COMPARATIVE GENETIC ANALYSIS IN INSULAR AND MAINLAND POPULATIONS OF THE FLORIDA COTTONMOUTH, AGKISTRODON

PISCIVORUS CONANTI

By

ANDREW WARNER ROARK

A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT

OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2003

Copyright 2003

by

Andrew Warner Roark

To my wife, Alison McCombe Roark, for her endless patience and support.

ACKNOWLEDGMENTS

I wish to express my gratitude to Harvey Lillywhite, my friend and committee

chair, and to Paul Moler and Max Nickerson for their advice, insight, and experience.

This project would not have been possible without them.

I thank Ginger Clark for her patience and willingness to teach at all hours of the

day, Ryan McCleary for his enthusiasm and encouragement, Kevin Kopf for his

invaluable field assistance, and Kenneth Krysko for his comments on this manuscript. I

also thank my wife, Alison Roark, for her support, companionship, assistance in the field,

comments on this manuscript, and for not forsaking me and this project following a nasty

encounter with a pelican.

iv

TABLE OF CONTENTS Page ACKNOWLEDGMENTS ................................................................................................. iv

LIST OF TABLES............................................................................................................. vi

LIST OF FIGURES .......................................................................................................... vii

ABSTRACT..................................................................................................................... viii

CHAPTER

1 HABITAT FRAGMENTATION AND THE FLORIDA COTTONMOUTH.................1

Habitat Fragmentation ..................................................................................................1 Island Biogeography.....................................................................................................3 Agkistrodon piscivorus as a Study Species...................................................................6 Seahorse Key ................................................................................................................8

2 COMPARATIVE GENETIC ANALYSIS.....................................................................12

Objective.....................................................................................................................13 Study Sites ..................................................................................................................14 Methods ......................................................................................................................16 Analyses......................................................................................................................18 Results.........................................................................................................................19 Discussion...................................................................................................................19

LIST OF REFERENCES...................................................................................................27

BIOGRAPHICAL SKETCH .............................................................................................34

v

LIST OF TABLES

Table page 1 AFLP selective amplification primers used in study. .....................................................26

2 Comparison of inter-population relationships. ...............................................................26

vi

LIST OF FIGURES

Figure page 1 Study sites .......................................................................................................................24

2 Preselective amplification program ................................................................................25

3 Selective amplification program.....................................................................................25

vii

Abstract of Thesis Presented to the Graduate School

of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science

COMPARATIVE GENETIC ANALYSIS IN INSULAR AND MAINLAND POPULATIONS OF THE FLORIDA COTTONMOUTH, AGKISTRODON

PISCIVORUS CONANTI

By

Andrew W. Roark

December 2003

Chair: Harvey B. Lillywhite Major Department: Zoology

Seahorse Key (Levy County, Florida) is an important rookery for wading birds and

supports an unusually dense population of Florida cottonmouth snakes (Agkistrodon

piscivorus conanti) that are dependent on the rookery. The snakes are entirely terrestrial

and scavenge on dead or rotting fish that are dropped or regurgitated by the nesting birds.

Sensitivity of snakes to seawater and the apparent differentiation of certain character

states relative to mainland populations suggest the cottonmouths on Seahorse Key

comprise a relatively isolated population. DNA fingerprinting was used to study genetic

divergences among the Seahorse Key and two mainland cottonmouth populations.

Amplified fragment length polymorphism (AFLP) analyses indicated little genetic

variation (Φst = 0.005 - 0.020) and small genetic distances (D = 0.003 - 0.012) among the

populations, although genetic distance was least between the two mainland populations.

The small differences measured suggest the population at Seahorse Key either

viii

experiences significant exchange with the mainland or has not been isolated for a long

evolutionary period. The highly conserved nature of reptilian DNA, rather than continual

gene flow, may also account for the overall lack of genetic divergence. Heterozygosity

estimates indicate that the Seahorse Key population has the lowest percent heterozygosity

of the three populations, potentially supporting the hypothesis that this population has

experienced decreased gene flow for an extended period of time. Consequently,

quantitative genetic distances and variation among populations, though very small, might

still be considered useful data in assessing the degree of isolation in this unusual insular

population of snakes.

ix

CHAPTER 1 HABITAT FRAGMENTATION AND THE FLORIDA COTTONMOUTH

The population of Florida cottonmouth snakes (Agkistrodon piscivorus conanti)

found on Seahorse Key in the Cedar Keys National Wildlife Refuge maintains a high

population density, exploits a seemingly novel food source, and appears to be

characterized by generally larger body sizes and darker coloration than mainland

cottonmouths (Wharton 1966, 1969; H. Lillywhite, unpublished data). Cottonmouths on

Seahorse Key are believed to be relatively sedentary, and thus these ecological and

morphological differences from cottonmouths on the Florida mainland may indicate that

this population is experiencing the effects of long-term geographic isolation.

Habitat Fragmentation

Habitat fragmentation, the division of ecosystems into “habitat islands,” is an

increasingly prevalent problem in wildlife conservation and management (Lande and

Barrowclough 1987, Templeton et al. 1990). Vast deforestation and expanding

agricultural land use are destroying natural areas throughout the world, leaving behind

fragmented habitats that resemble islands in their isolation, limited area, and distance

from each other (Mader 1984).

Habitat fragmentation may interfere with migration and consequently with gene

flow, leading to the eventual retardation and degradation of evolutionary processes and to

a high probability of eventual extinction within fragmented populations (Erwin 1991,

Frankham 1995). The magnitude of the effects of this fragmentation is critically

dependent on the degree to which dispersal of individuals between habitat islands is

1

2

limited (Templeton et al. 1990). Reduced potential for dispersal may heighten the

probability of inbreeding and genetic drift within isolated populations, especially when

population size is small (Bushar et al. 1998). Isolated populations are more susceptible to

inbreeding depression than are non-fragmented populations, because inbreeding may

cause more frequent expression of recessive deleterious alleles, and genetic drift may fix

these deleterious alleles in the population. Once deleterious alleles are fixed in an

isolated population, no mechanisms for their removal exist (Templeton et al. 1990).

Even without fixation of deleterious alleles, inbreeding and genetic drift ultimately

reduce genetic diversity within a population (Templeton et al. 1990, Bushar et al. 1998).

Just as fixed deleterious mutations are known to accumulate within isolated populations

until these populations can no longer survive (Muller 1964), cumulative loss of genetic

diversity has been linked to decline and extinction in wild populations (Westemeier et al.

1998). Genetic variability has thus been proposed as an indicator of a population’s

vulnerability to both natural and anthropogenic stressors (D’Surney et al. 2001).

As decreasing genetic diversity caused by inbreeding and genetic drift within one

species promotes extinction of local populations, the number of populations of that

species existing at the global level declines, and genetic variants unique to individual

populations are lost. In an “extinction ratchet,” the loss of each habitat island population

that cannot be recolonized is an irreversible step towards the total extinction of the

species. As the extinction ratchet decreases the total number of populations globally, the

rate of genetic diversity loss at the species level accelerates, thus potentially driving the

entire species toward extinction (Templeton et al. 1990).

