• TABLE OF CONTENTS
HIDE
 Title Page
 Acknowledgement
 Table of Contents
 List of Tables
 List of Figures
 Abstract
 Introduction
 Distribution and population phenetics...
 The phenetics and ecology of an...
 A search for the population asymmetry...
 References
 Biographical sketch














Group Title: population phenetics and behavioral ecology of the Florida scrub lizard (Sceloporus woodi)
Title: The population phenetics and behavioral ecology of the Florida scrub lizard (Sceloporus woodi)
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00099154/00001
 Material Information
Title: The population phenetics and behavioral ecology of the Florida scrub lizard (Sceloporus woodi)
Physical Description: x, 119 leaves. : illus. ; 28 cm.
Language: English
Creator: Jackson, James Frederick, 1943-
Publication Date: 1972
Copyright Date: 1972
 Subjects
Subject: Lizards -- Florida   ( lcsh )
Zoology thesis Ph. D
Dissertations, Academic -- Zoology -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis -- University of Florida.
Bibliography: Bibliography: leaves 113-118.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00099154
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000577401
oclc - 13977838
notis - ADA5096

Downloads

This item has the following downloads:

PDF ( 5 MBs ) ( PDF )


Table of Contents
    Title Page
        Page i
    Acknowledgement
        Page ii
    Table of Contents
        Page iii
        Page iv
    List of Tables
        Page v
        Page vi
    List of Figures
        Page vii
        Page viii
    Abstract
        Page ix
        Page x
    Introduction
        Page 1
        Page 2
    Distribution and population phenetics of sceloporus woodi
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
    The phenetics and ecology of an extraordinarily narrow hybrid zone
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
    A search for the population asymmetry property
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
    References
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
    Biographical sketch
        Page 119
        Page 120
        Page 121
Full Text














The Population Phenetics and Behavioral Ecology of
the Florida Scrub Lizard (Sceloporus woodi)










By

James Frederick Jackson


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


UNIVERSITY OF FLORIDA
1972


















ACKNOWLEDGEMENTS


I would like to thank Drs. Archie F. Carr, A. M. Laessle, and

E. S. Ford for valuable consultation during the preparation of this

manuscript. For advice and assistance with the illustrations, I

am indebted to Mr. Paul Laessle. Discussions of statistical

aspects of the study with Mr. William Ingram and Dr. David Hughes

were very much appreciated.



















TABLE OF CONTENTS



Page

ACKNOWLEDGEMENTS .................. . ii

LIST OF TABLES ..... . . . . . . .... v

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

ABSTRACT ... . . . . . . . . . ix

INTRODUCTION .... . . . . . . . . . 1

CHAPTER I: DISTRIBUTION AND POPULATION PHENETICS
OF SCELOPORUS WOODI ..... . . . . 3

Introduction .... . . . . . . . . 3

Methods . . . . . . . . ... . . . 4

Results .................. ..... 17

Discussion .................. . 52

Phenetic Affinities .............. 52
Adaptive Interpretation of Variation . . .. 53
Dispersal History . . . . . . . . 56
Comparison with Other Species . . . . . 65
Gene Flow and Differentiation . . . ... 66

CHAPTER II: THE PHENETICS AND ECOLOGY OF AN
EXTRAORDINARILY NARROW HYBRID ZONE . . .. 69

Introduction . . .... .............. . 69

Methods .................. ..... 71








iii









TABLE OF CONTENTS (Continued)


Page

CHAPTER II (Continued)

Results ........ . . . . . . . 75

Phenetics on the Ecotones . . . . .. 75
Interspecific Behavior . . . . . . 86
Habitat Selection . . . . . . ... 90
Food and Foraging Behavior . . . ... . . 92

Discussion .... . . . . . . . . 95

CHAPTER III: A SEARCH FOR THE POPULATION
ASYMMETRY PROPERTY . . . . . ... 104

Introduction . . . . . . . .... . 104

Methods . . ... ........... .. . . 105

Results . . . . . . . . . . . 106

Discussion .... . . . . . . . . . 1ll

REFERENCES .................. . . .. 113

BIOGRAPHICAL SKETCH . . . . . . . ... . 119

















LIST OF TABLES


Table Page

1 Locations of Scrubs . . . . . . . 7

2 Description of Characters Examined . . . .... .13

3 Mean Values of Sexually Dimorphic Characters . . .. 19

4 Population Means for Male Sceloporus woodi ...... 20

5 Population Means for Female Sceloporus woodi . . .. 22

6 Standard Deviations for Male Sceloporus woodi .... 24

7 Standard Deviations for Female Sceloporus woodi .. .. 26

8 Character Correlation Matrix for Male Sceloporus woodi 28

9 Character Correlation Matrix for Female
Sceloporus woodi .. . . . . . . . . 30

10 Coefficients of the First Six Canonical
Variates for Males . . . . . . . . 33

11 Coefficients of the First Six Canonical
Variates for Females .... . . . . . ..... 35

12 Generalized Distances (D2) Between Sceloporus woodi
Populations (Males) ... . . . . . . 41

13 Generalized Distances (D2) Between Sceloporus woodi
Populations (Females). ... . . . . . . 43

14 Mean Regional Generalized Distances Between
Populations . . . . . . . .... .... 51

15 Assignment of Temperature Data to S. woodi Populations 55

16 Estimated Ages of Florida Marine Terraces ...... 57









LIST OF TABLES (Continued)


