INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films the text direct^ from the orignal or copy submitted. Thus, some thesis and dissertation copies are in typewriter free, vdiile others may be fi"om any type of computer printer.
The quality of this reproduction is dependent upon the qualityr of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps. Each original is also photographed in one exposure and is included in reduced form at the back of the book.
Photographs included in the original manuscript have been reproduced xerograpliically in this copy. Higher quality 6” x 9” black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. UMI A Bell & Howell Infomiaticn Conqiany 300 North Zed) Road, Ann Arbor MI 48106-1346 USA 313/761-4700 800/521-0600
THE CONSERVATION GENETICS OF GREEN TURTLES {ÇHELONIA MYDASy. CONSEQUENCES OF PHILOPATRT AND MATING BEHAVIOR
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy In the Graduate School of The Ohio State University
By
Tlgeiin Peare
$ * $ $ $
The Ohio State University 1996
Dissertation Committee: Approved by Professor Patricia Parker, Adviser Professor Thomas Grubb Professor Thomas Hetherington Adviser Professor Allison Snow Department of Zoology m n Nimber: 97 1 0 6 4 3
UMI Microform 9710643 Copyright 1997, by UMI Company. All rights reserved.
This microform edition is protected against unauthorized copying under Title 17, United States Code.
UMI 300 North Zed> Road Ann Arbor, MI 48103 ABSTRACT
Genetic tools can provide information about behavior and demographics of endangered species with intractable life histories like the green turtle {Chelonia m ydas ). This project exam ined the degree of precision in natal philopatry and the mating system of turtles from two nesting populations (Tortuguero, Costa Rica and Melbourne, Florida. U.S.A) by applying high-resolution molecular techniques. In Tortuguero. multilocus minisatellite DNA fingerprinting revealed a significant negative correlation between genetic similarity of pairs of nesting females and the distance between their nest sites both within years (r^ = 0.273: P < 0.001) and between years (r^ = 0.578; P < 0.001). In addition. 12.3% of 122 female pairs scored for Tortuguero had genetic similarity values resembling those of mother-offspring pairs. In the Melbourne population, however, no relationship between genetic similarity and distance was found (r^ = 0.017; P = 0.075) and none of the 66 female pairs scored for this population had genetic similarity values that resembled those of first-order relatives.
11 These results indicate that females from Tortuguero exhibit low levels of dispersal from natal sites, and that nestmates return independently to nest near their natal sites. The lack of genetic structure in the Melbourne population suggests that females from this population may not return to natal sites with comparable precision. Higher levels of disturbance or mortality, lower quality of homing cues, or younger rookery age may also be responsible for the absence of distance-related local genetic structure within Melbourne. To characterize the mating system of green turtles, single-locus microsatellite technology was used to analyze paternity in 18 clutches from the two rookeries. Five of eight Tortuguero families and five of 10 Melbourne families contained more alleles among hatchlings than would be expected if each nest had a single father. A statistical approach (developed by Dr. Mark Irwin, Department of Statistics, The Ohio State University) estimated that the number of fathers represented in each Tortuguero nest ranged from one to three while the number for each Melbourne clutch was either one or two. These results suggest that green turtles from these breeding aggregations exhibit a promiscuous mating system. Hatching success data revealed that multiply-fertilized clutches produce a higher proportion of offspring than singly-fertilized clutches (Mann-Whitney U = 3, p = 0.028). This su^ests that females may
iii benefit firom mating with several partners through fertilization insurance, or the enhancement of sperm competition among ejaculates. The potential detrimental effects of genetic isolation due to strong female philopatry may be ameliorated by a mating system that features multiple paternity.
iv To my parents and my good teachers ACKNOWLEDGMENTS
This work would have been impossible without the help of many people and agencies. 1 thank my academic advisor, Patricia Parker, for support, ideas and constructive criticism. My other committee members, Tom Grubb, Tom Hetherington and Allison Snow provided additional encouragement emd insight. My co-workers in the laboratory, Nidia Arguedas, T. J. Jones, and Kim Lundy were extremely cooperative and supportive. Thcinks to José Diaz for always saying “No problem” no matter what the request. 1 am especially grateful to Julie Rieder for her friendship, cooperation, hours of discussion and for plodding through months of microsatellite woes with me. For suggestions and other assistance with lab work, I thank Hans EUegren and Colin Hughes. 1 also thank Lynn Kramer and Doug Warmolts at the Columbus Zoo for their assistance. 1 am indebted to Charlie Luthln who gave me my first break. 1 appreciate aU the help I received in the field from Mary Bell, Cathi Campbell, Ray Carthy, Llew Ehrhart, Elizabeth Giuliano, Laine Gonzales, Nancy Israel, Steve Johnson, Tracy Lenihan, Allison Leslie, Rebecca Morse, Ann Peterka, Jessy Rabenold, Andy Rabenold and Janet Winbome. Special thanks to Kennedy Champos cind Lloid Taylor for being bodyguards and “big brothers” during the 1993 season.
Vi Monica Guevara was exceptionally helpful during my 1991 and 1992 field seasons, her intelligence, humanity and flair for singing show tunes provided comic relief and gave me a better understanding of the conflict between human requirements and ecosystem limits. Mark Irwin invested a tremendous amount of time creating a computer program to analyze the microsatellite data, and was extremely patient and helpful in explaining the likelihood analyses. I also thank Brian Bowen for his ideas, suggestions, enthusiasm, and humor, and for standing as an exeunple of what one can accomplish with modesty, cooperation and integrity. I am especially grateful to my firiends who made graduate school a lot easier. Gentry Holloway. Jen Knight and Tania Oberyszyn collectively provided me with unconditional support, boundless energy and enthusiasm, recreation, amusement and excellent advise. Finally, I would like to thank my parents for their support and encouragement and Zydeco for teaching me that sleeping and playing in the park are sometimes more important than going to the lab. This study was supported by grants firom Sigma Xi. the American Museum of Natural History, Wildlife Conservation International, and the Caribbean Conservation Corporation awarded to me and funds firom the National Science Foundation #DEB-9322544. awarded to my adviser. Patricia Parker. My parents also provided me with a loan for the 1993 field season.
vu VITA
July 30. 1967 ...... Bom - Pointe-a-Plerre, Trinidad, West Indies
1989 ...... B. S. Natural Resources, Cornell University, Ithaca, New York
1990-199 1 ...... Teaching Associate, Departm ent of Biology Purdue University West Lafayette, Indiana
1991-1992 and 1995-1996 ...... Research Associate, Department of Zoology The Ohio State University
1993-1994 and 1996 ...... Teaching Associate, Department of Zoology The Ohio State University
viii PUBLICATIONS
Peare, T. and P. G. Parker. 1996. Local genetic structure within two rookeries of green turtles {Chelonia mydas ). Heredity (in press)
Peare. T. and P.O. Parker. 1996. The use of multilocus minisateUite DNA fingerprinting to assess local genetic structure in green turtle rookeries. In: Proceedings of the International Symposium on Sea Turtle Conservation Genetics. B. W. Bowen and W. N. Witzel, eds. NCAA Technical Memorandum. NMFS-SEFSC (in press).
Parker. P. G.. T. A. Waite and T. Peare. 1995. Paternity studies in animal populations. In: Molecular Genetic Approaches to Conservation. T. B. Smith and R. K. Wayne, eds. Oxford University Press.
FIELD OF STUDY
Major Field: Zoology
Focus: Conservation Genetics
IX t a b l e o f c o n t e n t s
ABSTRACT...... ii
DEDICATION...... v
ACKNOWLEDGMENTS...... vi
VITA...... vlU
LIST OF TABLES...... xiv
LIST OF FIGURES...... xv
CHAPTERS:
1. Green turtle natural history, causes of decline and project summary ...... I
Introduction to green turtle life history ...... 1 Description and distribution ...... 1 Feeding ecology and growth ...... 2
Migration ...... 3 Breeding behavior ...... 4 Incubation and hatching ...... 5 Decline and extinction of populations ...... 8 Description of project ...... 9
2. Local genetic structure within green turtle rookeries ...... 12
X Introduction ...... 12 Withln-population genetic structure ...... 12 Philopatry in green turtles ...... 14 Testing the precision of natal homing: A molecular approach ...... 15 Materials and Methods ...... 19 Field methods ...... 19 Laboratory methods ...... 20 Data analysis ...... 22 Relationship between intemest distance and genetic similarity ...... 22 Calibrating relatedness among femedes ...... 23 R esults ...... 24 Relationship between intemest distance and genetic similarity ...... 24 Calibrating relatedness among females ...... 25 D iscussion ...... 34 Precise natal philopatry ...... 34 Comparison of rookery substructure ...... 37 Human and natural disturbance ...... 37 Quality of homing cues ...... 39 Ability to detect relatives ...... 40 Relative age of nesting beaches ...... 43 Conservation implications ...... 44
3. The mating system of green turtles ...... 49
Introduction ...... 49 Genetic variation and persistance of populations ...... 49 Mating systems and genetic diversity ...... 51 Background on the green turtle ...... 53
XI Application of microsatellite technology to paternity analysis ...... 54 Materials and Methods ...... 55 Field methods ...... 55 Laboratory methods ...... 56 Characterization of alleles ...... 58 Analysis of paternity ...... 58 Statistical analysis ...... 59 Ability to detect multiple paternity ...... 61 Results...... 62 Population allele frequencies ...... 62 Paternity analysis ...... 67 Ability to detect multiple paternity ...... 73 Mating system and hatching success ...... 73 Discussion ...... 76 Interpretation of the mating system ...... 76 Advantages to multiple insemination ...... 78 Increase genetic diversity of clutches...... 79 Ensure fertilization of eggs ...... 79 Enhance sperm competition ...... 80 The relationship between hatchling success and mating strategy ...... 81 Comparisons with other marine turtles...... 82 Explanation for population differences in breeding behavior ...... 83 Consequences of multiple paternity ...... 85
Notes on Appendices ...... 87
APPENDIX A DNA Extraction ...... 88
APPENDIX B Dialysis of extracted samples ...... 90
APPENDIX C Reading samples on the spectrophotometer ...... 92 xii APPENDIX D Running a minigel ...... 94
APPENDIX E Running a digestion gel ...... 97
APPENDIX F Southern transfer ...... 100
APPENDIX G Hybridization for mlnlsatelUtes using Jeffreys’ probes 33.6 or 33.15 ...... 103
APPENDIX H Washing filters and producing autoradiograms ...... I l l
APPENDIX 1 PCR Reactions...... 113
APPENDIX J Solutions for polyacrylamide gels ...... 117
APPENDIX K Gel preparation. running and visualization ...... 119
APPENDIX L Data relative to Chapter 2: Matrices of genetic slmllarily values and intemest distances for nesting females ...... 125
APPENDIX M Data relative to Chapter 2: Genetic similarity scores for first-order relatives ...... 128
APPENDIX N Data relative to Chapter 3: Genotypes of nesting females ...... 131
APPENDIX O Data relative to Chapter 3: Genotypes of families ...... 134
LIST OF REFERENCES...... 144
X lll LIST OF TABLES
Table Page la. Results from patemlly analysis of Tortuguero families ...... 68 lb. Results from paternity anedysis of Melbourne families ...... 69
2a. Results from computer analysis of Tortuguero families ...... 70
2b Results from computer analysis of Melbourne families ...... 71 3 Hatching success information for singly- and multiply-sired clutches ...... 75
xiv LIST OF FIGURES
Figure Page
1. Map of field sites ...... 18
2. Relationship between intemest distance and genetic similarity values for pairs of females nesting within one season in Tortuguero ...... 26
3. Relationship between intemest distance and genetic similarity values for pairs of females nesting between seasons in Tortuguero ...... 27 4. Relationship between intemest distance and genetic similarity values for pairs of females nesting in Melboume ...... 28 5. Distributions of genetic similarity values for all females and for first-orderrelatlves in Tortuguero ...... 30
6. Distributions of genetic similarity values for all females and for first-order relatives in Melboume ...... 31
7. Distributions of genetic similarity values for “unrelated” females and for first-order relatives in Tortuguero ...... 33 8. Allele firequencies for Tortuguero females at the Cc 117 locus...... 63
XV 9. Allele frequencies for Tortuguero females at the Cm 3 locus...... 64 10. Allele frequencies for Melboume females at the Cc 117 locus...... 65
11. Allele frequencies for Melboume females at the Cm 3 locus...... 66
12. Probability of finding a second father ...... 74
xvi CHAPTER I
GREEN TURTLE NATURAL HISTORY. CAUSES OF DECLINE AND PROJECT SUMMARY
Introduction to Green Turtle Life History
Description and Distribution
Green turtles [Chelonia mydas) are the largest of the hard-sheUed marine turtles, weighing 136-158 kg. with an average carapace length of 100 cm or more (Pritchard. 1979). They are circumglobal in their distribution, but are Umited to tropical and subtropical regions where their feeding grounds and nesting beaches are located (Carr. 1952). Some individuals, have, however been observed wandering in both northern (English Channel) and southern (Polla Island. Chile) extremes (Hirth. 1980). The period following hatching is poorly understood because observations of first-year turtles in the sea are rare. As a result, this period has been called the "lost year". Some studies have, however, confirmed that individuals in their first few years are highly pelagic, often found at convergent zones or associated with currents carrying Sargassum mats (Carr, 1986; Carr and Meylan, 1980). Older juveniles euid adult green turtles tend to inhabit shallow feeding grounds and coral reefs (Carr and Ogren, 1960).
Feeding Ekzology and Growth
Green turtles are the only herbivorous marine turtle (Hirth, 1971; Hendrickson, 1980) with the seagrass Thalassis testudinum comprising 80% to 90% of the diet of adults and sub-adults (Mortimer, 1981; 1982). At seagrass beds in the Bahamas, individual feeding trurtles have been observed exhibiting fidelity to particular areas (Bjomdal, 1980). This behavior keeps patches cropped short and results in the continuous production of young shoots that contain six to 11% more protein than older blades (Bjomdal, 1980). At the feeding grounds, daily consumption of seagrass represents onfy 0.24 to 0.33% of the body weight, and the digestibility of protein firom seagrass is only 50% (Bjomdal, 1980). The low efficiency of nutrient acquisition in green turtles results in slow rates of growth (Bjomdal, 1982). Relative to other marine turtles, green turtles take a significantly longer time to reach sexual maturity (Bjomdal, 1982). Mean growth rates of immature individuals vary for different populations firom an estimated 0.04 cm per month in Bermuda (Bumett-Herkes, 1974) to 0.44 cm per month in Hawaii (Balazs, 1982). The estimated age of first-time breeders is 27 to 33 years in the Atlantic (Frazer and Ladner, 1986), 30 or more years in Australia (Llmpus and Walter, 1980) and from 9 to 48 years in Hawaii (Balazs, 1982).
Migration
Long-distance migrations are characteristic of green turtles (Carr et al. 1978) which may travel as far as 2,250 km (in the case of the Ascension Island turtles) to reach nesting beaches (Carr, 1975). Herbivoiy h£is been linked to the migratory behavior of this species. Feeding grounds (seagrass pastures) are generally located in shallow, coastal areas, and, during the non-breeding season, the distribution of juvenile and adult green turtles matches the distribution of these areas (Hirth, 1971). However, because good seagrass pastures are typically not cissociated with suitable nesting beaches, turtles must travel to other regions for breeding (Mortimer, 1982). Migrating turtles appear to foUow relatively direct routes and achieve estimated speeds of 22.5 km per day (Hirth and Carr, 1970) to 66 km per day (Schultz, 1975). Information from tag returns has revealed that female green turtles tend to return with fidelity to areas of the beach that they used in previous nesting events, both within a season and between seasons (Carr and Carr, 1972; Carr and Hirth, 1962; Mortimer and Portier, 1989; Johnson, 1994). These observations led researchers to suggest that females may be homing to their natal beaches rather than selecting the nearest, suitable rookery (Carr, 1967). Recently, this hypothesis has been confirmed through the comparison of maternally-transmitted mitochondrial DNA from different nesting populations (Meylan et al., 1990; Bowen et al., 1992; Allard et al., 1994). The significant levels of divergence among rookeries indicate that very little exchange of genetic material is mediated through female movements.
Breeding Behavior
Mating is believed to take place predominantly in the water surrounding the nesting beach (Mortimer and Carr, 1987; Llmpus et al., 1992); however, mating individuals have been observed during migration (Meylan, pers. comm.). Observations of breeding behavior in captive populations have revealed that females appear to have a two to four day "heat" period during which they are sexually receptive to the advances of males (Wood and Wood, 1980). Between three and five weeks after mating, females haul up onto the beach at night and dig a large body pit (Wood and Wood, 1980; Urlich and Parkes, 1978). They then use their rear flippers to dig a vase-like chamber, about 70 to 80 cm deep, within the body pit and deposit a large clutch of eggs into the nest. The average size of clutches varies from 160 at the Yemen rookery (Hirth 1971) to 89 in Aldabra (Frazier, 1971). Following egg deposition, females fill the egg chamber, cover the body pit and return to the ocean. Through the course of the nesting season, a female may return to the beach to lay one to seven additional clutches (Carr et al. 1978) at intervals of 12 to 15 days (MoU, 1979). In captivity, females do not appear to mate between nesting events, suggesting that sperm storage is possible within a season. Urlich and Parkes (1978) reported that one multiply mated female in a captive population laid her eighth clutch 110 days after her last mating event. The hatching success of this clutch was 85%, indicating that her mating activity at the beginning of the season was sufficient to provide sperm for the fertilization of successive clutches (Urlich and Parkes, 1978). At the end of the breeding season, females migrate back to feeding grounds and do not return to the nesting beach for another two to three years (Hendrickson 1958; Carr and Ogren, 1960).
Incubation and Hatching
During incubation, two of the most important factors affecting development and hatching success are temperature and gas exchange. Marine turtles have temperature-dependent sex determination with warmer temperatures producing females and cooler temperatures producing males (Mrosovsky and Yntema, 1980; Morreal et al., 1982; Spotila et al. 1987). This mode of sex determination makes the continuous production of equal sex ratios unlikely. Temperature fluctuations between seasons or between months within a season can result in hatchling groups with skewed sex ratios. Presumably, the conditions on successful nesting beaches ultimately produce an overall adult population sex ratio that permits breeding activity, despite fluctuations in temperatures. Temperature also plays a role In nest success and the duration of incubation. Temperatures th at are too high (e.g., 38°C) or too low (e.g., 20°C) result in the death of clutches (Bustard and Greenham (1968). The failure of nests at low temperatures is likely to be a factor limiting the distribution of green turtle nesting beaches to warmer regions. Within these areas, suitable beaches would be those that do not experience extremefy high temperatures. The iypical range in incubation temperatures tends to be between about 26°C and 32°C. Within this range, the duration of incubation is negatively correlated with temperature. At the Heron Island rookery. Incubation time ranges from 80 days at 27°C to 47 days at 32°C (Bustard and Greenham, 1968). During incubation, respiratory gases are exchanged between the egg mass and the surrounding beach. Gas exchange requirements increase as the metabolic activily of the eggs increases; however, the movement of air through the beach to the nest is limited, (Ackerman, 1980). If availability of respiratory gases is insufficient, growth of embryos is slowed or the entire clutch may die (Ackerman, 1980). Incubation conditions, such as moderately low temperatures and limited gas exchange, may not directly result in the death of a clutch. However, they tend to increase incubation time which places nests at a higher risk of mortality by increasing exposure to predation and other dangers. Female nest site selection, therefore, can have a significant effect on the hatching success of a clutch. Observations through glass-sided nests have provided information about hatching and emergence in green turtles (Carr and Hirth, 1962). The first individuals to hatch remain inactive until more of their nestmates have emerged firom their eggs. Movement originating firom the bottom of the pile stimulates activity in the top individuals which scrape sand firom the roof of the nest cavity. The sand works its way through the mass of moving hatchlings and is trampled into the floor of the nest cavity by the bottom turtles. Over the course of two to four days, the whole cavity rises through the sand and hatchlings emerge almost simultaneously at night (Carr and Hirth, 1962; Hendrickson, 1958). After emergence firom the nest, hatchlings orient in the direction of the brightest light source (the sea) and move rapidly to the surf (Mrosovsky and Kingsmill, 1985). Hatchlings may also find the sea by moving away firom elevated horizon silhouettes (Salmon et al, 1992). Once in the water, hatchlings orient towards oncoming waves in the near-shore environment (Salmon and Wÿneken, 1994), and then appear to be guided by geomagnetic cues as they travel farther firom the beach (Lohman and Lohman, 1994). The magnetic sense acquired by hatchlings may be the same system used to guide turtles in other life stages. Decline and Ehctinction of Green Turtle Populations
There are approximately 150 separate green turtle rookeries throughout the world (Sternberg, 1981). Since only a few of these remain that support significant numbers of nesting females, they have achieved endangered or threatened status throughout their range. The primary cause of the decline of green turtle populations has been the systematic over exploitation of e ^ s and adults (King, 1982). The fate of Caribbean green turtles is relatively well documented and provides an example of dramatic over exploitation by the early European explorers and settlers who used green turtles as a source of protein (Carr, 1956), Historically, the largest green turtle rookery was probably in the Cayman Islands. In 1505, on his final voyage to the Americas, Christopher Columbus wrote that these islands emd the surrounding waters were were so dense with turtles, they "look'd like little rocks" (Lewis 1940, Carr, 1956). For the next 150 years, Europeans visited the islands to capture turtles, and then full-scale exploitation began in the middle of the 17th century by the British colony in Jamaica (Lewis, 1940). Long (1774) reports that during the initial stages of this exploitation turtles were still so abundant that sailors lost in poor weather steered their boats based on the location of noises made by m asses of s wim m ing individuals. By 1688, more than 13,000 turtles were being captured annually for the colony in Jam aica (Lewis, 1940). The nesting population in the Cayman Islands plummeted, and despite the enactment of a law in 1711 prohibiting the collection of turtle eggs, this rookery was extinct by 1900 (King, 1982). Turtle fleets had moved to the feeding grounds at the Miskitu Cays in the late 1700's where 2,000 to 3,000 turtles were killed each year (Rebel, 1974 King 1982). Since 1970, this population has declined precipitously due to the slaughter of about 10,000 green turtles each year (King, 1982).
8 Over-exploitation in the Cayman Islands nesting population exemplifies the pattern of decline in other populations worldwide (e.g., de Silva, 1982; Frazier, 1982). In the Caribbean, only two substantial nesting populations still remain. This is a dramatic decline from the 10 large nesting aggregates formerly known to exist (Groombridge, 1982). Today, overharvesting is still a major problem despite laws to protect all green turtle life stages (King, 1982), and incidental catch in fish and shrimp nets results in high rates of mortality (N.R.C. 1990; Henwood and Stuntz, 1987). In addition, beach development has resulted in decreases in the sizes of some nesting populations (IVlortimer, 1982; Worth and Smith, 1976).
Description of Project
It is necessary to consider the population genetics of any species whose numbers are declining because low levels of genetic variation, resulting from phenomena such as genetic drift (Wright 1931), may limit the ability of populations to persist through time (Frankel and Soule 1980; AUendorf 1983). Extinction of a population can result from failure to adapt to environmental fluctuations. Theoretical models have been used to examine conditions under which this failure occurs, and these studies have attributed extinctions to a lack of sufficient variation to replace inferior genotypes with new ones at frequencies that are sufficiently high to respond to changing selective pressures (see Burger and Lynch. 1995; Lynch and Lande, 1993; Pease et al., 1989; Lynch et al.. 1991). In terms of threats to their immediate survival, declining populations may be as vulnerable to detrimental genetic consequences as they are to demographic stochasüciiy (Lande, 1994; Lande, 1995). Since natural selection acts on the variation present within a population, the most effective conservation plans are likely to be those that attempt to preserve a pool of genetic diversity as a buffer against fluctuating selective pressures. However, effective management of genetic resources requires information about the behaviors that affect the maintenance of genetic diversity. Confounding the ability to manage for the recovery of green turtle populations is a lack of knowledge about fundamental aspects of their biology. Their pelagic and migratory habits as well as their slow development to sexual maturity make it difficult to track individuals through time. In addition mating behavior is difficult to observe and male behavior is virtually unknown. In order to reach an understanding of philopatiy and mating behavior in green turtles, I have used hlgh- resolution genetic tools to examine local genetic structure and to analyze paternity of clutches within two nesting populations. Natal homing precision and mating systems are of critical interest because they determ ine the magnitude of gene flow and influence the level of genetic diversity within populations.
10 PLEASE NOTE
Page(s) missing in number only; text follows. Filmed as received.
11
UMI CHAPTER n LOCAL GENETIC STRUCTURE WITHIN GREEN TURTLE ROOKERIES
Introduction Within-Population Genetic Substructure
Genetic differentiation can occur within a continuous population if the effective dispersal rate of individuals is restricted (Wright, 1943a). Signihcant departure from randomly distributed genotypes over a continuous area may occur as a result of spatially distinct selective pressures within an environmentally heterogeneous habitat, or may be due to low rates of dispersal by members of the population occupying the range (Ehrlich and Raven, 1969; Rohlf and Schnell, 1971; Sokal and Wartenberg, 1983). There are several examples of within- population genetic divergence associated with discrete variations in habitat. For example, genetic subdivision occurs in plant populations where sections of the soil contain heavy metal wastes from mining activities (Jain emd Bradshaw, 1966; Antonovics and Bradshaw 1970). Genetic changes conferring resistance to the rodenticide, warfarin, have
1 2 been found in treated rats while adjacent, untreated individuals (separated by distances as smsdl as 5 meters) show no changes (Bishop and Hartley, 1976; Bishop et al., 1977; Bishop, 1981). Spatially distinct selective pressures have also been implicated as the cause of genetic subdivision within several other species including populations of Daphrtia (Welder, 1985), AnihoxanHmm plants (Snaydon and Davies, 1972), andBorriChia plants (Antifinger, 1981). In the absence of discrete differences in selection within a habitat, genetic divergence has been attributed to the short-range dispersal of pollen or propagules. Selander (1970) reported genetic subdivision within a population of mice inhabiting a bam and concluded that limited movement and mating between mice firom different areas within the bam was responsible for this divergence. Similar dispersal-related, local genetic structure has also been observed within populations of several taxa including mammals (e.g.. White and Svendsen, 1992), birds (e.g., Barrowclough and Coats, 1985), fish (e.g.. Chapman, 1989) and plants (e.g., Wright, 1943b). For orgcuiisms with short-dlstance dispersal firom natal sites, the shape of the area occupied by a population can have a strong effect on its genetic structure (Wright, 1943b). When organisms occupying a circular reinge have dispersal distances that are large relative to the diameter of the habitat, complete mixture of alleles is possible during each generation. The same dispersal distance in an elongated range of equal area will not spread alleles eis effectively; local structure can result (Wright 1943a, 1951; Rohlf and SchneU, 1971).