3

Unfortunately, it is not feasible to directly study the effect of habitat fragmentation

on dispersal and susceptibility to extinction for most species (Templeton et al. 1990,

Lawson and King 1996). Indirect investigation of a population’s degree of isolation (i.e,

using genetic techniques) has thus become necessary for species management and

conservation (Moritz 1994, Avise 1995, Frankham 1995, Bushar et al. 1998). Surveys

capable of determining genetic variability within and between potentially isolated

populations are now used to effectively identify populations at high risk of extinction

(Templeton et al. 1990, Palumbi and Baker 1994, Avise 1994).

Island Biogeography

Populations that have experienced limited genetic exchange for an extended period

of time due to geographic isolation are natural laboratories in which to study the long-

term ecological, physiological, and behavioral effects of habitat fragmentation. Islands,

as examples of fragmented and isolated habitats, are useful systems for investigating

patterns and mechanisms in ecology and evolution at both species and population levels

(MacArthur and Wilson 1967, Grant 1998). Genetic drift, combined with selection

pressures different from those experienced by mainland populations, facilitates trait

divergence in isolated populations (Galis and Metz 1998, Bonnet et al. 1999).

Furthermore, individual isolated populations have the potential to make each island

system distinctive in regard to the suite of traits displayed there.

In many cases, unoccupied ecological niches, combined with the founder effect,

have allowed populations on islands to evolve life-history traits vastly divergent from

those found in mainland relatives (Miller et al. 2000). Adaptations in island populations

are believed to occur more rapidly and with greater directionality due to both a small

gene pool and decreased gene flow between populations (Foster 1964). Characteristic

4

divergences attributable to population isolation frequently include differences in

behavior, morphology, color, longevity, and fecundity (Miller et al. 2000). Populations

experiencing isolation and decreased geographic area often maintain higher and more

stable population densities, higher individual survival rates, and reduced aggressiveness

and reproductive output relative to non-fragmented populations (Gliwicz 1980, Adler and

Levins 1994, Padmanabhan and Yom-Tov 2000). Variation in mean body mass between

isolated and mainland populations is also common (Adler and Levins 1994, McNab 1994,

Adler 1996, Anderson and Handley 2002). These differences can be attributed to both

the inverse relationship between geographic separation and rate of immigration (Soulé

1966) and the decrease in habitat diversity and density-depressing factors (i.e., predation

and interspecific competition) created by decreased geographical area and decreased

species richness (Adler and Levins 1994, Wiggins et al. 1998).

In reptiles, examples of characteristic divergences between insular and mainland

populations include differences in body size, coloration, scalation, reproductive output,

clutch size, size of offspring, and territorial behavior. Variation in mean body size

between insular and mainland reptile populations has been found in tiger snakes

(Notechis scutatus) on islands off the coast of south-eastern Australia (Schwaner and

Sarre 1988; Schwaner 1990); in side-blotched lizards (Uta stansburiana) on islands in the

Gulf of California, Mexico (Soulé 1966); in Chinese pit-vipers (Gloydius shedaoensis) on

Shedao island in northeast China (Li-xin et al. 2002); and in the eastern cottonmouth

snake (Agkistrodon piscivorus conanti) at the northern edge of its range near Hopewell,

Virginia, USA (Blem 1981, Blem and Blem 1995). Body sizes in reptiles on islands are

not dependent on island size, latitude, age, or distance to the mainland, but instead tend to

5

be bimodal based on prey size and abundance (Li-xin et al. 2002, Boback and Guyer

2003). However, mean population body size comparisons in reptiles are rare, presumably

because sampling is problematic and continuous growth in reptiles makes these

comparisons difficult (Soulé 1966). Morphological variation in coloration between

populations has been reported in cottonmouth snakes (Blem and Blem 1995), variation in

scale counts and dorsal scale size has been observed in tiger snakes (Schwaner 1990), and

variable scale serration has been reported in side-blotched lizards (Soulé 1966).

Reproductive output, offspring size, and offspring number also vary between

insular and mainland reptile populations. Total reproductive output is expected to

correlate directly with resource availability, and thus a high degree of variability is

expected among insular populations. A strong trade-off between offspring body size and

number of offspring produced also exists: larger offspring and reduced clutch sizes are

expected when large body size increases offspring viability, whereas small body size and

large clutch size are expected when offspring survival is low (Roff 1992). Large body

size in island population offspring is believed to be strongly favored when a larger gape

size allows the exploitation of a food source that could otherwise not be consumed at an

early age (Li-xin et al. 2002).

Large body sizes in adult reptiles may reduce the population’s vulnerability to

predation, thereby increasing the population’s survival rate and allowing the population

size to be sustained by a smaller number of offspring (Williams 1966, Bull and Shine

1979). Females of large size and better body condition have been shown to produce

larger offspring in water snakes (Nerodia sipedon, Weatherhead et al. 1999), Chinese pit-

vipers (Li-xin et al. 2002), and cottonmouth snakes (Blem 1981). Although not

6

statistically significant, this trend was further observed in smooth snakes (Coronella

austriaca, Luiselli et al. 1996) and European adders (Vipera berus, Andren and Nilson

1983). However, the opposite trend (large females producing numerous, small offspring)

was reported in the North American brown snake (Storeria dekayi, King 1993).

The primary behavioral difference between insular and mainland reptile

populations appears to be a heightened tolerance to proximity between individuals. If

food is the limiting environmental factor for insular population growth, then abundant

food availability may create a high population density (Soulé 1966). Soulé (1966) found

the population of side-blotched lizards on the island of San Pedro Mártir to maintain a

population density two to three times higher than mainland populations and to exhibit

decreased levels of territorial behavior. Territorial behavior is still expected to exist at

some level, however, as it prevents resource exhaustion by controlling population density

(Wynne-Edwards 1962). In terms of species abundance, reptiles and amphibians are the

vertebrate taxa most significantly impacted by island isolation, followed by mammals,

resident birds, and finally migratory birds (Harris 1984).

Agkistrodon piscivorus as a Study Species

The combination of unique characteristics and relatively high probability of

population extinction in island systems in general, and in snakes in particular (Dodd

1993), makes locating and researching established insular populations of reptiles and

other vertebrates important to conservation. The population of Florida cottonmouth

snakes (Agkistrodon piscivorus conanti) found on Seahorse Key (Levy County, Florida)

is a potentially useful system for the study of insular vertebrate populations experiencing

significantly reduced gene flow due to geographic or habitat isolation. This population

appears to exhibit behavioral, morphological, and reproductive character divergences

7

commonly used to identify isolated populations (i.e., darker coloration, larger adult body

sizes, greater tolerance of close proximity to others, and increased offspring body size

relative to mainland populations) (Wharton 1966), and could be useful to both

evolutionary and conservation biologists.

