Aspects of embryonic physiology and the evolution of viviparity in the lizard genus Sceloporus

MISSING IMAGE

Material Information

Title:
Aspects of embryonic physiology and the evolution of viviparity in the lizard genus Sceloporus
Physical Description:
xii, 145 leaves : ill. ; 28 cm.
Language:
English
Creator:
Demarco, Vincent Gerard, 1955-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 134-144).
Statement of Responsibility:
by Vincent Gerard Demarco.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001693334
notis - AJA5413
oclc - 25221960
System ID:
AA00003294:00001

Full Text











ASPECTS OF EMBRYONIC PHYSIOLOGY AND
THE EVOLUTION OF VIVIPARITY IN THE LIZARD GENUS SCELOPORUS
















By


VINCENT GERARD DEMARCO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1991

































To Beth,

coauthor of our most important work.















ACKNOWLEDGEMENTS


I would like to thank all of my friends who helped in

different ways. They are: Paul Andreadis, John Donald, Andy

Gannon, Yeong-Choy Kam, Mike Lacy, Greg Masson, John Matter,

John Paine, Brent Palmer, Steve Platania, Lou Somma, Mike

Thompson, and Kent Vliet. I would also like to thank my

committee members, Drs. John Anderson, Donald Caton, Louis

Guillette, Harvey Lillywhite, and Michele Wheatly. I

especially appreciate the many conversations I had with Lou

Guillette, the generous use of John Anderson's equipment, and

Michele's kind help with my research following my accident.















TABLE OF CONTENTS

ACKNOWLEDGEMENTS ...................... ................... iii

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

LIST OF FIGURES .......................................... viii

ABSTRACT .................................................... x

CHAPTERS

1. INTRODUCTION AND LITERATURE REVIEW ....................... 1
Egg Retention............................................ 4
Benefits and Costs of a Reproductive Mode................ 7
Viviparity and Parental Investment.....................10
Hypothesis Testing...................................... 1
Embryonic Development and Viviparity...................13
Outline of Studies .....................................19

2. METHODS ................................. ................ 24
Collection and Maintenance of Adult Lizards............24
Collection and Incubation of Eggs.......................25
Measuring Oxygen Consumption... .........................27
Statistical Analysis................................... 29

3. ESTIMATING EGG RETENTION TIMES .......................... 30
Methods .............................................. 31
Terminology .......................................... 31
Embryonic Stages at Oviposition ......................32
Total Embryonic Development Time ..................... 33
A Direct Method for Calculating %ERT .................34
An Indirect Method for Calculating %ERT .............. 35
Results.................................................36
Embryo Stages at Oviposition ..........................36
Directly Estimating %ERT in a. woodi .................37
Embryo Stage and Time to Hatching ....................38
Heterochrony and Embryo Stages .......................38
Discussion ...... .....................................44

4. RAPID EMBRYONIC DEVELOPMENT: AN ADAPTATION TO COLD
CLIMATE. ................................................ 52
Methods. ............................................... 56
Development Time in Oviparous Species ................56
Laboratory study ................................... 56
Field estimates of development time ................57
Development Time in a Viviparous Species .............57









Results................................. ...............59
Development Time in Oviparous Species ................59
Laboratory study ...................................59
Field observations on development time ............. 59
Environmental Variation Between Study Sites .......... 62
TEDT in a Viviparous Species ......................... 63
Discussion............................................. 63
Estimates of TEDT .................................... 63
TEDT and Hatchling Size .............................. 67
TEDT in a Viviparous Species ..........................70
Field Estimates of TEDT ..............................71
Implications for the Evolution of Viviparity .........73

5. EMBRYONIC METABOLISM ....................................78
Methods ................................................ 81
Embryonic Metabolism .................................81
Oviparous species ..................................81
Viviparous species .................................81
Statistical Analysis .................................82
Results ................................................ 83
Sceloporus woodi .....................................83
Sceloporus virgatus .................................. 84
Sceloporus scalaris ..................................85
Sceloporus iarrovi................................... 86
Discussion ............................................. 92

6. THE PHYSIOLOGICAL COST OF PREGNANCY ....................104
Methods... ............................................ 106
Adult Metabolism ................................... 106
Egg Metabolism ...................................... 107
Estimating MCP ...................................... 108
Results ......................... ......... .. ......... 109
Metabolic Pattern During Pregnancy .................. 109
Adult Metabolism .................................... 110
Embryo Metabolism ...................................... 110
Estimating MCP ...................................... 111
Discussion ............................................ 122
Pattern of Metabolism During Pregnancy .............. 122
Maintenance Cost of Pregnancy ....................... 122
Pregnancy and Standard Allometry .................... 125

7. FUTURE DIRECTIONS ............... ...................... 131

LIST OF REFERENCES ........................................ 134

BIOGRAPHICAL SKETCH ....................................... 145
















LIST OF TABLES


Table 3.1. Polynomial equations relating embryo stage
(y) and days before hatching (x) for Sceloporus wood,
S.virgatus, and a. scalaris embryos developing at
300C .................................................... 41

Table 3.2. Relative egg retention times for three
species of lizards in the genus Sceloporus for embryos
developing at 300C. Methods for calculating these
variables and abbreviations are described in the
text ..................................................... 42

Table 4.1. Modal (range) embryo stage at oviposition
(ESO) and mean total embryonic development time (TEDT)
at 300C in three oviparous species of Sceloporus.
Estimates of these variables based on field
observations are also reported (see text for
explanation).............................................64

Table 6.1. Allometry of oxygen consumption (cc 02 h-1)
as a function of body mass (g).......................... 116

Table 6.2. Pairwise comparisons of adjusted mean log
metabolic rate of allometric equations for pregnant
and nonpregnant a. jarrovi. Adjusted mean logs were
determined by an analysis of covariance. Adjusted
mean rates of metabolism antilogss, cc 02 h-1) are
shown in parentheses. See Table 6.1 for equations and
regression statistics................................... 117

Table 6.3. A comparison of estimates of the
instantaneous maintenance cost of pregnancy (MCP) in
late pregnant Sceloporus jarrovi, weighing between 10
and 30 grams, using the method of Beuchat and Vleck
(1990). Embryo metabolism was estimated by
substituting mean hatchling mass into the equations
describing male metabolism as a function of body
size .................................................... 120









Table 6.4. A comparison of estimates of the
instantaneous maintenance cost of pregnancy (MCP) in
late pregnant Sceloporus jarrovi weighing from 10 to
30 g using the method of Beuchat and Vleck (1990).
Embryo metabolism was estimated by substituting mean
hatchling mass into eq. 6.7............................. 121

Table 6.5. Mean ( 1 S.E.) mass-specific Vo2 (cc g-1
h-1) and mass (g) of adult Sceloporus iarrovi......... 129














LIST OF FIGURES


Figure 3.1. A frequency distribution of embryonic
stages at oviposition for Sceloporus woodi, a.
virgatus, and S. salaries ............................... 39

Figure 3.2. A time schedule of embryonic stages for
Sceloporus woodi, S. virgatus, and S. scalaris.............40

Figure 3.3. The relationships between embryo stage and
% total embryonic development time (TEDT) indicate
that there is little evidence for heterochrony in the
gross morphology of embryonic development in
Sceloporus woodi, -. vircatus, and S. scalaris...........43

Figure 4.1. Relationship between mean total embryonic
development time and mean hatchling mass in Sceloporus
woodi, S. virgatus, and S. scalaris and a. jarrovi.......69

Figure 4.2. A model for the evolution of viviparity in
the genus Sceloporus. The model shows that as ground
temperatures decrease, the probability of clutch
survivorship increases as egg retention time and
developmental rates increase............................. 76


Figure 5.1. Vo2 as a function of total embryonic
development time for Sceloporus woodi eggs. Curve
represents a fourth order polynomial regression...........87


Figure 5.2. Vo2 of 7 Sceloporus woodi eggs from 2
clutches as a function of day of incubation. Each
curve represents several measures on a single egg.
(a) Eggs from clutch 1. (b) Eggs from clutch 2...........88


Figure 5.3. Vo2 as a function of total embryonic
development time for Sceloporus virgatus eggs. All
points are independent and curve represents a fourth
order polynomial regression..............................89


Figure 5.4. Vo2 as a function of total embryonic
development time for Sceloporus scalaris eggs. All


viii









points are independent and curve represents a fourth
order polynomial regression..............................90


Figure 5.5. Vo2 as a function of total embryonic
development time for Sceloporus Jarrovi eggs. All
points are independent and curve represents a fourth
order polynomial regression..............................91


Figure 5.6. Embryonic patterns of Vo2 as a function of
% total embryonic development time (%TEDT) in 4
species of Sceloporus with different periods of egg
retention ...... ........................................96

Figure 6.1. A line graph of the pattern of rate of 02
consumption (cc 02 h -1) during pregnancy in seven
Sceloporus iarrovi. All rates of 02 consumption are
expressed as percentages of the maximum 02 consumption
of the individual. Line segments represent
interpolation between two data points for an
individual female. Open circles represent postpartum
rate of 02 consumption. (pp = postpartum) ...............114


Figure 6.2. Vo2 as a function of body mass in adult
Sceloporus jarrovi. Data for males are not plotted in
the top graph in order to avoid confusion. Regression
lines describing data do include values for males.
Regression statistics are shown in Table 6.1............115

Figure 6.3. Bar graph showing the adjusted mean
metabolism (ANCOVA, Table 6.2) for nonreproductive
(nr), postpartum (pp), early, middle, and late
pregnant a. jarrovi. Horizontal line used to compare
the metabolism of non-reproductive females to the
metabolism of females in other reproductive
conditions. Numbers above bars represent the percent
of non-reproductive metabolism..........................118

Figure 6.4. The relationship between peak maternal
metabolism as a function of the sum of postpartum and
litter metabolism. The equation is Vo2 preg = 1.15 Vo2
pp + litter 0.372 and the slope is greater than 1 (see
text for details) .......................................119

Figure 6.5. The relationship between peak embryonic
metabolism and mean hatchling mass in 5 species of
Sceloporus lizards ......................................126















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


ASPECTS OF EMBRYONIC PHYSIOLOGY AND THE EVOLUTION OF
VIVIPARITY IN THE LIZARD GENUS SCELOPORUS

By

Vincent Gerard DeMarco

August, 1991




Chairman: Louis J. Guillette, Jr.
Major Department: Department of Zoology


Relative egg retention times (RERT) were determined for

three oviparous lizards, Sceloporus woodi, S. virgatus, and

S. scalaris. These species represent different stages along

the oviparity-viviparity continuum. The method used to

determine RERT in these species was also used to determine

the modal RERT of all lizards. Modal egg retention time

corresponds to approximately 30% of development time. Thus,

prolonged egg retention may not be as widespread as has been

suggested.

The RERT of Sceloporus woodi, a. virgatus, and S.

scalaris are consistent with the cold climate hypothesis for

the evolution of viviparity: specifically, cold climate

selects for longer periods of egg retention. I found










evidence to support the hypothesis that more rapid

embryogenesis evolves in response to cold climates. Thus,

both rapid embryogenesis and prolonged egg retention are

adaptations to reduce the time embryos take to develop in

cold climates.

No data exist on the Vo2 of lizard eggs. I measured Vo2

of eggs of Sceloporus woodi, S. virgatus, S. scalaris, and S.

iarrovi. There are differences in the patterns of Vo2 during

development that are related to RERT. A peaked pattern of

metabolism was found in S. woodi and S. virgatus, whereas S.

scalaris and S. jarrovi have sigmoidal patterns. These

patterns are not consistent with hypotheses explaining

patterns of embryonic metabolism in birds and reptiles. It

had been suggested previously that peaked patterns are

correlated with long development times in reptiles. My

results strengthen that correlation; however, the underlying

mechanisms and causes for interspecific variation in the

patterns of metabolism in reptiles are still unknown.

Several authors have suggested that a substantial

component of the resting metabolism of late pregnant

squamates (16% to 40%) is due to the physiological costs

associated with litter maintenance. Increased maintenance

costs during pregnancy may be due to the increased load on

various physiological systems. I found that the maintenance

cost of pregnancy (MCP) during late gestation in Sceloporus

jarrovi is approximately 1.5%. I suggest that because others

apparently underestimated the metabolism of the litter in









their models, they overestimated the MCP. My results do not

support the hypothesis that litter maintenance costs are a

substantial component of reproductive effort in viviparous

species.


xii















CHAPTER 1
INTRODUCTION AND LITERATURE REVIEW



Two modes of reproduction are recognized among the

reptilian order Squamata (lizards, snakes, and

amphisbaenids). Oviparity, occurring in the majority of

species, is the condition in which females lay shelled eggs

containing partially developed embryos. Embryos complete

development in a nest cavity and hatch as free-living young.

Approximately one fifth of all squamates give birth to fully

developed live young, a condition defined as viviparity. As

far as is known, hatchlings that emerge as neonates from a

live-bearing female are as precocial as those from shelled

eggs.

Viviparous species are generally assumed to have evolved

from oviparous species (Weekes, 1935; Neill, 1964; Packard et

al., 1977; Shine and Bull, 1979). Perhaps the best

supporting evidence is the presence of an egg tooth in the

embryos of viviparous species. Oviparous embryos possess an

egg tooth that is used to cut through a leathery eggshell

composed of fibrous and calcareous layers. Although

viviparous embryos are not enclosed within a leathery

eggshell, they still possess the egg tooth.