17 Description of Characters in the Discriminant
Function ......... . . . . . . .

18 Results of the Discriminant Analysis . . . . .

19 Results of the Male Choice Tests . . . . . .

20 Results of the Female Association Tests . . . .

21 Characteristics of the Ground Surface in Three
Tracts of Sandhill Vegetation . . . . . . .

22 Percentage of Total Individuals (I) and Total
Weight (W) by Prey Taxa . . . . . .

23 Percentage of Foraging Time in Various Positions . .

24 Asymmetry Values (d) of Six Characters in
Sceloporus Populations . . . . . . . .

25 Average Asymmetry Values (d) of Hybridizing and
Non-hybridizing Sceloporus Populations . . . . .


72

76

89

91


93


94

98


107


110


Table


Page
















LIST OF FIGURES


Figure Page

1 Distribution of sand-pine scrub and Sceloporus
woodi in Florida. Stippling indicates large
continuous areas of scrub. Enclosed numbers = scrubs
containing woodi; circles = woodi populations
examined phenetically; squares = other woodi popu-
lations. Unenclosed numbers = scrubs apparently
lacking woodi. ...... . . . . . . .

2 Projection of population means for males on the
first three canonical variates; K3 vertical.
Nearest neighbors by generalized distance (D2)
are linked. . . . . . . . ... ..... 37
3 Projection of population means for females on the
first three canonical variates; K vertical.
Nearest neighbors by generalized distance (D2)
are linked. . . . . . . ...... ..... 39

4 For each character the vector shows the direction
and magnitude of its influence on the first two
canonical variates of the males. See Table 2 for
character code .. .. ..... . . . . . 47

5 For each character the vector shows the direction
and magnitude of its influence on the first two
canonical variates of the females. See Table 2
for character code. .... . . . . . . 49

6 Distribution of Z values for males of two non-
hybridizing populations (below) and three hybrid
zone populations (above). Means are marked by arrows;
asterisks mark means that differ significantly from
those of the non-hybridizing populations . . ... 77









LIST OF FIGURES (Continued)


7 Distribution of Z values for females of two
non-hybridizing populations (below) and three
hybrid zone populations (above). Means are
marked by arrows; asterisks mark means that
differ significantly from those of non-
hybridizing populations. . . . . . . . . 79

8 Map of the Alexander Springs ecotone. Sand-
pine scrub indicated by stippling; sandhill
vegetation indicated by unshaded areas.
Dark circles represent S. woodi individuals.
Open circles represent S. undulatus individuals.
Partially dark circles represent hybrids of
varying degrees. . . . ... ......... . 82

9 Map of the Forts Bear Hole ecotone. Sand-pine
scrub indicated by stippling; sandhill vegetation
indicated by unshaded areas. Dark circles represent
S. woodi individuals. Open circles represent S.
undulatus individuals. Partially dark circles
represent hybrids of varying degrees .... ...... .84

10 Map of the Lake Eaton ecotone. Sand-pine scrub
indicated by stippling; sandhill vegetation indicated
by unshaded areas. Dark circles represent S. woodi
individuals. Open circles represent S. undulatus
individuals. Partially dark circles represent
hybrids of varying degrees. .. . . . . . . 87

11 Distribution of prey individuals and prey volume
over prey length for S. undulatus and S. woodi. ... .96


Figure


Page









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


THE POPULATION PHENETICS AND BEHAVIORAL ECOLOGY OF
THE FLORIDA SCRUB LIZARD (SCELOPORUS WOODI)


By

James Frederick Jackson

June, 1972


Chairman: Dr. A. F. Carr, Jr.
Major Department: Zoology



Sceloporus woodi does not occupy all available sand-pine scrub

habitat in Florida but is found only on the Lake Wales and northern

Bartow Ridges, in the Ocala National Forest region, along the central

and southern Atlantic Coast, and on the southwestern Gulf Coast.

The poor dispersal is related to habitat restriction. Canonical

analysis clustered S. woodi populations according to phenetic

affinity more effectively than did principal components analysis.

The clusters correspond to the geographic regions above except that

Lake Wales Ridge woodi are differentiated into northern and

southern entities. Differentiation of populations within the

geographic regions is inversely related to the opportunity for gene

flow. Phenetic and paleogeographic considerations suggest an

origin of S. woodi in the Pliocene on the Lake Wales Ridge from a

southwestern progenitor and dispersal northward in the early

Pleistocene; dispersal occurred in the late Pleistocene from the

Ocala National Forest region to the Atlantic Coast and from the

Lake Wales Ridge to the Gulf Coast.









Sceloporus woodi and S. u. undulatus have different foraging

tactics for exploitation of similar food resources. These

differences are necessitated by differences in physiognomy between

the sand-pine scrub association inhabited by S. woodi and the long-

leaf-pine/turkey-oak association inhabited by S. undulatus. On

three ecotones between these plant associations, S. woodi and

S. undulatus hybridize in a very narrow zone. Habitat selection

by the two species minimizes the width of the zone. No ethological

reproductive isolating mechanisms were detected, and their failure

to be evolved is ascribed to the minimal overlap of the species'

distributions.

Among twenty S. woodi populations, levels of fluctuating

asymmetry in six characters were not correlated. Hybrid zone

populations showed levels of asymmetry intermediate between those

for non-hybridizing populations of the two species. These results

do not support a relationship between genetic coadaptation and

asymmetry level.


















INTRODUCTION


Biologists have long recognized that certain organisms are

especially apt subjects for particular kinds of biological investiga-

tions. Drosophila, Neurospora, and Zea in genetics, sea urchins and

amphibians in embryology, and Darwin's finches and horses in evolution

are well-known examples. Although to suggest placing the lizard

Sceloporus woodi in that select company would border on sacrilege,

it does lend itself to studies of problems that are important in

ecology and evolution.