Cottonmouth snakes are New World pit vipers of the monophyletic clade

Agkistrodon, in the family Viperidae, subfamily Crotalinae (Parkinson 1999). New

World pit vipers are currently believed to have moved into North America in a single

invasion (Krauss et al. 1996, Parkinson 1999, Parkinson et al. 2000, but see Burger 1971,

Gloyd and Conant 1990). Fossil records suggest that divergence between Agkistrodon

and the North American rattlesnake genera, Crotalus and Sistrurus, occurred no later than

the late Miocene era (10-12 million years ago; Conant 1990). The genus Agkistrodon

comprises three widely distributed species: the copperhead, A. contortrix, the

cottonmouth, A. piscivorus, and the cantil, A. bislineatus. Restriction fragment length

polymorphism and mitochondrial DNA sequence analyses suggest that A. contortrix is

the basal species of the genus and that A. bislineatus subsequently arose from A.

piscivorus (Knight et al. 1992, Parkinson et al. 1997, Parkinson 1999, Parkinson et al.

2000). Agkistrodon piscivorus and A. contortrix, along with most rattlesnake species,

resided in temperate ancestral habitats (Klauber 1972), whereas A. bislineatus diverged

into tropical ancestral habitats (Parkinson et al. 2000). Fossil records indicate that A.

piscivorus has been present in Florida since the Pleistocene era (Brattstrom 1953,

Auffenberg 1963).

Agkistrodon piscivorus has a broad head relative to its neck, a pair of facial pits, a

single row of ventral scales under the major part of the tail, and a single anal plate. It

8

reaches the largest size of any member of the genus Agkistrodon, especially in girth (may

exceed 1.8 meters in length and 4.6 kg in body mass). It is the only semi-aquatic member

of the genus, has the greatest number of scale rows (25) at midbody, and lacks a loreal

scale (Gloyd and Conant 1990). Cottonmouths are largely nocturnal during hot weather

but bask during daylight hours in the cooler months of the year. The typical cottonmouth

diet is generalized and includes carrion, small mammals, birds, snakes, amphibians,

turtles, and fish (Burkett 1966, Savitzky 1992). These snakes are opportunistic feeders

and take the prey item that is most readily available to them (Ernst 1992). The maximum

known lifespan for a cottonmouth is 21 years (Gloyd and Conant 1990).

The Florida cottonmouth (Agkistrodon piscivorus conanti) is found from Southern

Georgia to the Florida Keys and inhabits many coastal islands (Conant and Collins 1998).

This subspecies can be distinguished from other Agkistrodon piscivorus subspecies by its

11 to 16 dark dorsal cross-bands, dark cheek stripe, dark vertical stripes at the first

supralabials and edges of the rostral and adjacent prenasals, and pair of dark blotches

extending from the mental along the first four or five infralabial scales and the outside of

the chin shield (Gloyd and Conant 1990). Pattern and color variations are generally

slight, and although dorsal body markings may become subdued by a uniform black or

dark brown coloration in adults, some indication of facial markings is generally retained

(Gloyd and Conant 1990).

Seahorse Key

Insular populations of the Florida cottonmouth inhabiting the Cedar Keys of

Florida have been studied by Carr (1936) and, more extensively, by Wharton (1966,

1969), who focused his investigations on an unusually dense population of snakes

inhabiting Seahorse Key. This island, a crescent-shaped land mass just over 1.6 km in

9

length (0.67 ha in area), is the largest in the Cedar Keys National Wildlife Refuge (Levy

County, FL). It lies approximately 3.5 km offshore, is of sand dune origin, has a

maximum elevation of 15.85 meters, and consists of salt marsh, mangrove, and mixed

hardwood forest habitats. Seahorse Key is part of a wide continental shelf and is

separated from the mainland by water approximately 2 – 3.5 meters in depth. Given that

sea levels have continuously risen in Florida since the late Holocene (some 8,000 years

before present) at an estimated rate of 4 cm per 100 years over the last 3,000 years and 25

cm per 100 years prior to that (Wanless 1982), the separation of Seahorse Key from the

mainland can be roughly estimated at 3,300 – 3,900 years before present.

The island’s peripheral vegetation consists mostly of black mangrove (Avicennia

nitida) and salt marsh. The inland hammock consists primarily of dwarf palmetto (Sabal

minor), red bay (Persea borbonia), sand live oak (Quercus geminata) and Virginia live

oak (Quercus virginiana). The upland hammock and mangroves are heavily populated

by white pelicans (Pelecanus erythrorhynchos), brown pelicans (Pelecanus

occidentaliscarolinensis), nesting cormorants (Phalacrocorax auritus), osprey (Pandion

haliaetus), and wading birds like white ibis (Guara alba), snowy egret (Leucophoyx

thula), American egret (Casmerodius albus), as well as several heron species (Wharton

1966, 1969).

Seahorse Key has no permanent fresh water, and resident cottonmouths are thus

denied their characteristic semi-aquatic habitat. The primary food source for this

population comes from scavenging fish dropped or regurgitated by colonial wading birds

that nest in large numbers on the island, usually from March through September or

10

October. Marine foraging is not considered a possible feeding strategy due to saltwater

avoidance by cottonmouths (Wharton 1966).

During the years 1954-56, Wharton captured 545 cottonmouths on this island and

estimated its total population at 600 individuals. He noted that snakes occur throughout

the island but are concentrated in the area of highest nesting bird abundance, a low

peninsula at the west end of the island. Cottonmouth density was estimated at 56.1

individuals per hectare beneath these rookeries and 4.6 individuals per hectare on the

main ridge of the island, where rookeries are relatively scarce. Although Seahorse Key

snakes may face less pressure from predators relative to their mainland counterparts, food

availability varies seasonally with the presence of nesting birds, and an estimated 77%

are in danger of starvation during the winter months when rookeries are largely empty

(Wharton 1969). This strong dependence on nesting bird populations, seasonal food

supply, and high probability of starvation-induced mortality caused Wharton to refer to

this population as living at a “critical survival level” (Wharton 1969).

Cottonmouths are also present on the keys adjacent to Seahorse Key, but the

density of these snakes appears to correlate with the presence and numbers of nesting

birds, which are by far most common on Seahorse Key (H. Lillywhite, unpublished data).

These island populations of cottonmouths presumably rafted to the islands sometime in

the undetermined past and then became established or have been present since the islands

became disconnected from the mainland. Observations of coloration and behavior

suggest divergence between island and mainland snake populations (H. Lillywhite,

unpublished data). The occurrence of island snakes lacking one or both eyes (Wharton

1969; Roark and Lillywhite, unpublished data) could support this theory and indicate that

11

phenotypic divergence could be rooted in the genetic composition of these island

populations if this condition is congenital. Such phenotypic variation between

populations is frequently used to identify microevolutionary processes including gene

flow, natural selection, and genetic drift (Lawson and King 1996).

Preliminary field and laboratory data suggest Seahorse Key cottonmouths are

sensitive to salt water and avoid seawater (Wharton 1966, 1969; Roark and Lillywhite,

unpublished data), thereby potentially isolating island snakes from their mainland

counterparts. The sharing of refugia, which has been observed in the Seahorse Key

cottonmouth population, could also decrease population genetic diversity, as it facilitates

inbreeding within snake species that copulate upon emergence in the spring (Bushar et al.