It is widely known that oviparous squamates, unlike

chelonians and crocodilians, are capable of retaining eggs










and that embryonic development continues in utero. As such,

many squamates are considered to exhibit intermediate stages

in the transition from oviparity to viviparity (Shine and

Bull, 1979). Furthermore, the degree of egg retention is

highly variable among species (Shine, 1983a), among females

within a species (Sexton and Marion, 1974), and within a

single female (DeMarco, personal observation). Some species

oviposit eggs within a week or two of ovulation (e.g., Uta

stansburiana, Tinkle, 1967; Cnemidophorus tiaris, Anderson

and Karasov, 1988). In these species, embryos are relatively

undeveloped and spend a large fraction of the total

development time in the nest. Other species retain eggs for

a long period and oviposit eggs containing relatively

advanced embryos (Opheodrys vernalis, Blanchard, 1933;

Sceloporus scalaris, Newlin, 1976; Tvyhlops bibroni, Erasmus

and Branch, 1983; Sceloporus aeneus, Guillette and Lara

Gongora, 1986). The more advanced an embryo at oviposition,

the less time it will spend in the nest. In extreme cases,

embryos may hatch within a few days (Opheodrys vernalis,

Blanchard, 1933) or hours (Sphenomorphus fragilis, Greer and

Parker, 1979) of oviposition. Thus, the reproductive modes

of some squamates strain the definitions of oviparity and

viviparity. Furthermore, all species that give birth to live

young are referred to as viviparous, regardless of placental

complexity or the degree to which nutrients are added to eggs

following ovulation (Guillette, 1982a; Yaron, 1985). Two

placental nutritional patterns have been recognized and










represent extremes of a continuum (Blackburn et al., 1985).

Lecithotrophic viviparity, also known as ovoviviparity, is

the condition in which nutrients and energy for development

come from yolk. Matrotrophic viviparity occurs when the

mother supplies the nutrition required for development during

gestation.

Occasionally the terms gravid and pregnant are confused,

as are the terms oviduct and uterus. Gravid refers to the

condition in which an oviparous female is carrying eggs

within her reproductive tract. Viviparous animals carrying

young are referred to as pregnant. Oviduct refers to the

entire reproductive tract that is derived from the Mullerian

ducts (Guillette, 1987), whereas uterus refers to a distinct

portion of the oviduct where reptilian embryos spend the

majority of their time while in the female. Additionally,

gestation refers to the time eggs spend in the oviducts and

can be used for both oviparous and viviparous animals.

Finally, the term egg is used to describe the shelled eggs of

oviparous species, as well as the unshelled eggs in

viviparous species.

Over the last 80 years investigators have suggested that

viviparity has evolved numerous times in squamates (Gadow,

1910; Weekes, 1930, 1935; Sergeev, 1940; Fitch, 1970; Tinkle

and Gibbons, 1977; Shine and Bull, 1979; Blackburn, 1982;

Shine, 1985). Documentation of a separate origin of

viviparity requires that the two modes of reproduction occur

in closely related species within the same genus (Fitch,









1970; Tinkle and Gibbons, 1977; Packard et al., 1977;

Blackburn, 1982; Shine, 1985). For instance, Shine (1985)

suggested that within the lizard genus Sceloporus

(Phrynosomatidae), there have been four to six independent

origins of viviparity. Perhaps the clearest example in the

genus is found among three very closely related species

within the Sceloorus salaries species complex. Sceloporus

salaries is oviparous and S. goldmani is viviparous.

Sceloporus aeneus was found to be reproductively bimodal

(Guillette, 1981a), i.e., some populations are viviparous and

some are oviparous. The species has been divided into two

subspecies based on parity mode. Sceloporus a. bicanthalis

is viviparous, and S. a. aeneus is oviparous. Assuming this

phylogenetic scheme is correct, there have been at least two

separate origins of viviparity within this species complex.

Subsequent to Shine's analysis, Guillette and Smith (1985)

showed that S. a. bicanthalis and S. a. aeneus were separate

closely related species. These two closely related species

may represent one of the most recent origins of viviparity.

Despite the numerous independent origins of viviparity in

squamates, little is known concerning how this parity mode

evolves.


Egg Retention


The transition from oviparity to viviparity is a complex

process that must involve numerous morphological and






5



physiological changes (Weekes, 1935; Packard et al., 1977;

Shine and Bull, 1979; Yaron, 1985; Guillette, 1989). Because

of the apparent complexity of the changes required, it has

been hypothesized that the evolution of viviparity is a

gradual process (Packard et al., 1977; Shine and Bull, 1979).

The following evolutionary scenario has been suggested

(Packard, et al., 1977; Shine and Bull, 1979). In certain

environments, females that retain eggs longer leave more

offspring. If there is strong selection for increased egg

retention over many generations, then eventually embryos

complete development in utero and are born alive. During

this process, there is progressive thinning of the eggshell.

Ultimately, the eggshell is no longer secreted. Thus, it is

clear that during the transition from oviparity to viviparity

a species proceeds through a number of intermediate stages of

egg retention. Any environmental factor hypothesized to

select for viviparity must also select for prolonged egg

retention (Weekes, 1935; Packard, et al., 1977; Tinkle and

Gibbons, 1977; Shine and Bull, 1979). Obviously,

environmental factors that favor viviparity, but not the

intermediate stages, only explain the maintenance of

viviparity, but not its evolutionary origin. Additionally,

the model can only work if variation in egg retention time

has a genetic component.

The reproductive modes of squamates are viewed as a

continuum between oviparity and viviparity (Shine, 1983a).

Unfortunately, there are few quantitative data on absolute or










relative egg retention times in oviparous squamates to

compare variation in this trait. Most authors assess a

species' position on the oviparity-viviparity continuum by

noting the stage of embryonic development at oviposition

(Newlin, 1976; Shine, 1983a; Guillette and Lara Gongora,

1986). Dufaure and Hubert (1961) described 40 morphological

stages for embryos of the lizard Lacerta vivipara. These

stages have become the standard to evaluate developmental

stages of lizards. The problem with using embryonic stages

to estimate egg retention times is that stages do not occur

at equivalent fractions of the total period of development.

Although each morphological stage has been assigned an

integral numerical value, the values are not meristic. For

example, at a given incubation temperature it may take only

hours to develop to stage 27 from stage 26; however, it make

take a week of embryonic development to go from a stage 38

embryo to stage 39 embryo.

Shine (1983a) suggested lizards and snakes retain eggs

for approximately half the period of embryonic development

and prolonged uterine egg retention is the rule rather than

the exception. It is necessary to know what fraction of the

total development time it takes embryos to achieve each

embryonic stage. Assuming all lizards achieve each stage at

similar fractions of the total development time, it would

then be easy to determine the period of egg retention by

knowing embryonic stages at oviposition. This would permit a

more quantitative assessment of interspecific, intraspecific,










and individual variation in egg retention times. Such a

technique would be an invaluable tool for testing hypotheses

concerning the evolution of viviparity.




Benefits and Costs of a Reproductive Mode


On theoretical grounds, fitness is maximized when the

benefits of a specific mode of reproduction are greater than

the costs. Hypotheses concerning the selective forces

responsible for the transition from oviparity to viviparity

must take into account both the benefits and costs associated

with these modes of reproduction (Tinkle and Gibbons, 1977;

Shine and Bull, 1979). Cost-benefit analyses designed to

predict the kinds of species most likely to evolve viviparity

should take into account species characteristics, such as

physiological constraints (inability to reduce eggshell

thickness, embryonic intolerance to reorientation),

reproductive frequency (single or multiple clutches),

foraging mode (sit and wait, active forager), habitat

specialization (terrestrial, marine, arboreal), predator

avoidance behavior (crypticity, rapid escape), defensive

ability (venomous, aggressive), and thermoregulatory strategy

(heliothermic, thigmothermic) (Shine, 1985). For instance,

viviparity is unlikely to evolve in actively foraging species

that rely on speed to avoid predators (Fitch, 1970). In such










cases, retention of young would greatly increase the cost and

risk of activity.

There are a number of potential disadvantages ("costs")

to viviparity (Tinkle and Gibbons, 1977). Most are viewed in

terms of how pregnancy affects a female's survivorship, as

well as her ability to produce more offspring. First, most

viviparous species produce a single litter per year,

primarily because gestation periods are so long there are

fewer opportunities to produce a second clutch within a

single reproductive season (Fitch, 1970; Tinkle and Gibbons,

1977). Thus, it is argued that viviparity potentially

involves a reduction in offspring number over a lifetime

because clutch frequency is reduced. However, reductions in

fecundity no doubt occurred long before a species evolved

viviparity. Ancestral forms with prolonged egg retention

probably only produced one clutch per season, as well. The

reduction in fecundity is more likely to result from a

reduction in clutch size. Guillette (1981b) showed that

Sceloporus aeneus, a prolonged-egg-retaining species, had

larger clutches (by about 1 egg) than a sibling species, ..

bicanthalis, which is viviparous. Associated with the

reduction in offspring number is the potential loss of

genotypic diversity of offspring, i.e., females reproducing

once per season not only have fewer young, but may mate with

fewer males (Tinkle and Gibbons, 1977). In many species of

lizards and snakes, females mate with several males before

ovulation (Saint-Girons, 1985). Additionally, because the










weight of a litter increases throughout gestation as water

moves into the eggs, carrying offspring reduces a female's

mobility, thereby increasing her risk of predation (Fitch,

1970; Shine, 1980). Sinervo (1991) demonstrated reduced

mobility in gravid and immediately post-gravid Sceloporus

occidentalis. An increase in litter volume may also reduce a

female's ability to feed because there is little room in the

abdomen that is not occupied by the contents of the uterus

(Shine, 1980). By not feeding during pregnancy a female may

reduce the amount of energy she could take in that could be

invested in her own maintenance or in the production of

future offspring.

In contrast, viviparity would be beneficial in

environments where embryo survivorship is greater in utero

than it is in the nest, for instance environments that are

too cold, too wet, too dry, unpredictable, or subject to high

egg predation (Tinkle and Gibbons, 1977; Shine and Bull,

1979). Again, it should be emphasized that only those

environmental conditions that favor the intermediate stages

of egg retention as well would be the conditions where

viviparity would evolve. There may be many environments that

would favor viviparity over oviparity but would not be the

kinds of environments that would favor prolonged egg

retention (e.g., a marine environment).

Another advantage of live-bearing is that females are

freed from the task of finding suitable nest sites (Fitch,

1970). Nesting may be an energetically expensive process










(Rose, 1989), and may be risky if females are exposed to

predators (Blair, 1960). In summary, both environmental

factors and species characteristics will be important

considerations for models predicting how viviparity evolves.

Any model should take into account both the costs and

benefits of either reproductive mode.


Viviparity and Parental Investment


Parental investment is defined as any investment in

current offspring that increases an offspring's chance of

survival that is not invested in future offspring (Trivers,

1972). Among non-brooding squamates, parental investment in

offspring would be greater in viviparous species than in

oviparous species. The most obvious cost associated with

viviparity are the costs associated with a long gestation

period. During the time females are carrying a litter, they

are unable to invest in future offspring. The gestation

period of an oviparous species is considerably shorter, thus

females may begin to harvest resources for subsequent

reproductive events sooner than would viviparous females.

The cost of carrying offspring is also a form of parental

investment. The female must be able to move about and feed,

avoid predators, and thermoregulate at temperatures optimal

for embryonic development. The additional weight of the

litter means that all movements are energetically more

costly. There may also be physiological costs of supporting










a litter. For instance, some viviparous reptiles supply

nutrients to young during the gestation period (Blackburn et

al., 1984). Even if females supply no nutrients to offspring

during gestation, the female must maintain her own body so

that offspring survive. It has also been suggested that

there is a maintenance cost of supporting embryos (Guillette,

1982b; Birchard et al., 1984; Beuchat and Vleck, 1990); that

is, the resting metabolism of pregnant females is greater

than the resting metabolism of nonreproductive females.

Higher rates of metabolism are thought to be due to the

increased energetic expenditures by the circulatory and

excretory systems. Measurement of any of these costs would

be useful in helping to predict the kinds of environments or

the types of species in which viviparity is most likely to

evolve.


Hypothesis Testing


Squamates are ideal organisms with which to model the

evolution of viviparity because of the abundance of extant

forms that represent intermediate stages of this evolutionary

process. Intermediate forms are not known from birds or

mammals.

It is difficult to test hypotheses concerning events

that have taken place in the past (Shine, 1983b). To test

hypotheses for the evolution of viviparity, the comparative

approach is necessary. The following methodology could be










utilized. Predictions are made concerning the traits

possessed by individuals at various stages of the

evolutionary process and congeneric species believed to

represent the different stages are examined for the predicted

traits. The following are examples of the methodologies used

to test hypotheses concerning reptilian viviparity in the

context of the evolutionary scenario presented above.

It is hypothesized that viviparity is an adaptation to

cold climate (Weekes, 1935; Sergeev, 1940; Packard et al.,

1977; Shine and Bull, 1979; Guillette et al., 1980; Shine,

1985). In cold climates reproductive seasons are short.

Reptilian embryos develop slowly at cold temperatures and may

not have sufficient time to complete development in a cold

nest. Egg retention has evolved as a mechanism to reduce the

time it takes embryos to develop. It is assumed that because

females thermoregulate, they expose embryos to higher

temperatures than those occurring in underground nests.

Consequently, embryos spend less time developing in a cool

nest and hatch sooner.