13 Philopatiy in Green Turtles
Green turtles lay their clutches on virtually linear ocean beaches which can be many kilometers in length but only several meters wide. Nesting populations consist of adult females that have migrated hundreds or thousands of kilometers from feeding grounds to converge on particular beaches where they lay their e^s. If females exhibit precise homing to specific natal sites along a rookery, genetic structure may develop through isolation by distance. Life history characteristics of aU marine turtles make the direct study of effective dispersal very difficult. An estimated 27-33 years are required for female green turtles in the Atlantic to reach sexual m aturity (Frazer and Ladner, 1986), eind individuals move widely among different marine habitats during their development (Carr, 1980). These features make it virtually impossible to track individuals from the hatchling stage to the first nesting event in order to observe the realized dispersal distances from natal sites. Indirect methods have therefore been used to examine dispersal in this species. IVIeylan et al. (1990), Bowen et al. (1992) and Allard et al. (1994), using maternally transmitted mitochondrial DNA, found significant divergence among different nesting populations of green turtles. They concluded that females exhibit strong homing behavior to natal rookeries and hence contribute very little to the exchange of any genetic material between populations. Data from tag returns have revealed that female green turtles tend to return to specific sections of a rookery
14 to nest (Carr and Carr. 1972; Carr and Hirth, 1962; Mortimer and Portier, 1989; Johnson, 1994). These results provide evidence that females show homing precision to areas of the beach where they have previously nested.
Testing the Precision of Natal Homing: A Molecular Approach
Although these studies have revealed that females tend to return to natal beaches, and to sites they have used before, it is unknown whether the first nesting site is also near the natal site. If female turtles are returning to nest at natal sites, genetic structure can develop which may be identified through the use of high-resolution genetic techniques. Multilocus minisatellite DNA fingerprinting CJeffreys et al., 1995) provides a way to assess relatedness among individuals by producing individual-specific banding patterns (genoiypes). This technique is commonly used in patemiiy analysis (e.g., Burke and Bruford, 1987; Rabenold et al., 1990; Decker et al., 1993); however, it has more recently been used for population-level cinalyses (e.g., Gilbert et al., 1990; Parker and Whiteman, 1993; Triggs et al., 1992). The use of DNA fingerprinting to produce individual- specific banding patterns of nesting green turtles provides a way to determine the spatial distribution of genotypes within a rookery. For this study, multilocus minisatellite DNA fingerprinting was used to examine the local genetic structure along two nesting populations (Tortuguero, Costa Rica and Melbourne, Florida, U. S. A.;
15 Fig. 1) to assess the extent to which female green turtles exhibit withln-beach precision in natal philopatiy. If females return to their natal site to nest, then individuals that nest in a particular section of beach should be more closely related than individuals that nest several kilometers apart. Because spatial clumping of related individuals within a nesting season would also be expected if female kin groups nest together on randomly selected sections of the beach, the relationship between distance and genetic similariiy was compared for pairs of females nesting one or two years apart in the Tortuguero population. If individuals returned to their natal site to nest, then two turtles from two different years that nest in a particular section of beach should be closely related. If, on the other hand, females show no natal philopatiy. but nest near their kin, then turtles from different years that nest in the same area should not be closely related. To calibrate the relatedness among females nesting along each beach, the genetic similarity values of first-order relatives (mother-ofifspring pairs) were compared with the scores of adult females. Distance-related genetic structure along nesting beaches would indicate that populations are composed of several genetic neighborhoods or spatially discrete lineages. For populations of conservation concern, like those of endangered green turtles, such an organization may have serious genetic consequences if disturbemces are also spatially discrete; particular lineages may be disproportionately affected, causing the population to lose genetic diversity.
16 Figure 1. Map of Central America and Florida. U.S.A. Field sites are indicated with arrows. Tortuguero is located on the northeast coast of Costa Rica, and Melbourne is along the central Atlantic coast of Florida.
17 Melbourne /B each Florida
Gulf of Atlantic Ocean Mexico
00
Caribbean Sea
Tortuguero / Beach
Cosa Rica Materials and Methods
Field Methods
During the s um m ers of 1991, 1992 and 1993, a total of 98 blood samples (20-100^1) were collected from adult female green turtles nesting on the northernmost 8 km of Tortuguero beach, Costa Rica. In the summer of 1994, blood was collected from 50 additional green turtles nesting along 16 km of Melbourne beach, Florida, USA, between Sebastigm Inlet and Coconut Point Park. Blood was taken by intravenous sampling from either the dorsal cervical sinus using 18 gauge needles or the femoral vein using 23 gauge needles. Tortuguero beach is marked with stakes every 0.2 km and Melbourne beach every 0.1 km so the location of each nesting turtle was recorded based on its proximity to the nearest beach marker. Clutches laid by sampled females were marked and monitored throughout the summer. Incubation periods ranged from 56 to 78 days at Tortuguero Beach and 55 to 64 days at Melbourne Beach. Emerging hatchlings were placed in buckets lined with moist sand and covered with dark towels. Blood collection took place at shelters away from the beach to decrease the risk of over heating for hatchlings that emerged in the early morning, and to eliminate the need for headlamps on the beach when hatchlings emerged at night. Forty-/^l heparinized capillary tubes were used to collect 10-30 fi\ blood from hatchlings after venipuncture of the dorsal cervical sinus using 26 gauge needles. Blood
19 samples were collected from hatchlings representing 10 clutches in each population with a total of 430 hatchlings sampled in Tortuguero and 350 in Melbourne. Following blood collection, hatchlmgs were returned to their specific nest sites and allowed to crawl down to the surf. Each sample (from both populations) was stored in 1 ml of lysis buffer (100 mM Tris HCl, pH 8.0, 100 mM EDTA, 10 mM NaCl, 0.5% SDS; Longmire et al., 1988).
Laboratory methods
Samples were incubated overnight at 65°C with 25 fi\ proteinase K ( lOmg/ml). DNA was extracted from samples by two phenol extractions, two phenohClA (chloroformdsoamyl alcohol at 24:1) extractions and one CIA extraction. Samples were dialyzed for 3-10 hours in cold (4°C)TNE2 (10 mM Tris, 10 mM NaCl, 2mM EDTA pH 8.0). For each individual, Afig DNA was digested with 5X excess HaeUl at 37°C for 3-5 hours. The fragments produced by digestion were separated by size along an electrical gradient in a 0.8% agarose gel for 65 hours at 20V. Southern blotting (Southern, 1975) was used to transfer the DNA to a nylon membrane, to which it was fixed by UV cross-linking. Membranes were hybridized with Jeffreys' probe 33.15 (Jeffreys et al., 1985), which had been labeled by primer extension with a32pciCTP. Hybridizations were run overnight at 62°C in 1.5X SSC, 0.1%SDS, 5X Denhardt's solution, and 6% w/v dextran sulfate. Following hybridization, filters were washed four times at 62°C in 1.5X
2 0 SSC and 0.1% SDS, then exposed to X-ray film at -20°C for > 24 hours with intensifying screens (for protocols of lab methods, see Appendices A-H). DNA samples were randomly selected from Tortuguero and Melbourne adult green turtles and were run on four gels. For the Tortuguero rookery, three gels were run with samples from three partial families (mothers and between 11 emd 18 of their hatchlings). One additional gel was run that contained three Melbourne mothers each with four to six of their hatchlings. In each gel except the Melbourne family gel, DNA from one or two individuals was repeated two to three times. On the autoradiograms, horizontal lines were drawn that connected fl^ve or six of the identical bands in the repeated lanes. The autoradiograms were sliced apart between lanes and the horizontal lines were used to position the strips so that the five to 10 closest lanes could be scored adjacent to one another. This method was used to address the problem of decreased accuracy in band shairing estimates that is associated with increased distance between scoring lanes (Piper and Parker 1992). With the Melbourne family autoradiogram, pairs to be scored were no more than three lanes apart, so this film was left intact during scoring.
2 1 Data Analysis
Relationship Between Intemest Distance and Genetic Similarity
Genetic similarity values {proportion of bands shared) were calculated for dyads of nesting females as D = 2S / (2S + A + B). where S equals the number of bands shared by the two individuals under comparison, A is the number of bands exclusive to one, emd B is the number exclusive to the other (Lynch, 1990). The local genetic structure within each population was determined by examining the genetic similarity of femede pairs as a function of the distemce between their nest sites. For the Tortuguero population, this relationship was examined for turtles nesting in the same year ( 1991 ; n = 14) and for those nesting in different years (1991, n = 6; 1992, n = 5: 1993, n = 7). For the Melbourne population, turtles nesting in 1994 (n = 19) were analyzed. Due to the lack of independence of the data points (each female was scored against multiple other individuals) the Mantel test (Mantel, 1967) was used to evaluate whether genetic similarity values and distance were correlated for pairs of turtles nesting in Tortuguero and in Melbourne. Two symmetrical similarity matrices were used for each test (one for genetic similarity based on band-sharing values and a corresponding matrix of distances between nest sites); the significance of the relationship between the elements of the two matrices was assessed through permutational analysis (Schnell et al., 1985). This
2 2 anatysis randomly permutes the order of the elements of one matrix, while holding the other constant, and compares the correlation values for each of a specified num ber (we used 1000) of perm utations to the initial correlation using the original matrix. The computer program Numerical Taxonomy and Multivariate Analysis System (NTSYS-pc) was used for both the Mantel matrix correlation test and the permutational anafysis (Rohlf. 1990).
Calibrating Relatedness Among Females
To evaluate the genetic relatedness of nesting females firom each rookery, band-sharing scores were calculated for mother-ofifepring pairs and firom sibling pairs to generate distributions of genetic similarity values for first-order relatives. The only sibling scores used in first- order relative distributions were those firom singty-sired families firom Melbourne (see Chapter 2). Familes identified as having onty one father contained nestmate pairs that were full-siblings and thus represented first-order relatives. The distributions firom each rookery were used to examine whether any pairs of nesting females consisted of close genetic relatives. For both Tortuguero and Melbourne, the 95% confidence interval (mean + 1.96 x S.D.) was calculated for the distribution of first-order relative pairs and for the distribution of all female pairs. Female-female dyads with genetic similarity scores that feU above the lower 95% confidence limit for first-order relatives were identified as pairs that represented closely related individuals.
23 Using the vEilues for first-order relatives calculated from the mother-ofifspring pairs and sibling pairs, the expected mean genetic similarity value was estimated for unrelated individuals (Georges et al.. 1988). This expected mean was then compared with the mean calculated for all female pairs. A higher than expected value for mean genetic similarity would provide evidence that some pairs consisted of related females. Where a relationship existed between genetic similarity emd intemest distance, potential^ unrelated pairs were defined as those with intemest distances greater than a particular threshold value. The threshold value was determined by identifying the point in the relationship between intemest distance and genetic similarity above which high genetic similarity values no longer occurred.
R esults
Relationship Between Intemest Distance and Genetic Similarity
In the Costa Rica population, there was a significant negative correlation (Mantel matrix correlation r^ = 0.273: p < 0.001) between genetic similarity and intem est distance (Fig. 2). Thus, individuals nesting in the same area are more likely to be close relatives than those nesting farther apart. This indicates that there is little movement of alleles along the length of the beach, and suggests that females are returning to natal sites within beaches to nest.
24 The relationship between genetic similarity and Intemest distance was stronger for pairs of females that nested one or two years apart in Tortuguero (Fig. 3; Mantel matrix correlation r^ = 0.578; p < 0.001). Even between years, pairs of turtles nesting in the same parts of the beach have higher genetic similarity values than pairs nesting in different parts of the beach. In contrast, the Melbourne population showed no relationship between genetic similarity and intemest distance for pairs of nesting females (Fig. 4. Mantel matrix correlation r2 = 0.017; p = 0.075). The lack of distance-related genetic structure within Melboume failed to reveal precise natal philopatry by the females of this population (for data on intemest distances and genetic similarity scores, see Appendix L).
Calibrating Relatedness Among Nesting Females
The distributions of genetic similarity values for adult female pairs and for mother-offspring pairs from Tortuguero are shown in Figure 5 (see Appendix M for genetic similarity scores of first-order relatives). The 95% confidence interval for female pairs is 0.252-0.553. The upper end of this distribution for female pairs overlaps with the distribution for first-order relatives (95% C.I. = 0.496-0.765); 12.3% (15 of 122) of the scores for all female pairs are higher than the lower 95% confidence limit of the distribution for first-order relatives (0.496). In addition.
25 0.6
0.5- < s ■ < ( a 0.4-< s (A • U « MP 0.3- a î5M 0.2
0.1
INTERNEST DISTANCE (in kms)
Figure 2 . Relationship between intemest distance and relatedness for 60 pairwise comparisons of 14 green turtles nesting along Tortuguero Beach in 1991. Intemest distance and genetic similarity values showed a significant negative correlation (Mantel matrix correlation r^ = 0.273; p < 0.001)
26 0.6
0.5-
S I 0.4-
0.3-
0. 2-
0.1 0 1 2 3 4 5 6 7
INTERNEST DISTANCE (in kms)
Figure 3 . Relationship between intemest distance and relatedness for 62 pairwise comparisons of 18 green turtles nesting along Tortuguero Beach between 1991 and 1993. 25 pairs nested two years apart and 37 pairs nested one year apart. Intemest distance emd genetic similarity values showed a significant negative correlation (Mantel matrix correlation r^ = 0,578; p < 0.001)
27 0.6
0.5-
0.4- i P M 0.3- I
0 . 2 -
01 2 3 4 5 6 7 8 9 10 II 12 13 14
INTERNEST DISTANCE (in kms)
Figure 4 . Relationship between intemest distance and relatedness for 66 pairwise comparisons of 19 green turtles nesting along Melboume Beach in 1994. Intemest distance and genetic sin&arliy values were not significEuitly correlated (Mantel matrix correlation r^ = 0.017; p< 0.075)
28 none of these 15 pairs (with scores over 0.496) had intemest distances higher than 1.407 km (mean distance 0.737 km. S. D. = 0.370). These results indicate that several of the pairs of females scored for this analysis are composed of closely related individuals. No overlap was observed between the genetic similarity distributions from the Melboume nesting population (Fig. 6). For the female-female distribution, the 95% confidence interval was 0.233-0.467, and for the distribution of first-order pairs (see Appendix M for genetic similarity scores of first-order relatives), the interval was 0.524-0.740. None of the 66 Melboume female pairs had genetic similarity scores that were higher than the lower 95% confidence limit of the distribution for mother-offspring pairs (0.524) Based on the band-sharing scores from the distribution of first- order relatives, the expected mean genetic similarity value was calculated for unrelated pairs (Georges et al., 1988). For Tortuguero, this value, 0.327, is lower than the mean score for female pairs (mean = 0.402; S. D. = 0.077), indicating that the mean for remdomly selected pairs of nesting females is inflated by the presence of closely related individuals nesting near each other. To explore this possibility, each peiir of females was categorized as "near" or "far" based on the distance apart that members of the pair nested. The cutoff point used to distinguish "near" from "far" was identified as the point in the relationship beyond which the r^ value feU below 0.1 (at 2 km, r^ = 0.098).
29 0.30-
0.25- 95% C. I. for all female pairs -| 95% C. I. for flrst-order relatives > 0.20- u a M D 0.15- Q> • g k 0.10-
0.05-
0.00- I juUl r ' r'"r' i“ i‘
9000 OOOO 0090 OOOO OOOO 0090 GENETIC SIMEARITY
□ female pairs □ mother-oflsprlng pairs
Figure 5 . Distributions of genetic similarity scores for 41 mother- offspring pairs and for all 122 female pairs (n = 26) nesting along Tortuguero beach. The bars above each distribution identify the 95% confidence intervals (for mother-offspring pairs, 0.496-0.765; for all female pairs, 0.252-0.553).
30 0.3-
95% C. I. for ali female pairs 95% C. I. for flrst-order relatives I------1
0 . 2 - U g
kI 0 . 1-
0 . 0 - j J
9900 9909 9 9 0 9 0 9 9 9 GENETIC SIMILARnY
□ female pairs Q flrst-order pairs
Figure 6. Distributions of genetic similarily scores for 32 first-order relative pairs and for ali 66 female pairs (n = 19) nesting along Melboume Beach, The bars above each distribution identify tiie 95% confidence intervals (for first-order relatives, 0.524-0.740: for all female pairs, 0.233-0.467).
31 After eliminating all female pairs that had intemest distances less than 2 km. the mean genetic similarity values of the remaining, more distant pairs were calculated. These “unrelated” pairs had a mean score of 0.342 (± 0.041) which is close to the expected value for unrelated individuals, and is significantly lower than the mean for all female pairs (Mantel T = -10.321; p < 0.001). In addition, the 95% confidence interval for the distribution of female pairs nesting far apart (0.261-0.423) does not overlap with the distribution for first order relatives (Fig. 7). These analyses provide evidence that some of the pairs in the distribution of female-female genetic similarity scores represent closely related individuals and that these individuals tend to nest near one another. The estimated mean genetic similarity value for unrelated pairs in Melbourne was 0.331. which was just lower than the mean calculated for female pairs (mean = 0.350 ± 0.060). This slight discrepancy between the mean genetic similarity value for all female pairs and the expected mean for unrelated individuals may indicate the presence of one or more related pairs within the sample of females scored for Melbourne. However, because none of the female-female dyads had values that fell above the lower end of the distribution for mother- offspring scores, there is no evidence that any of the female pairs comprises very close relatives. The approach used to calibrate the relatedness of adults in this study does not provide a way to detect
32 0.30- 95% C. I. for "unrelated" pairs I------1 0.25-
95% C. I. for flrst-order relatives 0 . 20-
0.15-
0 . 10-
0.05-
0 . 00- I I " 1“ I " I . 1 “ I **i “ I ‘ ‘r
' O O O I ' o o o o o GENETIC SIMILARHY
□ "unrelated" females □ mother-ofisprlng pairs
Figure 7 . Distributions of genetic similarity scores for 41 mother- ofifspring pairs and for 46 female pairs (n = 26) nesting more than 2 km apart (identified as "unrelated" pairs) in Tortuguero. The bars above each distribution identify the 95% confidence intervals (for mother-ofispring pairs, 0.496-0.765; for "unrelated" females. 0.261-0.423).
33 pairs with intermediate levels of relatedness (e.g. half-siblings, cousins, etc.). Thus it is possible that the mean genetic similarity value for pairs of females in Melbourne represented a sample of females that consisted of a few pairs with moderate levels of relatedness.
D iscussion
Precise Natal Philopatry
The results of this study reveal a distance-related genetic structure in the Tortuguero population (Fig. 2) which persists (and actually strengthens) when comparing different individuals from separate nesting seasons (Fig. 3). Females migrate to the nesting beach every 2-3 years (Carr et al., 1978); therefore, if related females nest independently rather than synchronously, the probability of detecting these pairs is lower within a single year than among three different seasons. This higher probability of detecting related pairs over multiple seasons may account for the apparent difference in the r2 values observed for the two correlations (Fig. 2 and Fig. 3). The upper end of the distribution of genetic similarity values for all females overlaps with the distribution for first order relatives (Fig. 5). The mean genetic similarity value for pairs of females nesting more than 2 km apart is significantly lower than the mean for all female pairs (Fig. 6). Whüe these results do not rule out the possibility that
34 female green turtles may be returning in kin groups, they do indicate that turtles are homing to their natal sites to nest. Strong homing behavior by one or both sexes has been observed in many vertebrate and invertebrate taxa (for reviews see Papi. 1992; Schmidt-Koenig and Keeton, 1977), and cases have been reported of populations where individuals exhibit precise natal philopatry, sometimes following long developmental absences (e.g. salmon, Quinn and Ditman, 1990; yellow perch, Aalto and Newsome, 1990; toads, Waldman et al., 1992; boobies, Osorio-Beristain and Drummond, 1993; Atlantic puffins, Kress and Nettleship, 1988; least terns, Atwood and Massey, 1988; Scottish guillemots. Swan and Ramsay, 1983). The Laysan albatross provides one of the most striking examples of a highly philopatric species that is long-lived and has delayed sexual maturation. The Laysan albatross is a pelagic species that ranges widely throughout the North Pacific, and nests in colonies on Hawaiian atolls (Kenyon and Rice, 1962). Age at first reproduction is eight years for females and nine for males, and individuals live for 25 years or more (Van Ryzin and Fisher 1976). Females and males breeding for the first time nest an average of 26 meters and 15 meters respectively, from natal sites, and then return in subsequent seasons to the same nest (Fisher and Fisher 1969; Fisher 1976). The adaptive significance of such precision in natal homing remains obscure. However, nesting near natal areas may be favored under conditions
35 that include high dispersal costs and predictability of the rookery habitat (Switzer 1993). For green turtles, migration is an energetically expensive activity and excessive wandering in search of a suitable nesting area may be disadvantageous to females. For example, instead of laying additional clutches, female loggerhead turtles {Caretta caretta ) in Australia tend to resorb yolked eggs during their return to feeding grounds (Limpus, in press). If dispersal from the natal beach to a new rookery results in the production of fewer eggs due to increased energy demands, then dispersing individuals may be selected against. In a stable environment, a female whose mother was successful at a particular site should return to that area to nest because she is likely to experience a similar level of reproductive success. Thus, if the nesting beaches used by green turtles remain relatively stable over long periods of time, the fitness of females that home to their natal sites is likely to be higher than that of females attempting to find another (unproven) nesting area. The strong homing behavior characteristic of green turtle nesting populations does not appear to exist to the same degree in leatherback turtles {.Dermochelys coreacea ) which tend to move among nesting beaches more often (Dutton et al., 1994; Dutton et al., 1995). The higher rate of migration between rookeries in this species has been attributed to the fact that females tend to select nesting beaches that are often dynamic and relatively ephemeral, thus a female returning to her natal site is unlikely to find similar conditions even between nesting seasons. Leatherbacks have
36 therefore apparently responded to the absence of stable nesting areas by perlodlcedly dispersing to different sites.
Comparison of Rookery Substructure
In contrast to the Tortuguero turtles, the population of green turtles nesting at Melbourne beach showed no distance-related genetic structure (Fig. 4). These results s u r e s t that Melbourne turtles may not exhibit the same level of precision in natal homing as the Tortuguero turtles, and, as a consequence, sufficient gene flow occurs along the beach to mix alleles within this population. While these results reveal that green turtles in Tortuguero typically exhibit short-distance dispersal from natal sites, the factors influencing local genetic structure in the Melbourne rookery are unknown. The difference may be a result of differential precision in homing, or of demographic parameters affecting each rookery.
Human and Natural Disturbance
While female green turtles from both the Costa Rica and Florida rookeries exhibit strong philopatry to natal beaches (Allard et al., 1994), the level of within-beach homing precision may differ between the two nesting populations. A difference in natal philopatry between the two populations may be caused by the level of human disturbance that each experiences. The Melbourne population is exposed to high
37 levels of potential disturbance in the form of artificial lighting (see Witherington, 1992) Eissociated with beach development. However, there is much less development along Tortuguero beach, so disturbance to nesting turtles in this population is likely to be lower. If Florida turtles are forced to move farther along the beach to nest, the average natal dispersal distance will be elevated and any natural genetic structure may be disrupted. Tag return data firom green turtles nesting along Melbourne beach show that these females exhibit site fidelity witliin and between nesting seasons (Johnson, 1994). These data suggest that Melbourne turtles are capable of distinguishing among different sections of the beach; however, whether they are distinguishing natal sites firom other sites is unknown. It is possible that many first-time nesters abort attempts to return to their natal site due to disturbance firom human development, and instead nest in less developed areas that may be some distance away. If disturbance there remains low, females may remain faithful to this section of beach and return for subsequent nesting events, thereby exhibiting fidelity to their first nesting site. Natural disturbance may also eiffect the ability of females to nest in specific natal sites. The green turtles of Melbourne share this nesting beach with the second largest aggregation of nesting loggerhead turtles [Caretta caretta) in the world (Ross, 1982). Green turtles attempting to nest at Melbourne may suffer high levels of displacement ty male or female loggerheads, and therefore may
38 exhibit breakdowns in natal philopatry. The difference in genetic structure within the Tortuguero and Melbourne nesting populations may therefore result from differential homing precision that is caused not by human disturbance, but rather by natural displacement from loggerheads.