1998). The potential for reduced gene flow and increased inbreeding, combined with

recognized phenotypic differences relative to mainland populations, suggests that genetic

divergence from mainland cottonmouth populations may characterize the Seahorse Key

population. However, genetic analyses of the current degree of gene flow and genetic

variation within and between this island population and surrounding mainland

populations are required to determine the degree of isolation affecting this population

and, thus, its susceptibility to eventual extinction.

CHAPTER 2 COMPARATIVE GENETIC ANALYSIS

Geographically isolated populations are interesting to biologists because they have

a high potential for extinction as predicted by island biogeography theory (MacArthur

and Wilson 1967), may exhibit long-term ecological, physiological, and behavioral

effects of habitat fragmentation, and are likely to experience inbreeding depression and

reduced intrapopulation genetic diversity (Templeton et al. 1990). The cottonmouth

population residing on Seahorse Key, an island in the Cedar Keys National Wildlife

Refuge, appears to be a potentially useful system in which to study the characteristics of

an isolated population. However, before this system can be studied effectively, a

determination of whether this population is truly isolated must be made. Studies

attempting to determine phylogenetic relationships and divergences based on

morphological and physiological traits are often flawed because these traits are affected

by environmental factors and phenotypic plasticity (Avise 1995). Problems building

these types of phylogenies are further compounded in snakes because they possess few

variable morphological traits relative to other groups, retain primitive character traits, and

frequently exhibit evolutionary convergence between taxa (Knight et al. 1992).

Population isolation can, however, be measured with the use of molecular techniques

such as amplified fragment length polymorphism (AFLP, Vos et al. 1995) analysis.

The AFLP technique developed by Keygene BV (Wageningen, The Netherlands) is

a multilocus marker protocol wherein genomic DNA undergoes fragmentation through

double-restriction digestion and subsequent ligation to specific oligonucleotide adapters

12

13

that alter the restriction site to prevent reformation and post-ligation digestion. A subset

of fragments is then amplified by polymerase chain reaction (PCR) as selective primers

are added. Amplified fragments are separated and identified using polyacrylamide gel

electrophoresis. The AFLP technique can generate a number of potential markers per

genome ten times greater than can be accomplished using simple sequence repeats (SSR)

or random amplified polymorphic DNA (RAPD) techniques (D’Surney et al. 2001).

AFLP markers have gained popularity due to their relatively low cost (Giannasi et al.

2001), utility in the absence of prior sequence information (Busch et al. 2000; Negi et al.

2000), high multiplex ratio (Negi et al. 2000), and high degree of reproducibility (Ribeiro

et al. 2002). AFLP markers are also ideal for fluorescent dye labeling and gel-based

automated analyses (D’Surney et al. 2001). The primary deterrent to using AFLP

markers is their dominant nature, which does not allow direct measurement of

heterozygosity (Negi et al. 2000, although see Nei 1978 for an indirect measure of

heterozygosity).

Objective

The objective of this study was to test two hypotheses: 1) the Seahorse Key

cottonmouth population is more genetically distinct from select mainland populations

than these mainland populations are from each other, and 2) gene flow into the Seahorse

Key population is greatly reduced relative to that between mainland populations. The

investigation of these hypotheses was carried out in hopes of establishing a quantitative

measure of the degree of isolation of the Seahorse Key cottonmouth population to create

a foundation from which future comparative studies between this island population and

mainland populations might be launched. This research was also intended to create a

knowledge base about this system that, when combined with current and future

14

morphological, physiological, and genetic comparative studies of these island and

mainland snakes, has the potential to generate new hypotheses regarding the behavioral

and physiological effects of habitat fragmentation in this and possibly other vertebrate

species.

Study Sites

Three study sites (one island and two mainland) were sampled for genetic

comparison between November 2000 and October 2002 (Fig. 1). The first study site was

the island of Seahorse Key in the Cedar Keys National Wildlife Refuge, the second was a

coastal mainland site in the Lower Suwannee National Wildlife Refuge (Levy County,

FL), and the third was an interior mainland site in Paynes Prairie State Preserve (Alachua

County, FL). Both mainland sites were initially selected to approximate the geographic

area of Seahorse Key (0.67 ha) but required expansion beyond this size to allow

collection of adequate sample sizes. Approval for animal use was obtained from the

University of Florida’s Institutional Animal Care and Use Committee (IACUC; #Z074),

and approval for animal collection from preserve areas was obtained from the Florida

Department of Environmental Protection (#01040212) and the United States Fish and

Wildlife Service (#4151502005; #4151102002).

Samples were collected throughout Seahorse Key in a rough representation of the

population density variation found on the island, and were thus collected in greatest

numbers from areas rich in nesting bird rookeries. The correlation between snake and

nesting bird densities supports the theory that bird rookeries are the primary food source

for the adult cottonmouth population. Samples collected from Gardner Point and the

vegetation along the beach - areas where rookeries are common - comprised 40% and

36% of the total sample, respectively. Samples collected from the upland ridge and dock

15

area, where birds are less common, comprised 16% and 8%, respectively. Sample

collection from Gardner Point was completed in a single day, whereas sample collection

from the beach, ridge, and dock areas required numerous visits to the island. The

disproportionate amount of time spent sampling Gardner Point relative to the rest of

Seahorse Key probably indicates that cottonmouth population density is much greater

there.

The Paynes Prairie State Preserve study site is located approximately 97 km

northeast of Seahorse Key and was chosen for its close proximity to the University of

Florida, large size (85 km2), and for its centralized location between the eastern and

western coasts of northern Florida. Cottonmouths from Paynes Prairie State Preserve

were collected primarily from two distinct locations approximately 8 km apart along a

riverbed. The type locality of Agkistrodon piscivorus conanti is located 11 km SE of

Gainesville, FL (Gloyd 1969), in close proximity to samples collected from Paynes

Prairie State Preserve during this study.

The Lower Suwannee National Wildlife Refuge (214 km2) was selected as a study

site for its coastal mainland location and close proximity to Seahorse Key. The

Suwannee River, with an average flow rate of 304.7 m3/s (Nordlie 1990), is a possible

dispersal vector for cottonmouth snakes in that area, and thus, snakes from the Lower

Suwannee National Wildlife Refuge could potentially be frequent migrants to the islands

of the Cedar Keys. The Lower Suwannee National Wildlife Refuge study site is located

approximately 40 km north of Seahorse Key. With the exception of three individuals

found near park access roads, all cottonmouths from Lower Suwannee National Wildlife

Refuge were collected along nine miles of County Road 349.