There are two lines of evidence to support the cold

climate hypothesis. First, if viviparity evolves in cold

climates, then the majority of species occurring in a cold

climate should be viviparous. Indeed, it has been found that

the proportion of viviparous species is greatest at high

elevations (Weekes, 1933, Sergeev, 1940; Greer, 1968; Greene,

1970; Guillette, et al., 1980) and latitudes (Tinkle and

Gibbons, 1977). However, this evidence provides only










equivocal support for the hypothesis. Viviparity could have

evolved under different conditions, and subsequently a

species could have expanded its range into a cold climate

where few, if any, oviparous forms occur. Secondly, egg

retention is assumed to provide a thermal advantage that

permits more rapid development of embryos; therefore, in cold

climates, it would be predicted that, in the absence of

prolonged egg retention, embryos would develop too slowly and

would not hatch before the onset of lethal autumn

temperatures. Shine (1983b) provided evidence in support of

the assumption that egg retaining females have higher average

body temperatures than the average temperatures found in

nests. He found that egg retaining skinks from cool climates

in Australia did have higher average body temperatures than

the average temperatures of nests. Furthermore, he

determined that eggs could not hatch before the beginning of

lethal autumn frosts if the period of egg retention was

brief. Although the cold climate hypothesis has not been

tested directly, Shine (1983b) has shown that an underlying

assumption of the hypothesis is sound.


Embryonic Development and Viviparity


While in an underground nest, squamate embryos must

obtain water and oxygen from the surrounding substrate.

Prior to oviposition or parturition, females must supply

oxygen and water to developing embryos and remove excess










carbon dioxide (and possibly nitrogen) produced by the

embryos (Packard et al., 1977). Very little is known

concerning the impact of embryonic respiration and excretion

on the maternal organism (but see Guillette, 1982a; Birchard

et al., 1984; Beuchat and Vleck, 1990). Surprisingly, no

published data exist on embryonic respiration in lizards.

Nonetheless, there is some suggestion that embryonic

respiration and growth may influence the evolution of

viviparity in reptiles (Weekes, 1935; Packard et al., 1977;

Tinkle and Gibbons, 1977; Guillette, 1982a). The rate of

oxygen consumption of eggs from oviparous snakes increases

with advancing development (Dmi'el, 1970; Black et al.,

1984). Presumably, the embryos of viviparous species have

similar patterns of oxygen consumption as the embryos of

oviparous species. Guillette (1982a), Birchard et al.

(1984), and Beuchat and Vleck (1990) all showed that rates of

metabolism were higher in late pregnant lizards or snakes

than in the same postpartum females. Indeed, late pregnant

animals had rates of metabolism that were higher than

expected if the metabolism of the litter and the nonpregnant

female were taken into account. All these authors suggested

that the metabolic component not due to litter and somatic

metabolism represents a maintenance cost of supporting the

litter. Among the three species, the maintenance component

represented between 16% and 42% of the total resting

metabolism of a late pregnant female. These estimates are

much higher than those reported for sheep (Clapp, 1978).










Additionally, the eggs of reptiles with parchment

eggshells swell, during gestation and incubation, due to

water uptake (see Packard and Packard, 1986). Viviparous

embryos also take up water during gestation and may increase

in mass by 3 to 4 times when compared to freshly ovulated

eggs (Thompson, 1981). Water uptake has been shown to be

essential for embryonic development in squamates (Packard and

Packard, 1986). Thus, the implications of internal

development are that more advanced embryos require greater

amounts of oxygen, especially late in development, and that

the weight and volume of the clutch place an increasing

burden on the female. The ecological implications of

increased egg retention were discussed above.

Weekes (1935) was the first to suggest that the primary

selective pressure for the evolution of the placenta was gas

and water exchange. Further evidence of the importance of

efficient gas exchange is the extreme reduction or absence of

an eggshell in viviparous species (Weekes, 1935; Packard et

al., 1977). Some viviparous species possess rudimentary

eggshell membranes early in gestation; however, the eggs of

viviparous species are never calcified (Weekes, 1935). A

thick hydrated eggshell could act as a barrier to gas

exchange between mother and embryos (Packard et al., 1977).

Ackerman (1980) found that sea turtle embryos could tolerate

extreme hypoxia to a much greater extent than bird embryos.

Mortality increases and development slows in embryonic

turtles exposed to extreme hypoxic conditions. Guillette et










al. (1980) expanded the oxygen limitation hypothesis by

suggesting that the evolution of the placenta was a response

to low partial pressures of oxygen in high altitude

environments where viviparity was likely to evolve.

Eggshell thinning followed by loss of the eggshell

would be expected during the transition from prolonged egg

retention to viviparity. Nonetheless, there is little

quantitative evidence for gradual reduction in eggshell

thickness in prolonged egg retaining species. Alternatively,

selective pressures to reduce shell thickness may not act

until a species has almost evolved complete egg retention.

There is evidence to support this new hypothesis

concerning the reduction and loss of the eggshell. First,

Blanchard (1933) reported that the eggs of the prolonged-egg-

retaining smooth green snake, Opheodrys vernalis, are opaque

and white suggesting that the eggs of this species are

calcified and of "normal" thickness. In the most northern

populations of the species, eggs can hatch within four days

of oviposition; however, incubation times of 10 to 14 days

are more typical. An effort should be made to quantify

eggshell thickness in northern and southern populations of Q.

vernalis and in its southern relative, Q. aestivus, which

does not exhibit as prolonged a period of egg retention.

Additionally, Sceloporus scalaris and a. aeneus can retain

eggs that hatch in 2 weeks following oviposition (DeMarco,

personal observation; Guillette and Lara Gongora, 1986).

Eggshell reduction does not appear to have occurred in these









species despite prolonged egg retention. Evidence of

eggshell reduction has been documented in Sphenomorphus

fragilis (Guillette, in press). This oviparous species

oviposits eggs that hatch within hours. The eggshells do not

have a calcium layer and the fibrous shell membrane is only

10 p.m thick. The shell membrane, however, may not be reduced

compared to that in other closely related lizards that have

normal eggshells. Thus, species exhibiting incipient

viviparity may not have a calcium layer, but still may

possess a normal shell membrane. Because eggshell reduction

is not occurring long before the final stages of complete egg

retention, eggshell reduction is not gradual, as has been

proposed in the past (Packard et al., 1977).

Given that shell reduction does not occur in many

prolonged-egg-retaining species, chronic hypoxia may not

occur during the period of egg retention in squamates as some

have implied (Packard et al., 1977; Guillette et al., 1980).

Ackerman (1980) suggested reptilian embryos can tolerate

extreme hypoxia. This gives further support for the idea

that squamates could retain shelled embryos until development

is almost complete.

The preceding observations do not support Shine's (1985)

definition of viviparity. Shine considers species that

oviposit eggs that hatch within hours or days to be

viviparous (e.g., Saiphos equalis, Pseudechis porphyriacus,

and Trimeresurus okinavensis). These species, like

Sphenomorhpus fragilis, may still possess eggshell membranes.









Shine (1985) also considers the four-day incubation period

seen in Opheodrys vernalis as an example of viviparity

despite the presence of a normal eggshell. A four day

incubation period is only slightly greater than the

incubation period of Saiphos equalis, Pseudechis

porphyriacus, and Trimeresurus okinavensis). It follows that

oxygen limitation may not occur as a result of the presence

of the eggshell.

Eggs that have short incubation periods may still

require an eggshell to prevent rapid desiccation and death of

the embryo. There may be other means by which the potential

effects of hypoxia could be mitigated without having to

reduce the thickness of the eggshell. For instance, the

hemoglobin of squamate embryos, as in all other vertebrate

embryos examined, has a greater affinity for oxygen than that

of adult hemoglobin (Grigg and Harlow, 1981; Berner and

Ingermann, 1988). Thus, the tendency is for oxygen to be

transferred to the embryonic circulation even at a low P02.

Recently, it was found that the P50 of hemoglobin of late

pregnant brown snakes (Holland et al., 1990) is greater than

that of nonpregnant snakes. This means that as the embryos

require more oxygen, the female is better able to release it.

It would be interesting to know if this mechanism exists in

prolonged-egg-retaining females, such as Q. vernalis.











Outline of Studies


The studies reported here were designed to provide basic

information on metabolism and growth of Sceloporus embryos to

test ideas concerning the evolution of reproductive modes in

squamates. Most of the work presented involves comparisons

of embryonic traits in three oviparous and one viviparous

species in the lizard genus Sceloporus. The species were

chosen because of apparent differences in their positions

along the oviparity-viviparity continuum. The three

oviparous species are: Sceloporus woodi, a brief egg

retainer; a. virgatus, an intermediate egg retainer; and a.

scalaris, a prolonged egg retainer. The viviparous species

is a. Jarrovi, a species reported to have a Weekes's Type II

placenta that does not supply additional nutrients to embryos

during development (Guillette, et al., 1981). The

comparative nature of the studies and the close relationship

between the species make each study a natural experiment.

Although embryonic physiology has been the subject of much

speculation with regard to the evolution of viviparity in

reptiles, very little is actually known about embryos. Goals

and hypotheses are outlined below.










Hypothesis 1. Egg retention times can be predicted from

embryo stages at oviposition.



Prediction 1: Under similar laboratory conditions

the distribution of embryo stages at oviposition

will be different in S. woodi, a. virgatus, and S.

scalaris.



Prediction 2: The absolute and relative periods of

egg retention are different in S. woodi, S.

virgatus, and S. scalaris.



Prediction 3: The relative time it takes to develop

to each embryonic stage is similar in a. woodi, S.

virgatus, and a. scalaris.



The extent to which lizards retain eggs is

poorly known. Quantitative information on relative

egg retention times is vital for determining a

species position on the oviparity-viviparity

continuum. A method is designed to determine the

proportion of total development time it takes

embryos to achieve specific embryo stages. From

this, determination of inter- and intraspecific

variation in egg retention times can be based on

variation in embryo stages at oviposition.










Hypothesis 2. More rapid embryogenesis is an adaptation to cold

climates.



Prediction: Species inhabiting cooler climates have

longer embryonic development times than species

found in warmer climates.



More rapid embryogenesis, like egg retention

can reduce developmental time in species inhabiting

cold climates. Total embryonic development times,

the period from ovulation to hatching, are measured

at the same temperature in 3 oviparous and 1

viviparous species in the genus Sceloporus.



Hypothesis 3. The pattern of metabolism is related to the

length of the total developmental period.



Prediction: Species with longer development times

will have more peaked patterns of metabolism.



No data have been published on the patterns of

embryonic metabolism in oviparous lizards during

development. First, patterns of metabolism are

compared to other reptiles. Combined with

information on egg retention times, such data could

provide critical background information necessary

for future investigations concerning the oxygen










limitation hypothesis for the evolution of

viviparity.



Hypothesis 4. The physiological cost of pregnancy is high in

viviparous species due to the increased energetic load on

various physiological systems.



Prediction: Resting metabolism in Sceloporus

jarrovi will be greater than the sum of litter and

maternal somatic metabolism, indicating there is a

maintenance cost of supporting a litter.



Under the appropriate selective regime,

viviparity will evolve if the benefits associated

with this mode of reproduction outweigh the costs.

Costs associated with viviparity include loss in

fecundity, increased vulnerability of females to

predation, and decreased genotypic diversity of

offspring (Tinkle and Gibbons, 1977). Several

authors have reported substantial energetic costs

that are due to the physiological maintenance of the

litter. Beuchat and Vleck (1990) recently reported

that the maintenance cost of pregnancy in Sceloporus

iarrovi represents 16% to 29% of the total resting

metabolism of late pregnant females. In view of new

data on embryo metabolism in .. jarrovi, the






23



maintenance cost of pregnancy is reevaluated for

this species.














CHAPTER 2
GENERAL METHODS


Collection and Maintenance of Adult Lizards


All lizards were collected with a noose. During the

spring and summer, from 1986 to 1990, Florida scrub lizards,

Sceloporus woodi, were collected along fire trails bordering

sand pine scrub habitat in the Ocala National Forest, Marion

Co., Florida. Striped plateau lizards, a. virgatus, and

bunch grass lizards, S. scalaris, were collected in the

Chiricahua Mountains, Coronado National Forest, Cochise Co.,

Arizona. The latter two species were collected during the

last week of May and the first week of June in 1987 to 1989.

Specifically, a. virgatus were collected at several sites in

Cave Creek Canyon. All sites were within 4 km of the

Southwestern Research Station (SWRS), and had been used

previously by Ballinger (1979), and Congdon et al., (1979).

Sceloporus scalaris were obtained near Rustler Park (RP) at

sites used previously by Ballinger and Congdon (1981). Adult

male and pregnant Sceloporus jarrovi were collected in the

Dragoon (November, 1986 and March, 1987) and Chiricahua

Mountains (May, 1987, 1988, 1989), Cochise Co., Arizona.

Additionally, nonreproductive females were collected in the

Chiricahua Mountains in July, 1987. The elevation at both

sites is approximately 1700 m.