Quality of Homing Cues
Another possible reason for a difference in natal philopatry between the two populations is that the turtles from Tortuguero may be able to home with greater precision because better cues are available to them. Migrating females may use magnetic fields (Lohmann, 1992) or olfactory cues (Carr, 1967; Koch et al., 1969) to orient and return to natal beaches. It is possible that Tortuguero beach may provide high quality magnetic or olfactory information that allows turtles to distinguish among different sections of the nesting beach even after long absences. If cues of comparable quality are unavailable on Florida beaches, turtles nesting there would show greater natal dispersal. Despite the fact that Melbourne turtles tend to nest near their previous nesting sites (Johnson, 1994), the cues used by females to return to a specific nesting area may be ephemeral, chemging in strength or location over the course of only a few years. Femsdes may be capable of returning to a previous nesting site based on these cues, but then, during each nesting event, they may gather current
39 information for homing in subsequent seasons. The estimated time to reach sexual maturity for green turtles in the Atlantic is between 27 and 33 years (Frazer and Ladner. 1986). Once they have reached breeding age, females return to the rookery every 2 to 3 years to nest (Carr and Ogren, 1960). If olfactory, visual or magnetic cues along Florida's coast change more rapidly than those along Costa Rica's Atlantic coast, a first-time nester from Melbourne may not be capable of identifying her natal site, but a re-nester may recognize her previous nesting area. In contrast, a first-time nester returning to Tortuguero may find reliable cues that allow her to distinguish her natal site firom other parts of the rookery.
Abiütv to Detect Relatives
Alternatively, local genetic structure within each rookery may be influenced not by differences in within-beach homing precision, but rather by different levels of nest or hatchling mortality (B. Bowen, pers. comm.). The probability of detecting distance-related genetic structure within a rookery is enhanced when more relatives survive. Therefore, if hatchlings firom Tortuguero enjoy higher survivorship to sexuail maturity, and exhibit natal philopatry. there would be more relatives within a cohort nesting in the same section of beach. Lower levels of survivorship among nests and hatchlings within Melbourne could ultimately produce a nesting population with few close relatives, making local genetic structure more difficult to detect.
40 In this study, the distributions for first-order relatives were used to estimate the proportion of adults that resemble close relatives. This approach provided a way to examine the idea that Melbourne females may experience higher levels of mortality than Tortuguero turtles. Anedysis of the two populations revealed that 12.3% of female pairs in Tortuguero consisted of close relatives while none of the Melbourne pairs overlapped with the distribution for first-order relatives. The first possible explanation for this result is that Melbourne females do suffer higher levels of mortality than Tortuguero turtles, and so pairs of individuals that were former nestmates are rare. Therefore, the sample of females along this beach may not contain any pairs of individuals that represent close relatives, reducing the ability to detect a relationship between genetic similarity and intemest distance. The annual nesting population of green turtles at Tortuguero ranges between about 6,000 to 23,000 females (Carr et al., 1978) while the Melbourne population numbers only about 30 to 700 females per year (Johnson, 1994). Higher levels of mortality among turtles produced at Melbourne may be associated with this rookery's small size. Population comparisons of survivorship rates for different life stages can provide one way of evaduatlng whether higher levels of mortality have caused the absence of close relatives in the Melbourne rookery. While very few estimates of mortality for juvenile and adult green turtles are available (see N.R.C. 1990), reasonable measures of hatching success have been made for both the Tortuguero and
41 Melbourne rookeries. In Tortuguero, overall emergence success has been estimated at approximately 35% for natural nests (Fowler, 1979: Horlkoshl, 1989) and 51% for nests relocated to protective hatcheries (Carr and Hlrth, 1962). Emergence success for natural nests In Melbourne ranged from 75.2% In 1986 to 25.2% In 1989 with a mean over eight consecutive years ( 1985 to 1992) of 51.1% (Johnson, 1994; Ehrhart and Witherington, 1987). The larger production of emerging hatchlings per nest In Melbourne suggests that levels of mortality during this life stage do not contribute to the lack of closely related females nesting along this beach. However, current data on hatching success may not resemble past trends; thus, currently nesting females may have been produced in seasons where emergence success was much lower. At the same time, levels of mortality may be significantly higher for Melbourne turtles during other stages of development, which would influence the number of related adults that return to nest. A second interpretation of the absence of close relatives adong Melbourne Beach Is that mortality is similar In both populations, but that females may disperse farther from their natal sites. Thus, for this study, former nestmates may have been undetected simply because the distance between them was beyond the range of the 16 km sampled beach area. Further comparisons of Individuals nesting along a longer stretch of the Florida coastline may reveal the presence of close relatives which would support the Idea that levels of precision
42 in natal homing dlfifer between the Tortuguero and Melbourne nesting populations.
Relative Age of Nesting Beaches
Another natural feature that may have resulted in the differences in local genetic structure found between the Tortuguero and Melbourne nesting beaches is the age of each nesting colony. Less than 10,000 years ago. the Wisconsin glacier had not fully retreated, and the climate along the coast of Florida was significantly colder than it is today (Hedgpeth, 1954). It is unlikely that the Florida coast provided a suitable environment for sea turtle nesting before approximately 10,000 years ago. The nesting population along the coast of Florida, therefore, may be a relatively new one (Bowen et al. 1992). If the colonization of Florida by green turtles is a recent event, there may have been insufficient time for detectable levels of genetic structure to develop along a distance of only 16 km. However, an examination of the relationship between genetic similariiy and intemest distsmce for pars of females nesting along the entire coast of Florida may reveal that some level of genetic substructuring occurs within this rookery.
43 Conservation Implications
Strong natal philopatry in the Tortuguero population has produced genetic structure where different lineages of turtles are represented in different sections of the beach. The effects of disrupting such a system are unknown. However, costs associated with forced exclusion from natal sites may result in lower fitness for dispersing individuals. For example, green turtles are herbivorous emd are generally unable to find sufficient food along the nesting beach (Carr, 1982); they rely instead on stored fat as the source of energy for reproduction and activity during the nesting season (Bjomdal, 1982). A considerable amount of energy may be expended in repeated attempts to nest. For this reason, continued disturbance to turtles could result in depleted energy reserves due not only to multiple failed nesting attempts, but also to searching for a new, suitable section of beach. Energy diverted from reproduction to nest site selection can reduce fitness of affected turtles and ultimately decrease the net productivity of an already declining population. Decreases in the number of nesting females in a particular area have been observed following beach development (Mortimer, 1982; Worth and Smith, 1976). There is evidence that artificial lighting associated with such development is a major factor influencing the reduction of nesting activity (Witherington, 1992). If female green turtles have been sufficiently disturbed by development and photopollution along Florida's coast, this may have resulted in
44 increased natal dispersal distances and subjected individuals to the fitness costs of greater dispersal. If, on the other hand, effective dispersal firom natal sites is characteristic of females nesting in Florida, and this dispersal is due to natural factors such as the qualiiy of stable environmental cues, then the differences in the genetic structure between the two populations may simply reflect differences in breeding ecology. In a situation such as this, where individuals firom one population exhibit strong natal site fidelity, and individuals firom another tend to disperse farther, the population with dispersing individuals is likely to be less susceptible to any detrimental genetic consequences associated with spatially discrete disturbances. This is because the loss of genetic diversity (resulting firom the disproportionate reduction in fitness of particular lineages that are forced to disperse firom disturbed sites) is reduced if lineages are not concentrated in spatially discrete sections of the beach. For populations with females that exhibit precise natal homing, managers should consider the potentially negative consequences of programs that relocate many nests to a single artificial hatchery. Incubating embryos or emerging hatchlings may receive site-specific, chemical or magnetic information fi*om the beach environment. Euid later use these cues to return to specific sites along the rookery (Carr. 1967; Grassman and Owens. 1987). Nest relocation practices may therefore result in some hatchlings that are imprinted to the hatchery area.
45 If a majority of females return to specific natcii sites, then this could ultimately produce a situation where a large proportion of femedes in subsequent generations converge to nest in the area around the hatchery site. Concentrated in this way, the nesting females may be more susceptible to problems specific to that site, such as future beach development. An increased density of nesting activity in a particular section of the beach may also result in a higher risk of females digging up each other's clutches. Until more is known, a program that spaces hatcheries along the length of the beach may be more effective in maintaining an evenly distributed nesting population in future generations. If the lack of detectable levels of genetic structure within Melbourne is due to high levels of mortality causing the absence of closely related individuals, then managers of the Melbourne rookery should be especially concerned with attempting to reduce mortality pressures for the different life stages of Florida green turtles. Incidental catch in shrimp trawls is considered the most important factor depleting sea turtle populations along southeast coast of the U.S. (N.R.C. 1990). A conservative estimate of about 47,000 sea turtles are captured annually, resulting in the immediate death of at least 11,000 individuals, and up to 32,000 if comatose turtles do not survive (Henwood and Stuntz, 1987). Stranding data collected edong the Southeast coast of the U.S. have been used as an index of relative levels of annual mortality for juvenile and adult sea turtles. When these data are examined in relation to shrimping activity a positive
46 correlation is observed, which further demonstrates the impact of shrimp trawling on sea turtles. In response to this problem, turtle excluder devices (TED's) have been designed that are placed in existing shrimping nets to prevent the incidental catch of turtles. The intermittent use of TED's in U.S. waters between 1987 and 1989 resulted in a rapid reduction in the number of turtle strandings (Booker, 1990). For lo^erhead turtles in South Carolina, the number of strandings was 42% to 52% lower after the 1990 implementation of TED's (Crowder et al. 1994). Since 1994, TED's have been mandatory on all U.S. shrimp trawlers. The success of this law provides an example of how mortality at pelagic life stages can be reduced. If the number of close relatives along Melbourne beach is influenced by the juvenile and adult mortality caused by shrimp trawlers, the use of TED's may result in an increase in the survivorship of flrst-order relatives. This could eventually affect the ability to detect local genetic structure within the Melbourne rookery. An examination of the genetic structure within other populations could prove helpful in determining whether natural or human-related factors are influencing the local genetic structure of green turtle rookeries. For example, if distance-related genetic structure tends to be a feature of populations occupying undeveloped beaches, but is absent in populations from highly developed areas, this would suggest that local genetic structure in green turtle rookeries is more strongly influenced by dramatic human activities than by the qualiiy of
47 environmental cues or other aspects of the natural ecology of the nesting beach.
48 CHAPTER m
THE MATING SYSTEM OF GREEN TURTLES
Introduction
Genetic Variation and the Persistence of Populations
The immediate impacts of demographic stochasticity on small populations is often of major concern in the conservation of threatened species (Lande, 1988; Soule and Wilcox, 1980). However, when long term persistence through evolutionary time is considered, the preservation of genetic resources becomes more importeint (Frankel, 1974; Frankel and Soule, 1980). Even in relatively large populations of a few thousand breeding individuals, the fixation of new deleterious alleles may be as important as demographic stochasticity in causing the extinction of a population (Lande, 1994; Lande, 1995). As a result, researchers and managers have begun to address the question of whether threatened species exhibit sufficient genetic variation to allow populations to adapt to fluctuating environmental conditions (Beardmore, 1983; Chesser 1983; Eriksson et al. 1993). Low levels of
49 genetic variation have been reported as a consequence of small population size in insects (Britten et al., 1994), birds (Rave et al., 1994), and mammals (Gottelli et al., 1994; Hartl and Pucek. 1994; Taylor et al., 1994). Recent studies of several animal populations have observed a negative relationship between genetic variation and fitness characteristics such as body condition (Rhodes emd Smith, 1993), metabolic efficiency (Teska et al., 1990; Rodhouse et al., 1986), growth rate (Koehn and Gaffirey, 1984; Quattro and Vrijenhoek 1989; Liskauskas and Ferguson 1991), survival rate (Quattro and Vrijenhoek 1989; Liskauskas and Ferguson 1991) and reproductive success (McAlpine 1993). These studies have validated the concern over the impact of reduced genetic diversiiy on the persistence of populations. In addition to the potential long-term effects of reduced genetic diversity, there is also concern over the more immediate reductions in fitness that can occur at low population sizes due to inbreeding depression (Gilpin and Soule 1986; Frankham 1995; Senner 1980; Ralls and Bcdlou 1983). The level of inbreeding within a population is inversely related to the number of reproducing individuals, thus small populations experience more inbreeding even if mating is random (Falconer 1981). Breeding between close relatives has resulted in the production of offspring with lower viability (Greenwood and Harvey, 1978: Pray and Goodnight 1995; Frankham et al., 1993; Chen 1993) and more developmental impairments (Ralls et al., 1979; Roelke et al. 1993; Laikre and Ryman, 1991) than found in outbred offspring. In addition, inbreeding has been shown to adversely affect reproductive
50 functions (Green, 1968: Roelke et al. 1993; Wildt et al. 1987) and compromise the immune system (O'Brien and Evermann, 1988; Thome and WlUlams, 1988; O'Brien. 1989; HoUebecq and Haffray, 1994). Together, these studies su rest that Inbreeding In small populations can result In reduced reproductive success. Furthermore, there may be very little warning of an Impending extinction due to Inbreeding depression because populations may fall to reveal significant negative effects until Inbreeding reaches a critical level (Frankham, 1995).
Mating Systems and Genetic Diversity
An understanding of the genetics of declining populations requires knowledge about the amount of genetic diversity In addition to Information about how this variation Is maintained. The prevalent mating system of a population Is an important demographic characteristic that has been shown to Influence both fitness (measured by competitive ability of offspring) and levels of genetic diversity (Briton et al. 1994). The mating system describes the sex ratio of reproducing individuals and the variance In reproductive success. If reproductive success is highly variable among members of a population, then some lineages will be over-represented while others will be under-represented. The diversity and firequency of genotypes that will be present in subsequent generations are therefore directly affected by the mating system.
51 In some species the prevalent mating system is unknown; therefore, information about breeding sex ratios and the number of mates attained by each individual is unavEiilable. When mating behavior is difficult to observe, high resolution molecular techniques can often provide a way to analyze paternity and thereby identify prevalent breeding behaviors within a population. For example, a finding of single paternity within a clutch or litter would suggest that the female mated with a single male (thus the mating system could be identified as either monogamous or polygynous). while multiple paternity would suggest that the mating system was either promiscuous or polyandrous. While the consequence of mixed patemify can be increased genetic variation in future generations, the evolution of multiple mating by females may be a result of different selective pressures. Copulating with more than one male may be advantageous to females by allowing them to (1) extract resources firom multiple males; (2) mate with a high quality male if a female’s social mate is of low quality; (3) increase the genetic diversity of the clutch; (4) insure fertilization of an entire clutch of eggs (Birkhead and Moller 1992); or (5) increase sperm competition to enhance the production of more viable offspring (Knowlton and Greenwell 1984). The last three hypotheses may be particularly relevant for species in which individuals do not hold territories, may have ephemeral interactions with large numbers of individuals, and those with large clutch sizes. An exploration of breeding systems indirectly through the use of genetic techniques can be especialty useful in understanding the population dynamics of species with life stages
52 that are inaccessible or difficult to monitor, like most plants and marine organisms.
Background on the Green Turtle
The green turtle represents an example of an endangered species with an intractable life history that restricts opportunities for direct field observations. During their 27 to 33 year development to sexual maturiiy (Frazier and Ladner. 1986) green turtles move widely among different ocean habitats (Carr. 1980). Once they have reached breeding age. they migrate hundreds or thousands of kilometers between feeding grounds and nesting beaches (Meylan 1982). Mating occurs in the water and only females emerge onto the beach (every two to three years) to lay their clutches of about 100 eggs (Carr et al. 1978). Because male green turtles remain at sea while females go ashore to lay their eggs, nesting females and emerging hatchlings are the onty easily accessible members of the population. These life history traits hamper our ability to examine male behavior and to characterize the mating system of green turtles. This project has. therefore, developed an approach for analyzing paternity when samples firom only mothers and hatchlings are available. This approach involves the use of microsatellite technology in combination with a statistical analysis not only to detect multiple paternity but also to estimate the number of fathers contributing to a particular clutch.
53 Application of Microsatellite Technology to Paternity Analysis
Microsatellites are composed of short (2-6 bp) tandem repetitive units (e.g. (CA)n/(GT)n) that are scattered throughout the genome (Tautz 1989). At a particular microsatellite locus, variation among individuals is defined in terms of the number of copies of these repeat units. Because microsatellite technology provides a way to examine genetic variation among individuals by identifying and comparing codominant alleles at specific loci, it has tremendous potential for answering questions about behavior (e.g. Brockmann, 1994: Houlden et al.. 1996; Craighead et al.. 1995: Evans. 1993) evolution (e.g. FitzSimmons et al.. 1995: Meyer et al. 1995: Roy et al. 1994) and population genetics (e.g. Gottelli et al 1994: Taylor et al. 1994: Dallas et al. 1995: Paetkau et al.. 1995) This study used single locus microsatellite technology to examine whether green turtle clutches firom two populations (Tortuguero. Costa Rica, and Melbourne Beach. Florida. USA: Fig. 1) were characterized by single or mixed paternity. This analysis of paternity was performed in order to gain an understanding of the prevalent mating systems exhibited by each nesting a^regate of green turtles. In addition. I examined whether any relationship existed between the number of males represented in a female’s clutch and the proportion of eggs that produced hatchlings.
54 Materials and Methods
Field Methods
Field work for this study took place over the course of four green turtle nesting seasons in two different populations. During the summers of 1991, 1992 and 1993, blood samples were collected from 98 adult females green turtles nesting on the northernmost 8 km of Tortuguero Beach, Costa Rica. In the summer of 1994, 50 females were sampled from 16 km of Melbourne Beach, Florida, USA, between Sebcistian Inlet and Coconut Point Park. Blood (20 to 100 fxl) from adult females was taken by intravenous sampling from either the dorsal cervical sinus using 18 gauge needles, or from the femoral vein using 23 gauge needles. Eggs were counted as they were laid and nests were marked and monitored throughout their incubation periods. The number of emerging hatchlings produced for each monitored nest was recorded, and blood samples were collected from a total of 360 individuals (representing 8 clutches) at Tortuguero, and 300 individuals (representing 10 clutches) at Melbourne. Incubation periods ranged from 56 to 78 days on Tortuguero Beach, and 55 to 64 days on Melbourne Beach. Emerging hatchlings were placed in buckets lined with moist sand and covered with dark towels. Blood collection took place at shelters away from the beach to decrease the risk of overheating for hatchlings that emerged hi the early morning, and to
55 eliminate the use of headlamps on the beach when hatchlings emerged at night. Heparinized capillary tubes (40 fi\) were used to collect 10 to 30 fA of blood from hatchlings following venipuncture of the dorsal cervical sinus using 26 gauge needles. After blood collection, hatchlings were returned to their respective nest sites and allowed to crawl to the surf. Each blood sample (from adults and hatchlings of each population) was stored in 1 ml of lysis buffer (100 mM Tris, pHS.O, 100 mM EDTA,
lOmM NaCl, 0.5% SDS; Longmire et Ed.. 1988)
Laboratory Methods
Samples were incubated overnight at 65°C with 25 fA proteinase K ( 10 mg/ml). DNA was extracted from samples by two phenol extractions, two phenoLCIA (chloroformrisoamyl sdcohol at 24:1) extractions Eind one CIA extraction. Samples were dialyzed for 3 to 10 hours in 4°C TNE 2 (10 mM Tris. pH 7.4. 10 mM NaCl. 2 mM EDTA) to remove impurities. Final DNA concentrations for each sample were estimated using a spectrophotometer. Two primer sets (Cc 117 Eind Cm 3; FitzSimmons et al.. 1995) were used in PCR reactions to amplify specific microsatellite loci from the DNA samples. Each 15 fA PCR reaction consisted of 1.5 fA lOX PCR buffer (Gibco BRL). 1.5 mM dNTP's. 7.5 mM of primers. 0.5 U Taq polymerEise (Gibco BRL). 48 ng of template DNA and 46 mM MgCl 2 for reactions with primer set Cc 117 or 41 mM MgCl 2 for reactions with
56 primer set Cm3. PCR reactions were processed using a Perkin Elmer DNA Thermal Cycler 480. The "step down" method (H. EHegren, pers. comm.) was used for all reactions. This approach is characterized by an initial denaturing phase of 95°C for 2.5 mln. followed by a series of step cycles with sequentially decreasing annealing temperatures. The initial step cycle (95oC for 45s, 62°C for 1 mln, 72°C for 1 mln) was modified in sequential cycles by decreasing the annealing temperature by 1°C. After 7 step-cycles, an annealing temperature of 55°C was reached and malntedned throughout the rest of the reaction. Following the step- cycles, a final extension was performed at 72°C for 5 mln. Samples were then stored at 4°C. The fragments amplified through PCR were separated by size along an electrical gradient in a 7.5% polyacrylamide gel using a BIORAD Seqi-Gen II sequencing gel apparatus (21 x 40 cm). Gels were run at 20-30 watts for 2 to 3 hours. To visualize the fragments, gels were stained with ethldlum bromide (6^1 EtBr: 30 ml 1 X TEE) for 2 minutes, and then rinsed thoroughly with cold (4°C) dH20. Gels were then exposed to UV light, and photographs of the fragments were taken using a Fisher Biotech Electrophoresis Systems Photo-Documentation camera. Banding patterns of individuals were scored from these photographs.
57 Characterization of Alleles and Allele Frequencies
Adult females from each rookery (43 females from Tortuguero and 48 females from Melbourne) were screened with both primer sets to obtain genotypes for each individual. Alleles were categorized based on their relative size. The letter "A" identifed the largest allele found for each primer set in each population, "B", the second largest, and so on. Each gel contained one to seven ladders each comprising multiple previously characterized genotypes of different individuals within the population. The banding patterns of new individual turtles were compared with these ladders to identify their genotypes. Following the characterization of adult female genotypes, the allele frequencies were Ccdculated for each rookery, and these frequencies were assumed to be representative of the breeding aggregation for each rookery.
Analysis of patemify
Because the DNA of potential fathers was inaccessible in this study, male genotypes were not available for the direct analysis of paternity for each clutch. However, because offspring inherit microsatellite alleles in MendeUan fashion, the genotypes of hatchlings and mothers reveals information about the paternal genotypes. At each locus, each hatchling receives one allele from its mother eind one from its father. After eliminating all maternally derived alleles within a clutch, the remaining alleles are paternal in origin. For each locus, a
58 parent can contribute at most two different alleles to a clutch; therefore, any clutch with more than two paternally derived alleles provides evidence for multiple insemination.
Statistical Analysis
The banding patterns of mothers and offspring allowed the direct assessment of paternity as either single or multiple for each clutch: however, in order to estimate the number of fathers contributing to each clutch, a computer program was designed (by Dr. Mark Irwin, Department of Statistics, Ohio State University). This program used population allele frequencies to estimate whether the observed genotypic information from each family could be most parsimoniously explained by the contributions of specific numbers of fathers. The program was provided with allele frequencies calculated from the population of nesting females, the genotypes of family members (mothers and offspring) and the potential number of fathers to be considered. It then calculated the probability of observing the pattern of hatchling genotypes within each clutch for a particular number of fathers. While the number of fathers possible for a given clutch can be as high as the number of hatchlings produced (if each hatchling is sired by a different male) this pattern of paternity is unlikely. Therefore the number of fathers considered for each family ranged from two (for families with only one or two paternally derived alleles) to four (for multiply fertilized clutches).
59 For each father, the program considered every genotype possible from the pool of alleles identified within the adult female population (as well as the frequencies of these alleles) and calculated the probability of creating the observed offspring genotypes for each possible male genotype. If the most likely explanation for the distribution of hatchling genotypes involved the genotype of one male, then the clutch was categorized as exhibiting single paternity. In families where fertilization by one father failed to explain the pattern of genotypes among nestmates, the program examined all possible permutations for the cissignment of hatchlings to specific numbers of fathers. The probabilities associated with these permutations combined with the probabilities of potential paternal genotypes were used to calculate likelihood values for one, two, three or four fathers. There are several ways for nestmates to be attributed to a particular series of fathers. For example, in a clutch of four hatchlings, there is one way to assign a single male as the sole father in the clutch, 10 ways to distribute parentage among two males, six ways for three fathers to contribute, and one way for each hatchling to have a different father. The probabilities associated with each particular number of fathers are added; therefore, in the example of the four hatchling clutch, the 10 probabilities generated for the 10 different ways to have two fathers are added together for an overall likelihood that two fathers fertilized that clutch.
60 For the final interpretation, the relative values of the overall likelihood calculations for each number of fathers were expressed in terms of ratios of the likelihood for any particular number of fathers to the highest likelihood found among aU numbers of fathers considered. For example, if the overall probabilities for one, two, three and four fathers were 0.02, 0.03, 0.01, and 0.005 respectively, then the likelihood ratios would be 0.667 for one father, 1.0 for two, 0.333 for 3 and 0.167 for four. Therefore, in this example, the most likely number of fathers is two, and two fathers is 1.5 times more likely than one (the next most likely number of fathers). Using this approach, it was possible to evaluate the relative probabilities for each number of fathers considered for each clutch.
Ability to Detect Multiple Paternity
For each family, the number of hatchlings sampled ranged from 16 to 27. Because all hatchlings from each clutch were not included in the paternity analysis, it was possible that multiply fertilized clutches were undetected if the contribution of a second male was small relative to that of the primary male. This sampling problem could occur even with perfect genetic detectability of mixed paternity within a family. We evaluated the probability of finding two fathers for a range of possible allocations of hatchlings to each father. We calculated the probability of detecting the genetic contribution of a second male (E) as:
61 E = I - QN
Where Q is the proportion of hatchlings sired by the primary male and N is the number of hatchlings sampled. E was calculated for a series of possible allocations of hatchlings ranging from 100% to 75% attributed to the primary male. This was performed for two sample sizes (16 and 27) that represented the smallest and largest number of hatchlings sampled for the clutches examined in this study.
R esults
Population Allele Frequencies
While only two loci were used in this study, both revealed substantial polymorphism within each nesting population and this enhanced the ability to detect the genetic representation of multiple males within each family. Among the 43 adult Tortuguero females, 13 alleles were found at the Cc 117 locus and 8 were found at the Cm 3 locus. The alleles ranged in frequency from 0.026 to 0.205 for
62 0.50-
0.45-
0.40-
0.35- > Ü 0.30- R M 0.25- t» 0.20-
0.15-
O.IO-
0.05- iiü 1 0.00- B D E F G H I K
ALLELES (IN ORDER OF SIZE)
Figure 8 . Distribution of allele frequencies among Tortuguero females for locus Cc 117 (n = 39).