The scattered habitat of the lizard, together with the possibility

for dating the origin of particular areas of it, allows us to evaluate

the dispersal abilities of an extremely habitat-limited terrestrial

vertebrate. The physical isolation of many populations makes possible

the measurement of evolutionary divergence of such populations that

have occupied very similar environments for estimable time periods.

The influence of gene flow on differentiation of populations may be

gauged by comparing divergence among isolated populations with

divergence among populations genetically connected.

The presence of a congener, S. undulatus, in a habitat that

frequently borders that of S. woodi presents opportunities for the

study of the relation between habitat structure and foraging behavior,





2



of competition, and of the evolution of reproductive isolating

mechanisms. Finally, the existence of populations under presumably

different selection regimes allows an evaluation of the reality of the

population asymmetry property and of its correlation with genetic

coadaptation.

















CHAPTER I


DISTRIBUTION AND POPULATION PHENETICS OF SCELOPORUS WOODI



Introduction


The genus Sceloporus, which forms a conspicuous element of the

western North American vertebrate fauna, contains only one species not

found west of the Mississippi River: the Florida endemic, Sceloporus

woodi. Since its description by Stejneger (1918), little has been

published on the species. Barbour (1919), Jones (1927), and Lee and

Funderburg (1970) dealt with aspects of the geographic distribution;

Carr (1940) reported on the natural history; aspects of the physiological

ecology were investigated by Bogert and Cowles (1947) and Hutchison and

Larimer (1960), and reproduction was studied by Funderburg and Lee (1970).

Sceloporus woodi is, except in special circumstances, restricted

to a single plant association, the sand-pine scrub or Pinus clausa-

Quercus spp. Association (Laessle, 1942). The sand-pine scrub is a two-

layered fire subclimax in which Pinus clausa, often closely spaced, is

the sole overstory species. The understory is a dense thicket of largely

sclerophyllous shrubs dominated by Quercus chapmanii, Q. myrtifolia, Q.

virginiana maritima, and Serenoa repens. Herbaceous plants are rare and

the ground surface consists of bare sand, leaf litter, and lichens.









Scrub communities exist as scattered stands, large and small, surrounded

by other plant communities (see Davis, 1967, for mapping of the larger

scrubs); yet they are remarkable for floristic and physiognomic

uniformity from place to place. The disjunct nature of the habitat of

S. woodi offers interesting problems of geographic distribution and

variation. This study has sought to define the distribution of S. woodi

precisely by determining in which scrub localities it is present.

Phenetic variation within and among several populations is examined with

reference to affinities among populations and to the relation of these

affinities to certain environmental factors and to gene flow.



Methods


Approximately one hundred scrub communities were searched for

S. woodi; these are mapped in Figure 1 and their locations are listed

in Table 1. They probably constitute a substantial majority of those

extant in Florida and certainly include all the large ones. During

weather suitable for lizard activity, each locality was visited at least

twice if no lizards were seen, and some as many as four times.

Additionally, in view of the close relationship of S. woodi to S.

undulatus, the distribution of the latter species in the peninsula was

investigated.

Twenty populations were selected for phenetic analysis (Figure 1).

The sample for each population consisted of a minimum of twenty adults

(body length 40 mm). Nineteen characters were examined for each

individual (Table 2). Unless otherwise specified, the characters were

scored according to Smith (1939; 1946). Criteria for choosing the


























Figure 1. Distribution of sand-pine scrub and Sceloporus woodi
in Florida. Stippling indicates large continuous
areas of scrub. Enclosed numbers = scrubs containing
woodi; circles = woodi populations examined
phenetically; squares = other woodi populations.
Unenclosed numbers = scrubs apparently lacking woodi.









































































SCALE
0 20 40











TABLE 1. Locations of Scrubs



1. intersection of Forest Service Roads 77 and 88, Ocala National
Forest, Marion Co. Section 32, T. 11 S., R 25 E. elev 85

2. on State Road (SR) 314 3.5 mi. NE Eaton Creek, Ocala National
Forest, Marion Co. Section 7, R. 14 S., R. 25 E. elev. 80

3. near Central Tower, Ocala National Forest, Marion Co. Section 14,
T. 15 S., R. 25 E. elev. 150

4. intersection of Forest Service Roads 73 and 88, Ocala National
Forest, Marion Co. Section 36, T. 16 S., R. 25 E. elev. 100

5. Blue Lake, 4 mi. NW Winter Haven, Polk Co. Section 13, T. 28 S.,
R. 25 E. elev. 150

6. on SR 580 6.5 mi. E Haines City, Polk Co. Sections 19 and 20,
T. 27 S., R. 28 E. elev. 80

7. Flaming Arrow Boy Scout Camp, Heperides, Polk Co. Section 7,
T. 30 S., R. 29 E. elev. 115

8. on SR 64 8.7 mi. NE Avon Park, Polk Co. Section 35, T. 32 S.,
R. 29 E. elev. 95

9. Placid Lakes Subdivision, 3 mi. SW town of Lake Placid, Highlands
Co. Section 22, T. 37 S., R. 29 E. elev. 130

10. Red Hill, Archbold Biological Station, Highlands Co. Section 8,
T. 38 S., R. 30 E. elev. 210

11. on U. S. 27 6 mi. S Bair's Den, Highlands Co. Section 4, T. 39 S.,
R. 30 E. elev. 190

12. Naples, Collier Co. elev. 10

13. near Tommy Barfield School, Marco Island, Collier Co. elev. 55

14. on SR 528 2.5 mi. NW Cocoa, Brevard Co. Sections 18 and 19,
T. 24 S., R. 36 E. elev. 50

15. 0.3 mi. S Pineda, Brevard Co. Section 12, T. 26 S., R. 36 E.
elev. 30

16. 1 mi. N Sebastian, Indian River Co. Section 31, T. 30 S., R. 39 E.
elev. 35

17. 4 mi. N. Ft. Pierce, St. lucie Co. Section 17, T. 34 S., R. 40 E.
elev. 45









TABLE 1. Continued


18. 2 mi. S Jupiter, Palm Beach Co. Section 8, T. 41 S., R. 43 E.
elev. 15

19. several mi. W West Palm Beach, Palm Beach Co.