Females of all oviparous species were palpated at the

time of capture and determined to have either advanced

vitellogenic follicles or oviductal eggs. The palpation

technique was described by Cuellar (1971). Lizards of all

species were returned to the University of Florida and housed

in 40 (50 x 26 x 29 cm) or 80 (75 x 31 x 31 cm) liter

aquaria. Species were maintained separately. Two or three

females and a male were placed in each aquarium which

contained a 7.5 cm deep layer of sanitized sand box sand as

substrate. A 60 watt incandescent light bulb was placed 1 cm

above the substrate at one end of each aquarium. During the

day, the air temperature 1 cm above the substrate and 3 cm

from the lightbulb reached 390C, whereas the air temperature

at the far end of the aquaria 1 cm above the substrate was

270C. Within this range of temperatures, lizards could

control their body temperatures by shuttling in and out of

the thermal gradient created by the light bulb. Fixtures

containing vita light (Duro-Light Corp.) and fluorescent

tubes were suspended across the tops of the aquaria. All

lights were kept on a 13L/11D photoperiod. Drinking water

was provided in a 10 cm diameter petri dish and lizards were

fed crickets, mealworms, and waxworm larvae ad libitum.


Collection and Incubation of Eggs


Given a moist deep substrate, gravid sceloporine lizards

will readily construct a nest chamber and oviposit a clutch.










Depending on the requirements of the study, eggs were

obtained either by allowing females to naturally oviposit or

by surgical removal.

After oviposition or surgery, each clutch was incubated

separately in a 100 ml specimen cup with a tight fitting lid.

A small hole was placed in the lid to allow for gas exchange.

An attempt was made to maintain a similar substrate water

potential for all clutches. Initially, the container and lid

were weighed. One hundred grams of dried sanitized sandbox

sand and 5cc of distilled water were added to the container

and thoroughly mixed. The water potential of this sand/water

mixture is approximately -5 kiloPascals (Kam and Ackerman,

1990). The clutch was placed in a depression in the sand so

that the tops of the eggs were 1 to 2 mm below the surface.

Clutches were placed in a constant temperature cabinet set at

300C. Eggs were removed from each container every 7 to 10

days and the container rehydrated to its initial weight.

Typically, containers lost between 1 and 1.5 grams of water

over a 7 to 10 day period. Under these conditions eggs gain

weight constantly throughout the incubation period until a

day or two prior to hatching. Initially, a portable Wescor

thermocouple thermometer was used to measure substrate

temperatures in the nest containers. A thermocouple probe

was placed just below the surface of the sand of several

containers. Chamber temperature controls were adjusted to

maintain a temperature of 30.00C in nest containers.











Measuring Oxygen Consumption


Rates of oxygen consumption (Vo2) were determined using

closed respirometry. Respirometer chambers for adult lizards

were constructed from 1 L canning jars with tight sealing

lids. Before the study began chambers were tested and found

to be gas tight over 24 h, i.e., the initial and final

fractions of 02 were equal. Chambers were painted black so

that lizards could not see light or movements by the

investigators. To insure that the air in the chamber became

saturated with water vapor during the measurement period, 3

ml of distilled water was place in a small glass vial glued

to the floor of the chamber. Lizards were weighed and then

gently coaxed into the chambers. At approximately 1200 h

(EST), chambers with lizards were placed in an environmental

cabinet maintained at 300C (0.50C). Measurements were taken

between 2200 h and 0300 h the following morning. Immediately

before measurement each chamber was flushed with water

saturated air (at 300C) for 5 minutes. For each chamber, a

gas sample was collected in a 60 cc syringe following

flushing and the chamber was immediately sealed. The initial

fraction of 02 (Fio2) was determined on each gas sample. At

the end of each measurement period a second sample of gas

from each chamber was collected. From these second samples,

the fractions of 02 at the end of the measurement interval

(FEo2), were determined. Initial and final samples were










injected into an Ametek Applied Electrochemistry S-3A oxygen

analyzer (Pittsburgh) and an N-22M Sensor (volume of sensor

cell, 0.25 cc). Output from the 02 analyzer was fed to a

chart recorder. Water vapor and CO2 were removed from the

samples prior to entry of each sample into the 02 analyzer.

An empty chamber was treated in the same fashion as the

chambers containing lizards and served as a control. Oxygen

consumption (cc 02 h-1) was calculated using the following

equation from Vleck (1987):



Vo2 = V (FIo2 FE02) / (1 FEO2) t, (2.1)



where V is the initial volume (cc) of dry C02-free air in the

chamber at STP; FI02 and FEo2 are the fractions of 02 at the

start and end of the measurement period, respectively; and t

is the duration of the measurement period in hours.

Following several hundred measurements of Vo2 it was

determined that flushing the chambers with air for 5 minutes

always resulted in an FIo2 of 0.2094. Subsequently, it was

decided that empirical determination of FIO2 was not

necessary.

The Vo2 of eggs were measured in a similar manner as

that reported for adults, but with the following exceptions.

Chambers were constructed from 240 cc canning jars. Embryos

were allowed between 1 and 2 hours to equilibrate to 300C.

The measurement period for embryos began at approximately










2100 h and ended at approximately 0900 h the following

morning.


Statistical Analysis


Means are reported with the standard deviation (S.D.)

Standard errors are reported if groups being compared had

very different sample sizes. Unless stated otherwise, all

tests performed are two-tailed and a P < 0.05 is required to

determine significance. All analyses were performed on an

Apple Macintosh SE computer. Statistical software packages

included Statview and SuperANOVA (Abacus Concepts, Inc.).

All references to embryo stages refer to stages described by

Dufaure and Hubert (1961). Rates of 02 consumption (Vo2) are

reported as cc 02 h-.















CHAPTER 3
ESTIMATING EGG RETENTION TIMES



Differences in the proportion of development time that

eggs are retained among oviparous and viviparous squamates

have been viewed as a continuum from very brief egg retention

to complete egg retention (Shine, 1983a). In the past, the

length of the incubation period (Blanchard, 1933; Erasmus and

Branch, 1983) or an embryo's stage at oviposition (Shine,

1983a) have been used as indicators of relative egg retention

time (RERT) in squamates. Although these indicators lead to

qualitatively correct conclusions about RERT, it is still not

possible to determine the difference in the period of egg

retention among animals ovipositing eggs with embryos at

different stages. To date the only measures of absolute and

relative ERT come from Shine (1983b), for the skink, Anotis

macoyi. Eggs from this species were kept at the same

constant temperature during the entire developmental period,

from the approximate date of ovulation until hatching. Shine

found that this species oviposits eggs with embryos at stages

30 and 31 (stages from Dufaure and Hubert, 1961) and that

this represented approximately 40% of the total amount of

time it took embryos to develop to hatching. Having this

information for many populations and species would be of

great value to those interested in testing hypotheses










concerning environmental factors that select for a viviparous

mode of reproduction.

My primary goal was to develop a quantitative method to

estimate the proportion of time eggs are retained relative to

total development time. The method described here is based

on determination of the relationship between embryo stage and

time from ovulation of eggs that were incubated at constant

temperature. Three oviparous lizard species in the genus

Sceloporus were used: S. woodi, a brief egg retainer, ..

virgatus, an intermediate egg retainer, and a. scalaris, a

prolonged egg retainer.



Methods



Terminology


The following terminology is adopted to describe egg

retention and embryonic development times. Total embryonic

development time (TEDT) is a measure of the total number of

days it takes to develop from ovulation to hatching. Egg

retention time (ERT) refers to the number of days eggs are

retained in utero following ovulation. Relative egg

retention time (RERT), therefore, is the ratio ERT/TEDT

multiplied by 100%. Incubation time (IT) refers to the the

number of days it takes embryos to develop from oviposition

to hatching. Percent IT, therefore, is the ratio IT/TEDT

multiplied by 100%. In this study, ERT, IT, and TEDT were










determined in the laboratory at a single constant temperature

(30 oC). Any ratios calculated using these measures are also

a function of constant temperature conditions. It is not the

intention here to advocate restriction of the use of these

terms to laboratory studies done under constant conditions.


Embryonic Stages at Oviposition


To obtain a qualitative assessment of the degree of egg

retention in these oviparous species, embryonic stages at

oviposition were determined for clutches from 25 Sceloporus

woodi, 23 S. virgatus, and 20 S. scalaris females. Gravid

lizards of each species were collected and allowed to

oviposit naturally in laboratory aquaria. Species were

maintained separately and all animals were exposed to a

similar thermal gradient. Lizards were checked every 4 to

six hours for signs of oviposition. In most cases, clutches

were recovered within an hour or two of oviposition.

Immediately after a clutch was discovered, eggs were removed

and incubated. One egg from each clutch was opened within an

hour or two of oviposition and embryonic stage was

determined. Although there may be slight intra-clutch

variation in embryo size at oviposition, there is no

detectable variation in embryonic stage (DeMarco, unpublished

data). Consequently, only one embryo per clutch was

examined. A Kruskal-Wallis test was used to determine










whether differences existed in embryonic stage at oviposition

for these three species.


Total Embryonic Development Time


Lizards were palpated twice daily to determine the

approximate time (at most, within 12 hours) of ovulation.

Once a lizard ovulated, it was placed in a 40 liter aquaria

located in a glass fronted environmental cabinet. The

temperature in the cabinet was maintained at 300C and

lighting in the laboratory provided approximately 10 hours of

light each day. Additionally, a 20 watt fixture with a

fluorescent and an ultraviolet tube was placed over an

aquarium at least twice a week for 10 hours. Animals were

fed and watered as described in Chapter 2. Because S. woodi

are known to retain eggs for a relatively brief period of

time they were allowed to oviposit naturally. Given a moist

deep substrate, these lizards will readily construct a nest

chamber and oviposit a clutch. Only half the substrate in

each aquarium was kept moist in order to minimize the effects

of evaporative cooling of the substrate on lizard body

temperature. The number of days from ovulation to

oviposition was noted for each of the scrub lizards (x = 11.7

days, n = 10). Eggs were surgically removed from S. virgatus

and a. scalaris after a similar interval of egg retention.

It was thought that removing eggs prior to oviposition from

.. virgatus and S. scalaris would reduce the possibility of










confounding maternal or environmental effects (such as

differences in the hydric environment in utero versus that in

the incubation substrate) that might be related to longer or

shorter periods of egg retention. Since the majority of

eggshell formation occurs within 10 days of ovulation in all

three species (DeMarco, unpublished data), it was thought

that differing degrees of eggshell development would not be a

factor influencing embryonic development time. Despite the

fact that eggs were removed from gravid S. virgatus and a.

scalaris prior to oviposition, substrate conditions in the

incubator were the same as for S. woodi (see above).

After oviposition (S. woodi) or surgery (S. virgatus and

S. scalaris), each clutch was incubated as described in

Chapter 2.

The TEDT was noted for each clutch. Most clutches

hatched in synchrony; however, if clutchmates hatched over

more than one day, then the average time to hatch was used to

estimate TEDT. A one way analysis of variance (ANOVA)

followed by Scheffe's tests were used to compare TEDT between

the three species.


A Direct Method for Calculating RERT


In April, 1989, 10 female Florida scrub lizards,

Sceloporus woodi, were collected along sandy trails in the

Ocala National Forest, Marion Co., Florida. Females were

palpated and those having large vitellogenic follicles were










used. Eggs developed at 30 OC during the periods of egg

retention and incubation (Chapter 2). To calculate RERT, the

number of days from ovulation to oviposition (ERT) was noted

for each of the scrub lizards and was divided by the time

from ovulation to hatching (TEDT).


An Indirect Method for Calculating RERT


Although the method described above allows direct

determination of RERT its usefulness is limited to the

species for which it was determined. Alternatively, RERT

could be determined if both the proportion of TEDT taken to

achieve each developmental stage and embryo stages at

oviposition (ESO) were known. Assuming embryos of all

oviparous species of Sceloporus attain similar stages of

differentiation and growth at similar proportions of TEDT,

i.e., there are no gross developmental heterochronies, then

it is possible to estimate RERT in other congeners by knowing

their ESOs. The assumption that there is no heterochrony in

embryonic stages needs testing.

Eggs were obtained from field collected gravid females

of all three species. Additionally, clutches of eggs were

surgically removed from some females to obtain embryos at

pre-oviposition stages. All eggs were incubated at 30 OC

until hatching.

Eggs from all 3 species were opened at intervals

throughout their incubation periods to determine the










relationship between embryonic stage and time to hatching.

Embryos were sampled from 26, 23, and 20 clutches of

Sceloporus woodi, a. virgatus, and S. scalaris, respectively.

Time to hatching was determined for each embryo by

calculating the time to hatching of clutchmates. No attempt

was made to sample embryos prior to stage 24 because it is

technically more difficult and would have required much

larger numbers of lizards. Data were entered into different

regression models to find one that best describes the

relationship between the independent variable, time to

hatching, and the dependent variable, embryo stage. The

appropriate regression model was used to calculate time to

hatching for any stage between the stages examined in this

study (24 to 40). By dividing the stage-specific number of

days to hatching by the mean TEDT, I calculated the stage-

specific fractional time to hatching. I then determined the

proportion of TEDT it takes to reach each embryonic stage

from ovulation by subtracting the stage- specific fractional

time to hatching from one. The RERT was determined for each

female for which embryo stage at oviposition was known.



Results




Embryo Stages at Oviposition


The modal embryonic stages at oviposition (ESO) for -.

woodi, S. virgatus, and a. scalaris are 27, 31, and 38,










respectively (Fig. 3.1). The distributions of embryonic

stages at oviposition are significantly different among these

three species (Kruskal-Wallis: H = 55.3; df = 2; P < 0.0001).

A nonparametric test was used because embryonic stages are

ordinal data.

Laboratory and field data on ESO are similar. Eggs from

autopsied free ranging S. woodi had embryos at stage 28 or

less (DeMarco, unpublished data). Vinegar (1974) reported

that gravid S. virgatus in early July had stage 31 embryos,

which he suggested was the minimum stage at oviposition.