63 0.50
0.45-
0.40-
0.35-
ü 0.30
0.20
0.10
0.05-
0.00 C D E F ALLELES (IN ORDER OF SIZE)
Figure 9. Distribution of allele frequencies among Tortuguero females for locus Cm 3 (n = 39).
64 0.50
0.45-
0.40-
0.35- I 0.30- 0.25-
0 . 20 -
0.15- o.io-
0.05-
0.00 B D E F G H ALLELES (EST ORDER OF SIZE)
Figure 10. Distribution of allele frequencies among Melbourne females for locus Cc 117 (n = 46)
65 0.50
0.45-
0.40-
0.35-
ü 0.30- H â 0.25- g 0.20-
0.15-
0 . 10-
0.05-
0 . 00 - B C D E ALLELES (IN ORDER OF SIZE)
Figure 11. Distribution of allele frequencies among Melbourne females for locus Cm 3 (n = 46)
66 Cc 117 (Fig. 8) and 0.026 to 0.385 for Cm 3 (Fig. 9). The 48 females from Melbourne showed a similar polymorphism. The Cc 117 locus had 12 alleles ranging in frequency from 0.011 to 0.217 (Fig. 10), while 7 alleles (with frequencies from 0.033 to 0.457) were found at the Cm 3 locus (Fig. 11).
Paternity Analysis
Examination of family gel photographs from both rookeries showed that five of the eight Tortugero families contained more paternal alleles (at both loci) than would be expected if only a single father were involved (Table 1). It was therefore concluded that the mothers of these five clutches used sperm from at least two males to fertilize their clutches. For the Melbourne population, five of the 10 families had three paternally derived alleles at one or both loci (Table 2). Thus, initial results provided evidence of multiple paternity in 62.5% of the Tortuguero nests and 50% of the Melbourne nests. While the Tortuguero population showed a slightly higher proportion of multiply inseminated clutches than the Melbourne population, this difference was not significant (p = 0.66; two-tailed Fisher's exact test). The results of the likelihood analysis (using the Irwin computer program) of the genotypes for each Tortuguero family (Table 3) and each Melbourne family (Table 4) provide estimates of the number of fathers
67 Number and identity of paternal alleles Number of Family hatchlings Cc 117 Cm 3 Paternity
207 19 2(C. E) 2 (C, F) single
309 18 2 (C, K) 2 (C. E) single
312 27 3 (D, F. K) 3 (B, C. F) mixed
3 1 4 28 3 (F. H. J) 3 (E. F. G) mixed
O) 3 1 5 17 3 (G. 1. H) 3 (A, C. E) m ixed 00 3 1 6 18 3 (D. E. 0 ) 4 (B, C. E, F) mixed
3 1 8 18 4 (A, C, D, J) 3 (A, B, C) mixed
3 1 9 27 1 (D) 1 (C) single
Table 1. Results of microsatellite paternity emalysls for eight Tortuguero clutches. Five of the eight families contained more paternally derived alleles (at both loci) than would be expected if only a single father was responsible for fertilization. Number and identity of paternal alleles Number of Family hatchlings Cc 117 Cm 3 Paternity
403 23 3 (B, E and D or H) 1 (D) m ixed
4 0 4 22 2 (B, E) 3 (A, B, C) mixed
405 16 2 (B. J) 3 (A. C, D) m ixed
4 0 7 23 1 (F) 1 (E) single
4 0 8 23 1 (C) 2 (A, E) single s 4 1 7 22 2 (B. G) 3 (A, C and B or E) mixed
421 24 2 (A, F) 2 (B, E) single
426 23 3 (C. D. E) 3 (B.E.F) mixed
4 2 8 21 1 (D) 2 (A, F) single
4 3 0 23 2 (B. H) 2 (B,G) single
Table 2. Results of mlcrosatelllte paternity analysis for ten Melbourne clutches. Five of the ten families contained more paternally derived alleles (at one or both loci) than would be expected If only a single father was responsible for fertilization. For family 403 at locus Cell? and for family 417 at locus Cm 3, several hatchlings had genotypes that were identical to their mother; as a result, it was not possible to determine which allele within these hatchings was paternally derived. Likelihood ratios for each number of fathers F am ily 1 2 3
207 1.0000 0.0498 0.0049
309 1.0000 0.1014 0.0025
312 0.0000 0.0000 1.0000
314 0.0000 0.4356 1.0000
315 0.0000 0.0000 1.0000
316 0.0000 1.0000 0.3064
318 0.0000 1.0000 0.1031
319 1.0000 0.0002 < 0.0001
Table 3 . Results from the computer analysis of eight Tortuguero clutches. According to the program, a single father is most likely for families 207, 309 and 319; two fathers are most likely for fam ilie s 316 and 318; and three fathers are most likely for fam ilie s 312, 314 and 315. However, based on the distribution of genotypes among hatchlings, fam ily 314 probably has two fathers rather than three. Only one hatching contained a paternal band at each locus that was different from the paternal bands of its nestmates. This suggests that one male fathered one of the hatchlings in this sample, and another male fathered the remaining hatchlings.
70 Likelihood ratios for each number of fathers F am ily 1 2 3
403 0.0000 1.0000 0.0429
404 0.0000 1.0000 0.0221
405 0.0000 1.0000 0.0806
407 1.0000 0.0038 < 0.0001
408 1.0000 0.0009 < 0.0001
417 0.0000 1.0000 0.0740
421 1.0000 0.0009 < 0.0001
426 0.0000 1.0000 0.2017
428 1.0000 0.1842 0.0035
430 1.0000 0.0011 < 0.0001
Table 4 . Results from the computer analysis of eight Tortuguero clutches. A single father is most likely for families 407, 408, 421, 428 and 430; two fathers are most likely for families 403, 404, 405, 417 and 426.
71 for each clutch. Within the Tortuguero rookery, the estimated number of fathers in each clutch ranged from one (in families 207, 309 and 319) to three (in fam ilie s 312 and 315). According to the likelihood analysis for family 314, three fathers is 2.3 times more likely than two (Table 3); however, inspection of the distribution of genotypes among nestmates suggests that this family had two fathers. Only one hatchling contained a paternal beuid at each locus that was different from the paternal bands of its nestmates. If this hatchling is eliminated from the computer analysis, then a single father becomes the most likely pattern of paternity for family 314. However, because the genotype of this hatchling does not match with any other family, it does not represent an accidental contamination. The most likely situation for this family, therefore, is that one father was responsible for the fertilization of one of the eggs in this sample, and another male fertilized all of the remaining eggs. The Melbourne population had five clutches (403, 404, 405, 417 and 426) that were fertilized by two males each and five clutches with only a single father. The difference in the mean number of males that were responsible for the fertilization of each clutch in this sample was not significantly different between Tortuguero (mean = 1.88 ± 0.84) and Melbourne (mean = 1.50± 0.53; Mann-Whitney U = 50, p = 0.33).
72 Ability to Detect Multiple Paternity
Eight families from both rookeries showed single paternity. The ability to detect more than one father in these clutches may be affected by the hatchling sample size for each family. If the genetic contribution of a second male was as low as 5%. the probability of detecting that father would range from 56% to 75% (Fig. 12). Family 309 had the smallest sample size of hatchlings for any of the single- patemity clutches ( 18). If 95% of the eggs in this clutch were fertilized by the primary male, the probablity of detecting a second father would be 60%. This probability rises quickly to 85% when the contribution of the second male goes up to 10% and exceeds 98% when the contribution of the second male is 20%. Our ability to detect two fathers is relatively high for our family sample sizes, even when the contribution of the second male is low. As a result, it is likely that those eight clutches that were characterized as having been sired by a single male either did have only one father, or the contribution of the second male was too low to be biologically meaningful considering the high rates of mortality suffered by hatchlings.
Mating System and Hatching Success
For the analysis of hatching success, information from clutches of both populations was pooled (Table 5). The sizes (number of eggs) of
73 1.0
0.9-
N = 16 0 .8- N = 2 7
oI 0.7- M 0.6-
0.5-
0.4-
n 0.3-
0. 2-
O.L-
0.0 i d
HYPOTHETICAL PROPORTION OF EGGS FERTILIZED BY A SECOND MALE
Figure 12. Probability of finding a second father over a range of hypothetical levels of possible allocations of hatchlings ranging firom 100% to 75% attributed to the primary male. Probabilities were calculated for two sample sizes (16 and 27) that represented the smallest and largest number of hatchlings sampled for the clutches examined in this study.
74 Proportion of Number of eggs that Paternity Number of hatchling# produced (number of Familyegg# laid produced hatchlings fathers)
2 0 7 106 — ----- single
T 309 131 87 0.6641 single O R 312 ----- 72 ----- m ixed (3) T U 314 102 101 0.9901 mixed (2) G U 315 81 68 0.8395 mixed (3) E R 31 6 118 92 0 .7 7 9 7 m ixed (2) O 31 8 107 97 0 .9 0 6 5 m ixed (2)
319 119 83 0.6975 single
403 133 93 0.6992 mixed (2)
4 0 4 ----- 73 ... m ixed (2)
405 145 --- ... M m ixed (2) E L 40 7 126 85 0.6746 single B O 40 8 152 109 0.7171 single U R 41 7 ----- 103 ----- m ixed (2) N E 421 I l l 87 0 .7 8 3 8 single
426 102 88 0.8627 mixed (2)
428 113 ----- ... single
4 30 138 ----- ... single
Table 5. Hatchling success information for singly- and multiply- sired clutches from Tortuguero and Melbourne. The mean proportion of e ^ s that produced hatchlings was 0.85 ±0.10 for families with mixed paternity, and 0.71 ± 0.05 for families with one father. This difference was statistically significant (Mann-Whitney [/ = 3; p = 0.028.
7 5 single paternity clutches (mean = 124.50 ± 15.45) was slightly, but not significantly, higher than those with mixed paternity (mean = 112.71 ± 21.33; Mann-Whitney U = 39, p = 0.203). Similarly, the number of hatchlings produced per clutch was not significantly different between nests with single paternity (mean = 90.2 ± 10.64) and those with multiple paternity (mean = 87.44 ± 13.20; Mann-Whitney (/ = 21, p = 0.841). However, the proportion of eggs producing hatchlings was significantly higher for nests fertilized by more than one male (mean = 0.85± 0.10) than for nests with a single father (mean = 0.71 ± 0.05; Mann-Whitney U = 3, p =0.028).
D iscussion
Interpretation of the Mating System
Five of the eight nests examined firom Tortuguero were sired by more than one father, and five of 10 clutches showed mixed paternity in Melbourne. While it is premature to make strong conclusions about the mating system of a population based on paternity analysis of a small Scunple of clutches, this study provides some preliminary information about the breeding behavior of green turtles. The finding of mixed paternity in clutches from both Tortuguero and Melbourne suggests that females from both rookeries commonly mate with multiple males. If males also mate with several females, then promiscuity may be a feature of these breeding populations. If, on the
76 other hand, males mate with only one female each, then the mating system of green turtles from these two rookeries may be characterized by pofyandiy. To reach a thorough understanding of the mating system of green turtles, information from many loci would be ideal. The use of several primer sets would provide an opportunity to reconstruct paternal genotypes for each clutch. These reconstructed genotypes could then be used to compare among different clutches to determine whether a particular male is represented in severed females' nests. Multiple representation of peirtlcular males in clutches characterized by mixed paternity would eliminate polyandry as the prevalent mating system, and suggest that the breeding behavior was promiscuous. This type of analysis is beyond the scope of this study; however, multiple mating by a single green turtle male in a natural population has been reported (Booth and Peters 1972), providing evidence that some green turtle populations may exhibit promiscuity. Other observations of behavior and demographics also suggest that males mate multiply. First, males enthusiastically pursue mating opportunities with apparent disregard for the identity of a potential partner (Carr and Giovannoli, 1957; Booth and Peters, 1972). This suggests that males may be receptive to mating with more than one female if opportunities arise. In addition, female-biaised sex ratios are often typical of captive breeding programs (Wood and Wood, 1980; Ulrich and Parkes, 1978) and perhaps characteristic of some natural populations (Limpus et al., 1994). Under such conditions, males would
77 have to mate with several females each for all breeding females to produce fertilized eggs. Together these observations provide evidence that promiscuity is more likely than polyandry within these green turtle breeding populations.
Advemtages to Multiple Insemination
Evidence for multiply inseminated clutches has been found in several reptiles (Gibson and Falls 1975; Harry and Briscoe 1988; Galbraith et al. 1991; Hoggren and Tegelstrom 1995). The ability to produce clutches with mixed paternity may be enhanced in many species of reptiles because they can store sperm for long periods of time (Devine 1984). Hypothetical advantages accruing to a female that engages in multiple matings and fertilizes her clutches with sperm from more than one male include; (1) she may obtain resources from multiple males; (2) she may mate with a high quality male if her social mate is of low quality; (3) she may increase the genetic diversity of the clutch; (4) she may insure fertilization of an entire clutch of eggs (Birkhead and MoUer 1992); or (5) she may increase sperm competition to enhance the production of offspring with higher viability (Haig and Bergstrom 1995). Because there is no evidence that marine turtles hold territories over the breeding seeison or maintain social pair bonds between mates, the first two hypotheses do not seem appropriately applied to multiple mating in femede green turtles. However, one or more of the last three
78 hypotheses may explain the evolution of a mating system that is characterized by multiple fertilizations.
Increase Genetic Diversity of Clutches There may be advantages associated with laying clutches that are genetically diverse. If the probability of persisting through time is higher for lineages that contain sufficient genetic variation to respond to changing selective pressures, then selection in the past may have favored families with several fathers. Therefore, present females are more likely to be the daughters of multiply mating mothers, and, assuming that the tendency to mate multiply is heritable, the behavioral trait of multiple mating by females spreads.
Ensure the Fertilization of Eggs Another possible advantage associated with multiple copulation is fertilization insurance. Considering the enormous energetic demands associated with migration to the rookery, and with the nesting process (Bjomdal, 1982), the production of a large proportion of unfertilized eggs within a clutch would be extremely inefficient. At the same time, each green turtle clutch is very large (averaging about 100 eggs), so the probability of fertilizing an entire clutch may increase with multiple mating events. This idea can be tested by comparing the proportion of fertilized eggs in clutches characterized by single or mixed patemify. If fertilization success is the primary advantage associated with multiple
79 mating, then clutches with single paternity should have a higher proportion of unfertilized eggs than clutches with multiple paternity. Descriptions of breeding turtles in captive and natural populations have reported that pairs tend to copulate for long periods of time (Wood and Wood, 1979: Booth and Peters, 1972; Urlich and Peters) and that females that mated for longer periods of time produced clutches with a significantly higher hatching success (Wood and Wood, 1979). These observations su rest that fertilizationof an entire clutch may require large numbers of sperm, and provides support for the idea that females that mate multiply with one or more males may have higher fertilization success than those that mate singly.
Enhance Sperm Competition A third possible explanation for the evolution of multiple mating in female green turtles is that sperm competition among the ejaculates of different males ensures that the most competitive sperm are responsible for fertilization, and this enhances the competitive ability of offspring (Knowlton and Greenwell 1984; Haig and Bergstrom 1995). Evidence supporting this idea (in the form of hatching success and yearling survival) has been found in snakes (Madsen et al. 1992; LuiseUi 1995), lizards (Olsson and Madsen 1995), and squirrels (Murie 1995).
The Relationship Between Hatching Success and Mating Strategv To examine whether the number of males represented in a female’s clutch was related to hatchling production, I used hatching success
80 data collected from 12 of the 18 clutches analyzed in this study. In clutches with multiple paternity, a significantly higher proportion of e^ s produced emerging hatchlings than in clutches with single paternity (Mann-Whitney (/ = 3; p = 0.028). In this study, unhatched e^ s were not examined for evidence of fertilization. Because a nest can contain several unhatched but partially developed embryos, the number of unhatched e^ s is not necessarily equivalent to the number of unfertilized eggs (Wyneken et al. 1988). However, if emergence success provides a relative index of fertilization success, then these hatching success data may support either the fertilization insurance or the sperm competition hypothesis despite the fact that precise calculations of the proportion of fertilized eggs for each clutch are unavailable. These results, in either case, suggest that multiple paternity is associated with a higher efficiency in the production of hatchlings. However, in the absence of information about fertilization ratios for these clutches, it was not possible to distinguish whether inter-ejaculate sperm competition or fertilization insurance best explains the evolution of multiple mating in green turtle clutches from Tortuguero and Melbourne. Although the proportion of eggs producing hatchlings was higher in multiply fertilized clutches (Mann-Whitney 17 = 3; p = 0.028), the number of hatchlings emerging from each nest type was not different (Mann-Whitney U = 21; p = 0.841) due to the slightly (but not significantly) larger number of eggs deposited in single paternity clutches (Mann-Whitney U = 39; p = 0.203). Whether multiply mating
81 females have higher annual (or lifetime) reproductive success depends on the number of clutches they lay compared to females that mate with a single male each. Thus, information about the number of clutches laid by females exhibiting both mating strategies is needed to fully evaluate the influence of multiple paternity on female reproductive success.
Comparisons with Other Populations and Species
The first attempt to characterize the mating system of a sea turtle species by analyzing patemify used allozymes to examine clutches of Australian loggerhead turtles [Çarettacaretta; Harry and Brisco 1988). The results firom this study revealed multiple paternity in 20% of the clutches examined and the authors suggested that this population may exhibit a promiscuous mating system. Multilocus minisatellite DNA fingerprinting has also been used to determine the relatedness of nestmates within three green turtle clutches firom Tortuguero (Parker et al. 1996). Half siblings were found within one clutch, providing preliminary evidence that multiple patemify was a feature of clutches firom this nesting beach. With the development of microsatellite primer sets for sea turtle species (FitzSimmons, 1995), there has been a recent increase in research seeking to characterize the mating system by analyzing paternity of clutches firom several marine turtle species (FitzSimmons et al. (in press) on Australian green turtles; Rieder et al. (in press) and
82 Dutton et al. (in press) on leatherbacks [Dermochelys coreacea), and Kichler et al. (in press) on Kemp's ridleys [Lepldochelys kempi)]. Of these studies, only the last has revealed preliminary evidence of mixed paternity at one locus.
Explanation for Population Differences in Breeding Behavior The results from our work corroborate the previous suggestion of multiple paternity in Tortuguero green turtle clutches (Parker et al. 1996) and provide an interesting contrast to the observations of single paternity in the Australian green turtle population (FitzSimmons et al. in press) One hypothesis for the differences in mating behavior between the Atlantic and the Australian nesting populations involves the relative abundance of males. In marine turtles, sex is determined by the temperature at which e ^ s incubate (Mrosovsky and Yntema, 1980; Morreale et al, 1982; Spotila et al. 1987). Therefore, the relative frequency of females and males within a population is affected by prevalent incubation conditions on the nesting beach (warmer temperatures produce females and cooler temperatures produce males). Extreme temperatures for several years on a nesting beach can result in the production of only one sex, thereby influencing the breeding sex ratio in subsequent nesting seeisons (Mrosovsky 1980). Sex ratio information is lacking for the Melbourne rookery; however, in Tortuguero, the proportion of green turtle hatchlings that were female has been estimated at 67% in 1980 (Spotila et al. 1987), 40% in 1988, and 10% in 1989 (Horikoshi, 1992). While there is
83 substantial variation in the sex ratios of hatchlings produced at this rookeiy, it is unlikely that males are in short supply; thus, females may have the opportunity to mate with several partners. The breeding sex ratio in Australia may be skewed in favor of females (Limpus, 1994), producing a situation where females may not encounter many potential mates. If searching for additional medes represents a significant energy cost to females, they may settle for only one mating event. There is evidence that female loggerhead turtles resorb clutches; this may represent a substantial energy savings during their migration back to the feeding grounds (Limpus, in press). The cost associated with searching for more males (potential reduction in the number of eggs produced) may be higher than the benefits (genetic variation, fertilization insurance or enhanced sperm competition) associated with multiple matings; thus, in Australia, female fitness may be maximized by mating with a single male. In contrast, if males are more abundant in the Tortuguero and Melbourne breeding populations, female fitness may increase with multiple matings. As information accumulates about the mating systems and sex ratios of different populations of marine turtles it wül be possible to reach a better understanding of the environmental influences on breeding behavior.
84 Consequences of Multiple Paternity
Female green turtles return faithfully to their natal beaches to nest (Meylan et al. 1990, Bowen et al. 1992, Allard et al. 1994), and can exhibit precision in within-beach philopatry (Peare and Parker in press). However, the genetic distances separating different nesting populations (measured by divergences of mtDNA sequences) are relatively small, suggesting that some currently distant breeding aggregations were connected as recently as 0.5 to 1.0 million years ago (Bowen et al. 1992). Carr ( 1978) suggested th at wandering by gravid females occurs occasionally emd can lead to the colonization of new beaches. At the same time, geological events and fluctuations in climate modify beaches and coastal areas used by green turtles; if conditions change significantly, females m ust seek new nesting areas (Bowen, 1992). Because a colonization event may be accomplished by only one, or perhaps a few gravid females, the breeding behavior of individuals that exhibit breakdowns in philopatry can influence the population genetics of a new rookery. Colonizing females that have mated with multiple males produce clutches with genetically diverse progeny. If these progeny return to breed at their natal beach they would comprise a nesting population with more genetic resources than would be expected firom progeny of a singly-sired clutch (B. Bowen, pers. comm.). The evolutionary potential of a nesting agrégation may therefore depend on the mating behavior of founding females.
85 As populations decline, breeding behavior can also play an Important role in determining the rate of loss of genetic diversity. The mating system describes the variance in reproductive success for females and males. In a population where females are not excluded from nesting by the behaviors of other females, genetic diversity can be elevated by increasing the number of males contributing to the next generation (Sugg and Chesser, 1994). True promiscuity would maximize the number of paternal genotypes that are represented in the gene pool and thereby help maintain levels of genetic variation within small populations. Therefore, the potential detrimental genetic effects of isolation due to female philopatry may be ameliorated by a mating system that features multiple paternity, and thus maximizes the preservation of paternal genotypes.
86 NOTES ON APPENDICES
Appendices A through K are protocols written for the Parker laboratory by Tigerin Peare, Julie Rieder and Kim Lundy. Appendices L through O are data sets generated for this project
87 APPENDIX A DNA EXTRACTION
E^raction isolates DNA firom blood or tissue samples. 1. Put about 500^1 of each blood sample into autoclaved Eppendorf tubes.
2. Pipet 20-30 fd of ProK into each sample and shake the tubes.
3. If blood has been stored in PBS, or if it is whole blood, pipet 30/zl of SDS into each tube and shake them again. 4. If the sample has been stored in a lysis buffer (eg. Longmire's Solution) you do not have to add SDS because it is already in the preservative.
5. If you do not know in what the blood sample was stored, add 30/zl of SDS. 6. Place samples in the incubator at 65°C for 4-12 hours. Return at legist once to shake the samples during incubation. When samples have been incubated for a sufficient period of time, they should be red-brown in color, and less viscous than the original sample. 7. Once the samples have been incubated, take them to the hood to perform the extractions
8. Get phenol from the glass door refrigerator, and set the pump volume at 0.5ml (OR to the equivalent volume of the sample you are extracting, so use onty 0.3ml if your sample is only that size). Lift and push down the column a few times until bubbles have been removed from the dispenser tube.
9. Squirt phenol into each sample, and shake each sample VIGOROUSLY for at least 30 seconds, or until the phenol is mixed well into the sample.
88 10. Put the samples into the centrifuge, making sure that the tubes are balanced both In number and In sample volume, and spin them down for 5 minutes.
11. When centrifuging is complete remove tubes and return to the hood. 12. Using disposable plastic transfer pipettes, suck out the TOP (clear) portion of the sample and squirt It Into a new. appropriately labeled tube. 13. Dump the rest of the sample (I.e., the darker bottom part) into the bottie labeled "WASTE".
14. If necessary, repeat steps 9-13. Otherwise, go on to step 15.
15. After the samples have gone through two phenol extractions, they are ready for phenol/CIA extractions. 16. To the extracted (clear) part of each sample, add 0.25 ml phenol, and 0.25ml CIA.
17. Shake each sample vigorously, then centrifuge again for 5 m inutes. 18. Draw off the top of each sample and squirt It Into a new tube. Dump the remaining portion Into the "WASTE" bottle.
19. Repeat steps 16-18 with your samples 20. Once you have completed the phenol and phenol/CIA extractions, do one final extraction with CIA alone to clean up the sample. 21. Add 0.5 ml CIA to each sample, shake and centrifuge for 5 minutes.
NOTE: During this extraction step, the layers separate very quickly.
22. Suck off the top layer of solution for each sample (put them In new tubes), and discard the bottom layer. The samples are now ready for dialysis or ethanol precipitation.
89 APPENDIX B
DIALYSIS OF EXTRACTED SAMPLES
Dialysis serves to clean up the DNA samples by removing protein and any remaining traces of phenol from the extraction process. 1. Dialysis tubes are stored in 70% ethanol in the refiigerator. Take 12 (or however many samples you have) numbered tubes out of the container using clean gloves, emd rinse them thoroughly with distilled water several times. It is important to rinse off all of the ethanol, and to keep the tubes wet at all times.
2. From the drawer to the left of the sink, get 12 (or however many samples you have) orange dialysis clips numbered 1-12 and 12 clips without numbers.
3. Place a non-numbered clip on one end of each piece of tubing and put them in a small beaker of distilled water. (If there are not enough clips for two per sample,then carefully tie a tight knot in one end of each of the dialysis tubes) tubes MUST be kept submerged in water, so Just allow them to soak until you are ready to use them). 4. Prepare the dialysis solution by mixing 50ml of 40X TNE2 and 1950 ml distilled water in a large (3500ml) beaker.
5. For each sample, take one diafysis tube firom the beaker, and squeeze any excess water out of it. Using a disposable pipette, draw the organic phase (top part) off of the sample and squirt it into the dialysis tubing.