20. S. W. 4th Avenue, Boca Raton, Palm Beach Co. Section 30, T. 47 S.,
R. 43 E. elev. 30

21 Miami, Dade Co.

22. 1 mi. N. Hallandale, Broward Co.

23. Pompano Beach, Broward Co.

24. Lake Worth, Palm Beach Co.

25. Juno Beach, Palm Beach Co.

26. Johnathan Dickinson State Park, Martin Co.

27. Vero Beach, Indian River Co.

28. Ti-Co Airport, 12 mi. N. Cocoa, Brevard Co. elev. 35

29. north end of Estero Island, Lee Co.

30. 6.5 mi. NNW town of Lake Placid, Highlands Co.

31. on U. S. 27 1 mi. SW DeSoto City, Highlands Co. Section 16,
T. 35 S., R. 29 E. elev. 95

32. South Florida Jr. College, 1.6 mi. SW Avon Park, Highlands Co.
Section 34, T. 33 S., R. 28 E. elev. 140

33. on U. S. 27 7 mi. N Avon Park, Polk Co. Section 30, T. 32 S.,
T. 28 E. elev. 100

34. on U. S. 27 4 mi. S SR 60, Polk Co. Section 35, T. 30 S., R. 27 E.
elev. 140

35. on U. S. Alt-27 2.5 mi. S Frostproof, Polk Co.

36. at Audubon Center, Babson Park, Polk Co.

37. Haines City, Polk Co.

38. Davenport, Polk Co.

39. 22.8 mi. S Minneola, Polk Co.









TABLE 1. Continued



40. on SR 545 1.5 mi. N SR 532, Osceola Co. Section 35, T. 25 S.,
R. 27 E. elev. 100

41. Lake Thomas-Lake Sears, 3 mi. SW Winter Haven, Polk Co.

42. Eagle Lake, 5 mi. SW Winter Haven, Polk Co.

43. near intersection of SR 561 and SR 448, Lake Co.

44. Tavares, Lake Co.

45. on SR 44 2 mi. W Grand Island, Lake Co. Section 1, T. 19 S.,
R. 25 E. elev. 70

46. 5 mi. N Eustis, Lake Co.

47. 2 mi. S. Mayport, Duval Co. Section 38, T. 1 S., R. 29 E. elev. 10

48. on SR 210 1.5 mi. E U. S. 1, St. Johns Co. Section 1, T. 5 S.,
R. 28 E. elev. 60

49. on SR 100 3 mi. W Flagler Beach, Flagler Co. Section 39, T. 12 S.,
R. 31 E. elev. 35

50. on SR 40 0.7 mi. W Ormond Beach, Volusia Co. Section 21, T. 14 S.,
R. 32 E.

51. near Mainland High School, Daytona Beach, Volusia Co. Section 39,
T. 15 S., R. 32 E. elev. 55

52. Lake Como, Putnam Co.

53. 1 mi. S Crescent City, Putnam Co.

54. 0.7 mi. N DeLeon Springs, Volusia Co.

55. on SR 44 7 mi. W New Smyrna Beach, Volusia Co.

56. 2 mi. S Barberville, Volusia Co.

57. 1 mi. S Orange City, Volusia Co. Section 22, T. 18 S., R 30 E.
elev. 75

58. DeBary, Volusia Co. Section 3, T. 19 S., R. 30 E. elev. 80

59. Osteen, Volusia Co. Section 12, T. 19 S., R. 30 E. elev. 40

60. on SR 46 2.5 mi. NW Geneva, Seminole Co. Section 8, T. 20 S.,
R. 32 E. elev. 25









TABLE 1. Continued



61. Oviedo, Seminole Co. Section 14, T. 21 S., R. 31 E. elev. 75

62. Merritt Island-Cape Kennedy Air Force Base, Brevard Co.

63. on SR 46 10 mi. W Sanford, Lake Co.

64. on SR 44 2.5 mi. NE Cassia, Lake Co.

65. on Dist. 5-7878 0.6 mi. S SR 450 near Umatilla, Lake Co. Section 8,
T. 18 S., R. 27 E. elev. 155

66. intersection of SR 44A and Dist. 4-6585 about 7 mi. E Eustis,
Lake Co. Section 36, T. 18 S., R. 27 E. elev. 135

67 intersection SR 445 and Alexander Springs Creek, Lake Co.