Newlin (1976) observed that gravid a. scalaris collected in

the field had embryos with well developed limbs, eyes,

digits, and claws, and scales with light brown pigmentation

(developmental stage 37).



Directly Estimating RERT in S. woodi


The range of ERTs for 10 a. woodi whose embryos were

maintained in an incubator at constant temperature from

ovulation to hatching was 9 to 15 days, which represents

between 14.6% to 24.4% of TEDT. The mean ( 1 S. D.) ERT was

11.7 ( 1.95) days, whereas the mode was 13 days. The mean

( 1 S.D.) TEDT for these same 10 lizards was 61.6 ( 2.88).

Because embryo stages at oviposition were recorded for only 6

of the 10 females in this sample, the results of this method

can not be compared to those of the indirect method.











Embryo Stage and Time to Hatching


The relationship between embryo stage and time to

hatching (days) are best described by second order polynomial

equations (Table 3.1). Because the relationship between

embryo stage and time to hatching is known, incubation times

(IT) can be calculated by substituting ESO into the equations

in Table 3.1. Subtracting minimum and maximum IT from the

mean TEDT yields maximum and minimum egg retention times

(ERT) (Table 3.2). By dividing mean ERT by mean TEDT, the

fraction of time eggs are retained in the oviducts can be

calculated. The modal RERT (determined by using the modal

ESO) for embryos developing at 30 OC is 12.7%, 29.8%, and

73.7% for S. woodi, a. virgatus, and a. scalaris,

respectively (ranges shown in Table 3.2).


Heterochrony and Embryo Stages


It was suggested earlier that if no heterochrony existed

in the timing of embryonic stages then the method used here

to determine RERT, could be used, at least, for other species

of Sceloporus and, at most, for all other lizards. Embryos

achieve similar stages at similar proportions of their

respective TEDTs (Fig. 3.3). These results support the

hypothesis that there is no heterochrony with respect to

gross morphological appearance among these three species of

Sceloporus.



















10

H S. woodi
8 U S. virgatus

SS. scalaris
c 6-
w

4


2
ZII .



24 25 26 27 28 29 30 31 32 33 34 35 36 37 38


EMBRYONIC STAGE AT OVIPOSITION



Figure 3.1. A frequency distribution of embryonic stages at
oviposition in Sceloporus woodi, S. virgatus, and S.
scalaris.






















40

W 38
0 S m a
36-

O 34 Aa 3
Z A nA S. wood
S 32- A S. virqatus
A A/a S. scalaris
i 30- no
UJ r / []
28

26
-50 -40 -30 -20 -10 0

TIME TO HATCHING (DAYS)



Figure 3.2. A time schedule of embryonic stages for
Sceloporus woodi, a. virgatus, and a. scalaris. Hatching
occurs on day 0. The curves shown are based on the
equations listed in Table 3.1. All embryos were
incubated at constant 300C. Each point on the graph
represents the embryonic stage of a single embryo at
autopsy. Time to hatching was determined by calculating
the number of days between autopsy and hatching of
clutchmates.



















Table 3.1. Polynomial equations relating embryo stage (y)
and days before hatching (x) for Sceloporus woodi,
S.virgatus, and S. scalaris embryos developing at 300C.



species equation N N r2
clutches eggs

a. woodi y = 40.012 + 0.082x 26 57 .96
0.003x2

S. virgatus y = 40.095 + 0.087 x 23 75 .98
0.004x2

s. calaris y = 39.605 0.052x 20 46 .97
0.007x2

















*C





r 0

Co





ON 0.*
Oi C



O -c
0 0







N 4)


OC b

0 rO.




o o
-H
CO -)


*)


0 o



c) 0
Sn a
C 0


4 Co-





to
u 3 *















O aO
a- o0 u
0 4-)



CA 0


M 00
041
dP U-















-) 0
0 0
0 o) 0


4-) -H -d
0E 0 0


0)







4-) >iC



0 -(D0


0



0-Q



E-1




E-4







C--t





























cdm
-4
0

a











o









E-4
H-



















0
ro -c







0


0
ri




















4-4
a
E-
Q-Q









01
CO








t1









1-


(

) H

+1


H 0
cI- oI

H CN ro

U)


iZn
(n





cH O
04
t N^







OD id 00 0


LO H
C)













ro I -D
-0








0 0
Hd rH1


In m o
0N (N (4


CN


CN


-I
04



0


L'O
(n
r-I
m
o




0
e


aO,
a,'
cv)
OD
co
m










-H1



Ml


oI-


0)




















Cr
'-)
C>H


w

(4





















-t-


+1


+1





















40
AEM
38- Am
a S. woodi
S 36 A S. vir-atus a
0 36 S. scalaris

I 34-

0 "
32- M



UU
S 30- u

28 U

26
0 10 20 30 40 50 60 70 80 90 100



% TEDT




Figure 3.3. The relationships between embryo stage and
percent total embryonic development time (%TEDT) indicate
that there is little evidence for heterochrony in the
gross morphology of embryonic development. The points
shown were generated using the equations found in Table
3.1 and subsequent calculations described in the text.











Discussion


The degree of embryonic development at oviposition

varies among oviparous lizard species from a small embryonic

disk (Chameleo lateralis, Blanc, 1974) to an almost

completely developed embryo capable of hatching within a week

or two (Sceloporus scalaris, this study (Fig. 3.1, Table

3.1); Sceloporus aeneus, Guillette and Lara Gongora, 1986).

Some lizards are known to oviposit thin shelled eggs with

embryos that hatch within a few hours (Sphenomorphus

fragilis: Greer and Parker, 1979). Guillette (in press) uses

the term "incipient viviparity" to describe this condition.

Variation in developmental stages at oviposition is

hypothesized to be due to differences in the proportion of

total embryonic development time (TEDT) eggs are retained in

the oviducts (Neill, 1964; Packard et al., 1977; Shine and

Bull, 1979). In the next chapter, evidence is presented to

show that variation in embryo stage at oviposition (ESO) is

also due to species differences in rates of development. To

date, determining the embryonic stage at oviposition (ESO) or

the length of the incubation period have been the basis for

estimating RERT.

Shine (1983a, b) suggested that the amount of time

squamate embryos spend in utero is approximately half the

TEDT, and the embryonic stage at the halfway point in

development is stage 30. These suggestions were contradicted










by data presented in one of these studies (Shine, 1983b).

Shine determined that the RERT of Anotis macoyi was 40%. The

method used by Shine (1983b) was similar to the method

described above for a. woodi, except that ovulation times

were estimated for A. macoyi. Furthermore, ESO in A. macoyi

varied between stages 30 and 32. For Sceloporus embryos,

stage 30 and 32 occur when 25 and 35 percent of development

has been completed, respectively. Halfway through

development an embryo is stage at 35 (Table 3.2). Shine may

have overestimated TEDT in Anotis macoyi. This could have

occurred if lizards had not yet ovulated when they were

placed in the incubator. Shine estimated ovulation dates

instead of actually observing them as was done here.

Alternatively, Shine's estimates of TEDT could be accurate

and Anotis macoyi achieves embryo stages at different

proportions of TEDT than Sceloporus embryos. Nonetheless, my

results and those of Shine (1983b) for A. macoyi do not

support the conclusion that stage 30 embryos represent the

midpoint of TEDT.

A more complete analysis of the available data on

embryonic stages at oviposition in lizards may provide us

with a better estimate of RERT for lizards. Ideally, the

analysis should include data from a large number of species.

The sample should be unbiased in that it should contain

realistic proportions of brief, intermediate, and prolonged

egg retaining species. Unfortunately, there is very little

information on ESO in lizards and no way of knowing whether










the available data represent an unbiased sample. The only

data set available is from Shine (1983a) and consists of a

list of ESOs for 38 species of lizards and 23 species of

snakes. To create a frequency distribution of ESOs for

lizards a decision must be made concerning what single

embryonic stage best represents ESO for each species or

population. The modal ESO is probably the best statistic to

represent a sample. Unfortunately, modes can not be

calculated for any of the species in the sample because most

reports are based either on examination of less than 3

clutches or an unknown number of clutches. Nonetheless,

Shine's data set is the only information available on ESO in

lizards and, as such, is an invaluable starting point for

analysis.

The modal ESO for the 38 species of lizards is stage 31.

The data were manipulated in the following manner. If a

range of ESOs was reported for a species the midpoint of the

range was used. The midpoint may fall closer to the mode

than either of the extremes of the range. In 7 cases no

range was given and ESOs were considered minimal for the

species or population. These data were not manipulated in

any way. If stage 31 is a fair estimate of ESO for lizards,

i.e., the sample size is adequate and the sample is unbiased,

then lizards retain eggs for approximately 31% of TEDT. This

estimate is based on substitution of stage 31 into each of

the 3 equations in Table 3.1 and averaging the RERTs. This

new estimate of the RERT is considerably less than Shine's










(1983a,b) and may change the view that prolonged egg

retention is the typical reproductive mode among oviparous

squamates. Conversely, the sample this estimate is based on

may be highly biased in favor of brief egg retainers.

Nonetheless, this is the only sample available on which to

base an estimate of RERT in lizards. My estimate of RERT for

oviparous lizards could change as more species are added to

the existing data set. If there is considerable heterochrony

among taxonomic groups of squamates in the proportion of time

it takes embryos to achieve specific embryo stages, then new

estimates of RERT may apply only to sceloporines (at least)

or phrynosomatids (at most). Similar studies on lizards in

different taxonomic groups would be necessary to determine

whether heterochrony exists for embryonic stages.

Currently, there is no objective application of the

terms "brief" or "prolonged" ER. Furthermore, applying

current terminology to all amniotes is confusing. In the

past, authors have tended to refer to all squamates as

prolonged egg retainers. There is no doubt that some

squamates are prolonged egg retainers. Clearly, 20% of the

squamates exhibit complete egg retention or viviparity

(Shine, 1985). Of the oviparous species, however, prolonged

egg retention may be rare.

Another possible reason for the perception that

squamates are prolonged egg retainers is that they oviposit

eggs with embryos that are more advanced than embryos

oviposited by birds, chelonians, and crocodilians. Embryo










stage at oviposition is not a good character to compare among

reptilian orders. This is because in utero development in

turtles and crocodilians is arrested at the gastrula stage

(Ewert, 1979). Thus, the absolute number of days eggs are

retained by turtles and crocodilians is not related to stage

of development. It is possible that ERTs in some turtles and

crocodilians may be similar to ERTs in the majority of

oviparous squamates. Ewert (1979) estimated ERT in several

species of turtles to be at least two weeks, and suggested

that the RERT for turtles may vary between 11% and 26% of

TEDT. These estimates may only roughly estimate RERT because

temperature was not controlled during the period of egg

retention. It is clear that turtles are brief egg retainers.

If Ewert's estimates of RERT for turtles are accurate, then

it the majority of lizards may have RERTs that are only

slightly longer (5% to 10%) than turtles.

Of all the amniotes, birds are the briefest egg

retainers. Not only are eggs retained in the reproductive

tract for a brief period (approximately 24 hours), but they

are expelled directly after shell deposition. Birds are

unique among amniotes in that they exhibit sequential

ovulation. A single follicle is ovulated, fertilized,

shelled and oviposited within a 24 h period. This sequence

continues until all the eggs in a clutch have been laid. A

brief period of ER is a necessary consequence of this system

of egg production. The TEDT of a chicken embryo is 22 days

and ERT is 1 day; therefore, the RERT of the domestic chicken










is 4.5%. Avian embryos develop to the blastula stage by

oviposition (Avery, 1974).

Brief-egg-retaining reptiles may necessarily hold eggs

for the minimal period of time required for eggshell

deposition. The majority of shelling occurs within 8 to 10

days of ovulation in S. woodi, S. virgatus and .. scalaris

(DeMarco, unpublished data). I suggest the reason there are

few instances of squamates ovipositing eggs with very early

embryos is because the time required to produce the eggshell

is greater than the time required to complete the earliest

stages. The lack of data concerning the timing of eggshell

deposition and embryonic development schedules has probably

led to the perception that lizards are exceptionally long egg

retainers. Thus, the many species that oviposit clutches

soon after shelling has been completed exhibit only a brief

period of egg retention beyond the time eggs must spend in

the oviducts for shell deposition.

When referring to the period of egg retention,

investigators should make every practical attempt to relate

egg retention time to total embryonic development time for

embryos developing at a constant temperature. I propose that

squamates be considered either brief, intermediate, or

prolonged egg retainers and that the determination be made on

the basis of the trimester in which the mean, median, and

mode ESO occurs. A problem arises with these definitions if

some females oviposit eggs with embryos at developmental

stages within an earlier trimester and some females oviposit










eggs with embryos in the next trimester. A determination of

the degree of ER would be based on the trimester in which two

of the statistics occurred. For instance, the ESOs for 24

Sceloporus virgatus ranged from stage 30 to stage 35 (Fig.

3.1). The mode of this distribution is stage 31, a stage

that is achieved at the end of the first trimester. Both the

mean and the median ESO occur between stages 32 and 33 which

occur in the second trimester. Therefore, S. viraatus would

be considered an intermediate-egg-retaining species. Using

the above criteria to determine the degree of ER in a.

scalaris is more difficult. The mean and median ESO of a

sample of 20 females occurs between stages 36 and 37. Stage

36 occurs at the end of the second trimester and stage 37 at

the beginning of the third trimester. The mode ESO for this

species is stage 38 which represents 75% of TEDT. Because

the median splits the distribution of ESOs between the second

and third trimesters and the mode occurs in the third

trimester, this species should be considered a prolonged egg

retainer.