16

Methods

Cottonmouth samples for DNA analysis from Seahorse Key (25), Paynes Prairie

State Preserve (25), and Lower Suwannee National Wildlife Refuge (22) were collected

in numbers large enough to ensure an adequate sampling of each population’s genetic

structure. Shed skins were collected and stored in paper when possible. Fresh tissue

samples were obtained via scale clipping (when live animals were encountered) or tail tip

severance (when freshly killed animals were encountered) and were saved in buffer

solution of saturated NaCl, 25 mM EDTA (pH 7.5) and 20% DMSO (modified from

Amos & Hoelzel 1991).

Live snakes were captured using a steel snake hook and were safely restrained within

a plastic tube of appropriate diameter while scale clipping took place. A single scale of

keratinized epidermis was excised from the aseptically swabbed posterior end of the animal

(in accordance with IACUC permit #Z074). Styptic powder was applied to the clipped

area with a moistened, cotton-tipped applicator prior to release to minimize any bleeding.

No noticeable change in behavior was observed following scale clipping, and an individual

snake that was spotted beneath a particular tree multiple times before capture was observed

there weeks after the procedure.

Of the Seahorse Key DNA samples, 24 were collected as shed skins and one as a

tail tip. Of Lower Suwannee DNA samples, 15 were collected as tail tips and seven were

collected as scale clippings. Of the Paynes Prairie DNA samples, ten were collected as

scale clippings, nine as tail tips, and six as shed skins. PCR techniques allow for the

effective use of extremely limited quantities of DNA in genetic analyses (Busch et al.

2000), so these methods of collection yielded acceptable quantities of DNA.

17

DNA was isolated from shed skins using a modified CTAB protocol (Clark 1998)

and from tissue samples using standard phenol/chloroform methodology (Hillis et al.

1996). Isolated DNA was then amplified using PCR. All molecular techniques were

conducted in the University of Florida’s Biotechnologies for the Ecological,

Evolutionary, and Conservation Sciences Laboratory in the Interdisciplinary Center for

Biotechnology Research (ICBR). Isolated DNA was then amplified using PCR.

Purified cottonmouth DNA samples were digested with a 6 base-pair rare-cutter

restriction enzyme, EcoRI, and a 4 base-pair common-cutter restriction enzyme, MseI.

Digested DNA was preselectively amplified (Fig. 2) and then selectively amplified (Fig.

3) using three adapter-specific primer combinations (EcoRI-ACC + MseI-ACTT, EcoRI-

ACC + MseI-ACAA, and EcoRI-ACC + MseI-ACTA; see Table 1). These three primer

combinations were selected from sixteen available primer combinations following

preliminary tests (Roark, unpublished data). Primer sets produced sufficient DNA

fragments of appropriate size and number for analysis.

Selectively amplified DNA fragments were separated by electrophoresis in a 5%

polyacrylamide gel. The resulting banding pattern was imaged using a Typhoon

fluorescence scanner (Amersham Pharmacia Biotech, Uppsala, Sweden) in the University

of Florida’s ICBR Molecular Services Core, and the presence or absence of fluorescent

AFLP marker bands was scored using the default settings for FragmeNT (Version 1.1,

Molecular Dynamics) software. Manual scoring of bands corrected for obvious software

algorithm errors, and only bands that could be scored unambiguously were used for

analyses. The highest and lowest 10% of molecular weight bands were also excluded

because bands of low weight may include products from the interaction of primers alone,

18

and bands of great weight may be strongly influenced by differences in reaction

conditions and DNA quality (Bagley et al. 2001). Band sizes were determined by the

FragmeNT software based on comparison to DNA standards (Invitrogen Low DNA Mass

Ladder; Invitrogen 50 bp DNA Ladder, Carlsbad, CA, USA).

Analyses

Bands were classified as either monomorphic (found in all samples) or

polymorphic (found in some but not all samples). Polymorphic bands were binomially

scored for each snake with a ‘1’ for the presence of a band and a ‘0’ for the absence of a

band at each polymorphic locus. All monomorphic bands were omitted from the

analyses.

Thirty-three polymorphic loci were analyzed using Tools For Population Genetics

Analysis© (Version 1.3) software (Miller 1997). The software calculated population

subdivision using Wright’s F-Statistic (Φst) (Weir & Cockerham 1984) and a 95%

confidence interval (based on 5000 bootstrap replicates). The accumulated number of

gene differences per locus (Genetic distance, D, Nei 1972), the proportion of genes that

are common to both populations being compared (Genetic identity, I, Nei 1972), and

Nei’s unbiased heterozygosity estimate (1978) were also calculated.

The rate of gene flow between populations was assessed using Wright’s F-Statistic

to estimate the number of migrants between populations per generation (Nm). Nm was

calculated based on Φst = 1/(1+4Nm) (Wright 1943, 1969). Low rates of gene flow (less

than 1 individual per generation) will indicate that the differences seen between

populations may be attributable to genetic drift, whereas high rates of gene flow (more

than 1 individual per generation) will suggest this scenario unlikely and point to

19

mechanisms other than genetic drift as the source of this variation (Slatkin 1987, Avise

1994).

Results

Apparent differences in phenotype and behavior in Seahorse Key cottonmouths

relative to mainland cottonmouths were not reflected in the assessment of genetic

variation within and between populations. When the genetic relationship between the

Seahorse Key and Lower Suwannee populations was assessed, Φst = 0.020 with a 95%

C.I. of 0.028 - 0.012, genetic identity (I) = 0.989, genetic distance (D) = 0.012, and Nm =

12.4. The Seahorse Key and Paynes Prairie relationship showed Φst = 0.009 with a 95%

C.I. of 0.017 – 0.002, I = 0.995, D = 0.005, and Nm = 29.2. Finally, in the Lower

Suwannee and Paynes Prairie relationship, Φst = 0.005 with a 95% C.I. of 0.0115 - 0.000,

I = 0.997, D = 0.003, and Nm = 48.8. Nei’s (1978) unbiased heterozygosity estimate

indicated an average heterozygosisty of 35.6% among alleles in the Seahorse Key

population, 43.9% among alleles in the Lower Suwannee population, and 39.1% among

alleles in the Paynes Prairie population (Table 1).

Discussion

The close proximity of all Φst values to zero indicates a near-complete absence of

interpopulation genetic variation and suggests that significant gene flow exists between

all three populations. Genetic identity values near 1 and genetic distance values near 0

indicate that nearly 100% of genes present are shared by these populations. Estimated

numbers of migrants per generation range from 12.4 between Seahorse Key and Lower

Suwannee to 48.8 between Paynes Prairie and Lower Suwannee, as calculated directly

from Φst values. This result further suggests a minimal amount of genetic divergence

between populations.

20

These results, however, do not necessarily mean that the Seahorse Key

cottonmouth population is not somewhat isolated. Survival of harsh shifts in

environmental conditions, such as those associated with isolation in a novel habitat, is

believed to be largely dependant on plasticity (Bonnet et al. 1999) rather than genetic

divergence in most systems. It has been theorized that differences between island and

mainland populations are probably short-term responses to increased longevity and high

population densities on islands rather than genetic drift or directional selection following

the isolation incident (Adler and Levins 1994).

