Zeh, J.A., D.W. Zeh, and M.M. Bonilla. 2003. Phylogeography of the harlequin beetleriding pseudoscorpion and the rise of th e Isthmus of Panamá. Molecular Ecology 12: 2759-2769. Zeitum Lopez, S. 1991. Amazonía boliviana . Primera parte. Producciones Gráficas Visión, La Paz, Bolivia. 295 p. Zhang, D.-X., and G.M. Hewitt. 1996. Nuclear integrations: challenges for mitochondrial DNA markers. Trends in Ecology and Evolution 11:247-251. Zucoloto, R.B., L.M. Verdade, and L.H. Coutinho. 2002. Microsat ellite DNA library for Caiman latirostris , Journal of Experimental Z oology (Mol. Dev. Evol.) 294: 346351.

PAGE 247

227 BIOGRAPHICAL SKETCH Robert E. Godshalk was born on August 10, 1949, in Newark, New Jersey. The young herpetologist brought home a steady collect ion of local frogs, lizards and snakes while growing up in the Midwest. He devel oped an interest in Latin America during undergraduate studies at the University of Colorado. In 1974, Robert began a 3 yr. herpetologist position in a Venezuelan na tional park within a Smithsonian/Peace Corps program. His research included surveys of cloud forest herptiles and studies of caiman in the lowland savannas. The following year, he traveled throughout Co lombia leading a research field team for a malaria in wildlife project by Tulane Un iversity. Robert returned to Venezuela with World Wildlife Fund support for a survey of the endangered Orinoco crocodile. This led to establishment of a national park and form ation of a species recovery program that continues today. His experience was further enriched by a 14 month trip through wild lands and national parks of seven Andean nations. Since returning to the US in 1980, he ha s broadened Latin American experience with frequent trips. He also obtained his Na turalist Guide certification for the Galapagos National Park, worked as a whitewater rive r guide in Wyoming and Colorado, as an explosives expert, 12 years as an Emerge ncy Medical Technician (EMT) and as a professional ski patrolman, among other employment. Robert entered the Graduate School at the University of Florida and spent time in Bolivia collecting data for his Master of Science thesis and planning a caiman

PAGE 248

228 conservation program with Bolivian author ities. He later conducted CITES funded caiman population surveys in preparation for a sustainable use program for caiman in Bolivia. During that time he collected blood samp les as the basis for this dissertation. He has served as an IUCN Crocodile Specialis t Group consultant for the governments of Bolivia and Paraguay in planning caiman cons ervation programs. In the future, he hopes to continue to be involved with projects dealing with community-based sustainable wildlife conservation. He will also continue in the exciting field of molecular research with emphasis on conservation appli cations and wildlife management.