68. on U. S. 441 2 mi. SE Zellwood, Orange Co. Section 35, T. 20 S.,
R. 27 E. elev. 140

69. Forest City, Orange Co. Section 18, T. 21 S., R. 29 E. elev. 120

70. Altamonte Springs, Seminole Co. Section 24, T. 20 S., R. 29 E.
elev. 95

71. 1.7 mi. E Ocoee, Orange Co. Section 17, T. 22 S., R. 28 E.
elev. 160

72. Orlovista, Orange Co. Section 25, T. 22 S., R. 28 E. elev. 130

73. intersection Skylane Drive and Turkey Lake Road about 4 mi. SW
Orlovista, Orange Co. Section 14, T. 23 S., R. 28 E. elev. 175

74. on Vineland-Apopka Road 0.2 mi. S Dr. Phillips, Orange Co.
Section 34, T. 23 S., R. 28 E. elev. 125

75. intersection I-4 and SR 535 about 9 mi. NW Kissimee, Orange Co.
Section 27, T. 24 S., R. 28 E. elev. 115

76. 0.5 mi. E Ashton, Osceola Co. Section 8, T. 26 S., R. 31 E.
elev. 75

77. Indian Lake Estates, Osceola Co. Section 3, T. 31 S., R. 30 E.
elev. 100

78. on U. S. 441 1.5 mi. N Yeehaw Junction, Osceola Co. Section 10,
T. 32 S., R. 34 E. elev. 75

79. on SR 68 3 mi. E U. S. 441, Okeechobee Co. Section 7, T. 35 S.,
R. 36 E. elev. 60









TABLE 1. Continued



80. on SR 70 10 mi. E Okeechobee at St. Lucie-Okeechobee Co. line
Section 31, T. 36 S., R. 37 E. elev. 55

81. Deer Lake, Goldhead State Park, Putnam Co. Section 36, T. 7 S.,
R. 23 E. elev. 200

82. 2 mi. N Grandin, Putnam Co. Section 32, T. 8 S., R. 24 E. elev. 150

83. Florahome, Putnam Co.

84. on SR 310 3 mi W U. S. 19, Putnam Co.

85. on SR 315 1.4 mi. N Orange Creek, Putnam Co.

86. Way Key, Cedar Keys, Levy Co. Section 30, T. 15 S., R. 13 E.
elev. 25

87. on SR 24 7 mi. E Cedar Keys, Levy Co. Section 2, T. 15 S., R. 13 E.
elev. 20

88. on 1-75 4 mi. N SR 484, Marion Co. Section 23, T. 16 S., R. 21 E.
elev. 95

89. on U. S. 441 6 mi. E Leesburg, Lake Co. Section 22, T. 19 S.,
R. 25 E. elev. 80

90 on SR 490 5 mi. E Lecanto, Citrus Co. Section 6, T. 19 S., R. 19 E.
elev. 140

91. Weeki Wachee, Hernando Co. Section 1, T. 23 S., R. 17 E. elev. 50

92. on SR 580 0.5 mi. E U. S. 19, Pinellas Co. Section 30, T. 28 S.,
R. 16 E. elev. 85

93. airport, Plant City, Hillsborough Co. Section 36, T. 28 S.-
R. 21 E. elev. 150

94. Surveyors Lake, 8 mi. SE Bartow, Polk Co. Section 27, T. 30 S.,
R. 26 E.

95. 2.5 mi. NE Balm, Hillsborough Co. Section 18, T. 31 S., R. 21 E.
elev. 120

96. near DeSoto National Monument, Manatee Co. Section 19, T. 34 S.,
R. 17 E. elev. 20

97. on SR 64 8 mi. W Manatee-Hardee Co. line, Manatee Co. Section 3,
T. 35 S., R. 21 E. elev. 100





12



TABLE 1. Continued



98. on SR 764 1.3 mi. E U. S. 17 near Cleveland, Charlotte Co.

99. Mossy Head, Walton Co. Section 23, T. 3 N., R. 21 W. elev. 250

100. 1 mi. NW Sunnyside, Bay Co. Section 4, T. 3 N., R. 17 W.
elev. 30












TABLE 2. Description of Characters Examined*




1. Dorsal scales
2. Scales around midbody
3. Femoral pores (sum of both sides)
4. Subdigital lamellae of fifth hind toe (sum of both sides)
5. Circumorbitals (sum of both sides)

6. Supraoculars in lateral rows (sum of both sides)
7. Supraoculars in right medial row
8. Parietals and frontoparietals
9. Frontals
10. Prefrontals and frontonasals

11. Right foreleg: length from bent wrist to elbow
12. Right thigh: length from midline to knee
13. Right shank: length from knee to ankle
14. Right hind foot: length from heel to base of second toe
15. Hind fourth toe: length from base excluding claw (sum of both sides)

16. Length of longest auricular lobule (sum of both sides)
17. Interparietal length
18. Interparietal width
19. Head: length from snout to posterior edge of interparietal


*1-10 counts; remainder linear measurements; 11-15 and 19 by dial
calipers to nearest 0.1 mm; 16-18 by ocular micrometer.









characters included (1) expression in a quantitative form to allow

analysis by multivariate techniques, (2) sampling a wide portion of

the phenotype, and (3) inclusion of both characters whose differences

from one population to another were likely to be explicable in terms

of ecology, and others of more obscure adaptive function. Since the

majority of the characters were sexually dimorphic and the sex ratio

in every sample was not one, separate analyses were done for males

and females.

The first ten characters, being counts, could be compared directly

between populations, but the metric characters needed prior treatment.