It should also be pointed out that slow mutation rates leading to small population

divergences are assumed to be correlated with slow metabolic rates and relatively long

life spans. Organisms with low metabolic rates and long life spans tend to have longer

nucleotide regeneration times and thus may have a slower rate of molecular evolution

(Martin and Palumbi 1993). As an ectothermic vertebrate with a potential life span of

approximately 20 years, A. piscivorus almost certainly possesses one, if not both, of these

characteristics. Previous studies of intrapopulation genetic variability in cottonmouths

support this potential explanation. Low levels of genetic variation have been found

between populations of A. piscivorus piscivorus at the northern edge of its range in

southeastern Virginia (I = 0.964; Merkle 1985), and populations of A. piscivorus

leucostoma located east of Memphis, TN, showed no difference in DNA content when

flow cytometry was employed (Tiersch et al. 1990).

Given that previous studies have suggested limited genetic variability throughout

the range of this species (Merkle 1985, Tiersch et al. 1990), trends observed within the

low Φst values, small genetic distances, and high genetic identity scores should be

21

considered when addressing the stated hypotheses. Although gene flow appears

uninterrupted between populations in this study, the hypothesis that the Seahorse Key

cottonmouth population is more genetically distinct from the mainland populations than

these populations are from each other is supported by the finding that the genetic

distances and Φst values between the Seahorse Key population and either of the mainland

populations are larger than the genetic distance and Φst value between the two mainland

populations. The estimated rate of migration between the mainland populations is also

much higher than the estimated rate of migration between either of the two mainland

populations and the Seahorse Key population. Genetic identity scores also indicate that

the two mainland populations have a slightly higher percent genetic similarity to each

other than either does to the Seahorse Key population.

Similarly, the hypothesis that gene flow into the Seahorse Key population is greatly

reduced relative to that between mainland populations is supported by the finding that Φst

value is smallest, and consequently the migration rate is greatest, between the two

mainland populations when compared to the Φst values and migration rates between

either mainland population and the Seahorse Key population. Also, the lowest estimated

heterozygosity was found in the Seahorse Key population (35.6% versus 39.1% and

43.9%), possibly indicating reduced dispersal ability in this population relative to the

mainland populations.

Although these small genetic differences may allow us to begin theorizing about

general trends in the ecologies of the study populations, they are insufficient to establish

a quantitative measure of the extent to which the Seahorse Key cottonmouth population is

isolated. Variations in Φst, genetic distance, and genetic identity values between

22

populations are miniscule, and bootstrapping indicates that the Φst values’ 95% C.I.s do

overlap. Nei’s unbiased heterozygosity estimate may offer the most reliable quantitative

measurement of the Seahorse Key cottonmouth population’s geographical isolation, and

thus a genetic investigation utilizing a co-dominant marker system that allows the direct

measurement of heterozygosity (i.e., microsatellites) could be beneficial.

The final goal of this research was to establish a knowledge base about the

Seahorse Key system that, when combined with current and future research in this

location, could assist in the generation of new hypotheses regarding the effects of habitat

fragmentation in this and possibly other vertebrate species. Through this study, it has

become apparent that significant genetic divergence between this island population and

mainland populations has not occurred. Whether this lack of divergence exists because

the Seahorse Key population is not geographically isolated, because it has not been

isolated for a sufficient period of time to affect the genetic composition of its members,

or because the microevolutionary processes associated with population isolation work

slowly in this particular taxon can not be determined from these data.

However, the knowledge that the Seahorse Key population is not genetically

distinct from mainland populations is inherently valuable. Keogh et al. (2003) suggested

that a lack of genetic diversity between insular and mainland snake populations may have

positive implications for species conservation because the loss of an isolated population

may not significantly reduce genetic diversity within the entire species. This statement

does not mean, however, that continued monitoring and management of these systems is

not required, since the loss of insular populations such as the Seahorse Key cottonmouths

could affect prey populations or open niches for invasive or introduced species capable of

23

unbalancing the entire ecosystem (Keogh et al. 2003). Furthermore, the variations that

were detected in Φst values, genetic distances, genetic identities, and estimated

heterozygosities between these populations occur in a manner suggesting that

microevolutionary processes commonly associated with genetic isolation could be in their

early stages in the Seahorse Key cottonmouth population. In light of the pattern of slight

divergences demonstrated by this study, these results serve to heighten awareness of the

need for a non-genetic study of the interpopulation migration rates of island and mainland

cottonmouth snakes in and around the Cedar Keys National Wildlife Refuge.

24

Figure 1. Study sites.

25

1) 72°C – 2 minutes – Initial incubation 2) 94°C – 20 seconds – denaturing 3) 56°C – 30 seconds – annealing 4) 72°C – 2 minutes – extension 5) 72°C – 2 minutes – Final extension 6) 60°C – 30 minutes – Final incubation 7) 4°C – Storage

20 cycles

Figure 2. Preselective amplification program. Figure 3. Selective amplification program.

1) 94°C – 2 minutes – Initial denaturation

2) 94°C – 20 seconds - denaturation 3) 66°C – 30 seconds - annealing 4) 72°C – 2 minutes – extension

5) 94°C – 20 seconds – denaturation 6) Decrease 1°C per cycle – 30 seconds – annealing 7) 72°C - 2 minutes – extension 8) 94°C – 20 seconds – denaturation 9) 56°C – 30 seconds – annealing 10) 72°C – 2 minutes – extension 11) 60°C – 30 minutes – Final incubation 12) 4°C - Storage

20

9

26

Table 1. AFLP selective amplification primers used in study.

Name Sequence: 5’ – 3’ EcoRI-ACC GACTGCGTACCAATTCACC MseI-ACTT GATGAGTCCTGAGTAACTT MseI-ACAA GATGAGTCCTGAGTAACAA MseI-ACTA GATGAGTCCTGAGTAACTA

Table 2. Comparison of inter-population relationships.

Populations Φst C.I. I D Nm SHK-LSNWR 0.020 0.028 - 0.012 0.989 0.012 12.4

SHK-PP 0.009 0.017-0.012 0.995 0.005 29.2 LSNWR-PP 0.005 0.012-0.000 0.997 0.003 48.4

LIST OF REFERENCES

Adler, G. H. 1996. The island syndrome in isolated populations of a tropical forest rodent. Oecologia 108:694-700.

Adler, G. H., and R. Levins. 1994. The island syndrome in rodent populations. The Quarterly Review of Biology 69:473-487.

Amos, B., and A. R. Hoelzel. 1991. Long-term preservation of whale skin for DNA analysis. Report of the International Whaling Commission 1991:99-103.

Anderson, R. P., and C. O. Handley. 2002. Dwarfism in insular sloths: biogeography, selection, and evolutionary rate. Evolution 56:1045-1058.

Andren, C., and G. Nilson. 1983. Reproductive tactics in an island population of adders, Vipera berus (L.), with fluctuating food resource. Amphibia-Reptilia 4:63-79.

Auffenberg, W. 1963. The fossil snakes of Florida. Tulane Studies in Zoology 10:131-216.