The relative absence of prolonged-egg-retaining species

could be a reflection of any of a number of factors. First,

knowledge of embryo stage at oviposition is limited to only a

few dozen of the over 3750 species of lizards (Bellairs,

1986). Second, there may be morphological or physiological

constraints that make it difficult to increase the period of

egg retention (Packard, et al., 1977; Tinkle and Gibbons,

1977; and Guillette et al., 1980; Guillette, 1982). Once






51



crucial adaptations have arisen enabling a species to retain

eggs for longer periods, there may follow an intense period

of rapid evolution toward viviparity. Therefore, at any one

point in geological time, few prolonged egg retainers may

actually exist. Although, among reptiles, there are hundreds

of species of brief egg retainers and live bearers, we only

know of a handful of species that oviposit eggs with embryos

in the last trimester of development.
















CHAPTER 4
RAPID EMBRYONIC DEVELOPMENT: AN ADAPTATION TO COLD CLIMATE



Cold environments present several challenges to

reptilian embryos. First, development at cold temperatures

is prolonged (Sexton and Marion, 1974; Muth, 1980; and Shine,

1983 ). At high altitudes and latitudes, the time when nest

temperatures are suitable for development may be shorter than

the time it takes embryos to complete development. Because

partially developed reptilian embryos are not known to

overwinter, unhatched embryos never hatch (Shine, 1983).

When embryos are able to complete development in a cold

environment, hatching may occur after an optimal time for

hatchling survivorship (Shine and Bull, 1979). Second,

embryos developing at low temperatures may experience greater

mortality and morbidity. Low temperatures are known to

result in higher incidences of morphological (Fox et al.,

1961; Vinegar, 1974), and behavioral abnormalities (Burger,

1990).

Reptiles cannot reproduce successfully in cold climates

unless embryos are buffered against the negative consequences

of incubation at cooler temperatures. Partial or complete

incubation of embryos within the female, i.e., prolonged egg

retention or live bearing, have been hypothesized to be such










mechanisms (Weekes, 1935; Neill, 1964; Packard et al., 1977;

Shine and Bull, 1979). In cold environments, gravid or

pregnant individuals can raise their body temperatures, as

well as the temperatures of embryos, by basking. Thus, egg-

retaining females can maintain embryos at higher average

temperatures than those found in terrestrial nest cavities.

When the eggs of a prolonged egg-retaining female are

oviposited they contain more advanced embryos that hatch

sooner than eggs that had not been retained as long.

Furthermore, retained embryos may be protected from the

deleterious effects of frosts because females would choose

warmer underground refugia at night (Shine and Bull, 1979).

The high proportion of viviparous squamates occurring in cold

climates (Tinkle and Gibbons, 1977; Shine and Berry, 1978;

Guillette et al., 1980) may not prove that live bearing

evolved in cold climates, but attests to the success of this

mode of reproduction in such environments.

Cold climates could simultaneously select for a number

of traits that improve the chances of successful

reproduction. In addition to maternal egg retention, embryos

could evolve greater tolerance to low temperatures and more

rapid rates of development at non-lethal temperatures (Shine

and Bull, 1979). These traits would enhance embyro survival

by reducing development time, embryonic mortality from short

term exposure to extreme cold temperatures, and the

possibility of developmental abnormalities due to low

incubation temperatures There is little evidence to suggest










that greater tolerance to low temperatures and more rapid

embryogenesis have evolved in squamates inhabiting cold

climates. Fitch and Fitch (1967) found little variation in

low temperature tolerances in several species of reptiles

from Kansas. Conversely, the range of optimal development

temperatures is lower in some temperate and montane reptiles

(Sceloporus undulatus: Sexton and Marion, 1974; Coluber

constrictor: Burger, 1990) than in some tropical or xeric

species (e.g. Iguana iguana: Licht and Moberly, 1964;

Dipsosaurus dorsalis: Muth, 1980). This comparison is

interesting, but equivocal, because it does not clearly

demonstrate that low temperature tolerance of embryos results

from selection by cold temperatures. In some groups

embryonic temperature tolerances could be constrained by

phylogeny rather than a consequence of selection by

temperature. Furthermore, there is no evidence for the

evolution of more rapid embryonic development in squamates

found in cold climates. Shine (1983) reported similar

incubation times in low and high altitude populations of

Lampropholis guichenoti which suggests that rates of

development have not responded to cold temperature in this

species. In contrast, there are at least seven species of

turtles that have shorter incubation times in northern

populations and it has been suggested that colder

environments at higher latitudes select for more rapid

development (Ewert, 1979). Embryonic adaptation to cold

temperatures has been reported in several species of anurans










(Moore, 1949; Zweifel, 1968). These observations neither

support adequately, nor reject firmly the hypothesis that

greater tolerance to cold temperatures and more rapid

embryogenesis have evolved in squamates in response to cold

climates.

I tested the hypothesis that lizard species inhabiting

cold climates (high altitude) have evolved more rapid rates

of embryonic development then species found in warmer

climates (low altitude). Development rates can be estimated

indirectly by measuring the total time it takes embryos to

develop at constant temperature (Shine, 1983). Assuming eggs

sizes are similar for the species compared, it could be

concluded that species with shorter development times have

more rapid rates of development than species with longer

development times.

Measuring development times in lizards is problematic.

Lizards are known to retain eggs and, unlike turtles and

crocodilians, embryos continue development in utero,

temperature permitting. Thus, without knowing developmental

stages at oviposition, measures of incubation time, the

period from oviposition to hatching, would be of little

comparative value. In squamates, measures of the entire

period of embryonic development, from ovulation to hatching,

is preferred. In some lizards the approximate time of

ovulation can be determined by frequent palpation of females

that possess large ovarian follicles (Cuellar, 1971).










I compared total embryonic development times at the same

constant temperature in 3 oviparous species in the lizard

genus Sceloporus to test the hypothesis that cold climates

select for more rapid development. The oviparous species

occur along an altitudinal and thermal gradient. The species

are, S. woodi, S. virgatus, and B. scalaris, which can be

characterized as having a brief, intermediate, or prolonged

period of egg retention, respectively. This protocol

minimizes potentially confounding effects of different

incubation temperatures or egg retention times among these

species. After measuring development times in these

oviparous species I measured development time in a viviparous

congener, a. jarrovi. This species is montane and occurs at

both intermediate and high elevations.



Methods




Development Time in Oviparous Species



Laboratory study

The method used to determine total embryonic development

time (TEDT) (ovulation to hatching) in S. woodi, S. virgatus,

and S. salaries at 300C is described in the Methods section

of Chapter 3. The temperature chosen for this experiment had

to satisfy several requirements. First, the temperature had

to result in high hatching success. Second, because a










portion of embryonic development took place in utero a

temperature was required that permitted gravid females to

remain active and feed. Gravid females will not feed at some

lower constant temperatures, consequently their physical

condition could deteriorate. Finally, a temperature that is

ecologically relevant for each species is desirable. I

suggest that meets these requirements. Although average

developmental temperatures may be lower than 300C in these

species of Sceloporus, this temperature is undoubtedly within

the range of temperatures in which eggs are exposed under

natural conditions.


Field estimates of development time

Studies of the demography or reproductive cycle were

obtained from the literature and used to estimate TEDT and

egg retention (ER) time in the field (Sceloporus woodi,

Jackson and Telford, 1974; S. virgatus, Vinegar, 1975; and S.

scalaris, Newlin, 1976, Ballinger and Congdon, 1981). These

authors estimated the time of ovulation, oviposition, and

hatchling emergence using censusing techniques, such as mark-

recapture or periodic sampling and autopsy of individuals.



Development Time in a Viviparous Species


In early November, 1989, 19 female S. jarrovi were

collected from the same locality as S. virgatus. At this

time, S. jarrovi are either late vitellogenic or have

recently ovulated and are pregnant. Four females were










autopsied on November 11, 1989. Three of the four lizards

were pregnant and had eggs with blastodiscs, indicating

recent ovulation. One lizard had large vitellogenic

follicles in both ovaries. A stigma, a small avascular ring

through which the ovum passes at ovulation, was present on

each follicle, indicating that these follicles would soon

rupture from the follicle. In this species, it is more

difficult to be certain of a female's reproductive condition

(vitellogenic versus pregnant), so it was assumed that the

remaining females were either very late vitellogenic or very

early pregnant.

On November 11, the remaining 15 females were placed in

a large walk-in incubator (4 x 3 x 2.5 m). Chamber

temperature was adjusted so that the temperature of a

covered 1-liter Ehrlenmeyer flask filled with tap water was

300C. The chamber contained two large piles of cement blocks

and lizards were allowed to run freely throughout the

chamber. Water and food were provided ad lib. A photoperiod

of 13L /11D was provided. When hatchlings appeared they were

collected and weighed. Post-partum (PP) females, recognized

by the presence of a collapsed abdomen and considerable

weight loss, were removed immediately and the date of

parturition recorded. The mean time to parturition from

November 11, 1989 of this group of females was used as an

estimate of TEDT for this species. Because the exact

ovulation dates are not known and because autopsy of 4

lizards revealed that lizards had either enlarged ovarian










follicles or recently ovulated, lizards with the briefest

TEDTs probably underestimate this variable, while those that

have the longest TEDTs overestimate TEDT.



Results




Development Time in Oviparous Species



Laboratory study

Total embryonic development time (TEDT) was

significantly different among the three oviparous species

(ANOVA: F = 162.9; df = 2, 26; P < 0.0001). TEDT ( 1 s.d.)

in Sceloporus woodi, a. virgatus, and a. scalaris was 61.6 (

2.88, N = 10), 55.3 ( 1.42, N = 10), and 45.6 ( 0.85, N =

9) days, respectively. Scheffe's tests indicate that all

pairwise species comparisons of TEDT are different.


Field observations on development time

Jackson and Telford (1974) and DeMarco (unpublished

data) found that ovulation of the first clutch of the season

for a. woodi from North-Central Florida (elevation = 24 m)

begins in late March. Females begin oviposition in mid-

April. The period of egg retention in the field is

approximately 15 to 20 days for the first clutch of the year.

Hatchlings are first seen in late June. A minimum estimate

of TEDT for the first clutch of the year is 90 days. Mean

monthly ground temperatures 10 cm below the surface, the










depth at which this species most likely oviposits, are below

300C in April, May, and June (Jackson and Telford, 1974). It

is not surprising that TEDT is long for the early season

clutch Total embryonic development time of subsequent

clutches could be shorter because ground temperatures 10 cm

below the soil surface average 30.10C in July and August. In

late May, 1991, a clutch of S. woodi eggs was discovered

buried in a sand pushup in front of a gorpher tortoise

(Gopherus pol~yhemus) burrow (personal observation). The

sand was very moist and the clutch was located 10 cm below

the surface. The area adjacent to the tortoise mound was

devoid of trees and large shrubs that could act to shade the

mound from the sun.

Vinegar (1971) estimated that the time from ovulation to

oviposition in a. virgatus in the Chiricahua Mts., Arizona

(elevation = 1650 m), was 27 to 31 days (late May to late

June). Estimates were based on both autopsy of females

throughout the reproductive period and mark-recapture of

individuals. Hatchlings first appeared in late August. A

minimum estimate of TEDT is 90 days. Rose (1989) has

observed several .. virgatus ovipositing. Nests were placed

under rocks located in open sunny areas. Two clutches took

longer than 85 days to hatch. Assuming a 30 day period of

egg retention, the TEDT of some clutches may exceed 110 days.

Newlin (1970) estimated that a. scalaris from the

Chiricahua Mtns., Arizona (elevation 2600 m) retained

eggs for at least 50 days (late May to late July). His










estimate was based on autopsy of females throughout the

reproductive period. Hatchlings first appear in early

September. A minimum estimate of TEDT is 100 days, but many

clutches probably take longer.

Comparing estimates of development time in the field for

S. arrovi with those of the oviparous species is confounded

by differences in reproductive cycles. All the oviparous

species are spring breeders and oviposit in the spring or

summer. Courtship, mating, and pregnancy in S. jarrovi occur

in the fall, before the animals enter hibernacula in

November. On sunny winter days adults will bask for

approximately 4 hours per day (Congdon et al., 1979) and

raise their body temperatures to 32 or 330C at a low

elevation site near the Southwestern Research Station (SWRS).

Tinkle and Hadley (1973) found that body temperatures of

basking lizards during winter were only 250C at higher

elevations. Females collected in mid-March near the SWRS

contain embryos at stage 30 or less, which suggests that

embryos complete less than a third of development (see

Chapter 3) before females disperse from the hibernaculum.

Parturition begins about a month and a half later in late May

and early June. A minimal estimate of TEDT is 7 months at

the SWRS and 8 months at a higher elevation site, Rustler

Park. If a. jarrovi ovulated in the early spring instead of

fall I would predict that TEDT in the field would be about

equal to the minimum estimate for TEDT in S. scalaris (100

days). This suggestion is based on the observation that










slightly more than two thirds of embryonic development in S.

iarrovi at the SWRS site takes approximately 70 to 75 days

(mid-March to late May). At Rustler Park (RP) the onset of

spring occurs later (Ballinger, 1979), embryos may be less

developed at the beginning of spring, and nighttime

temperatures are cooler. Despite the generally cooler

climate at Rustler Park, parturition begins just two weeks

later than at the SWRS. This is quite remarkable and may be

because high elevation lizards achieve similar daytime body

temperatures for a similar number of hours each day as

lizards at lower elevations (Ballinger, 1979).