In organisms with continuous growth, consideration of metric characters

is hindered by sample differences in mean body size that arise from

collection at different times and the consequent differences in mean

age between the samples. Though ratios are sometimes employed to deal

with this problem (Metter, 1967), they suffer from the same difficulty

-- unequal representation of age classes -- since the ratio of a

structure to body length will differ for different body lengths unless

the bivariate trend line passes through the origin; further, the

distribution of ratios is frequently not normal. Consequently, the

difficulty is best overcome by adjustment to a common body size (Steel

and Torrie, 1960). Soule (1967a), because his clustering procedure

worked with means, adjusted sample means. In the present study, the

clustering procedures required input of individual values and adjust-

ment of these was accordingly necessary. Analyses of covariance were

carried out for each metric character regressed linearly on body length

for raw and log-log transformed data (Dixon, 1968; BMD 01R), and as a

second degree polynominal for raw data (Dixon, 1968; BMD 05R). Adjust-

ment of values for each character was made by the regression method









which minimized the deviation about the regression line; this method

was linear regression of raw data for all characters except 11, 12, and

13 of the males, for which regression on body length was most effective

when the polynomial was used. Comparison of regression coefficients

among the samples often revealed differences. Most of these were felt

to be illusory, however, and to be due to some samples comprising a

relatively narrow range of body sizes. Thus, following the recommenda-

tion of Sokal (1965), adjustment of all samples to the overall mean body

size was made by the within-groups regression.

Several methods of clustering samples according to phenetic affinity

are available. The usual numerical taxonomic method, clustering from

a matrix of interpopulation correlation coefficients (Sokal and Sneath,

1963), was not chosen because it ignores the variance-covariance

structure of the data; this is likely to be important when the samples

are very similar phenetically. Soule felt this approach did not

very successfully cluster his intraspecific samples.

Employment of multivariate statistics in phenetic studies is

becoming standard (Metter and Pauken, 1969; Ingram and Tanner, 1971;

Atchley, 1971). Certain multivariate techniques cluster samples by

providing optimal representation in three dimensions of samples

actually in n(>3)-dimensional space. The data may be considered from

a single universe and subjected to a principal components analysis

(Seal, 1964) which creates, from the original variable axes, new

orthogonal axes corresponding to the major trends of variation in the

data. The samples are then plotted on the most important of these

principal components (Schnell, 1970; Johnston and Selander, 1971).

Principal components for the data of this study were extracted using

the program BMD 03M (Dixon, 1968).









Seal (1964), however, recommends canonical analysis as the method

most appropriate for describing the affinities among several multi-

variate samples. Canonical analysis constructs new orthogonal axes,

termed canonical variates, on which the among-sample variance-

covariance matrix A is maximized with respect to the within-sample

variance-covariance matrix W. Each canonical variate (Yi) is the sum

of the products of the original variables (X.) and canonical coefficients

(V); Y. = V. X + V. X + V. X The sets of coefficients are
j 11 12 2 ip p
the eigenvectors associated with the solutions (eigenvalues) of the

equation [W-A XII = 0 (Cooley and Lohnes, 1962). Each eigenvalue

and its corresponding canonical variate represent an identifiable

fraction of the total variation. Canonical analysis of the data was

done with program BMD 07M (Dixon, 1968).

Sample means or individuals are plotted on those canonical variates

which account for the greater fractions of total variation. The

relative importance of each original variable to a particular canonical

variate is shown by standardization of the canonical coefficients

through multiplication by the standard deviations of the corresponding

original variables (Rees, 1969). The standardized coefficients of

each original variable are plotted on the canonical axes and summed to

a resultant vector which visually indicates the influence of the

original variable on that combination of canonical variates. A plot

on the three most important canonical variates accounts for only part

of the total variation; serious distortions present in the plot may be

revealed by linking each sample to its nearest neighbor in character

hyperspace. These linkages are generalized distances (D2). The gen-

eralized distance between two samples was obtained by multiplying each









coefficient from the discriminant function for the samples by the

difference between the sample means for the corresponding character,

and then summing the products.

Since the generalized distance is the most complete estimate of

phenetic, the presumably genetic, affinity, it is interesting to

compare average D2's for sets of populations. Four sets were defined

geographically: populations 1, 2, 3, and 4 (Ocala National Forest);

populations 5, 6, 7, and 8 (Bartow and Northern Lake Wales Ridges);

populations 9, 10, and 11 (Southern Lake Wales Ridge); and populations

14, 15, 16, 17, 18, 19, and 20 (Atlantic Coast). Within each set the

average distance between population mean vectors was calculated as the

arithmetic mean of all such distances.



Results


As shown in Figure 1, Sceloporus woodi has been found only in

scrubs of four regions: the Ocala National Forest (1-4) and certain

localities immediately to the south (43-46); the Lake Wales Ridge

(6-11 and 30-40) and, to the west, the northern portion of the Bartow

Ridge (5, 41, and 42); the southwest Gulf Coast in Collier and Lee

Counties (12, 13, and 29); and the Atlantic Coast from the vicinity of

Titusville south to Miami (14-28). It is conspicuously absent from

scrubs in and near Orange County northeast, east, and southeast of

Lake Apopka (68-75); from the Atlantic Coast north of Titusville; and

from scrubs north and west of the Oklawaha River (81-91). Sceloporus

undulatus occupies sandhill vegetation [Pinus australis-Quercus laevis

Association (Laessle, 1942)] and drier hammocks in northern and









central Florida except for seven islands of sandhill vegetation completely

surrounded by scrub in the Ocala National Forest.

Thirteen of the characters examined proved to be sexually dimorphic

(Table 3). Of the dimorphic scale characters, males tend to have more

femoral pores but fewer dorsal scales, prefrontals and frontonasals,

parietals and frontoparietals, scales around midbody, and lateral

supraoculars than females. This may be a reflection of the supposed

role of femoral pores in intraspecific communication and of the more

slender body form of the males. Males have longer forelegs, shanks,

thighs, hind feet, heads, auricular lobules, and interparietal scales.