6. Using an orange clip number that matches the sample number, close ofif the top of the dialysis tube. Leave a very small bubble of air in the top of the tube.
7. Put the 12 samples and a large stir-bar into the beaker of dialysis solution, and place It on a stir-plate in the glass-door refirigerator. Set the stfir-plate to 2 or 3.
8. Over the course of 4 to 12 hours, replace the old dialysis solution with firesh solution 1-3 times, first replacing the solution after 1.5 -
90 2 hours.
9. Once dialysis is complete, prepare 12 Eppendorf tubes with the original identification codes of the samples you have just extracted. 10. Transfer each sample from the dialysis tube to its appropriate Eppendorf tube in the following manner:
—Hold the sample in the dialysis tube so that the orange clip is positioned up. then, using a razor blade, carefully cut off the top part of the dialysis tube right below the orange clip, (or carefully open the clip) and open the tubing without squeezing out any of the sample. —Turn an open Eppendorf tube upside-down, and place it over the top of the cut dialysis tube —Turn the dialysis tube upside-down, allowing the DNA sample to dreiin into the Eppendorf tube, and then squeeze any remaining sample from the dialysis tube.
NOTE: If you are not comfortable with this method, you can use a disposable pipet to draw out the sample and transfer it to the eppendorf tube. Be sure to get all of the sample from the dialysis tubing into the sample tube.
11. The samples are now extracted and clean, and ready to read on the spectrophotometer.
91 APPENDIX C: READING SAMPLES ON THE SPECTROPHOTOMETER
Samples must be read to determine their concentrations of DNA. 1. Turn on the spec, (ask how) and allow it to warm up for 15-30 minutes. 2. While the spec, is warming up, prepare samples for reading:
—Fill 12 labeled tubes with I ml distilled water each. —Add 25 fi\ from each DNA sample to the appropriate tube of water.
Note: If the volume of any of your samples is low (below 500//l), you will have to use a smaller amount. Check the notecard above the spec or ask someone. —Shake tubes very well, or vortex them. 3. Using a disposable plastic transfer pipet, put each sample into a separate CLEAN cuvette.
4. Prepare one cuvette that holds only distilled water. Use this one to zero the spec, at each setting.
5. Set the wavelength to 260 nm (the wavelength that reads the concentration of DNA) and put the water cuvette into the cuvette slot.
6. Using the dial at the bottom left side on the front of the spec., set the reading at 0.000.
7. Take out the water cuvette, and read each of your samples, one at a time, making sure to record these values on your extraction form.
8. After making the 260 readings, reset the spec, for 280 nm readings (for concentrations of proteins), by turning the lightwave knob to 280, and re-zeroing the machine with the water cuvette. Record the reading for each sample.
9. After taking the 260 and 280 reading for each sample, dump it out,
92 rinse the cuvettes with water, and repeat the process for the next set of samples.
10. Calculate the concentration of DNA for each of your samples (260 reading X 2) and the yield (concentration X volume), and record these on your extraction form.
Note: You will have to use a different formula if you used a sample smaller than 25/fl for your reading. Check the card above the machine or ask so m eone.
11. Rinse the cuvettes several times with CIH2 O, then 70% ethanol, and place them on a paper towel to drain.
93 APPENDIX D
RUNNING A MINIGEL
A minigel is done to determine whether any breakage has occured in the DNA sample, and to test that concentration estim ates are accurate. 1. Take all samples out of the fridge at least 1 hour before they will be used, and gently shake them to ensure that the samples are relatively homogeneous.
2. Prepare a 0.8% agarose gel following the recipe card above the pH meter. —You should use the smallest gel tray possible for the number of samples you are running. This usually requires only a 100 ml gel, so you would measure 0.80 g of agarose, place it into a small (250 ml) Erlenmeyer flask, and add 100 ml of IX TBE (Be sure you use IX, not lOX.). 3. Put a small stir-bar into the flask, and cover the top with foil. Place the flask on a heater/ stir-plate and set the heater on 5 and the stirrer on 3. The agarose will take several minutes to melt into solution. 4. Once the agarose has COMPLETELY MELTED (solution is clear and there are no flecks of things suspended in the liquid, usually just before it boils), put the flask on a cold stir-plate in the deli flridge and let it cool down to 50°C. This will take 10-20 m inutes. Do not allow the gel to cool below 50“C!! 5. While the agarose is cooling down, prepare a gel tray. Minigels are run in the smaller trays. Use 70% alcohol to cleeui the tray, then tape up the open sides, and carefully place 1 to 4 rows of teeth in the tray (whatever you need for the gel size and sample number).
6. When the agarose has cooled down to 50°C, slowly pour it into the gel tray. If there are any bubbles or bits of dust in the liquid, suck them out with a disposable plastic transfer pipet. Allow the agarose to set (at least 30 minutes). 7. While the agarose is setting, get a "form K" or some other data
94 sheet, and write in the following column headings:
ID ext. date conc. use forf i g1 B. J. II dH20 Total
"ID " is the sample identification code ""Ex. date"' is the date each sample was extracted "Cone. " is the concentration of each sample "Use for 1 fig" is the target amount of each sample needed for each lane. This is calculated by dividing the target amount by the concentration of the sample (e.g., if a sample has a concentration of 0.39 fig/fd, the quantity you should write down in the "use for 1 fig" column is 1/0.4 = 2.5 fj\). "B.J. n" is Blue Juice II, and refers to the dye used in the minigel. This should be between 10% and 20% of the total volume for each lane (e.g., if the total volume for each lane is 20,^1, the amount of Blue Juice H needed is at least 2 fi\). ""dH20" is the distilled water, and is calculated by subtracting the ""use for 1/fg" and the "B.J. 11" am ounts from the "Total" (e.g., using the above calculations, "dH20"" would be 20 - (2.5 - 2) = 15.5 fi\). "Total" is tiie am ount of solution (DNA + water + dye) th a t will be put into each well of the minigel. This is usually 20 fi\ for minigels. 8. Once everything has been calculated and written down on the form, you can begin to measure out the components into autoclaved Eppendorf tubes in the following order: Water, DNA, Blue Juice n. 9. Spin down the samples in the centrifuge (use the "momentary" button) to get the dye into the rest of the solution, smd heat the samples in the incubator for 5 minutes at 65oC.
10. While the samples are being heated, prepare the gel for loading by removing the row of teeth from the gel (rock it carefully several times, and gently pull it out). 11. Without touching the gel, itself, take off the tape from both ends, then slide the gel tray into a gel box with approximately 1 liter of 1 X TBE (The TBE should ju st cover the top of the gel). 12. Get the heated samples from the incubator, and flick them down in the centrifuge. They are now ready to load into the gel.
13. Set an Oxford pipetor for 20 fil and carefully load each well.
NOTE: If the pipette tip is too deep in the well, it may puncture the bottom of the gel; if it is not deep enough, the sample may float out. Get ALL of the
95 sam ple into the well.
18. After all of the samples have been loaded, place the lid on the box and attach the electrodes to the box and the power source. Be sure that the red wire is plugged into the positive outlet on the power source and to the post nearest to the bottom of the gel box (remember to "run to red"). The black wire should be attached to the negative outlet and the post near the top of the gel box.
19. Set the voltage at 80. and allow the gel to run for approximately 1-2 hours.
20. Turn off the power source, unplug the electrodes, and carefully remove the gel tray from the box.
21. Place the vat of ethidium bromide solution on Big BiU. Be extremely careful when using this solution as it is a very powerful mutagen. Carefully slide the gel off of the tray into the ethidium bromide staining solution. Replace the foil cover on the pan, set Big BiU for slow rotation (about 50), and aUow the gel to stain in the solution for approximately 30 minutes.
—Check on your gel every few minutes, as the rotator plate tends to speed up the longer it runs. 22. Once the gel is stained, slide it back onto the gel tray and transfer it to the light box. Close the lid, turn the box on and wait for it to zero. Press the "M" button, look to be sure the DNA is visible, then take a picture (ask for instructions).
96 APPENDIX E
RUNNING A DIGESTION GEL
In a digestion gel, DNA samples are cut u[ith an enzyme at specific base sequences, and the resulting fragments are separated along an electrical gradient in an agarose gel.
1. Take all samples out of the fridge at least 1 hour before they will be used, shake them gently to ensure that the samples are relatively homogeneous, then flick them down in the centrifuge.
2. Prepare the gel by following the recipe card above the pH meter. We use 0.8% agarose gels. Usually, you will use a total volume of 200 ml (17 lanes) or 300 ml (26 lane). If your total volumes are higher thcui about 50/fl, you will need a larger gel to ensure that the wells will be deep enough.
3. Put a small stir-bar into the flask, and cover the top with foü. Place the flask on a heater/stir-plate and set the heater on 6 and the stirrer on 2-3. The agarose will take several minutes to melt into solution. 4. Once the agarose has COMPLETELY MELTED (solution is clear and there are no flecks of things suspended in the liquid), put the flask on another stir-plate and let it cool down to 50°C. This will take 10- 20 minutes.
5. While the agarose is cooling down, prepare a gel tray. Use 70% alcohol to clean the tray, then tape up the open sides, and carefully place 1 row of teeth in the tray. 6. When the agarose has cooled down to exactly 50°C, slowly pour it into the gel tray. If there are any bubbles or bits of dust in the liquid, suck them out with a disposable plastic transfer pipet. Allow the agarose to set (at least 30 minutes).
7. While the agarose is setting, get a "form K" or some other data sheet, and write in the following column headings:
ID ext. date conc. use for 5 jug Hae m Buffer B.J. H dH20 Total
"ID" is the sample identification code "Ex. date" is the date each sample was extracted
97 "Conc." is the concentration of each sample "Use for 5 fig” is the target amount of each sample needed for each lane. This is calculated by dividing the target amount by the concentration of the sample (e.g.. if a sample has a concentration of 0.39 [ig/nl, the quantity you should write down in the "use for 5 //g" column is 5/0.4 = 12.5 fi\). If any one of the samples is very dilute, you will need to adjust your target values for all of the samples. "Hae m" is the most common restriction en^mie used. 2 fi\ is put into each tube. "Buffer" m ust be 10% of the total volume. "B.J. H" is Blue Juice n. eind refers to the dye used in the minigel. This should be between 10% and 20% of the total volume for each lane (e.g., if the total volume for each lane is 40/zl, the amount of Blue Juice n needed is at least 4 fi\). "dH20" is the distilled water, and is calculated by subtracting the "use for 5/6g", the "Hoe IH", the "Buffer" and the "B.J. II" amounts from the "Total" (e.g., using the above calculations, "dH20" would be 40 - (12.5 + 2 + 4 + 4) = 17.5 /^l) "Total" is the am ount of solution (DNA + enzyme + buffer + water + dye) th at will be put into each well of the minigel. This can be anywhere from 40 fi\ to 65 fi\ for a digestion gel. 8. Once everything has been calculated and written down on the form, you can begin to measure out the components into autclaved Eppendorf tubes in the following order: Water, buffer. DNA, and Hae m.
NOTE: The Blue Juice is NOT added until just prior to loading.
9. Spin down the samples in the centrifuge (use the "momentary" button) to get the enzyme into the rest of the solution. 10. Heat the samples in the incubator for 3 hours at 37°C. 11. When the samples have finished incubating, remove them from the heating block and spin them down in the centrifuge.
12. Add the Blue Juice to each of the samples, and spin them down again to get the dye into the solution. 13. Place the samples back in the incubator for 5-10 minutes at 65°C.
14. While the samples are being heated, prepare the gel for loading by removing the row of teeth from the gel (rock it carefully several
98 times, and gently pull it out).
15. Without touching the gel, itself, take off the tape from both ends, then place the gel tray in a gel box with approximately 2 liters of 1 X TBE (The TBE should ju st cover the top of the gel).
16. Get the heated samples from the incubator, and flick them down in the centrifuge. They are now ready to load into the gel. 17. Set an Oxford pipetor for the total volume and carefully load each well. NOTE: If the pipette tip is too deep in the well, it may puncture the bottom of the gel; if it is not deep enough, the sample may float out. Get ALL of the sample into the well.
18. After all of the samples have been loaded, place the lid on the box and attach the electrodes to the box and the power source. Be sure that the red wire is plugged into the positive outlet on the power source and to the post nearest to the bottom of the gel box (remember to "run to red"). The black wire should be attached to the negative outlet and the post near the top of the gel box. 19. Set the voltage at 20, and allow the gel to run for 65 hours. 20. After 65 hours, turn off the power source, unplug the electrodes, and carefully remove the gel tray from the box. 21. Place the vat of ethidium bromide solution on a mechanical rotator. Be extremely careful when using this solution as it is a very powerful mutagen. Carefully slide the gel off of the tray into the ethidium bromide staining solution. Replace the foil cover on the pan, set rotator for slow stirring (about 50), and allow the gel to stain in the solution for approximately 30 minutes. 22. Once the gel is stained, slide it back onto the gel tray and transfer it to the light box. Close the lid, turn the box on and wait for it to zero. Press the "M" button, look to be sure the DNA is visible, then take a picture (ask for instructions).
99 APPENDIX F
SOUTHERN TRANSFER
Once the digestion gel has been stained with ethidium bromide and photographed, it is ready for Southern Transfer. In this step, the DNA is denatured, neutralized and transfered to a stable nylon m em brane.
1. Slide the gel oflF of the light box onto a gel tray, then transfer it to a pyrex dish (large enough to allow the gel to lie flat) with enough Wash # 1 (denaturing wash) to just cover the top of the gel. Put the dish onto the rotator and let it rotate slowly for 15 minutes. 2. After 15 minutes, carefully drain off the wash solution into the sink, and replace it with fresh Wash # 1 solution. Let it rotate for another 15 minutes.
3. Pour off the Wash #1 solution and rinse the gel with distilled water for about a minute.
4 . Pour off the water and replace it with Wash #2(neutralizing wash); let the gel rotate slowly on the rotator for 30 minutes.
5. Drain off the old Wash #2. replace it with more Wash #2. and put it on Big Bill for 30 minutes. 6. While the gel is going through the washes, cut Watman paper and nylon membrane for the Southern Transfer. You will need:
"One piece of Watman paper that is the size of the Southern Trsmsfer platform plus about 2.5 inches on two opposite sides —Two pieces of Watman paper that are the size of the Southern Transfer platform —One piece of nylon that is the same size as the gel —Two pieces of Watman paper that are the same size as the gel —Two even stacks of brown single fold paper towels
7. Take the long piece of Watman paper, and fold down the sides at 2.5 inches, or so that it sits flat on the gel platform with the tabs hanging down into the buffer.
8. Place this piece on the Southern Transfer platform, pour some lOX SSC on it. and roll out the air bubbles and any excess solution by
1 0 0 gently rolling a 10ml pipette across the paper using only slight pressure. Be careful not to wrinkle or make the paper rough. This paper serves as a wick. 9. Place the two papers that are the size of the platform on top of the first piece, wet them with 10 X SSC and roll out air bubbles and excess solution. Besure that there are no air bubbles between the pieces of paper.
10. Slide the denatured, neutralized gel onto the platform, square i t , and gently push out any air bubbles which may have formed between the gel and the paper (this is VERY important). Then place the nylon membrane on top of the gel, making sure to remove any of the air bubbles by gently pushing them out to the edges using. 11. Take the two gel-sized pieces of Watman paper and put them on top of the nylon, wet them with lOX SSC and roll out the bubbles and excess solution.
12. To prevent evaporation of the buffer and to form a barrier between the paper above and below the gel, cover the buffer tank with plastic wrap from the edges of the dish up the the edges of the gel. It is very im portant to have the plastic touching each entire edge of the gel, but not covering any of it. This is to ensure that all of the solution, and therefore the DNA, goes through the nylon membrane. Then place the two even piles of paper towels on the top of the gel + paper pile. 13. Get a pyrex dish and put it on top of the piles of paper towels. Place a weight (a bottle of some solution) on the pyrex dish, and allow the Southern Tremsfer to go overnight. 14. The next day, peel off the paper towels until the top layer of Watman paper is exposed. 15. Pick up the paper + nylon + gel + paper "sandwich" (down to the wick paper) and turn the whole tWng upside-down.
16. Peel off the pieces of paper that are now on top, so that the gel is exposed. 17. With a pencil, draw horizontal lines on the nylon membrane where the outlines of the gel wells are. 18. Peel off the gel (which is now much thinner) eind on the top of the
1 0 1 nylon membrane, write the gel ID code and mark leme 1 (which will be on the top right side now)
19. Rinse the nylon in 2X SSC. blot it dry, and place it in the drawer of the Bioslink cross-linker. Set the machine to read 00.30 J/cm2 by pressing the appropriate double arrow button (third from the left) three times.
20. When the nylon is in the drawer (DNA side up) and the Bioslink is set, press the "enter" button and wait for the machine to count down to 00.00.
21. Take the nylon out of the Bioslink and put it into an envelope made of clean paper In the hot-room cabinet. It is now ready for hybridization. Dry nylon filters in the envelope may be stored indefinitely at room temperature before hybridization.
102 APPENDIX G
HYBRIDIZATION FOR MINISATELLITES USING JEFFREY'S PROBES 33.6 OR 33.15
Before doing a hybridization, you must be certified by radiation safety and authorized by Dr. Parker.
Hybridizations may be done on new filters following Southern transfer or on previously hybridized filters after the old probe is stripped off.
PREPARATION: 1. First thing in the morning, preheat the hybridization oven to 62°C, remove the Sephadex solutions (buffer without beads and with beads) firom the refirigerator so they may reach room temperature and turn on the heat block in the hood. It should be set at 37°C, so don’t touch the control, just check the temperature. You should £ilso label an autoclaved eppendorf tube for the probe you are using, and label a large centrifuge tube (10 or 15 ml, blue-capped ones) with a “P” on the lid. It is often useful to mark some of the gradations on this tube, as it will be used to collect the probe and to estimate the concentration for dilution.
2. Prepare the pre-hyb/ hybridization solution (1.5% SSC, 0.1% SDS, 5X Denheurdts, 6% w/v dextran sulfate). —Follow the recipe card above the scale, making enough solution for 20ml - 25ml per tube. You can put up to 3 (maybe 4) filters in each tube. —Add these ingredients to a small Erlenmeyer flask with a stir-bar in the following order: 1. dextran sulfate 2. dH 20 3. SSC 4. SDS
3. Place the flask on a stir plate without heat, allow the dextrem sulfate to dissolve, and then add the Denhardt’s solution. It is stored firozen and should be thawed only enough to remove the desired amount, then immediately returned to the fireezer. 4. About cm hour or so before you will do the labeling reaction, take the radiation out of the fireezer, take the pig out of the yellow container if it’s still in one, and set it in the hood behind the shield.
103 STRIPPING PREVIOUSLY HYBRIDIZED FILTERS: If some of the filters have been previously hybridized, the old probe must be stripped off before rehybridizing. If all of the filters are new, you can skip this part.Strip wash should be prepared ahead of time ( 1970ml dH20, 10ml 20X SSC, 20ml 10% SDS). Each filter should be stripped at least twice, and the solution must be lOC’C when it is poured over the filters. 1. Heat 500ml of strip wEish in an Erlenmeyer flask using a microwave or hot plate. It is often convenient to heat more than one flask at a time.
2. While the solution heats (about 5 mins.) unwrap the firozen filters and lay them in a single layer in pyrex dishes, DNA side up. Keep them covered with the plastic wrap; lay itright onto the filter. It is important that the filers do not get dry. Place only 2 or 3 in a dish, overlapping as little as possible if at all. DO NOT stack the filters as the boiling wsish must come in contact with the entire DNA side of the fiilter.
3. When the wash starts boiling, check the temperature. If it is 100°C, pour it directly onto the filters in one dish (remove the plastic first). If needed use a glass rod to gently move the filters around so that they do not overlap. Once again, it is absolutely necessary for the solution to be 100®C when it is poured or the old probe will not be completely removed. 4. While the filters are cooling in the first wash, heat the second batch. You can actually begin heating the second batch before the first boils, as it takes several minutes. Let the filters in solution cool to about 60° or 50°C before pouring the second wash. 5. When the second wash is 100°C, carefully pour the first wash off into the hot sink and repeat the wash procedure. The second wash must also be 100°C when it is poured. 6. If the filters had very bad background problems on the previous hybridization, you may want to repeat the stripping procedure a third time. It is also useful to strip the mesh spacers from time to time, especially if you are having background problems. 7. Let the filters cool in the strip wash until they are put into the hybridization solution PRE HYBRIDIZATION:
104 During prehybridization, the non-DNA areas of the nylon are blocked so that the probe only hybridizes to its recognition sequence in the DNA and not to the nylon.
1. Do the following in a pyrex dish:
—Wet a nylon spacer with a small amount of 1.5X SSC —Place a nylon filter, DNA side up, on the mesh and center it. —Place another spacer on top of the filter, covering the entire filter with the spacer, and push out emy air bubbles. Use a little SSC if necessary. —Place another filter and spacer and filter on this etc., until you have assembled all of the filters for the first tube. —Push out any remaining air bubbles, carefully roll up the mesh/filter stack and put it into a clean hybridization tube.
2. Put the cap on the tube and roll it on the table to unroll the filter stack so that it adheres to the inside of the tube.
3. Add the prehybe. solution and place the tube in the hybridization oven. Add about 18-22ml of solution to each tube; reserve about 5ml for diluting the probe. Balance the first tube with an empty one and turn on the rotor,
4. Repeat the procedure for each tube. Be sure to balance the tubes on the rotor. Check to see that the filters are not rolling up in the tubes causing them to pull off of the sides of the tubes. The easiest way to avoid this is to look at the tube as you put it on the rotor. The outside edge of the mesh/filters should be going down
5. Allow the filters to prehybridize for 3 - 5 hours.
LABELING REACTION: Here, copies of the probe DNA sequence are radiolabeled by primer extension with 32P dCTP (32P-labeled deoxycytosine-TP).
NOTE: Follow the directions for the kit you are using. Check the hood sash for the appropriate volume of aqueous probe to add to the labeling reactions. (Both kits require SO^ug of template DNA, but our calculations overshoot this amount slightly.)
105 Directions for Amersham random primer labeling kit: 1. Dilute the probe: —Dilute probe in dH20 to a total volume of 28 ;d in an autoclaved microcentrifuge tube. —Heat at 100“C (in a beaker of boiling water) 5 minutes, then chill on ice. 2. Set up the labelling reaction by adding the following to the denatured probe (keep it on ice): — 10/d buffer (Solution 1. blue cap) —5/d primer (Sol’n 2. black cap) -Add 5/d 32P dCTP —2/d (Sol’n 3, red Üd) 3. Incubate at 37“C for 30 minutes, then separate in sephadex column. Directions for Boehringer Mannheim random primer labeling kit:
1. Dilute the probe: —Dilute the probe with dH20 to a total volume of 9/d in an autoclaved microcentrifuge tube. —Heat at 100°C (in a beaker of boiling water) for 10 minutes, then chill on ice. 2. Set up the labelling by adding the following to the denatured probe (keep it on ice): —3/zlnucleotide mixture —2/dreaction mixture (tube #6) -5/fl 32P dCTP — 1/d enzyme (tube #7) 3. Incubate at 37®C for 30 minutes, then separate in sephadex column.
COLLECTION OF THE LABELED PROBE USING A SEPHADEX COLUMN:
The column is used to separate the probe from unincorporated nucleotides. The sephadex beads contain tiny holes in which the unincorporated nucleotides get stuck, reducing competition for the probe which then moves more quickly through the column allowing it to be isolated and collected separately from the nucleotides.
Alternatively, we have begun using Pharmacia spin-columns to separate the probe from the unincorporated nucleotides. If you are using a spin
106 column, carefully follow the instuctions provided by the manufacturer.
Building the sephadex column:
1. While the probe is incubating, build the sephadex column.Use a disposable 5ml pipette. If the top is not straight, break it off just below the constricted part. Put a very small piece os glass wool in the pipette and use a glass to gently push the wool down to the tip. Place a short about 2 - 3cm) piece of tygon tubing on the tip of the pipette and clamp it.
2. Secure the pipette to the ringstand in the hood and place the waste beaker under the tubing. Be sure that the standing straight.
3. Gently mix the sephadex bead solution. Don't shake it. Be sure that it remains mixed throughout the process. 4. Using a disposable plastic pipette, squirt some buffer solution (no beads) into the pipette. Add about 2ml or so.
5. Use a separate pipette for the bead solution, and begin adding this solution to the column, using the buffer to keep the beads wet and to wash them down the sides of the pipette. Be sure to keep the beads wet with buffer at all times. Do not allow air bubbles or cracks to form in the column or it wiU not work.
6. When the column is about 3/4 of the way up the pipette, add some buffer and let the beads settle a bit so that you can see a clear line at the top of the column and then the buffer head. Continue to add small amounts of beads and buffer altematefy, still keeping the beads wet, until the column is a few centimeters from the top of the pipette. If you have a big buffer head, unclamp the tubing and allow the column to pack down as the buffer runs through, then keep adding beads and buffer alternately.
7. When thec o lu m n is about 1 - 1.5cm fromthe top (let the beads settle so you can see the clear buffer head), unclamp the tubing and keep adding buffer to keep the beads wet. If the column packs down, add a little more beads and buffer so that the top of the column is about 1cm from the top of the pipette. 8. At this point. 30 minutes has probably passed and it is time to denature the probe, so clamp the tubing, and add a few drops of buffer to keep the column wet. Collection and dilution of the probe:
107 1. Remove the probe from the heat block and quickly flick it down in the little centrifuge 2. Set the appropriate eppendorf pipette to the total volume of the probe solution (SO/il for Amersham kit, 20fi\ for B.M. kit). You will use this to deliver the probe onto the top of the column.