Environmental Variation Between Study Sites


From 1973 to 1976 Ballinger (1979) measured ambient

temperatures at localities within a kilometer of both Arizona

sites used in this study. He concluded that temperatures

were lower at the high altitude site (2,500 m) and that the

growing season for S. jarrovi was between 170 and 175 days

(late April to early or mid-October), whereas the growing

season at the low elevation site (1,675 m) was between 220

and 225 days (early April to mid-November). In north-central

Florida, the growing season is probably the longest of the

three sites. Sceloporus woodi become vitellogenic in late

February and eggs continue hatching through early November

(Jackson and Telford, 1974). Although the growing season for










,. woodi has not been specifically determined it is probably

from early March to late November (> 250 days).



TEDT in a Viviparous Species


Of the 15 Sceloporus jarrovi placed in the environmental

chamber, eleven gave birth to normal offspring, two died, and

two were not pregnant. Mean ( 1 s.d.) TEDT is 56.6 days (

3.3). Total embryonic development time is significantly

different among the four species (ANOVA: F = 74.3; df = 3,

36; P < 0.0001). Scheffe's test was used for all possible

comparisons between S. jarrovi and the oviparous species.

TEDT in a. jarrovi is only shorter than TEDT in a. woodi.

There is no difference in TEDT between a. iarrovi and S.

viraatus, whereas scalaris has a shorter TEDT than S.

jarrovi.



Discussion


Estimates of TEDT

The hypothesis that more rapid embryogenesis is an

adaptation to cold climate is supported by this study.

Development times decrease with increasing altitude in

different oviparous species occurring at different altitudes

(Table 4.1). It is also evident that TEDT decreases with

increasing egg retention time in these 3 species of

Sceloporus.



















4-) 0)



>
-,-
4) 0
00

0 44 W
-11
41 4)
-H
CoO






400



0 0 0
r4





O 44
4-) 44 V)
4-) 0E-


0 co a




C -0










00 4E-)
C O











1:4
>0 0 -
-H a









0 > c
SO M














>10 U


4 4-) -
4) 0







S>1i
0 440
00 *
- 4l 4-)
























04-1 0'C

-H OX


*W E- 0
M (4.)
.0

























'4 -H 0 4)
(UPO --
C MIO
C 4-1 1











0 4- 0 1
a 1 X)rtl










E^ -p a 4-


0.0
o 0
D4


04)
a




+1
o
C>




(N
v


a (


o4


0 -i




,-
W 4-1















m



a,-
O
0 H
r4 -H
444



0



4-4




Ca


Irx (a





(0
41


44






r4 0


o00


+1


U)
In
+1


LO (mi Co


Cl



CN


UC)


0
04 U)O


0 0
-4 1-4


D

4-4-)

-I 4-)
fum


Un



+1


o
(I)
CO






o
CM)

0
C0










-I










The hypothesis that more rapid embryogenesis evolves in

response to cold temperatures rests on the assumption that a

species possesses a slow rate of embryonic development prior

to its occurrence in cold habitats. Supporting evidence is

lacking. Several species of Sceloporus could be used to test

this assumption (e.g., S. undulatus, S. occidentalis, S.

variabilis, and a. graciosus). They are wide ranging and

populations can be found at low and high altitude.

Failure to find intraspecific variation in TEDT among

populations at different altitudes would not necessarily

disprove that more rapid development is an adaptation to cold

climate. Environments affecting populations at different

altitudes can act to change seasonal and daily activity

periods, as well as affect the duration of the growing season

(Ballinger, 1979; Grant and Dunham, 1990). Changes in

activity periods and growing seasons can lead to changing

demographics and the evolution of different life-history

traits. Certain combinations of traits that permit

populations to exist in colder climates may not, at first,

require the evolution of more rapid embryogenesis.

In this study, TEDT was measured at a single temperature

(300C), a temperature widely used for incubating reptilian

eggs because it results in high hatching success in many

species. Ideally, TEDT and embryo mortality should be

measured at several different constant temperatures; however,

several observations suggest that 300C is within the range of

temperatures that are optimal for embryonic development among










species in the genus Sceloporus. Sexton and Marion (1974)

incubated eggs of S. undulatus from Missouri at 25, 30, and

350C and found that hatching success was highest at 25, and

300C. Evidence from studies of the thermoregulatory behavior

of two viviparous species, a. iarrovi and S. cyanogenys, also

suggest that optimal development temperatures may be close to

30 OC (Garrick, 1974; Beuchat, 1986). Pregnant lizards

maintain body temperature (Tb) near 320C, which is lower than

Tbs found in nonpregnant lizards and males. Over a large

percentage of the time it takes embryos to develop,

viviparous females may be regulating daytime Tb at

temperatures that are near optimal for embryonic development,

but below temperatures that are optimal for adult activity.

In the field, nest temperatures will fluctuate on both a

daily and seasonal basis (Packard and Packard, 1985).

Average nest temperatures may be several degrees lower than

300C, however, there is evidence to suggest that female

Sceloporus select nest sites that, at times, approach or

exceed 300C. Jackson and Telford (1974) showed that average

monthly ground temperatures in North Florida, 10 cm below the

surface, ranged between 24.9 and 30.10C during the months in

which eggs of a. woodi were incubating. Rose (personal

communication) found the average nest temperatures of S.

virgatus to be approximately 250C. She also found that

maximum nest temperatures could reach 300C for short periods.

Guillette and Lara Gongora (1986) reported average mid-day

nest temperatures in the montane lizard, S. aeneus, between










28.6 and 30.10C. aeneus and S. scalaris are closely

related members of the oviparous scalaris complex of lizards

(Guillette and Smith, 1985), have similar periods of egg

retention, and occur at similar elevations and in similar

open habitats that consist primarily of bunch grasses

(Guillette, personal communication). Daily and seasonal

ground temperatures may be similar in these two species.

Although these observations support the assertion that 300C

is within the range of ecologically relevant incubation

temperatures, it is clear that development at 300C occurs for

a very small proportion of the time embryos are in the nest.

There is an obvious need for detailed studies of the thermal

and nesting ecology of these species.



TEDT and Hatchling Size


An alternative explanation for the measured differences

in TEDT is that TEDT may scale to egg or hatchling size.

Hatchling mass is different in the three oviparous species

(ANOVA: df= 2,37, F = 16.9, p = .0001). The mean ( 1 S.D.)

mass of 85 hatchling E. jarrovi from the 11 females used to

measure TEDT was 0.596 ( 0.046). This species was not

included in the ANOVA because clutch means were not

available. There is no relationship between TEDT and

hatchling mass in these 4 species (Fig. 4.1). Concluding

that there is no relationship between TEDT and hatchling size

in this genus based on a sample with four species is not










warranted. These data do suggest, however, that within any

given hatchling size-class there may be species with

relatively long TEDTs and species with relatively short

TEDTs. On the other hand, scaling relationships usually

illustrate general trends in traits being scaled against mass

when mass varies by several orders of magnitude. It is

doubtful that we could demonstrate an affect of egg or

hatchling mass on development times in the genus Sceloporus

if there is so little variation in egg or hatchling mass

within this taxonomic group. Additionally, analysis of

covariance can not be used to determine whether a species

TEDT is a function of egg or hatchling size because TEDT is

not related to egg or hatchling size within species. I

suggest that egg or hatchling mass explains little of the

variation in TEDT found in this study. Incubation times of

similar sized bird eggs can vary by several hundred percent

(Rahn and Ar, 1974). Although the relationship between TEDT

and egg or hatchling size is not known for reptiles it is

likely that a scaling relationship does exist in this group.

It is also likely that there is a great deal of variation in

TEDTs among similar sized eggs and that the exponent, b, of

the allometric relationship is very small, like that found in

birds (Rahn and Ar, 1974). Given a small exponent, egg or

size would explain only a small fraction of the observed

interspecific variation in TEDT in Sceloporus lizards.





















70 m mBdi
710
65 iarrovi
SI 18 11

S O 01011
55 8 18

U virgatus
50 9

45 i 12
scalaris
40. .. .
0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70

MEAN HATCHLING MASS (g)




Figure 4.1. Relationship between mean ( 1 s.d.) total
embryonic development time (TEDT) and mean ( 1 s.d.)
hatchling mass (HM) in Sceloporus woodi, S. virgatus, and
S. scalaris and S. jarrovi. The regression equation
(line not shown) is TEDT = 43.9 + 24.4 HM; df = 3, r2 =
0.15, p = .61. Numbers beside error bars represent the
number of clutch means in the sample.












TEDT in a Viviparous Species


Given that TEDT decreases as egg retention increases in

three oviparous species of Sceloporus (Table 4.1), it was

predicted that a viviparous species, which retains eggs the

longest, would have the shortest TEDT. However, if a

viviparous species had a longer than expected TEDT, this

would not necessarily falsify the hypothesis that more rapid

embryonic development was not important during the

transitional stages from oviparity to viviparity. On the

other hand, it could indicate that physiological constraints

associated with the respiratory and circulatory systems in

viviparous species exist and result in a longer TEDT.

Sceloporus iarrovi has an intermediate TEDT. If the

long TEDT found in S. woodi represents an ancestral condition

for this trait, then it appears that S. jarrovi has an

accelerated rate of embryonic development compared to a brief

egg retaining species. The fact that TEDT in S. jarrovi is

greater than TEDT found in the prolonged egg retaining

species, a. scalaris, and no different from TEDT in the

intermediate egg retainer, a. virgatus, may indicate that

TEDT is prolonged due to physiological constraints (e.g., gas

exchange) that prohibit the rate of embryonic development to

proceed at its genetic potential. Problems associated with

gas exchange between maternal and embryonic circulations are

hypothesized to have led to the evolution of a primitive










placenta (Weekes, 1935; Guillette, et al., 1980). Measures

of uterine arterial oxygen tension and the critical oxygen

tension of the late term fetuses within could help answer

this question. There is little evidence for oxygen limitation

in fetal placental mammals, as near term fetal lambs have

rates of metabolism that are similar to those of neonates

(Clapp, et al., 1971).



Field Estimates of TEDT


Reptilian nest temperatures are correlated with air

temperatures and can vary both seasonally and diurnally (see

review in Packard and Packard, 1986). Although nest

temperatures can be high in montane species of Sceloporus

(Guillette and Lara Gongora, 1986), the duration of favorably

warm nest temperatures that promote rapid development is

probably least in montane species and greatest in low

elevation species on both a seasonal and daily basis.

Sceloporus woodi were collected at low elevation in

peninsular Florida and have the longest TEDT. In north-

central Florida the thermal environment of nest sites chosen

by Florida scrub lizards may be, on average, more favorable

for rapid embryonic development on both a daily and seasonal

scale than the nesting environments of the other two species.

There may be no selective pressure to decrease development

time or egg retention time in this species. Sceloporus

virgatus were collected in shaded creekbeds at lower










elevation (~1650 m) in the Chiricahua Mtns. and have an

intermediate TEDT. Average nest temperatures (~250C, see

above) may be intermediate compared to the other two species.

Development time has been reduced and egg retention time

increased compared to S. woodi. Sceloporus scalaris were

collected on open grassy slopes at high elevations (-2600 m)

in the Chiricahua Mtns. in Arizona. Daily fluctuations in

nest temperature (5.2 to 30.10C) are probably similar to

those reported for the closely related a. aeneus (Guillette

and Lara Gongora, 1986). Traits such as prolonged egg

retention and more rapid embryonic development evolve to

compensate for low average nest temperatures that may be

typical of high elevation sites.

I have estimated the minimum TEDT for the first clutch

of the season for a. woodi to be 90 days. Minimum TEDT for

S. virgatus is 90 days and that for S. scalaris is 100 days.

Intraspecific variation in field TEDTs may be considerable

(Rose, 1989) and this may be due to genetic differences among

females for this trait and/or differences in thermal

environments between nests. Interestingly, these estimates

of minimal TEDT are remarkably similar despite the fact that

these species occur at very different elevations with

different thermal characteristics. Without the thermal

advantage resulting from increased egg retention and more

rapid rates of embryonic development the eggs of S. virgatus

and S. scalaris would probably not hatch before autumn. It

would be very useful to quantify the independent effects more










rapid development and ER have on reducing the time it takes

embryos to develop.



Implications for the Evolution of Viviparity


This study raises a question concerning why selection

has operated on more than one set of alleles, each of which

has the similar effect of reducing development time. For

instance, why doesn't selection act exclusively on alleles

controlling egg retention time or on those controlling more

rapid development. Temperature is one of the most important

physical variables affecting organisms. Temperature

potentially affects all aspects of an organism's biology,

therefore, it is no surprise that there may be more than one

set of alleles responding to cold temperature, each set

contributing to a change in a single phenotype.

It has been suggested that the evolution of a massive

acceleration in embryonic development rate could preclude

further increases in egg retention thereby preventing the

evolution of viviparity (Shine, 1983a). In order to

demonstrate that such a massive acceleration in embryonic

development has occurred it would be expected that

interspecific variation in ESO would be due primarily to

differences in rate of development and secondarily to

differences in the proportion of time eggs are retained.