The longer limbs of males are probably connected with their greater

tendency to territorial activity.

Character means and standard deviations for the twenty populations

are presented in Tables 4-7. Tables 8 and 9 show the correlations

between characters. It is clearly desirable that intercharacter

correlations be low so that informational redundancy is avoided, and

such is generally the case for the characters used; the highly

correlated limb length characters are an exception.

For the males the first three principal components accounted for

45, 10, and 6 percent, respectively, of the total variation, or 61

percent in combination. The first three principal components for the

females contained 63 percent of the total variation, or 45, 11, and

7 percent individually. When the population means were plotted on the

first three principal components, occurrence was found between phenetic

affinity and geographic proximity, but less distinctly than on the

canonical variates to be described below.














TABLE 3. Mean Values of Sexually Dimorphic Characters






Character* Male Mean Female Mean Probability


Dorsal scales 40.58 41.05 <0.1

Scales around midbody 41.89 42.01 <0.1

Femoral pores 34.74 33.76 <0.001

Lateral supraoculars 26.98 27.77 <0.1

Parietals-frontoparietals 4.78 5.02 <0.01

Prefrontals-frontonasals 5.91 6.22 <0.02

Right foreleg length 7.86 7.73 <0.01

Right thigh length 13.89 13.55 <0.001

Right hind foot length 7.09 7.04 <0.02

Auricular lobule length 1.67 1.61 <0.001

Interparietal length 2.78 2.67 <0.001

Head length 9.98 9.74 <0.001


*Adjusted to 47.65 mm snout-vent length













TABLE 4. Population Means for



Populations**
1 2 3 4 5 6 7
Character (10) (10) (10) (10) (10) (13) (10)

1. Dorsal scales 40.2 40.5 40.4 39.6 41.4 43.2 42.9

2. Scales around midbody 41.7 41.8 42.2 42.0 42.2 42.5 43.2

3. Femoral pores 31.6 31.8 32.1 32.3 33.8 33.3 34.2

4. Subdigital lamellae 26.7 26.8 27.1 27.9 26.6 27.3 27.4

5. Circumorbitals 22.4 23.8 21.8 22.2 21.0 23.2 21.9

6. Lateral supraoculars 31.3 28.9 26.3 29.1 25.7 28.8 29.6

7. Right medial supraoculars 5.5 5.6 5.5 5.5 5.5 5.3 5.4

8. Parietals-frontoparietals 4.7 4.9 4.8 4.8 4.5 4.5 5.2

9. Frontals 2.5 3.1 3.2 3.0 2.2 3.1 3.0

10. Prefrontals-frontonasals 7.0 6.2 5.8 5.8 5.3 5.5 5.8

11. Right foreleg 7.9 8.0 8.0 7.8 7.6 7.7 7.8

12. Right thigh 13.9 14.0 13.7 14.0 13.8 13.5 13.4

13. Right shank 12.5 12.6 12.5 12.7 12.3 12.5 12.5

14. Right hind foot 6.9 7.1 7.1 7.1 6.5 7.0 6.8

15. Hind fourth toe 22.8 23.4 23.2 22.9 21.1 22.1 21.8

16. Auricular lobule 1.9 1.9 1.9 2.0 2.0 2.0 2.0

17. Interparietal length 3.1 3.0 3.0 3.0 3.1 3.1 3.1

18. Interparietal width 3.0 2.9 2.9 3.0 3.2 3.0 3.0

19. Head length 9.9 9.9 9.7 10.0 9.8 9.8 10.0


*Means of values adjusted to 47.65 mm snout-vent length













Male Sceloporus woodi*


Populations**
8 9 10 11 12 13 14 15 16
(10) (9) (11) (10) (6) (10) (10) (9) (14)

40.3 41.9 42.2 42.0 41.8 38.7 39.1 39.4 41.4

39.9 44.0 42.4 44.0 40.8 39.7 41.5 41.8 41.8

34.5 37.9 37.9 38.1 34.3 36.4 36.2 35.3 36.5

29.5 27.4 28.2 28.2 27.7 28.6 28.1 29.2 28.7

20.2 20.7 21.7 21.6 23.7 21.5 21.6 21.7 22.1

24.0 27.1 27.2 25.5 26.5 23.2 26.3 25.2 26.4

5.8 5.3 5.5 5.6 5.8 5.2 5.1 5.2 5.4

4.1 4.8 5.1 5.1 5.2 4.3 4.5 4.4 4.9

3.1 2.4 2.2 2.2 2.7 3.0 2.3 1.9 2.4

5.3 5.8 7.2 6.0 7.0 6.2 5.1 5.3 5.5

8.1 7.8 7.8 7.9 8.0 8.0 7.8 7.8 7.9

13.9 13.8 14.1 14.0 13.8 14.1 13.9 13.8 14.0

13.1 12.6 12.8 12.7 13.3 12.9 12.4 12.3 12.3

7.5 7.2 7.0 7.0 7.4 7.3 7.2 7.2 7.1

25.4 24.0 23.7 23.4 24.4 24.3 23.3 22.9 23.3

2.0 1.7 1.8 1.9 1.9 1.8 1.8 2.0 1.8

3.1 3.0 3.1 3.0 2.9 2.9 3.2 3.4 3.0

2.9 2.8 3.0 3.2 2.8 2.9 3.0 2.9 3.0

10.0 10.1 10.2 10.1 10.1 9.9 10.0 10.3 9.8


17 18 19 20
(11) (12) (13) (9)