3. Unclamp the tubing on the column and let the buffer fall just to the top of the column and add some more buffer. 4. Draw the probe solution out of the tube and deliver it to the top of the column just as the buffer drops to the level of the column. Add a small drop of buffer to keep the beads wet and to help get the probe into the column. Keep adding small drops of buffer, just enough to keep the column wet, while using a survey meter to check that the probe has gone into the column. 5. Once the probe solution has gotten completely into the column, start adding larger amounts of buffer. You must continually add buffer throughout the separation process to keep the probe moving through the column. NEVER let the column get dry. 6. Continue using the survey meter to keep track of the “hot" band as it moves down the column. The band wiU get wider and then you wiU begin to see and hear two peaks with the survey meter as the unincorprated nucleotides becom trapped by the beads and labelled probe moves through the colurrm. 7. The bands wül get farther apart and the two peaks will become clear. You can use the suvey meter to estimate the percentage of nucleotides incorporated into the probe. This is usuaSy higher for 33.15.
8. When the probe is at the tip of the pipette, just before the tubing, place the collection tube (the large, marked centrifuge tube) under the tubing and collect the probe. Use the survey meter to assess when the probe has passed from the pipette, and switch to the waste tube, as you do not want to collect any of the unincorporated nucleotides with the probe.
9. Denature the probe by placing it in a beaker of boiling water for five minutes. Quickly flick it down in the centrifuge and immediately place it in a beaker of ice.
108 10. Use the reserved hybridization solution to dilute the probe to a convenient volume, usuaify about 2ml per tube. DENATURING THE PROBE AND LOADING IT INTO HYBRIDIZATION TUBES:
Whichever method of separation is used, the probe must be denatured and diluted before adding it to the filter tubes.
1. Denature the probe by placing it in a beaker of boiling water for five minutes. Quickly flick it down in the centrifuge and immediately place it in a beaker of ice. Once the probe is denatured, it must be kept on ice!
2. Use the reserved hybridization solution to dilute the probe to a convenient volume, usually about 2ml per tube. If you used the Pharmacia spin-column, you can measure out the appropriate amount of hybe solution into a large tube, put that tube on ice, and add the probe to it. Still keep everything on ice.
3. Remove the tubes firom the oven one at a time and place them in the tube rack. DO NOT dump the prehybe solution, as it is used for the hybridization also. 4. Use a disposable plastic pipette to deliver an equal amount of the probe solution (about 2 ml) to each tube. Carefully squirt the probe down the center of the tube so that It goes into the solution at the bottom and not directly onto the filters. 5. Replace the cap and put the tube back in the oven. Repeat for each tube.
6. Check to be sure that none of the tubes are leaking and that none of the mesh/filters are rolling up in the tube. 7. Allow the the filters to hybridize overnight.
CLEAN UP AND HOT SINK RECORDS: This part is very important to ensure that nothing is left contaminated emd that all waste is redorded and disposed of properly. 1. Discard solutions and rinse liquid waste containers in the hot sink.
2. Rinse the tube racks and wipe the Incite shield with ethanol and a paper towel.
109 3. Discard all solid waste, including gloves, towels and the entire column, into the hot waste container. 4. Check with a survey meter to be sure that the oven handle, etc. is clean. You can use ethanol and paper towels to wipe it off.
5. This may seem like overkill, but it is a regulation that nothing hot be put into the regular waste stream, and it is good to check with the survey meter just in case a little drop of something was spilled or splashed.
6. FÜl out the hot sink record properly according to the amount of incorporation and probe used. This is very important. Also report any spills you have, like if some hybe solution splashes in the hood.
7. Record the amount and estimate the activity using the survey meter, then tell the person in charge of radiation, so they check the hood, etc. When dry, change the paper in the hood.
110 APPENDK H WASHING FILTERS AND PRODUCING AUTORADIOGRAMS
After hybridization is done (the next morning) the filters must be washed, wrapped, and loaded into a cassette with film to produce the autorad, 1. Remove the tubes firom the oven and place in the tube rack. Pour off the hybridization solution into the hot sink and fill the tube about 1/3 full with Jeffireys probe wash which has been preheated to 50°- 55°C. 2. Return tubes to the oven, turn on the rotor again, and allow the filters to wash at 62°C for 30 minutes. 3. Repeat steps 1 and 2 three more times for a total of four 30 minute washes. 4. At the end of the fourth wash, reverse the direction of each tube in the oven so that the filters roll up and pull off the sides of the tube. This will make them eeisier to remove.
5. Remove the tubes one at a time (put in an empty tube if necessary to keep the oven balanced), and pour off the wash solution into the hot sink. Dump the filters out into a pyrex dish. 6. One at a time, place the filters on lab bench soaker paper (hood paper). Allow the filters to dry slightly and wrap them with plastic wrap while still damp. Do not let the filters dry out and do not to leave any air bubbles in the wrapping. Check each filter with the survey meter and decide how long to expose the film (there is not reeüly a rule here, it Just takes some practice). 7. Tape each wrapped filter to the cardboard in the film cassette. Now go to the darkroom and load each cassette with a piece of x-ray film (bend the upper left comer to be able to orient it later) on top of the wrapped filter, then a screen on top of the film. Close the cassettes
111 before turning on the light! It is best to avoid exposing the top of the cassettes to light because they will leak some. It usually helps to place them in a box so the first to be developed are on top. Just put the whole box in the fireezer. 8. Place the cassettes in the box in the -20®C fi’eezer until the film is developed. Do not allow the wrapped filters, or loaded cassettes to sit at room temperature, as condensation will develop which wÜl cause black spots on the autorad. 9. When the film h£is been exposed for the appropriate time, take the cassettes to the x-ray developer in the Bio. Sci. building. Carry the cassettes in a box or other fight-proof container. It is still important not to allow the cassettes to be exposed to fight, especieilly sunlight, and not to let them sit out of the fireezer for too long before developing. 10. After the film is developed, mark each autorad with its unique ID, then label lane 1, the date, probe used and how long the film was exposed, and “with screen” or “no screen”. Remove each of the wrapped filters firom the cardboard, seal it in plastic bag (don’t put them all in one bag) marked with the ID, probe and “not stripped” (or “stripped” if you immediate^ remove the probe and do not re-hybe the filter). Place the filters in the -20°C fireezer in the appropriate slot. Frozen filters will keep indefinitely and can then be stripped and reprobed later. 11. Allow the cassettes to dry completely, then close them and put them away in the drawer.
112 APPENDIX I PCR REACTIONS
MATERIALS:
Perkin Elmer PCR Machine (thermal cycler) gloves autoclaved Eppendorf tubes (0.5ml and 1ml tubes) fine black Sharpie pipettors and tips Eppendorf trays m ineral oil dH 20 bufier dNTFs primer sets Taq DNA polymerase MgC12 tray of ice
PROGRAMMING MACHINE:
Program machine prior to mixing reaction. Decide on denaturing, annealing and extension temperatures and create four linked files. See manual for details. 1. initial denaturing cycle 2. general program consisting of about 30 cycles of denaturing, annealing and extension 3. long extension soak cycle 4. fined soak cycle that maintains samples at 4°C (acts as a refirigerator)
113 DILUTE SAMPLES:
Dilute extracted samples to 10ng//d concentrations using UV sterilized TE. Calculation: (concentration of sample x 1000) - 10 = # of fi\ of TE added to lO^al of sample to get 10ng//d sample.
NOTE: Diluting samples may not be necessary. We have had some luck with full strength samples.
REACTION: The reaction consists of a series of components (some of these components need to be diluted from stock solutions with UVed TE):
COMPONENT CONC GENERAL______EXAMPLE d H 2 0 bring reaction to vol. L d p l lOxbuffer 10% of reaction 1.5/fl dNTP’s ImM each base 10% of reaction 1.5/fl primer 1 5/fM 5% of reaction 0.75/fl p rim er 2 5/fM 5% of reaction 0.7 5 /d Taq 5U/fil 0.1/fl per sample O .lpl MgC12 lOniM varies 4.6/fl template DNA 10ng//d 20-60 ng 4.8/fl
NOTE: For trial reactions, use 15^1 total volumes for reactions. Scale components to this volume. For optimization, adjust MgC12, am ount of template and temperature cycles of PCR reaction. Decreasing MgC12 or raising the temperature of a reaction usually increases the specificity of a reaction. Using a step-down program for temperature cycle will also increase specificity (file #83 is an example of a step-down program). And remember, an optimized reaction may be different for various primer sets and species. Prior to a reaction, you may also place all plastic materials (pipettors, pipette tips, Eppendorf tubes and racks) and dH20 that will be used for the reaction in the UV box for sterilization. Although for microsatellites, this does not seem to be necessary.
1. Turn machine on to allow it to warm up. 2. Remove lOxbuffer, MgC12 , and dNTP’s from freezer to thaw and samples and primers from the refrigerator to come to room temperature.
3. Decide on a reaction recipe for a single sample. Multiply this by the number of samples you need to amplify and add one for a negative control and/or error measure. 4. While wearing gloves, set up a series of 0.5ml autoclaved Eppendorf tubes, one for each sample. 5. Before labeling, drop a single drop of mineral oil into the bottom of
114 each tube and cap each tube. 6. Label each tube with a black Sharpie with the date and the ID of the sample.
7. Add template to each tube. Set aside these prepared tubes. SiASTER TUBE:
1. In a I ml Eppendorf tube, mix the master reaction. Add all of the components except Taq and template DNA. Shake this reaction several times to ensure thorough mixing. Flick on centrifuge and set on ice.
2. Remove Taq from freezer and keep on ice. (You wiU need a tray of ice to keep all of your samples cool.) Measure out the appropriate volume of Taq and dispense into the master tube stirring continuously. Flush pipette to be sure all of the Taq has been transferred to the m aster mix.
NOTE: Taq is both “sticky” and heavier than the master mix so careful mixing is required here. Do not shake the tube at this time because Taq will create bubbles and this will make it difficult to draw all of the volumes you need for the reaction. Alternatively, a second error tube can be factored into the master mix and compensate for the lost volume due to m ixing.
3. Once the Taq is thoroughly mixed, place master tube back on ice, put original tube of Taq back into freezer.
4. Pipette specified amount of master mix into each sample tube. As soon as a portion of the reaction is dispensed into a sample tube, place on ice. 5. Once all tubes have a portion of master mix, flick them in microcentrifuge emd immediately place in PCR machine.
6. Choose file number, press enter, press start and remember to rescue your samples two and a half hours later or when convenient.
115 REFERENCES:
Erlich, Henry A. ed. 1992. PCR Technology: Principles and applications for DNA amplification. W.H. Freeman and Co.: New York.
Innis, Michael A.. David H. Gelfand. John J. Sinsky, and Thomas J. White eds. 1990. PCR Protocols: A Guide to Methods and Applications. Academic Press, Inc.: San Diego.
116 APPENDK J SOLUTIONS FOR POLYACRYLAMIDE GELS
30% STOCK SOLUTION: Wear gloves and a dust mask to handle the dry chemicals. Work under a hood if possible. Bis-acrylamide (N.N’-Methylene-bis- Acrylamide) is extremely light and can be easily disturbed into the air. Neurotoxin absorbed through the skin. Effects are cumulative.
A crylam ide 29g 1 45g bis-acrylamide Ig 5g d H 2 0 add to 100ml add to 500ml
Heat at 37°C to dissolve.
Storage in autoclaved amber bottles at 4oC for 2-3 months recommended.
Be sure to rinse all contaminated glassware with (IH 2 O and collect the waste. This waste should be polymerized by adding APS and TEMED and disposed of in hazardous waste bucket. 7.5% WORKING SOLUTION:
30% Stock solution 2 4 .9 4 m l 4 9 .8 8 m l 124.7m l d H 2 0 6 4 .3 6 m l 128.72m l 3 2 1 .8 m l lOXTBE 1 0 .0 0 m l 2 0 .0 0 m l 5 0 .0 0 m l APS 0 .7 0 m l 1.40m l 3.5 0 m l
100.00ml 200.00ml 500.00ml
Solution can be deaerated under vacuum if necessary. (Not usualty needed)
Store in autoclaved amber bottles. Prepare fresh working solution
117 every two-three weeks. Just prior to pouring the gel, TEMED should be added in a ratio of 35/d TEMED : 100ml of working solution. This will catalyze the process of polymerization.
BE SURE TO ALLOW 7.5% SOLUTION TO COME TO ROOM TEMPERATURE BEFORE ADDING TEMED AND POURING GEL! IF USED COLD, SMALL BUBBLES WILL FORM AFTER POLYMERIZATION.
APS SOLUTION = 10% AMMONIUM PERSULFATE [(NH4)2S208]:
Ammonium persulfate I g dH20 add to 10ml
Store in large centrifuge tubes at 4oC for up to two weeks. PLUG SOLUTION:
Used to seal the bottom of the glass plates. Polymerization of this solution is rapid (within 30 secs). Add APS before adding TEMED. Once TEMED is added, swirl to mix and immediately saturate Wattman strips.
Stock solution 10ml 20ml APS 80/d 160/fl TEMED 250//1 SOOfil
118 APPENDIX K
GEL PREPARATION, RUNNING AND VISUALIZATION
MATERIALS: BIORAD Sequi-Gen U Nucleic Acid Sequencing Cell COMPONENTS: Universal base Stablilizer bar IPC (Integral plate/chamber) Outer plate Clamps (2) Gel temperature mdicator Spacers (2) Comp Safety covers and cables Casting tray Sealing strips (Wattman paper) Gray cushions Power supply Alconox detergent 70% ethanol Rain-X Kim-wipes Kaydry sheets IxTBE 7.5% polyacrylamide working solution APS TEMED Manufactured ladder of choice Samples pipettors and tips ethidium bromide cold dH20 Polaroid camera film PLATE PREPARATION:
119 1. Gently wash the plates with Alconox detergent (both the outer plate and the inside plate of the IPC chamber — only need to do this with new plates). Rinse thoroughly with water. Clean with 70% ethanol and Kim-wipes (lint-free wipers). Be sure surface is extremely smooth, free of lint and polyacrylamide.
2. Apply several coats of Rain-X (does same job as Sigma-cote) to both inner surfaces of plates. Add an extra coat to the inner IPC plate (this will make the gel stick to the outer plate instead of the IPC plate).
NOTE: ONLY APPLY RAIN-X WHEN ABSOLUTELY NECESSARY (APPROX. ONCE EVERY FOUR USES OF PLATES). After 5-10 appUcations of Rain-X. plates need to be stripped with I CM NaOH to remove Rain-X build-up. If gel begins to shift while rig is running and causing current to arc, suspect Rain-X build-up and strip plates immediately.
ASSEMBLING RIG:
1. Lie the clean, polished IPC chamber glass inner plate side up on a bench top that is protected with Kaydry sheets.
2. Place one spacer along each of the long edges of the IPC plate. Be sure spacers are clean and smooth. Place the outer plate (Rain-X side in) on top of the EPC unit. Clamp into place with the provided clamps. Make sure the two plates, spacers, and clamps line up with each other on the bottom before clamping in place.
PLUG PREPARATION: 1. Clean the casting tray. Place two Wattman paper sealing strips on top of the gray cushion in the bottom of the tray.
2. Mix plug solution (20ml working solution, 500ul TEMED and 160ul APS) £md immediately pour the plug solution onto the strip. Saturate the strip with plug solution. 3. Quickly place the rig standing upright into the casting tray and allow the solution to polymerize (it will polymerize in about 30 seconds). Be sure to keep a bead of solution along both sides of the bottom of the gel to ensure a solid plug. Plug should be drawn between plates about 1.5cm by capillary action.
4. Once the solution has polymerized completely (1-2 minutes), the gel can be poured.
120 POURING THE GEL:
1. Fill IPC chamber about halfway with water. If polyacrylamide is spilled into this chamber while pouring, it will be unable to polymerize and can be poured down the drain before gel is mounted in rig.
2. Add I7.5ul of TEMED to 50ml of ROOM TEMPERATURE working solution in a Nalgene Unitary Wash Bottle. Swirl gently to mix. 3. With the casting tray still attached to the bottom of the rig, tilt the rig onto its bottom right comer. Place the nozzle of the bottle at the juncture of the two plates. Squeeze the bottle gently and slowly squirt the gel solution between the plates at the top right com er. Avoid bubbles by squirting the solution slowly and maintaining a fairly steep tilt to the plates (about 500 angle with working surface).
4. As the gel solution begins to fill the space between the plates, slowly flatten the angle of the plates. As the level reaches the top of edge of the inner plate, begin to follow the level of solution along the juncture of the two plates until reaching the left edge of the gel rig. FUI until a bead of gel solution is sitting along the top of the shortest plate. 5. Lay the plates down stiU slightly elevated (use a mbber stopper to give the plates a slight angle). 6. Gradually ease the teeth flat edge into the gel (backwards) between the plates to provide a straight edge for loading. 7. Once gel is poured and teeth are set, EMPTY NALGENE WASH BOTTLE into a beaker and RINSE IMMEDIATELY WITH LOTS OF WATER. If the polyacrylamide is left in the bottle, it will clog the nozzle and the bottle wiU be ruined.
8. Allow the gel to polymerize for at least an hour. It's best to aUow it to cure for several hours or even overnight. (If you let the gel cure overnight, cover the top edge in plastic wrap so it doesn't dry out.)
MOUNTING THE RIG:
I. Remove the casting tray and sealing strips from the bottom of the rig, remove the teetii from the top of the gel, and pour water from IPC chamber into sink.
121 2. Set the rig into the universal base by aligning the notches in the clamps with the ridges in the base. Clamp in place with the stabilizing bar. Do not leave rig unattended without stabilizing bar in place.
3 FiU the top cheunber 1cm from the top and the bottom chamber with 350-450ml of IxTBE (Do not fill bottom chamber with more than 500 ml of IxTBE).
4. With a disposable pipette, clear the bubbles from between the two plates above the straight edge of the gel.
5. Set the teeth between the two plates just barely into the gel (just enough to keep the samples from leaking out). If the gel is puckered by the teeth, the samples will run like smiles.
6. Attach the safety covers with cables onto the rig. Plug the cables into the power source and allow the assembly to warm up at 20 watts for 20-30 minutes.
LOADING THE GEL: 1. While the gel is warming, prepare each sample by using 4ul of sample to lul of Blue Juicell (Sambrook, Fritsch and Maniatis, 1989; p6.12).
NOTE: Avoid drawing oil by placing pipette tip directly into center of sample through oil layer, touch the bottom of the tube with tip to break connection with oil, move back slightly and draw sample slowly. Do not lean pipette tip against wall of tube or you wiU be unable to break oil connection. Also, to conserve plastics, a series of Eppendorf tubes can be labeled and used over and over again to hold sample and blue juice prior to loading. Rinse these tubes with water and ethanol and allow to dry completely before using them for the next series of samples.
2. Also load a manufactured ladder or a homemade ladder from several samples already identified to gauge the sizes of fragments in a product. Use \fi\ of manufactured ladder, 2/il dH20, and Ifil of Blue Juicell for a marker. 3. Turn off the power source. Remove the safety covers. 4. Load each sample into a separate well. When loading samples, use a pipette to slowly release the sample between the two plates. Be sure there aren’t any bubbles in the pipette tip or in the well.
122 5. Replace the safety covers and cables and run the gel at a constant 20W for about 2.5 hours (long enough to separate the fragments of interest). DISMANTLING THE GEL RIG:
1. Turn off the power source, dismantle the cables and dump the buffer out of the IPC chamber.
2. Lay the rig flat on a bench and pry the IPC chamber and outer plate apart slowly with an even force. The gel should stick to the outer plate (it is stickier than the plate on the IPC chamber which has an extra coat of Rain-X). 3. Cut the gel with a razor blade (be sure not to scratch the surface of the plate) to remove gel that does not contain the fragments of interest. 4. Cover the remaining gel with a sheet of plastic wrap and place a piece of plexiglas large enough to hold the gel onto of the sheet of plastic wrap. 5. Transfer the gel from the outer plate to the plastic wrap by flipping the rig sandwich, separating the plexiglas plate from the outer plate, and peeling the plastic wrap back and teasing the gel away from the plate by squirting dH20 between the gel and the outer plate until the gel falls away. The gel should now be lying on the plastic wrap.
STAINING THE GEL: 1. With the gel on top of the plastic wrap and plexiglas, place in a tray that will catch ethidium bromide waste. 2. Prepare a staining solution (20ml of dH20 £ind 6ul of ethidium bromide). 3. Pour the solution over the portion of the gel where fragments are expected. Allow to sit for 2-4 minutes. 4. Rinse the ethidium bromide away with a small amount (50ml) of cold (4oC) dH20. Collect this waste and store in EtBr wsiste bottle. 5. Then thoroughly rinse the gel with cold dH20 (approximately 800ml-1000^). Be careful not to pour too quickly or the gel will
123 slide firom the plastic wrap and may tear.
VISUALIZING FRAGMENTS:
1. Transfer the rinsed gel to a UV light box by lifting the plexiglas platform and carrying to the box. 2. Place absorbent towels to catch runofif. Carefully slide the gel off of the plastic wrap. This is the tricky step and just takes practice to keep gels firom tearing. Using plenty of water helps.
3. Turn on the light box. Fragments of DNA will glow hot pink.
4. Photograph the products with the short hood on the polaroid camera. Be sure photographs are developed and have been successful before destroying gel. 5. Be sure to store any ethidium bromide waste as hazardous waste and dispose of properly. Clean work space and frequently monitor with a UV wand to be sure ethidium bromide contamination is not occurring. Once gel has been photographed, dispose of gel in ethidium bromide waste bucket.
REFERENCES: Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular Cloning: A Laboratory Manual, Second Edition. Cold Spring Harbor Laboratory Press: New York.
124 APPENDIX L
Data relative to CHAPTER II Matrlcies of genetic similarity values and intemest distances for nesting females from Tortuguero and Melbourne
125 Genetic similarity values and internest distances for Tortuguero females
Pemmie# firom 1901 #o#»on Female# firom 1092 #ea#on Female# from 1003 #ea#on r 1 r 1 r IK 116 IG 1.1 111 1C IF IB 108 111 IL 112 IE lA 210 223 207 228 201 313 340 347 318 306 346 307 IK X 0.0780.400 0.310 0.0400.421 0.304 1 10 0.0 X 0.4800.000 0.420 0001 0.302 0.007 0.2040.024 0.444 0.408 0.380 0.372 0.4780.578 IG 1.2 0.0 X 0.018 0.4000.4000.314 0.400 0.333 0.000 0.480 0.480 0.314 0.340 0.4200.320 0.408 0,4880.417 0.528 IJ 0.0 0.0 1.2 X 0.033 0.4780203 0.4100.300 111 0.0 1.2 2.2 0.0 X 0038 0.2800.3800.4230.313 0.480 0.421 0.340 0.318 0.0000.542 0.348 0.480 0,471 0.3400.408 1C 0.0 0.0 1.2 0.0 0.0 X 0.333 0.432 0.3800.333 0.404 0.0000.380 IE 0.2 4.0 0.8 0.4 0.8 X 0.3480.3810.423 0.304 0.300 0.070 IB 0.4 1.0 0.2 0.8 0.2 0.0 X 0.4070.304 0.400 0.408 108 2.0 2.0 2.0 3.8 1.8 X 0.370 0.3080.307 0.408 111 0.2 0.0 4.4 1.4 4.2 2.4 X 0.317 0.3030.070 IL 0.2 0.0 0.0 1.8 4.2 X 0.448 0.4000.402 0.0410.3000.300 0.0400.444 112 1.0 2.2 0.4 1.0 0.8 1.2 3.0 0.4 0.8 X 0.270 0.340 0.080 0.0000.2000.2000.510 0.000 IE 0.4 0.2 1.0 1.0 0.4 2.0 X 0.027 0.400 lA 0.4 1.0 X 210 0.0 2.0 1.0 2.2 X 0.013 0.304 0.470 0.405 223 0.2 0.0 1.8 X 0.485 0.2440.270 0.300 207 0.0 1.0 X 0.270 0.4000.444 228 4.8 4.2 0.2 4.4 X 0.228 0.040 0.3180.3200.320 0.370 201 1.4 0.0 X 0.4810.4400.375 0.082 313 1.4 0.0 0.4 0.8 1.4 1.0 0.0 X 340 2.0 4.0 3.0 4.8 2.0 0.0 X 347 0.8 0.0 X 318 1.4 0.0 0.4 0.8 1.4 1.0 0.0 0.0 X 306 0.2 0.8 1.2 0.2 1.4 0.0 0.8 0.4 0.0 0.0 X 346 0.2 1.0 0.0 0.0 0.8 3.2 2.4 X 307 0.8 1.2 0.0 0.0 X Genetic similarity values and intemest distances for Melbourne females
436 448 428 402 412 408 454 452 446 429 444 409 445 410 451 440 431 449 447 436 X 0.316 0.356 448 3.7 X 0.419 0.397 428 5.5 0.2 X 0.428 0.279 0.400 0.465 402 12.0 2.8 X 0.367 0.296 0.361 0.413 0.333 0.393 0.333 0.429 0.321 0.444 412 8.5 11.3 X 0.355 0.375 0.200 0.339 408 1.9 0.9 10.4 X 0.459 0.400 0.431 0.327 454 2.2 5.0 6.3 4.1 X 0.333 0.340 0.310 0.339 452 4.2 7.1 3.3 0.8 X 0.286 0.258 0.314 0.367 446 2.7 7.7 3.6 4.4 X 0.302 0.415 0.393 0.393 429 0.4 3.7 2.9 7.3 X 0.364 0.305 0.305 0.386 444 1.4 0.6 5.0 2.3 X 0.346 0.327 0.440 0.481 409 3.3 1.4 5.9 3.9 X 0.385 0.393 0.357 0.346 445 4.4 2.9 0.6 3.3 X 0.415 0.400 0.346 0.386 410 12.2 10.9 8.6 4.7 8.0 X 0.452 0.400 0.327 451 2.6 1.0 4.9 1.6 9.6 X 0.328 0.305 0.329 440 4.8 2.7 0.6 7.4 1.0 X 0.255 0.333 0.280 431 13.2 9.0 10.6 8.4 X 0.233 0.321 449 11.9 9.3 7.1 1.3 X 0.235 447 9.0 4.2 4.2 2.9 X APPENDK M
Data relative to CHAPTER n Genetic simUarily scores for first-order relatives in Tortuguero and Melbourne. Values are listed adjacent to codes for pairs of individuals being compared. For each family, the numbers refer to hatchling identifications and “M” refers to the mother of the family. For the Tortuguero scores, only mother-ofifspring pairs were scored in each family. For Melbourne, mother-ofifspring pairs as well eis known fiill- sibling pairs (in femilies 407 and 428) were scored.