Avise, J. C. 1994. Molecular markers, natural history, and evolution. Chapman and Hall, New York, NY.

Avise, J. C. 1995. Mitochondrial DNA polymorphism and a connection between genetics and demography of relevance to conservation. Conservation Biology 9:686-690.

Bagley, M., S. Anderson, and B. May. 2001. Choice of methodology for assessing genetic impacts of environmental stressors: polymorphism and reproducibility of RAPD and AFLP fingerprints. Ecotoxicology 10:239-244.

Blem, C. R. 1981. Reproduction of the eastern cottonmouth Agkistrodon piscivorus (Serpentes: Viperidae) at the northern edge of its range. Brimleyana 5:117-128.

Blem, C. R., and L. B. Blem. 1995. The eastern cottonmouth (Agkistrodon piscivorus) at the eastern edge of its range. Journal of Herpetology 29:391-398.

Boback, S. M., and C. Guyer. 2003. Empirical evidence for an optimal body size in snakes. Evolution 57:345-351.

Bonnet, X., D. Bradshaw, R. Shine, and D. Pearson. 1999. Why do snakes have eyes? The (non-)effect of blindness in island tiger snakes (Notechis scutatus). Behavioral Ecology and Sociobiology 46:267-272.

27

28

Brattstrom, B. H. 1953. Notes on a population of leaf-nosed snakes Phyllorhynchus decurtatus perkinsi. Herpetologica 9:57-64.

Bull, J. J., and R. Shine. 1979. Iteroparous animals that skip opportunities for reproduction. American Naturalist 114:296-316.

Burger, L. W. 1971. Genera of pitvipers (Serpentes: Crotalidae). University of Kansas Press, Lawrence, KS.

Burkett, R. 1966. Natural history of the cottonmouth moccasin, Agkistrodon piscivorus (Reptilia). University of Kansas Publication 17:435-491.

Busch, J. D., M. P. Miller, E. H. Paxton, M. K. Sogge, and P. Keim. 2000. Genetic variation in the endangered southwestern willow flycatcher. The Auk 117:586-595.

Bushar, L. M., H. K. Reinert, and L. Gelbert. 1998. Genetic variation and gene flow within and between local populations of the timber rattlesnake, Crotalus horridus. Copeia 1998:411-422.

Carr, A. F., Jr. 1936. The gulf-island cottonmouths. Proceedings of the Florida Academy of Sciences 1:86-90.

Clark, A. 1998. Reptile sheds yield high quality DNA. Herpetological Review 29:17-18.

Conant, R. 1990. The fossil history of the genus Agkistrodon in North America. Pages 539-543 in H. K. Gloyd and R. Conant, editors. Snakes of the Agkistrodon complex: a monographic review. The Society for the Study of Amphibians and Reptiles, Saint Louis, MO.

Conant, R., and J. Collins. 1998. Peterson field guides: reptiles and amphibians (Eastern/Central North America). Houghton Mifflin Company, New York, NY.

Dodd, C. K. J. 1993. Strategies for snake conservation. Pages 363-393 in R. A. Seigel and J. T. Collins, editors. Snakes: ecology and behavior. McGraw Hill, New York, NY.

D'Surney, S., L. Shugart, and C. Theodorakis. 2001. Genetic markers and genotyping methodologies: an overview. Ecotoxicology 10:201-204.

Ernst, C. H. 1992. Venomous reptiles of North America. Smithsonian Institution Press, Washington, DC.

Erwin, T. L. 1991. An evolutionary basis for conservation strategies. Science 253:750-752.

Foster, J. B. 1964. Evolution of mammals on islands. Nature 202:234-235.

Frankham, R. 1995. Conservation genetics. Annual Review of Genetics 29:305-327.

29

Galis, F., and J. A. J. Metz. 1998. Why are there so many cichlid species? Trends in Ecology and Evolution 13:1-2.

Giannasi, N., S. Thorpe, and A. Malhotra. 2001. The use of amplified fragment length polymorphism in determining species trees at fine taxonomic levels: analysis of a medically important snake, Trimeresurus albolabris. Molecular Ecology 10:419-426.

Gliwicz, J. 1980. Island populations of rodents: their organization and functioning. Biological Review 55:109-138.

Gloyd, H. K. 1969. Two additional subspecies of North American crotalid snakes, genus Agkistrodon. Proceedings of the Biological Society of Washington 82:219-232.

Gloyd, H. K., and R. Conant. 1990. Snakes of the Agkistrodon complex: a monographic review. Society for the Study of Amphibians and Reptiles, Saint Louis, MO.

Grant, P. R. 1998. Epilogue and questions. Pages 305-320 in P. R. Grant, editor. Evolution on islands. Oxford University Press, Oxford.

Harris, L. D. 1984. The fragmented forest: island biogeography theory and the preservation of biotic diversity. The University of Chicago Press, Chicago, IL.

Hillis, D. M., C. Moritz, and B. Mable. 1996. Molecular systematics. Sinauer Associates Incorporated Publishers, Sunderland, MA.

Keogh, S. J., I. A. W. Scott, M. Fitzgerald, and R. Shine. 2003. Molecular phylogeny of the venomous snake genus Hoplocephalus (Serpentes, Elapidae) and conservation genetics of the threatened H. stephensii. Conservation Genetics 4:57-65.

King, R. B. 1993. Determinants of offspring number and size in the brown snake, Storeria dekayi. Journal of Herpetology 27:175-185.

Klauber, L. M. 1972. Rattlesnakes: their habits, life histories and influence on mankind. University of California Press, Berkeley, CA.

Knight, A., L. D. Densmore, and E. D. Rael. 1992. Molecular systematics of the Agkistrodon complex. Pages 49-70 in J. A. Campbell and E. D. Brodie, editors. Biology of the pitvipers. Selva, Tyler, TX.

Krauss, F., D. G. Mink, and W. M. Brown. 1996. Crotaline intergeneric relationships based on mitochondrial DNA sequence data. Copeia 1996:763-773.

Lande, R., and G. F. Barrowclough. 1987. Effective population size, genetic variation, and their use in population management. Pages 87-123. in M. E. Soulé, editor. Viable populations for conservation. Cambridge University Press, Cambridge.

30

Lawson, R., and R. B. King. 1996. Gene flow and melanism in Lake Erie garter snake populations. Biological Journal of the Linnean Society 59:1-19.

Li-xin, S., R. Shine, Z. Debi, and T. Zhengren. 2002. Low costs, high output: reproduction in an insular pit-viper (Gloydius shedaoensis, Viperidae) from north-eastern China. Journal of Zoology, London 256:511-521.

Luiselli, L., M. Capula, and R. Shine. 1996. Reproductive output, costs of reproduction, and ecology of the smooth snake, Coronella austriaca, in the eastern Italian Alps. Oecologia 106:100-110.

MacArthur, R. H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton University Press, Princeton, NJ.

Mader, H. J. 1984. Animal habitat isolation by roads and agricultural fields. Biological Conservation 29:81-96.