Although the differences in development rates among the

oviparous species are substantial, clearly more rapid










development on a scale capable of neutralizing selection for

increased ER is not indicated by this study. At 300C, S.

woodi retain eggs between 9.0 and 15.0 days which represents

14.6% to 24.3% of TEDT and a. scalaris retain eggs between

17.8 and 33.7 days which represents 39.0% and 73.9% of TEDT

(Chapter 3). If the maximum proportion of embryonic

development in -. scalaris (73.9 % at 300C) could occur in

the maximum number of days eggs are retained by woodi (15

days at 300C), then TEDT in a. scalaris would be 20.3 days at

300C. Although the rate of embryonic development does

explain some interspecific variation in ESO it is clear from

this analysis that other factors, e.g., egg retention, play a

major role.

Among squamates it is probably not possible to evolve

rates of development that are so rapid that they preclude the

evolution of viviparity in the coldest climates. On the

other hand, it is very likely that more rapid rates of

development may lower the threshold temperature which would

select for the evolution of viviparity (Fig.4.2). This

suggestion seems relevant to the hypothesis that viviparity

evolves in cold climates. As a species expands its range

into cooler environments females with shorter periods of ER

are selected against. However, because selection is acting

on development rate as well, development rate becomes a

variable in the equation explaining the relationship between

environmental temperature and RERT.










Do increased ERT and more rapid embryogenesis evolve

simultaneously or independently over the evolutionary history

of a species? It is possible that certain gene combinations

regulate both phenotypes simultaneously, therefore, as a

species evolves prolonged ER, more rapid embryogenesis also

evolves. Evidence for simultaneous evolution is limited to

the consistent negative correlation between TEDT and RERT in

three species of Sceloporus. This evidence only shows that a

change in one phenotype corresponds with a change in the

other phenotype in three species. Still, there is no

evidence of a potential genetic or environmental mechanism to

explain this correlation. Increasing the species sample size

would be a valuable test for this hypothesis. If the

correlation was consistent across a large number of species

we could then conclude there is strong evidence for

simultaneous evolution of increased ER and more rapid

embryogenesis. However, the presence in some intermediate or

prolonged egg retaining species of longer than expected TEDTs

would be evidence that these two phenotypes are not always

correlated. Shorter than expected TEDTs found in brief or

intermediate-egg-retaining species would also be evidence

that these phenotypes are not always correlated. From this

sort of evidence, it could be concluded that the genotypes

controlling these phenotypes were evolving independently.



































low


high


MEAN GROUND TEMPERATURE (C)








Figure 4.2. A model for the evolution of viviparity in the
genus Sceloporus. The model shows that as ground
temperatures decrease, the probability of clutch
survivorship increases as egg retention time and
developmental rates increase. (BER, brief egg retention;
IER, intermediate egg retention; PER, prolonged egg
retention; AED, accelerated embryonic development.)










Selection would act on one trait at one point in geological

time and on the other trait at another time.

For nearly a century, the primary focus of discussions

on the evolution of viviparity in reptiles has been on

environmental factors that select for increasing maternal egg

retention time in oviparous species (reviewed in Shine,

1985). Given our knowledge of the latitudinal and

altitudinal distribution of oviparous and viviparous reptile

species, most investigators, including this one, support the

hypothesis that prolonged egg retention and viviparity have

evolved in response to cold climates. However, the

successful invasion of cold habitats by reptiles may no

longer be viewed as being due solely to maternal egg

retention. This study focuses on the similar role an

embryonic adaptation has on reducing the time it takes

embryos to develop in cold climates.















CHAPTER 5
EMBRYONIC METABOLISM



Three distinct patterns of embryonic metabolism

have been described for birds and are correlated with

the degree of development of offspring at hatching (C.

Vleck et al., 1979; C. Vleck et al., 1980). Embryonic

oxygen consumption (VO2) increases exponentially

throughout the incubation period in altricial species.

In some precocial species, Vo2 increases during the

first 80% to 90% of the incubation period; after which,

the rate of increase slows until pipping. The third

pattern is considered a variation in the precocial

pattern and occurs in some ratite birds. This pattern

is characterized by a peak in VO2 followed by a decline

until hatching (Hoyt et al., 1978; D. Vleck et al.,

1980). The patterns of growth rate among avian embryos

are also correlated with the degree of development at

hatching. Absolute growth rate continues to accelerate

exponentially in altricial embryos, whereas growth rate

decreases after peaking in precocial embryos (C. Vleck

et al., 1980).

The patterns of embryonic metabolism have been

documented in several species of turtles (Ackerman,

1981a; Gettinger et al., 1984; M. Thompson, 1989; Webb,










et al., 1986), crocodilians (M. Thompson, 1989;

Whitehead and Seymour, 1990), and snakes (Dmi'el,

1970). The pattern of embryonic metabolism is unknown

for any lacertilian. As far as is known all reptilian

hatchlings are precocial. Thus, if the pattern of

metabolism is correlated with the degree of development

at hatching as it is in bird embryos, then it is

expected that reptiles would have either peaked or

sigmoid patterns of embryonic metabolism. With the

exception of snakes this is true. Snake embryos have

an exponential pattern of VO2 (D'miel, 1970), which is

the same pattern found in altricial bird embryos.

Attempts to correlate embryonic patterns of

metabolism with eggshell type are also inconsistent

(Whitehead, 1987). The eggshells of reptiles are

characterized into two types (Packard et al., 1977.

Calcareous eggshells, also known as brittle eggshells,

are very similar in structure to avian eggshells.

These eggs are surrounded by inner and outer fiber

layers underlying a thick hard calcareous layer

composed of calcite crystals. The calcareous layer is

penetrated by numerous pores through which gases can

move. Eggs with calcareous eggshells are true cleidoic

eggs. All crocodilians, many chelonians, and a few

lizards (Dibamids, and geckos in the subfamilies

Gekkoninae and Sphaerodactylinae) possess this type of

eggshell. Parchment eggshells, also known as flexible










eggshells, are typical of most squamates and turtles.

These shells are characterized by multiple thin layers

of fibers overlain by a thin layer of calcite crystals.

The crystal layer is relatively thin and disorganized

compared to that in calcareous eggshells. Unlike

calcareous eggs, parchment-shelled eggs can swell

tremendously during incubation.

The patterns of embryonic metabolism in 6 species

with calcareous eggshells (2 crocodilians and 4

chelonians) are all peaked (Lynn and von Brand, 1945;

Gettinger et al., 1984; Webb et al., 1986; Thompson,

1989); however, reptiles with parchment-shelled eggs (4

turtles and 7 snakes) have either peaked (Lynn and von

Brand, 1945; Gettinger et al., 1984), sigmoid

(Ackerman, 1981a), or exponential (Clark, 1953; D'miel,

1970; Black et al., 1984) patterns of embryonic

metabolism. Whitehead (1987) noted that species with a

peaked pattern of embryonic metabolism tended to have

longer incubation periods. Data from more species is

required to confirm such a suggestion.

This study examines the patterns of embryonic

metabolism in Scloporus woodi, a brief egg retainer,

S. virgatus, an intermediate egg retainer, S. scalaris,

a prolonged egg retainer, and S. iarrovi, a viviparous

species. In this study, I compare the patterns of

embryonic VO2 in lizard species that vary along the

oviparity-viviparity continuum and to assess whether










possible differences in metabolic patterns are

correlated with the degree of egg retention or other

traits (or environments) correlated with egg retention

time. Alternatively, there may be no differences in

the patterns of VO2, which would suggest that the

metabolic patterns of metabolism in this genus are

conservative.


Methods



Embryonic Metabolism


Oviparous species

The methods used for collecting and incubating S.

woodi, 5. virgatus, and S. scalaris eggs are described

in Chapter 2. The technique used to measure Vo2 of

embryos is also described in Chapter 2.



Viviparous species

Pregnant S. jarrovi were killed by decapitation

and eggs were gently teased from the oviducts and

floated into a watch glass filled with anoline ringers

(Guillette, 1982b). The eggs of S. jarrovi are shell-

less and extraembryonic membranes are transparent.

Eggs were maneuvered into a small cup shaped vessel

constructed out of filter paper. Embryos were then

placed in respirometer chambers that had been placed in










the environmental cabinet an hour earlier. Vo2 was

measured in the manner described in Chapter 2. Only

embryos with no apparent damage to their external

extraembryonic membranes were measured. Following

measurement of V02, larger embryos were prodded with a

blunt probe in order to stimulate movement. Typically,

this caused embryos to wiggle vigorously. This

behavior was an indication that the embryo was alive

and healthy. Subsequently, embryos were dissected free

of their yolk and extraembryonic membranes. Embryonic

stage was noted and yolk free wet and dry masses were

measured.


Statistical Analysis


Patterns of VO2 are determined by plotting the

relationships between the independent variable, %TEDT,

and the dependent variable, V02 (cc 02 h-1). First,

V02 was measured on individual eggs. Subsequently,

embryos were dissected free of the egg. Embryo stage

was noted and embryo wet and dry mass was measured.

Direct calculation of %TEDT for an individual embryo

was not possible if TEDT was not known for that

embryo's clutch. Alternatively, %TEDT was estimated

using the method described in Chapter 3. Actual and

estimated TEDTs were used to form the regressions.










Data were fitted to linear and polynomial

regressions. In all cases, fourth order polynomials

provided the highest coefficients of determination.

Points used to generate the equations are independent

with respect to individuals, but not with respect to

clutches. It would not have been practical to use a

single egg per clutch to generate these relationships.

Nonetheless, at any embryo stage, intraclutch variation

in V02 is probably similar to interclutch variation in

this variable; therefore, I treat all data as

independent.

VO2 of seven a. woodi eggs was measured repeatedly

to see if the pattern of Vo2 in individuals is similar

to that found for a large sample of eggs.


Results



Sceloporus woodi


The pattern of metabolism in a. woodi is peaked

(Fig. 5.1). For the first 60% of incubation Vo2

increases slowly. Afterwards VO2 increases rapidly,

peaks, and then declines during the last five days

before hatching. The following polynomial regression

equation relates V02 and %TEDT:










V02 = -0.045776 + 7.43 10-3 (x) 2.802 10-4 (x2)

+ 4.3559 10-6 (x3) 2.1684 10-8(x4)

(5.1)



(n = 62, r2 = 0.94, and P < 0.0001). Peak embryo

metabolism, VO2[max], predicted from eq. (6.1) is 0.106

cc 02 h-1 and occurs at 89% of TEDT. The V02 of

embryos just prior to hatching, V02 [pre-h], is 0.094 cc

02 h-1 and is 18% lower than V02[max]. For the most

part, the pattern of metabolism of the 7 eggs for which

V02 was measured repeatedly was similar (Fig. 5.2).

Curiously, 4 of the 7 eggs had a higher VO2 on the day

of hatching than on the day prior to hatching. On the

day of hatching all seven embryos were partially out of

the egg during the measurement period, i.e., they

appeared to be in a similar stage of the hatching

process. Thus, I can not correlate differences in

embryo metabolism during hatching to different stages

of hatching (but see discussion).


Sceloporus vircatus


Although slightly different from S. woodi, the

pattern of metabolism in S. virgatus is peaked, but

less so than S. woodi (Fig. 5.3). The following

polynomial regression equation relates VO2 and %TEDT:










VO2 = -0.052847 + 6.204 10-3 (x) 1.825 10-4

(x2)

+ 2.7396 10-6 (x3) 1.3817 10-8 (x4)

(5.2)



(n = 101, r2 = 0.93, and P < 0.0001). Vo2[max],

predicted from eq. 6.2, is 0.118 cc 02 h-1 and occurs

at 88% of TEDT. VO2[pre-h], is 0.100 cc 02 h-1 and is

15% lower than V02[max].


Sceloporus scalaris


The pattern of metabolism in S. scalaris is

transitional between a peaked and sigmoid pattern (Fig.

5.4). The following polynomial regression equation

relates V02 and %TEDT:



Vo2 = -0.057136 + 7.604 10-3 (x) 2.479 10-4

(x2)

+ 3.5362 10-6 (x3) 1.6528 10-8 (x4)

(5.3)



(n = 115, r2 = 0.92, and P < 0.0001). VO2[max],

predicted from eq. 6.3, is 0.114 cc 02 h-1 and occurs

at 94% of TEDT. VO2[pre-h], is 0.107 cc 02 h-1 and is

6% lower than V02[max].











Sceloporus iarrovi


The pattern of embryonic metabolism in S. iarrovi

is sigmoid (Fig. 5.4). The following polynomial

regression equation relates VO2 and %TEDT:



V02 = -0.084745 + 1.078 10-2 (x) 3.830 10-4 (x2)

+ 5.7537 10-6 (x3) 2.7217 10-8 (x4)

(5.4)



(n = 92, r2 = 0.96, and P < 0.0001). V02[max], predicted

from eq. 6.4, is 0.199 cc 02 h-1 and occurs at 96% of

TEDT. VO2[pre-h] is 0.195 cc 02 h-1 and is 2% lower

than V02[max]. The pattern of VO2 is more plateaued

then peaked as indicated by the slight difference

between V02[max] and V02[pre-h]
































00

S 0.08 4
0.06
0.04 -

0.02 *

0.00--
10 20 30 40 50 60 70 80 90 100


%TEDT





Figure 5.1. V02 as a function of total embryonic
development time (TEDT) for Sceloporus woodi eggs.
Curve represents a fourth order polynomial
regression.























C>


50 60 70 80 90 100


3

0


50 60 70 80 90 100


% INCUBATION





Figure 5.2. Vo2 of 7 Sceloporus woodi eggs from 2
clutches as a function of day of incubation. Each
curve represents several measures on a single egg.
(a) Eggs from clutch 1. (b) Eggs from clutch 2.


A
clutch 1