128 Genetic sim ilarity values for first-order relatives in Tortuguero
FAMILY 207 FAMILY 206 genetic genetic pair similarity pair similarity value value M, 1 0.679 M. 1 0.679 M.2 0.585 M, 2 0.646 M,3 0.605 M, 3 0.667 M.4 0.694 M. 4 0.679 M.5 0.717 M .5 0.621 M.6 0.694 M. 6 0.615 M,7 0.625 M, 7 0.567 M.8 0.636 M, 8 0.490 M.9 0.744 M. 9 0.621 M.IO 0.692 M. 10 0.677 M. 11 0.542 M. 11 0.593 M. 12 0.612 M. 13 0.706 FAMILY 108 M. 14 0.780 genetic M. 15 0.666 pair similarity value M. 16 0.679 M, 1 0.542 M. 17 0.696 M. 2 0.603 M. 18 0.656 M. 3 0.652 M, 4 0.545 M. 5 0.627 M. 6 0.500 M. 7 0.500 M. 8 0.625 M. 9 0.698 M. 10 0.571 M, 11 0.596 M. 12 0.520
129 Genetic similarity values for first-order relatives in Melbourne
FAMILY 407 FAMILY 426
genetic genetic pair similarity pair similarity value value 1.2 0.595 M. 13 0.682 I. 4 0.556 M. 23 0.585 2. 4 0.522 M. 17 0.585 2. 6 0.619 M, 1 0.591 2. 7 0.667 4 .6 0.718 FAMILY 428 4. 7 0.743 6, 7 0.722 genetic pair similarity value FAMILY 417 M. 6 0.683 M. 7 0.622 genetic M, 4 0.591 pair similarity value M. 8 0.558 M. 7 0.647 M. 2 0.611 M, 6 0.613 M. 1 0.714 M. 4 0.588 1. 2 0.615 M. 3 0.649 1. 8 0.698 M. 5 0.667 2. 8 0.636 M. I 0.621 4. 8 0.634 7. 8 0.627 4. 7 0.708 4. 6 0.571 6, 7 0.583
130 APPENDIX N
Data relative to CHAPTER m
Genotypes of nesting females from Tortuguero and Melbourne at both Cc 117 and Cm 3 loci. For Tortuguero, 39 females were screened, and 46 females were screened for Melbourne. Alleles (represented as letters at each locus) are named based on their size thus “A” is the largest allele at each locus, “B” is second largest, and so on. NOTE: “A" in locus Cc 117 is not the same as “A” in Cm 3 Hetero^gotes show two different alleles (letters) at a particular locus, while homo^gotes have the same letter repeated twice.
131 Genotypes of Tortuguero fem ales
L o c u s L o c u s Locus Locus Female Cc 117 Cm 3 Female Cc 117 Cm 3 IE EE EE 338 GH CF IF EE CE 340 HI E F IG II CE 342 BB FH IH BD CE 344 FK BE 18 KK DD 347 EHCE 207 BG CB 349 GK FF 303 HH CE 350 FF B C 305 B E DH 351 HI CE 307 LL AE 352 EK CC 309 AA BC 353 BECF 314 GH DF 355 FHCE 315 EH BF 357 II AB 316 D I CC 358 H J CF 317 EK CE 360 HH C C 318 FF BF 361 EE FG 319 IK BC 362 CC C F 321 F J FG 324 CD CC 326 EK CC 332 BE CE 333 CE CE 335 FH FF 337 BE CC
132 Genotypes of Melbourne fem ales
L ocus L ocu s L o cu s L ocus F em a le C c 1 1 7 Cm 3 F em a le C c 1 1 7 Cm 3 402 FI AB 431 D D BC 403 DH AB 432 D G BB 404 CE BB 433 H J CE 405 C J BE 434 HH B E 407 FE BE 435 D F BB 408 CC BE 436 CH EF 409 DH AD 437 CD AB 410 CD BE 438 D G CG 411 E J AB 439 KK B E 412 IJ B B 440 D F EF 415 DH EF 441 GH B B 416 D J BB 442 F J B E 417 El B E 444 D J AB 418 D F BC 445 GG D D 421 BG BG 446 H J B D 422 EE AE 447 CD B B 424 HI BE 448 D K B B 425 D H BE 449 F J B C 426 AD CC 450 HI B E 427 BG CC 451 IJ B E 428 DG B F 452 B I EG 429 B K BF 453 B E AB 430 FG AC 454 CD B E
133 APPENDIX O
Data relative to CHAPTER m
Genotypes of family members for eight clutches in Tortuguero and 10 clutches in Melbourne at both Cc 117 and Cm 3 loci.
134 Family genotjrpes: Tortuguero
Tortuguero Family: 207 Tortuguero Family: 309
Mother Genotype: Mother Genotype Cc 117: B G Cc 117: A A Cm 3: B C Cm 3: B C Hatchling Genotype# Hatchling Genotypes Locus Locus Locus Locus Cc 117 Cm 3 I.D. Cc 117 Cm 3 1 B E B F I AK c c 2 B E BC 2 AC BC 3 C G CF 3 AK CC 4 EG CC 4 AC CC 5 EG BC 5 AC BC 6 BC B F 6 AC BC 7 BE CC 7 AC B C 8 BE CF 8 AK CC 9 BE BC 9 AC CC 10 BC B F 10 AK B E II B C B F II AC CC 12 B C BF 12 AC BC 13 EG CC 13 AK B C 14 CG CF 14 AK B C 15 C G CC 15 AK B C 16 EG BC 16 AC B E 17 B E CC 17 AC B E 18 BE CF 18 AK B C 19 C G B F
135 Family genotypes:Tortuguero, cont.
Tortuguero Family: 312 Tortuguero Family: 314
Mother Genotype Mother Genotype Cc 117: CD Cc 117: G H Cm 3: C C Cm 3: D F Hatchling Genotypea Hatchling Genotypes Locus Locus Locus Locus IJ). Cc 117 Cm 3 IJ). Cc 117 Cm 3 1 DK BC I FH FF 2 DD CG 2 FH EF 3 CF CC 3 GF D F 4 DD CC 4 GH E F 5 DD BC 5 G H D E 6 DD C G 6 G J FG 7 DD CC 7 G F D F 8 DD CC 8 FH D E 9 CF CG 9 FH FF 10 DD BC 10 GH D E II DD CC II HH FF 12 DF BC 12 HH D E IS D F BC 13 FH DE 14 DD CC 14 FH FF 15 DD CB 15 FH DE 16 D D CC 16 G H FF 17 DK CG 17 FH D E IB C F B C IS FH FF 19 D F CC 19 G H DE 20 DF BC 20 GH E F 21 CF BC 21 FG D E 22 DK CG 22 HH D F 23 D K CG 23 FH FF 24 DD BC 24 G H EF 25 DF CG 25 FH E F 26 CF BC 26 FI E F 27 C F CC 27 G H D E 28 FH D F
136 Family genotypes:Tortuguero, cont. Tortuguero Familsr: 315 Tortuguero Family 316 Mother Genotype Mother Genotype CC 117: EH Cc 117: DI CMS: BF Cm 3: C C Hatchling Genotypes Hatchling Genotypes Locus Locus Locus Locus 1.0 . Cc 117 Cm 3 I.D. Cc 117 Cm 3 1 HG B E I DI CF 2 G E EF 2 D I CC 3 HG C F 3 G I C F 4 HG EF 4 DG CE 5 HE CF 5 G I CF 6 HE EF 6 DD CF 7 HI AB 7 D E CC 8 HH BE 8 DD CC 9 EG B E 9 G I C F 10 E I CF 10 DI BC II EG FF II DE BC 12 EG BE 12 DD CF 13 EG B C 13 DD CF 14 H I C F 14 DE CC 15 EH BC 15 El BC 16 H I C F 16 D D CC 17 EG EF 17 DI C F 18 D I C F
137 Family genotypes: Tortuguero, cont. Tortuguero Family: 318 Tortuguero Family: 319 Bfother Genotype Mother Genotype Cc 117: F F Cc 117: IK Cm 3: BF Cm 3: B F
Hatchling Genotypea Hatchling Genotypea Locus Locus Locus Locus I.D. Cc 117 Cm 3 I.D. Cc 117 Cm 3 1 C F AF 1 D I BC 2 CF AF 2 DI CF 3 AF B F 3 DK CF 4 DF BC 4 D K B C 5 AF B B 5 D I BC 6 J F B C 6 D I C F 7 DF B B 7 D K BC 8 CF AB 8 DK BC 9 DF B B 9 DK B C 10 DF B B 10 DK C F 11 D F B B 11 DK C F 12 DF AF 12 DI CF 13 DF B F 13 D K B C 14 DF B F 14 D K C F 15 D F B B 15 DI BC 16 AF B B 16 D K BC 17 J F AB 17 D K B C 18 CF AB 18 DK BC 19 D I C F 20 D K BC 21 DK C F 22 D K B C 23 D I BC 24 D I C F 25 D I BC 26 D I BC 27 D K C F
138 Family genotypes: Melbourne Melbourne Family: 403 Melbourne Family: 404 Mother Genotype Mother Genotype Cc 117: D H Cc 117: CE Cm: AB Cm3: BB Hatchling Genotypes Hatchling Genotypes Locus Locus Locus Locus ID. Cc 117 Cm 3 ID. Cc 117 Cm 3 I B H AB I EE AB 2 B H B D 2 BC BC 3 B D B D 3 EE BC 4 B D AB 4 CE B B 5 DH AD 5 BC AB 6 DH B D 6 C E BC 7 B D AB 7 EE B B 8 D E B D 8 E E AB 9 B H AB 9 B C B B 10 DH AB 10 B C AB II B H BF II C E AB 12 DE AF 12 B C B C 13 D E B D 13 E E B C 14 B D B D 14 B C BC 15 B H AB 15 C E BC 16 B D AB 16 EE BC 17 B H AB 17 BC B B IS B D AB 18 EE AB 19 B D B D 19 E E B B 20 B D B D 20 EE B B 21 DH B D 21 EE B B 22 DE B D 22 CE B B 23 B H AB
139 Family genotypes: Melbourne, cont.
Melbourne Family: 405 Melbourne Family: 407 Mother Genotype Mother Genotype Cc 117: C J Cc 117: FE Cm 3: B B Cm 3: B E Hatchling Genotypes Hatchling Genotypes Locus Locus Locus Locus I.D. Cc 117 Cm 3 I.D. Cc 117 Cm 3 1 J J AB 1 FF EE 2 B J DE 2 DE B E 3 B J B D 3 FF B E 4 B J B D 4 DF B E 5 J J B D 5 D F E E 6 B J B D 6 DE B E 7 BC BC 7 FF B E 8 J J B D 8 EF B E 9 BC D E 9 D F E E 10 BC AB 10 DF B E 11 J J CE 11 E F E E 12 BC DE 12 E F B E 13 J J CE 13 D E EE 14 J J B C 14 DF B E 15 BC BE 15 DE EE 16 BC B D 16 G F EE 17 G F B E 18 D F E E 19 FF EE 20 FF B E 21 FF B E 22 EF BE 23 FF B E
140 Family genotypes: Melbourne, cont.
Melbourne Family: 408 Melbourne Family: 417 Mother Genotype Mother Genotype Cc 117: CC Cc 117: E I Cm 3: B B Cm 3: B B Hatchling Genotypes Hatchling Genotypes Locus Locus Locus Locus ID. Cc 117 Cm 3 ID. Cc 117 Cm 3 I cc AE I EG AB 2 cc AE 2 B E BE 3 cc AE 3 B I BE 4 cc AB 4 G I BE 5 cc AE 5 B E BE 6 cc B E 6 B I AB 7 cc AE 7 EG CE 8 cc E E 8 EG BE 9 cc B E 9 B E BE 10 cc AB 10 B E AB II cc AE II B I AB 12 cc B E 12 B E AB IS cc B E 13 B E AB 14 cc AE 14 B E C E 15 cc E E 15 G I AB 16 cc EE 16 EG CE 17 c c EE 17 B I AB IS cc AB 18 B I BE 19 cc AE 19 EG AB 20 cc AE 20 EG BE 21 cc AB 21 B I AB 22 cc EE 22 B 1 AB 23 cc AE
141 Family genotypes: Melbourne, cont.
Melbourne Family: 421 Bfelboume Family: 426 Mother Genotype Mother Genotype Cc 117: B 6 Cc 117: AD Cm 3: B G Cm 3: C C
Hatchling Genotypea Hatchling Genotypea Locus Locus Locus Locus I.D. Cc 117 Cm 3 I.D. Cc 117 Cm 3 l AB B E 1 AD CF 2 AB BE 2 AD C E 3 AG B E 3 AE B C 4 FG B B 4 AD C F 5 B F B E 5 D D CF 6 FG B B 6 D D CE 7 FG EG 7 AD C E 8 AG EG 8 DD C E 9 FG BE 9 AD CF 10 FG B B 10 C D BC 11 AB B G 11 C D BC 12 B F B E 12 D A C F 13 B F B E 13 DD C F 14 AG B E 14 D D C F 15 FG E G 15 D D CE 16 AG EG 16 D A C F 17 BF B B 17 D A CF 18 AB EG 18 DD C E 19 AG EG 19 D D C E 20 FG B B 20 D A C E 21 AB B G 21 D A C F 22 AG EG 22 D A CE 23 AB B B 23 D D CE 24 B F B G
142 Family genotypes: Melbourne, cont.
Melbourne Family: 428 Melbourne Family: 430 Mother Genotype Mother Genotype Cc 117: D G Cc 117: G F Cm 3: BF Cm 3: AC
Hatchling Genotypes Hatchling Genotypes Locus Locus Locus Locus ID. Cc 117 Cm 3 I.D. Cc 117 Cm 3 1 DG BF 1 BF BC 2 D D AB 2 BF CG 3 D D BF 3 BF BC 4 D D B F 4 B F CG 5 DG B F 5 G H AB 6 D D AB 6 B F BC 7 DG FF 7 B G CG 8 DG B F 8 FH CG 9 DG FF 9 FH BC 10 DD B F 10 G H AG 11 DG B F 11 FH AB 12 D D B F 12 G A AB 13 D D BF 13 B G CG 14 DG AB 14 B G AB 15 D D AB 15 BF AG 16 D D AB 16 BG CG 17 D D FF 17 BF BC 18 D D B F 18 B G BC 19 DG B F 19 BG AG 20 DG FF 20 FH AG 21 DG B F 21 FH AG 22 FH AG 23 B F B C
143 LIST OF REFERENCES
Aalto, S. K. And Newsome, G, E. 1990. Additional evidence supporting demie behaviour of a yellow perch [PereaJlauescens ) population. Canadian Journal of Fisheries and Aquatic Science, 47: 1959-1962. Ackerman, R. A. 1980. Physiological and ecological aspects of gas exchange in sea turtles. American Zoologist, 20: 575-583. Allard, M. W., Miyamoto, M. M., Bjomdal, K. A., Bolten, A. B. and Bowen, B.W. 1994. Support for natal homing in green turtles from mitochondrial DNA sequences. Copeia, 1994: 34-41.
AUendorf, F.W. 1983. Isolation, geneflow and genetic differentiation among populations. In C.M. Schonewald-Cox, S.M. Chambers, B. MacBryde and W.L. Thomas. Genetics and Conservation. A Reference for managing Wild Animal and Plant Populations. The Benjamin/Cummings Publishing Company, Inc. Menlow Park, California.
Antifinger, A. E. 1981. The genetic basis of microdifferentiation in natural and experiment^ populations of Borrichiafnitescens in relation to salinily. Evolution, 35: 1056-1069. Antonovics, J. and Bradshaw, A. D. 1970. Evolution in closely adjacent plant populations, vm. Clinal patterns of a mine boundary. Heredity, 25: 349-362.
Atwood, J. L. And Massey, B. W. 1988. Site fidelity of least terns in California. Condor, 90: 389-394. Balazs, G. H. 1982.Growth rates of immature green turtles in the Hawaiian archipelago. In K.A. Bjomdal (ed.) Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. pp. 117-125.
144 Barrowclough, G. F. and Coats, S. L. 1985. The demography and population genetics of owls, with special reference to the conservation of the spotted owl [Strixocctdentalls ). In: Gutierrez, R, J. Euid Carey, A. B. (eds.) Ecology and management of the spotted owl in the PactSc Northwest. U. S. Forest Service, Portland, Oregon, pp. 74-85.
Beardmore, J. A. 1983. Extinction, survival and genetic variation. In: Genetics and conservation: a reference for managing wild animal and plant populations. C. M. Schonewedd-Cox, S. M. Chambers, B. MacBryde and W. L. Thomas, (eds.) Benjamin /Cummings Publishing, Inc. Menlow Park, California.
Birkhead, T. R. and A. P. Moller. 1992. Sperm Competition in Birds. Academic Press, London.
Bishop, J. A. 1981. A neo-Darwinian approach to resistance: examples from mammals. In: Bishop, J. A. and Cook, L. M. (eds.) Genetic Consequences of Man-Made Change. Academic Press, London, pp. 37-51.
Bishop, J. A. and Hartley, D. J. 1976. The size and age structure of rural populations of Rattus norvegicus containing individuals resistant to the anticoagulant poison warfarin. Journal of Animal Ecology. 45: 623-646. Bishop, J. A., Hartley, D. J. and Partridge, G. G. 1977. The population dynamics of genetically determined resistance to warfarin in Rattus norvégiens from mid-Wales. Heredity, 39: 389- 398.
Bjomdal, K. A. 1980. Nutrition and grazing behavior of the green turtle, Cheloniamydas. Marine Biology, 56: 147-154. Bjomdal, K. 1982. The consequences of herbivoiy for the life history pattem of the Caribbean green turtle, Chelonia mydas. In K.A. Bjomdal (ed.) Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. pp. 19-26. Booker, W. C. and L. M. Ehrhart. 1990. Aerial surveys of marine turtle carcasses in National Marine Fisheries Service Statistical Zones 28, 29, and 30; August 1987 to 31 December 1989. in T. H. Richardson, J. I. Richardson and M. Donnelly (compilers) Proceedings of the Tenth Annual Workshop on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFC 278.
145 Booth J. and J. A. Peters. 1972. Behavioral studies on the green turtle [Chelonia mydas) in the sea. Animal Behaviour, 20: 808- 812. Bowen. B. W., Meylan A. B., Ross, J. P., Limpus, C. J., Balazs, G. H., and Avise, J. C. 1992. Global population structure and natural history of the green turtle [Chelonia mydas ) in terms of matriarchal phylogeny. Evolution, 46: 865-881. Britten, H. R., P. R. Brussard and D. D. Murphy. 1994. The pending extinction of the UnconpahgreJritillary butterfly. Conservation Biology, 8: 86-94.
Briton, R. K., D. A Nurthen, D. A. Briscoe and R. Frankham. 1994. Modelling problems in conservation genetics using Dmsophilia : Consequences of harems. Biological Conservation, 69: 267-275. Brockmann, H. J. 1994. Sperm competition in horseshoe crabs [Llmuluspolyphemus). Behavioral Ecology and Sociobiology, 35: 153-160. Burger, R and M. Lynch. 1995. Evolution and extinction in a changing environment: A quantitative-genetic analysis. Evolution, 49: 151-163.
Burke, T. and Bruford, M. W. 1987. DNA fingerprinting in birds. Nature, 327: 149-152 Bumett-Herkes, J. !974, Returns of green sea turtles [Chelonia mydas Linnaeus) tagged at Bermuda. Biological Conservation, 6: 307-308.
Bustard, R. H. and P. Greenham. 1968. Physical and chemical factors affecting hatching in the green sea turtle [Chelonia mydas). Ecology, 49: 269-275. Carr, A. F. 1952. A Handbook of Turtles. Cornell University Press, Ithaca, New York. Carr, A. F. 1956. The Windward Road. New York, Alfred Knopf Inc. Carr, A. F. 1967. So Excellente a Fishe: A Natural History of Sea Turtles, Scribner, New York. Carr, A. F. 1975. The Ascension Island green turtle colony. Copeia, 1975: 547-555.
146 Carr, A. F. 1980. Some problems of sea turtle ecology. American Zoologist. 20: 489-498. Carr. A. F. 1982. Notes on the behavioral ecology of sea turtles. In: Bjomdal, K. (ed.) Biology and Conservation of Sea Turtles. Smithsonian Institution Press. Washington. D.C. pp. 19-26.
Carr, A. F. 1986. New Perspectives on the Pelagic Stage of Sea Turtle Development. NOAA-TM-SEFC-190. Miami. Florida. Southeast Fisheries Center. National Marine Fisheries Service. Carr. A. F. and Carr. M. H. 1972. Site fixity in the Caribbean green turtle. Ecology. 53: 425-429. Carr. A. F.. Carr. M. H. and Meylan. A. B. 1978. The ecology and migrations of sea turtles. 7. The west Caribbean green turtle colony. Bulletin of the American Museum of Natural History. 162: 1-46.
Carr. A. F. and L. Giovanoli. 1957. The ecology and migrations of sea turtles. 7. Results of fieldwork in Costa Rica. 1955. American Museum Novitates. 1835: 1-32. Carr. A. F. and Hirth, H. 1962. The ecology and migrations of sea turtles. 5: Comparative features of isolated green turtle colonies. American Museum Novitates, 2091: 1-42. Carr. A. F. and A. B. Meylan. 1980. Evidence for passive migration of green turtle hatchlings in sargassum. Copeia. 1980: 366-368. Carr, A. F. and L. Ogren. 1960. The ecology and migrations of sea turtles, 4. The green turtle colony in the Caribbean Sea. Bulletin of the American Museum of Natural History. 121: 1-48. Chapman, R. W. 1989. Spatial and temporal variation of mitochondrial DNA haplotype firequencies in the striped bass [MoronssaxatiUs) 1982 year class. Copeia, 1989: 344-348.
Chen, X. 1993. Comparison of inbreeding and outbreeding in hermaphroditic Arianta arbustorium (L.) (land snail). Heredity, 71:456-461.
147 Chesser, R. K. 1983. Isolation by distance: relationship to the management of genetic resources. In: Genetics and conservation: a reference for managing wild animal and plant populations. C. M. Schonewald-Cox, S.M. Chambers, B. MacBryde and W. L. Thomas (eds.). Benjamin /Cummings Publishing, Inc. Menlow Park California.
Craighead L., D Paetkau, H. V. Reynolds, E. R. Vyse and C. Strobeck. 1995. Microsatellite analysis of paternity and reproduction in Arctic grizzly bears. Journal of Heredity, 86: 255-261.
Crowder, L. B., S. R. Hopkins-Murphy and J. A. Royle. 1994. The effect of Turtle-Excluder Devices on loggerhead sea turtle strandings in South Carolina. In K. A. Bjomdal, A. B. Bolten, D. A. Johnson and P. J. Eliazar (compilers). Proceedings of the Fourteenth Annual Symposium on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFSC-351. pp 32-33. Dallas, J. F., B. Dod, P. Boursot, E. M. Prager and F. Bonhomme. 1995. Population subdivision and gene flow in Danish house mice. Molecular Ekzology, 4: 311-320. Decker, M. D., Parker, P. G., Minchella, D. J., and Rabenold, K. N. 1993 Monogamy in black vultures: genetic evidence from DNA fingerprinting. Behavioral Ecology, 4: 29-35. de Silva, G. S. 1982. The status of sea turtle populations in east Malaysia and the south C hina sea. In K.A. Bjomdal (ed.). Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. pp. 327-338. Devine. M. C. 1984. Potential for sperm storage in reptiles: behavioral and physiological consequences. In R.L. Smith (ed.). Sperm Competition and the Evolution of Animal Mating Systems 509-521. Academic Press, New York.
Dutton, P., S. K. Davis and D. W. Owens. 1994. Genetic population survey of leatherbacks based on mtDNA. In K. A. Bjomdal, A. B. Bolten. D. A. Johnson and P. J. Eliazar (compilers). Proceedings of the Fourteenth Annual Symposium on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFSC-351. p. 36.
148 Dutton, P.. D. W. Owens and S. K. Davis 1995. Natal homing: Do leatherbacks do it too? In J. A. Keinath, D. E. Bernard, A. A. Musick and B. A. Bell (compilers). Proceedings of the Fifteenth Annual Symposium on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFSC-387. p.79.
Dutton, P. Using molecular genetics to identify individuals and infer mating behavior in leatherbacks. In R. Byles, compiler. Proceedings of the Sixteenth Annual Symposium on Sea Turtle Conservation and Biology. NOAA Technical Memorandum NMFS- SEFSC (in press) Ehrlich, P. R. and Raven P. H. 1969. Differentiation of populations. Science. 165: 1227-1232. Ehrhart, L. M. and B. E Witherington. 1992. Green turtle Chelonia mydas (Linnaeus). In P.E. Moler (ed.). Rare and Endangered Biota of Florida, Volume m. Reptiles and Amphibians. University of Florida Press, Gainesville. FI. Eriksson, G., G. Namkoong and J. H. Roberds. 1993. Dynamic gene conservation for uncertain futures. Forest E)cology and Management, 62: 15-37.
Evans, J. D. 1993. Parentage analysis in ant colonies using simple sequence loci. Molecular E)cology, 2: 393-397. Falconer, D. S. 1981. Introduction to quantitative genetics. 2nd edition, Longman, London.