Martin, A. P., and S. R. Palumbi. 1993. Body size, metabolic rate, generation time and the molecular clock. Proceedings of the National Academy of Sciences 90:4087-4091.

McNab, B. K. 1994. Resource use and the survival of land and freshwater vertebrates on oceanic islands. The American Naturalist 144:643-660.

Merkle, D. A. 1985. Genetic variation in the eastern cottonmouth, Agkistrodon piscivorus piscivorus (Lacepede) (Reptilia: Crotalidae), at the northern edge of its range. Brimleyana 11:55-61.

Miller, M. P. 1997. Tools for population genetic analysis (TFPGA), version 1.3: A Windows program for the analysis of allozyme and molecular population genetic data. Distributed by author.

Miller, R. A., R. Dysko, C. Chrisp, R. Seguin, L. Linsalata, G. Buehner, J. M. Harper, and S. Austad. 2000. Mouse (Mus musculus) stocks derived from tropical islands: new models for genetic analysis of life-history traits. Journal of Zoology, London 250:95-104.

Moritz, C. 1994. Defining "evolutionarily significant units" for conservation. Trends in Ecology and Evolution 9:373-375.

Muller, H. J. 1964. The relation of recombination to mutational advance. Mutation Resource 1:2-9.

Negi, M., A. Singh, and M. Lakshmikumaran. 2000. Genetic variation and relationship among and within Withania species as revealed by AFLP markers. Genome 43:975-980.

Nei, M. 1972. Genetic distance between populations. American Naturalist 106:283-292.

31

Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583-590.

Nordlie, F. G. 1990. Rivers and springs. Pages 392-422 in R. L. Myers and J. J. Ewel, editors. Ecosystems of Florida. University of Central Florida Press, Orlando.

Padmanabhan, P., and Y. Yom-Tov. 2000. Breeding season and clutch size of Indian passerines. Ibis 142:75-81.

Palumbi, S., and C. S. Baker. 1994. Contrasting population structure from nuclear intron sequences and mtDNA of humpback whales. Molecular Biology and Evolution 11:426-435.

Parkinson, C. L. 1999. Molecular systematics and biogeographical history of pitvipers as determined by mitochondrial ribosomal DNA sequences. Copeia 3:576-586.

Parkinson, C. L., S. M. Moody, and J. E. Ahlquist. 1997. Phylogenetic relationships of the “Agkistrodon complex” based on mitochondrial DNA sequence data. Pages 63-78 in R. S. Thorpe, W. Wuster, and A. Malhotra, editors. Venomous snakes: ecology, evolution, and snakebite. Oxford University Press, London.

Parkinson, C. L., K. R. Zamudio, and H. W. Greene. 2000. Phylogeography of the pitviper clade Agkistrodon: historical ecology, species status, and conservation of cantils. Molecular Ecology 9:411-420.

Ribeiro, M. M., S. Mariette, G. G. Vendramin, A. E. Szmidt, C. Plomion, and A. Kremer. 2002. Comparison of genetic diversity estimates within and among populations of maritime pine using chloroplast simple-sequence repeat and amplified fragment length polymorphism data. Molecular Ecology 11:869-877.

Roff, D. A. 1992. The evolution of life histories. Chapman & Hall, New York, NY.

Savitzky, B. A. C. 1992. Laboratory studies on piscivory in an opportunistic pitviper, the cottonmouth, Agkistrodon piscivorus. Pages 347-368. in A. L. Campbell and E. D. Brodie JR, editors. Biology of the pitvipers. Selva, Tyler, TX.

Schwaner, T. D. 1990. Geographic variation in scale and skeletal anomalies of tiger snakes (Elapidae: Notechis scutatus-ater complex) in southern Australia. Copeia 1990:1168-1173.

Schwaner, T. D., and S. D. Sarre. 1988. Body size of tiger snakes in southern Australia, with particular reference to Notechis ater serventyi (Elapidae) on Chappell Island. Journal of Herpetology 22.

Slatkin, M. 1987. Gene flow and the geographic structure of natural populations. Science 236:787-792.

32

Soulé, M. 1966. Trends in the insular radiation of a lizard. The American Naturalist 100:47-61.

Templeton, A. R., K. Shaw, E. Routman, and S. K. Davis. 1990. The genetic consequences of habitat fragmentation. Annals of the Missouri Botanical Garden 77:13-27.

Tiersch, T. R., C. R. Figiel Jr., R. M. Lee III, R. W. Chandler, and A. E. Houston. 1990. Use of cytometry for the effects of environmental mutagens: baseline DNA values in cottonmouth snakes. Bulletin of Environmental Contamination and Toxicology 45:833-839.

Vos, P., R. Hogers, and M. Bleeker. 1995. AFLP: a new technique for DNA fingerprinting. Nucleic Acids Research 23:429-435.

Wanless, H.R. 1982. Sea level is rising – so what?. Journal of Sedentary Petrology 52: 1051-1054.

Weatherhead, P. J., G. P. Brown, M. R. Prosser, and K. J. Kissner. 1999. Factors affecting neonate size variation in northern water snakes, Nerodia sipedon. Journal of Herpetology 33:577-589.

Weir, B. S., and C. C. Cockerham. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38:1358-1370.

Westemeier, R. L., J. D. Brawn, S. A. Simpson, T. L. Esker, R. W. Jansen, J. W. Walk, E. L. Kershner, J. L. Bouzat, and K. N. Paige. 1998. Tracking the long-term decline and recovery of an isolated population. Science 282:1695-1698.

Wharton, C. H. 1966. Reproduction and growth in the cottonmouths, Agkistrodon piscivorus Lacépède, of Cedar Keys, Florida. Copeia 2:149-161.

Wharton, C. H. 1969. The cottonmouth moccasin on Sea Horse Key, Florida. Bulletin of the Florida State Museum 14:227-272.

Wiggins, D. A., A. P. Moller, M. F. L. Sorensen, and L. A. Brand. 1998. Island biogeography and the reproductive ecology of great tits Parus major. Oecologia 115:478-482.

Williams, G. C. 1966. Adaptation and natural selection. Princeton University Press, Princeton, NJ.

Wright, S. 1943. Isolation by distance. Genetics 31:114-138.

Wright, S. 1969. Evolution and genetics of populations Vol. 2: the analysis of gene frequencies. University of Chicago Press, Chicago, IL.

33

Wynne-Edwards, V. C. 1962. Animal dispersion in relation to social behavior. Hafner Press, New York, NY.

BIOGRAPHICAL SKETCH

Andrew Warner Roark was born on December 6, 1976, in Charlotte, NC, to Roger

and Annette Roark. He earned the degree of Bachelor of Science in biology with a

concentration in medical humanities from Davidson College, Davidson, NC, in May

1999. Following graduation, Andrew began conducting research in the Laboratory of

Developmental Neurobiology, National Institutes of Child Health and Human

Development, National Institutes of Health, Bethesda, MD. In August 2000, he began

graduate studies in the Department of Zoology, University of Florida.

34