Fisher, H. I. 1976. Some Dynamics of a breeding colony of Laysan albatrosses. Wilson Bulletin, 88: 121-142. Fisher, H. I. And Fisher, M. L. 1969. The visits of the Laysan albatross to the breeding colony. Micronesica, 5: 167-215.
FitzSimmons, N. N. Multiple Paternity in green turtles? In R. Byles, compiler. Proceedings of the Sixteenth Annual Symposium on Sea Turtle Conservation and Biology. NOAA Technical Memorandum NMFS-SEFSC (in press)
FitzSimmons, N. N., C. Moritz, S. S. Moore. 1995. Conservation and dynamics of microsatellite loci over 300 m illio n years of marine turtle evolution. Molecular Biology and Evolution, 12: 432-440.
149 Fowler, L. E. 1979. Hatching success and nest predation In the green sea turtle, Cheloniamydas, at Tortuguero, Costa Rica. E)cology, 60:946-955. Frankel, O. H. 1974. Genetic conservation: Our evolutionary responsibility. Genetics, 78: 53-65.
Frankel, O. H. and M. E. Soule. 1980. Conservation and Evolution. Cambridge University Press, Cambridge, England.
F ra n k h a m , R. 1995. Inbreeding and extinction: A threshold effect. Conservation Biology, 9: 792-799.
Frankham , R., G. J. Smith and D. A. Briscoe. 1993. Effects on hetero^goslty and fitness of Inbreeding with and without selection on fitness In Drosophila melanogaster. Theoretical and Applied Genetics, 86: 1023-1027. Frazer, N. B. and Ladner, R. C. 1986. A growth curve for green turtles IChelonJa mydas) In the U.S. Virgin Islands. Copeia, 1986: 798- 802. Frazier, J. 1971. Observations on sea turtles at Aldabra AtoU. Philosophical Transactions of the Royal Society of London. B, 260: 373-410.
Frazier, J. 1982. Status of turtles In the central western Indian Ocean. In K. A. Bjomdal (ed) Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. pp.385=390.
Galbraith, D. A., B. N. White, R. J. Brooks and P. T. Boag. 1993. Multiple paternity In clutches of snapping turtles [Chelydra serpentina ) detected using DNA fingerprinting. Canadian Journal of Zoology, 71: 318-324.
Georges, M., Lequarre, A-S., Castelll, M., Hanset, R., and Vassart, G. 1988. DNA fingerprinting in domestic animals using four different mlnlsatelllte probes. Cytogenetics and Cellular Genetics, 47: 127- 131.
Gibson, A. R., and J. B. Falls. 1975. Evidence for multiple Insemination In the common garter snake, Thamnophis sirtaMs. Canadian Journal of Zoology. 53: 1362-1368.
150 Gilbert. D. A., Lehman. N.. O'Brien. S. J.. and Wayne. R. K. 1990. Genetic fingerprinting reflects population differentiation in the California Cheuinel Island fox. Nature. 344: 764-766.
Gilpin. M. W. and M. E. Soulé. 1986. M in im u m viable populations: processes of species extinction. In: Conservation biology - the science of scarcity and diversity. M. E. Soulé. ed. Sinauer Associates. Sunderland. Massachusetts.
Gottelli. D., C. Sillero-Zubiri. G. D. Applebaum. M. S. Roy. D. J. Girman. J, Garcia-Moreno. E. A. Ostranders and R. K. Wayne. 1994. Molecular genetics of the most endangered canid: The Ethiopian wolf. Canis simensis. Molecular Ecology. 3: 301-312. Grassmem. M. emd D. Owens. 1987. Chemosensory imprinting in juvenile green sea turtles. Chelonia mydas. Animal Behavour. 35: 929-931.
Green. E. L. (ed.) Biology of the Laboratory Mouse. 2nd edition. New York. Dover Press.
Greenwood. P. J. and P. H. Harvey. 1978. Inbreeding and dispersal in the great tit. Nature 27: 52-54. Groombridge. B. 1982. The green turtle. The lUCN Amphibia and Reptdlia Red Data Book Part 1. lUCN. Gland. Switzerland, pp. 151-180.
Haig, D. and C. T. Bergstrom. 1995. Multiple mating, sperm competition and meiotic drive. Journal of Evolutionary Biology. 8: 265-282. Harry. J. L. and D. A. Brisco. 1988. Multiple paternity in the loggerhead turtle [Carettacaretta] Journal of Heredity. 79: 96-99. Hartl. G. B. zmd Z. Pucek. 1994. Genetic depletion in the European bison [Bisonbonasus). Conservation Biology. 8: 167-174. Hedgpeth, J. W. 1954. An introduction to the zoogeography of the northern Gulf of Mexico with reference to the invertebrate fauna. Publication of the Institute of Marine Science. University of Texas. 3: 111-211. Hendrickson. J. R. 1958. The green sea turtle. Cheloniamydas (Linn.), in Malaya and Sarawak, Proceedings of the Zoological Society of London. 130: 455-535.
Hendrickson. J. R. 1980. The ecological strategies of sea turtles. American Zoologist 20: 597-608.
151 Henwood T. A. and W. E. Stuntz. 1987. Analysis of sea turtle captures and mortaliiy during comercial shrimp trawling. Fisheries Bulletin, 85: 813-817.
Hirth, H. F. 1971. Synopsis of biological data on the green turtle Chelonia m ydas (Linnaeus)1758. FAO Fisheries Synopsis, 85: 1:1- 8:19. Hirth, H. F. 1980. Cheloniamydas. In: Catalog of Américain Amphibians and Reptiles 249: 1-4.
Hirth, H. F. and Carr, A. F. 1970. The green turtle in the Gulf of Aden and Seychelles Islands. Verhand. Konin. Nederl. Acad. Weten. Natur. 58: 1-41.
Hirth, H. F. and D. A. Samson. 1987. Nesting behavior of green turtles (Chelonia m ydas ) at Tortuguero, Costa Rica. Caribbean Journal of Science. 23: 374-379.
Hoggren, M. and H. Tegelstrom. 1995. DNA fingerprinting show within-season multiple paternity in the adder (Vipembems]. Copeia, 1995: 271-277.
HoUebecq, M. G. and P. Haffray. 1994. Genetic improvement of the common carp, Cyprinuscarpio, L.: State of the art. Bulletin Française de la Peche et de la Pisciculture, 333: 93-124. Horikoshi, K. 1992. Sex ratio of green turtle hatchlings in Tortuguero, Costa Rica. In M. Salmon, and J. Wyneken (compilers) Proceedings of the Eleventh Annual Workshop on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFSC-302. pp. 59-60. Houlden, B. A., P. England and W. B. Sherman. 1996. Paternity exclusion in koalas using hypervariable microsatellites. Journal of Heredity, 87: 149-152
Jain, S. K. and Bradshaw, A. D. 1966. Evolutionary divergence among adjacent plant populations. I. The evidence and its' theoretical analysis. Heredity, 20: 407-441. Jeffreys, A. J., Wilson, V., and Thein, S. L. 1985. Hypervariable 'minisatellite' regions in human DNA. Nature, 314: 67-73. Johnson, S. 1994. Reproductive ecology of the Florida green turtle (Chelonia mydas). Masters Thesis, University of Florida.
152 Kenyon, K. W. And Rice. D. W. 1962. Homing in Laysein albatrosses. Condor, 60: 3-6.
Kichler, K., S. K. Davis, D. W. Owens R. Marquez. Population bottleneck: How much genetic damage has man done to the Kemp's ridley? In R. Byles (compiler). Proceedings of the Sixteenth Annual Symposium on Sea Turtle Conservation and Biology. NOAA Technical Memorandum NMFS-SEFSC (in press)
King, F. W. 1982. Historical review of the decline of the green turtle and the hawksbill. In K.A. Bjomdal (ed.). Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. ppl83-188. Knowlton, N. emd S. R. Greenwell. 1984. Male sperm competition avoidence mechanisms: the influence of female interests. In R.L. Smith (ed.) Sperm Competition and the Evolution of Animal Mating Systems. Academic Press, Orlando Florida, pp. 61-84.
Koch, A. L., Carr, A. F. and Ehrenfeld, D. W. 1969. The problem of open-sea navigation: the migration of the green sea turtle to Ascension Island. Journal of Theoretical Biology, 22: 163-179. Koehn, R. K. and P. M. Gaffirey. 1984. Genetic heterozygosity and growth rate in MytUus edulis. Marine Biology, 82: 1-7. Kress, S. W. And Nettleship, D. N. 1988. Re-establishment of Atlantic puffins [Fratercula arctica ) at a former breeding site in the Gulf of Maine. Journal of Field Ornithology, 59: 161-170. Laikre, L. and N. Ryman. 1991. Inbreeding depression in a captive wolf {.Canis lupus ) population. Conservation Biology, 5: 33-40. Lande, R. 1988. Genetics and demography in biological conservation. Science, 241: 1455-1460.
Lande, R. 1994. Risk of population extinction from the fixation of new deleterious mutations. Evolution, 48: 1460-1469. Lande, R. 1995. M utation emd conservation. Conservation Biology, 9: 782-791. Lewis, B. 1940. The Cayman Islands and marine turtle. Bulletin of the Institute of Jamaican Science Serial, 2: 56-65.
153 Limpus. C. Caretta caretta fecundity: Long distance migration causes reduction in eggs laid. In R. Byles (compiler). Proceedings of the Sixteenth Annual Symposium on Sea Turtle Biology and Conservation (in press). Limpus, C. J., P. J. Couper and M. A. Read. 1994. The green turtle, Chelonia mydas, in Queensland: Population structure in a warm temperate feeding area. Memoirs of the Queensland Museum, 35: 139-154. Limpus, C. J., J. D. MiDer, C. J. Peuramenter, D. Reimer, N. McLachland and R. Webb. 1992. Migration of green [Chelonia m ydas ) and loggerhead [Caretta caretta ) turtles to and from eastern Australian rookeries. Wildlife Research, 19: 347-358. Limpus, C. J. and D. G. Walter. 1980. The growth of immature green turtles [Chelonia mydas ) under natural conditions. Herpetologica, 36: 162-165. Liskauskas, A. P. emd Ferguson, M. M. 1991. Genetic variation and fitness: A test in a naturalized population of brook trout [SaivelirmsfontinaUs ). Canadian Journal of Fishery and Aquatic Science, 48: 2152-2162. Lohmann, K. J. 1992. How sea turtles navigate. Scientific American. 266: 100-106. Lohmann, K. J. and C. M. F. Lohmann. 1994. Aquisition of magnetic directional preference in lo^erhead sea turtle hatchlings. Journal of Experimental Biology, 190: 1-8. Long, E. 1774. The History of Jam aica, or General Survey of the Ancient and Modem State of that Island. T. Loundes Inc. London
Longmire, J. L, Lewis, A. W., Brown, N. C., Buckingham, J. M., Clark, L. M., Jones, M. D., Meincke, R. L., Meyne, J., Ratcliff, R. L., Ray, F. A., Wagner, R. P. and Moyzis, R. K. 1988. Isolation and molecular characterization of a highly polymorphic centromeric tandem repeat in the family Falconidae. Genomics, 2: 14-24. Luiselli, L. 1995. The mating strategy of the European adder, Vipera berus. Acta Oecologica, 16: 375-388. Lynch, M. 1990. The similarity index and DNA fingerprinting. Molecular Biology and Evololution, 7: 478-484.
154 Lynch, M., W. Gabriel and A. M. Wood. 1991. Adaptive and demographic responses of plankton populations to environmental change. Limnology and Oceanography, 36: 1301-1312. Lynch, M. and R. Lemde. 1993. Evolution and extinction in response to environmental change. In P. M. Kareiva, J. G. Kingsolver and R. B. Huey (eds.). Biotic interactions and global change. Sinauer. Sunderland, Massachusetts, pp. 234-250.
Madsen, T., R. S. Shine, J. Loman and T. Hakansson. 1992. Why do female adders copulate so frequently? Nature, 355: 440-441.
Mantel, N. 1967. The detection of disease clustering and a generalized regression approach. Cancer Research, 27: 209-220. McAlpine, S. 1993. Genetic hetero^gosity emd reproductive success in the green treefrog, Hyla cinerea. Heredity, 70: 553-558.
Meyer, E., P. Weigland, S. P. Rand, D. Kuhhnan, M. Brack and B. Brinkman. 1995. Microsatellite polymorphisms reveal phylogenetic relationships in primates. Journal of Molecular Evolution, 41: 10-14
Meylan, A. 1982. Sea turtle migration: Evidence from tag returns. pp.91-100. In K.A. Bjomd^ (ed) Biology emd Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. Meylan, A. B., Bowen, B. W. and Avise, J. C. 1990. A genetic test of the natal homing versus social facilitation models for green turtle migration. Science, 248, 724-727. Moll, E. O. 1979. Reproductive cycles and adaptations. In M. Harless and H. Morlock. Turtles: Perspectives and Research New York, Wiley. 305-331.
Morreale, S. J., G. J. Ruiz, J. R. Spotila and E. A. Standora. 1982. Temperature-dependent sex determination: Current practices threaten conservation of sea turtles. Science, 216: 1245-1247.
Mortimer, J. A. 1981. The feeding ecology of the West Caribbean green turtle, Cheloniamydas, in Nicaragua. Biotropica, 13: 49-58. Mortimer, J. A. 1982. Feeding eccology of sea turtles. In K. A. Bjomdal (ed.). Biology and Conservation of Sea Turtles. Smithsonian Institution Press, Washington, D.C. pp. 103-109.
155 Mortimer, J. A. and A. F. Carr. 1987. Reproduction and migrations of the Ascension Island green turtle [Cheloniamydas ). Copeia. 1978: 103- 113.
Mortimer J. A. and Portier. K. M. 1989. Reproductive homing and Intemestlng behavior of the green turtle [Chelonia m ydas ) at Ascension Island. South Atlantic Ocean. Copeia. 1989: 962-977. Mrosovsliy. N. 1980. Thermal biology of sea turtles American Zoologist 20: 531-547.
Mrosovsky. N. and S. F. Klngsmlll. 1985. How turtles find the sea. Zeltschrlft ftier Tlerpsychologle. 67: 237-256. Mrosovslqr. N. and C. L. Ynetema. 1980. Temperature dependence of sexual differentiation in sea turtles: ImpUcations for conservation practices. Biological Conservation. 18: 271-280.
Murle. J. O. 1995. Mating behavior of Columblem ground squirrels. Canadian Journal of Zoolology. 73: 1819-1826.
National Research Council. 1990. Decline of the Sea Turtles: Causes and Prevention. National Academy Press. Washington. D.C.
O'Brien. S. J. and J. F. Everman. 1988. Interactive Influence of Infectious disease and genetic diversity of natural populations. Trends In Ecology and Evolution. 3: 254-259. O'Brien. S. J.. J. S. Martenson. M. A. Elchelberger. E. T. Thome and F. W. Wright. 1989. Genetic variation and molecular systematics of the black-footed ferret. In U. S. Seal. E. T. Thome. M. A. Bogan and S. H. Anderson (eds.). Conservation Biology and the Black- Footed Ferret.
Olsson. M. and T. Madsen. 1995. Female choice on male quantitative traits In lizards-why is It so rare? Behavioral E)cology and Socloblology. 36: 179-184.
Osorio-Berlstaln. M. And Drummond. H. 1993. Natal dispersal and deferred breeding In the blue-footed booby. Auk. 110: 234-239. Paetkau. D.. W. Calvert. I. Stirling and C. Strobeck. 1995. MlcrosateUlte analysis of population structure In Canadian polar begurs. Molecular Ecology. 4: 347-354. Papl. F. 1992. Animal Homing. Chapman and Hall. London.
156 Parker, P. G., T. A. Waite, and T. Peare. 1995. Patemlly studies in a n im a l populations. In T. B. Smith and R. K. Wayne (eds.). Molecular Genetics in Conservation, pp. 413-423.
Parker, P. G. And Whiteman, H. 1993. Geneüc diversity in fragmented populations of Clemmys guttata and Chrysemys picta margtnata as shown by DNA fingerprinting. Copeia: 1993, 841- 846.
Peare, T. and P. G. Parker. 1996. Local genetic structure within two rookeries of [Chelonia m ydas ) green turtles. Heredity, (in press)
Pease, C. M., R. Lande, and J. J. Bull. 1989. A model of population growth, dispersal and evolution in a changing environment. E)cology. 70: 1657-1664.
Piper, W. H., And Parker Rabenold, P. 1992. Use of firagment-sharing estimates from DNA fingerprinting to determine relatedness in a tropical wren. Molecular Ecology, 1: 69-78. Pray, L. A. and C. J. Goodnight. 1995. Genetic variation in inbreeding depression in the red flour beetle Triboliumcastaneum. Evolution, 49: 176-188.
Pritchard, P. C. H. 1979. Encyclopedia of Turtles. T.F.H. Publications, New Jersey.
Quattro, J. M. and R. C. Vrijenhoek. 1989. Fitness differences among remnant populations of the endangered Sonoran topminow. Science, 245: 976-978.
Quinn, T. P. And Dittman, A. H. 1990. Pacific salmon migrations and homing: Mechanisms and adaptive significance. Trends in Ecology and Evolution, 5: 174-177.
Rabenold P. P., Rabenold K. N., Piper W. H., Haydock J. H., Zack S. W. 1990. Shared paternity revealed by genetic analysis in cooperatively breeding tropical wrens. Nature, 348: 538-540. Ralls, K. and J. Ballou. 1983. Extinction: lessons from zoos. In: Genetics and conservation: a reference for managing wild animal and plant populations. C. M. Schonewald-Cox, S. M. Chambers, B. MacBryde and W. L. Thomas, (eds.). Benjamin /Cummings Publishing, Inc. Menlow Park California.
157 Ralls. K. K. Brugger and J. Bedlou. 1979. Inbreeding and juvenile mortaliiy in small populations of ungulates. Science, 206: 1101- 1103.
Rave, E. H., R. C. Fleischer, F. Duvall and j . M. Black. 1994. Genetic analysis through DNA fingerprinting of captive populations of Hawaiian geese. Conservation Biology, 8: 744-751. Rebel, T. P. 1974. Sea Turtles and the Turtle Industry of the West Indies, Florida and the Gulf of Mexico. Coral Gables, Florida, University of Miami Press.
Rhodes, O. E. Jr., and L. M. Smith. 1993. Relationships between genetic variation and carcass components in wintering American wigeons. Auk, 110: 354-360. Rieder, J. P., P. G Parker, J. R. Spotila and M. E. Irwin. The mating system of the leatherback turtle: A molecular approach. In R. Byles (compiler) Proceedings of the Sixteenth Annual Symposium on Sea Turtle Biology and Conservation. NOAA Technical Memorandum NMFS-SEFSC- (in press)
Rodhouse, P. G., J. H. McDonald, R. I. E. Newell and R. K. Koehn. 1986. Gamete production, somatic growth and multiple-locus enzyme hetero^gosity in Mytllus edulis. Marine Biology, 90: 209- 214.
Roelke, M. E., J. S. Martinson, and S. J. O'Brien. 1993. The consequences of demographic reduction and genetic depletion in the endangered Florida panther. Current Biology, 3: 340-349.
Rohlf, F. J. 1990. NTSYS-pc: numerical taxonomy and multivariate analysis system. Version 1.60. Applied Biostatistics, Setauket, New York. Rohlf, F. J. And SchneU, G. D. 1971. An investigation of the isolation-by-distance model. American N atur^st, 105: 245-324. Ross, J. P. 1982. Historical decline of loggerhead, ridley, and leatherback sea turtles. In K. A. Bjomdal (ed.). Biology and Conservation of Sea Turtles, pp. 189-196
Roy, M. S., E. Geffen, D. Smith, E. A. Ostrander and R. K. Wayne. 1994. Patterns of differentiation and hybridization in North American wolfiike canids, revealed by analysis of microsateUlte loci. Molecular Biology and Evolution, 11: 553-570.
158 Salmon. M., and J. Wyneken. 1994. Orientation by hatchling sea turtles: mechanisms and Implications. Herpetologlcal Natural History, 2: 13-24.
Salmon, M., J. Wyneken, E. Fritz, and M. Lucas. 1992. Seaflnding by hatchling sea turtles: role of brightness, sihouette and beach slope as orientation cues. Behavior 122: 56-77.
Schmidt-Koenig, K. And Keeton, W. T. 1978. Animed Migration, Navigation and Homing. Springer-Verlag, Berlin. SchneU, G. D., Watt, D. J. And Douglas, M. E. 1985. Statistical comparison of proximity matrices: appUcations in animal behaviour. Animëd Behaviour, 33: 239-253. Schulz, J. P. 1975. Sea turtles nesting in Surinam. Zoologische Verhandelingen, uitgegeven door het Rljksmuseum van Natuurlijke Historié te Leiden, 143: 1-144
Selander, R. K. 1970. Behavior and genetic variation in natural populations. American Zoologist, 10: 53-66. Senner, J. W. 1980. Inbreeding depression and the survival of zoo populations. In: Conservation biology - the science of scarcity and diversity. M.E. Soule, ed. Sinauer Associates, Sunderland, Massachusetts. Snaydon, R. W. And Davies, M. S. 1972. Rapid population differentiation in a mosaic environment. U. Morphological variation in Anthoxanthumodoratum. Evolution, 26: 390-405.
Soked, R. R. And Wartenberg, D. E. 1983. A test of spatial autocorrelation analysis using an isolation-by-distance model. Genetics, 105: 219-237. Soule, M. E. and B. A. WUcox. 1980 (eds.). Conservation Biology: An Evolutionary-Ekzological Perspective. Sinauer Associates, Sunderland, Massachusetts. Southern, E. M. 1975. Detection of specific sequences among DNA fragments separated by gel electrophoresis. Journal of Molecular Biology, 98: 503-517.
159 Spotila, J. R., E. A. Standora, S. J. Mcrreaie and G. J. Ruiz. 1987. Temperature dependent sex determination in the green turtle [Chelonia mydas ): Effects on the sex ratio on a natural beach. Herpetologica, 43: 74-81. Sternberg, J. 1981. The worldwide distribution of sea turtle nesting beaches. Center for Environmental Education, Washington, D C. S u ^, D. W. and R. K. Chesser. 1994. Effective populations sizes with multiple paternity. Genetics, 137: 1147-1155. Swann R. L. And Ramsay, A. D. K. 1983. Movements from and age of return to an expanding Scottish guillemont colony. Bird Studies, 30: 207-214. Switzer, P. V. 1993. Site fidelity in predictable and unpredictable habitats. Evolutionary Ekzology, 7: 533-555. Taylor, A. C., W. B. Sherman, and R. K. Wayne. 1994. Genetic variation of microsatellite loci in a bottlenecked species: The northern hairy-nosed wombat, Lasiorhimis krejffUL Molecular Ecology, 3: 277-290. Tautz, D. 1989. Hypervariability of simple sequences as a general source for polymorphic DNA markers. Nucleic Acids Research, 17: 6463-6471. Teska, W. R., M. H. Smith and J. M. Novak. 1990. Food quality, heterozygosity and fitness correlates in Peromyscuspolionotus. Evolution, 44: 1318-1325.
Thome, E. T. and E. S. Williams. 1988. Disease and endangered species: the black-footed ferret as a recent example. Conservation Biology, 2: 66-74. Triggs, S. J., Williams, M. J., Marshall, S. J. And Chambers, G. K. 1992. Genetic structure of blue duck [Hymenolaimus malcorhynchos ) populations revealed by DNA fingerprinting. Auk, 109: 80-89. Urlich, G. F. and A. S. Parkes. 1978. The green sea turtle [Chelonia m ydas ); further observations on breeding in captivity. Journal of Zoology, London, 185: 237-251. Van Ryzin, M. T. And Fisher, H. I. 1976. The age of Laysan Albatrosses, Diomedea immutabills, at first breeding. Condor, 78: 1-9.
160 Waldman, B., Rice. J. E., And Honeycutt, R. L. 1992. Kin recognition and incest avoidance in toads. American Zoologist, 30: 18-30. Wayne, R. K., D. A. Gilbert, N. Lehman, K. Hansen, A. Eisenhawer, D. Girman, R. O. Peterson, L. D. Mech, P. J. P. Gogan, U. S. Seal and R. J. Krumenaker. 1991. Conservation genetics of the endangered Isle Royale gray wolf. Conservation Biology, 5:41-51. Welder, L. J. 1985. Spatial and temporal genetic heterogeneity in a natural Daphnia population. Journal of Plankton Research, 7: 101-123.
White, M. M. and G. E. Svendsen, 1992. Spatial-genetic structure in the easten chipmunk. Tamias striatus. Journal of Mammalogy, 73:619-624. WHdt, D. E., M. Bush, K.L. Goodrowe, C. Packer, A. E. Pusey, J. L. Brown. P. Joslin, and S. J. O'Brien. 1987. Reproductive and genetic consequences of founding isolated Hon populations. Nature, 328: 328-331. Witherington, B. E. 1992. Behavioral responses of nesting sea turtles to artificial lighting. Herpetologica, 48: 31-39.
Wood, J. R. and F. E. Wood. 1980. Reproductive biology of captive green sea turtles Chelonia mydas. American Zoologist, 20: 499- 505.
Worth, D. F., And Smith, J. B. 1976. Marine turtle nesting on Hutchinson Island, Florida, in 1973. Florida Marine Research Publications, 18: 1-17.
Wright, S. 1931. Evolution in Mendelian populations. Genetics, 16: 97-159.
Wright, S. 1943a. Isolation by distance. Genetics, 28: 114-138. Wright, S. 1943b. An analysis of local variability of fiower color in Linanfhusparryae. Genetics, 28: 139-156. Wyneken, J., T. J. Burke, M. Salmon and D. K. Pedersen. 1988. Egg failure in natural and relocated sea turtle nests. Journal of Herpetology, 22: 88-96
161