39.7 37.8 40.0 38.9

42.4 40.5 42.2 40.8

35.9 33.6 34.5 34.3

26.9 27.0 27.8 27.7

23.7 23.5 21.5 20.4

26.1 26.1 29.5 25.2

5.4 5.1 5.5 5.2

4.7 4.9 5.1 5.0

2.4 2.3 2.9 2.6

6.0 6.0 6.4 5.2

7.8 7.8 7.9 7.8

14.2 14.2 13.8 13.8

12.6 12.4 12.4 12.5

7.2 7.1 7.1 7.0

23.8 23.2 23.3 22.0

1.7 1.8 1.6 1.8

3.2 3.2 3.0 3.1

3.0 3.0 2.8 3.0

10.1 10.3 9.9 9.8


**Sample sizes in parentheses













TABLE 5. Population Means for



Populations**
1 2 3 4 5 6 7
Character (10) (10) (10) (10) (10) (7) (10)

1. Dorsal scales 40.9 41.1 41.2 40.6 42.1 44.4 43.2

2. Scales around midbody 42.4 42.1 43.8 41.4 41.4 43.0 42.6

3. Femoral pores 32.2 32.3 31.1 32.3 30.7 32.0 32.8

4. Subdigital lamellae 27.0 26.6 28.2 26.5 26.1 27.8 27.3

5. Circumorbitals 23.9 22.9 23.1 23.6 19.7 23.3 21.9

6. Lateral supraoculars 29.8 29.4 30.2 30.4 23.9 28.6 25.9

7. Right medial supraoculars 5.4 5.6 5.6 5.7 5.4 5.6 5.5

8. Parietals-frontoparietals 5.0 5.1 5.2 5.1 4.6 4.7 5.1

9. Frontals 2.7 3.0 3.5 3.3 2.1 3.1 2.9

10. Prefrontals-frontonasals 7.0 6.1 6.1 6.5 5.6 5.7 6.1

11. Right foreleg 7.8 8.2 8.2 8.2 7.8 7.8 8.1

12. Right thigh 14.0 14.2 14.1 14.3 14.1 13.7 13.9

13. Right shank 12.6 12.9 13.0 12.7 12.3 12.2 12.6

14. Right hind foot 7.1 7.3 7.2 7.3 6.7 6.9 7.0

15. Hind fourth toe 23.2 23.8 23.2 23.4 21.5 22.3 22.7

16. Auricular lobule 1.8 1.9 2.0 2.0 2.1 1.8 2.0

17. Interparietal length 3.0 2.8 3.0 2.9 3.1 3.0 3.0

18. Interparietal width 3.0 3.0 2.9 3.1 3.1 2.9 2.8

19. Head length 9.9 9.7 9.9 9.8 10.0 9.8 9.9


*Means of values adjusted to 50.52 mm snout-vent length














Female Sceloporus woodi*


Populations**
8 9 10 11 12 13 14 15
(10) (11) (11) (10) (14) (10) (10) (11)


16 17 18 19 20
(14) (9) (8) (13) (11)


41.6 42.3 42.4


41.8 42.2 38.2 39.6 40.6 41.3 39.2


39.9 40.5 38.3


44.2

36.8

27.3

21.2

27.5

5.8

5.4

2.5

6.1

8.2

14.2

13.0

7.2

23.9

1.9

3.1

3.1

10.3


40.6

35.3

28.6

23.2

29.6

5.6

5.8

2.7

7.8

8.2

14.2

13.3

7.5

24.6

1.9

3.0

3.1

10.4


39.0

33.6

28.2

17.7

21.1

5.1

4.5

2.5

5.9

8.4

14.5

13.3

7.5

24.5

1.8

3.0

2.9


41.7

35.8

29.2

23.1

26.6

5.2

4.9

2.5

5.5

8.1

14.2

12.9

7.3

23.8

1.9

3.2

3.0


42.0

33.5

29.3

22.2

26.3

5.1

4.0

2.0

5.5

8.0

14.5

12.7

7.2

23.3

1.9

3.3

3.2


40.2

34.7

30.2

21.6

28.4

5.7

4.6

2.6

6.1

8.0

14.0

12.9

7.5

24.8

1.9

3.2

2.9

10.4


40.8

34.6

28.2

22.0

26.0

5.3

4.7

2.7

6.1

8.1

14.5

12.7

7.3

24.0

2.0

2.9

3.1


41.8

33.6

28.6

25.2

29.8

5.1

5.8

2.8

5.5

8.1

14.5

12.6

7.3

23.6

1.7

3.1

3.0


42.8

33.5

27.5

23.2

33.0

5.7

5.3

3.2

6.7

8.3

14.5

12.8

7.1

23.4

1.6

2.9

3.0


44.5

35.8

27.4

22.0

27.9

5.5

4.7

2.6

6.3

8.1

14.2

13.0

7.4

25.0

1.9

3.0

2.9

10.3


**Sample sizes in parentheses


43.7

36.2

28.2

21.7

28.4

5.7

5.4

2.4

6.4

8.5

14.9

13.4

7.5

24.4

1.9

3.2

3.1

10.5


10.0 10.1 10.4 10.2 10.0


10.3 10.2 10.5













TABLE 6. Standard Deviations



Populations*
1 2 3 4 5 6 7
Character (10) (10) (10) (10) (10) (13) (10)

1. Dorsal scales 1.87 2.01 1.35 1.96 2.17 1.96 2.23

2. Scales around midbody 1.57 2.15 2.30 2.00 1.55 1.85 1.75

3. Femoral pores 1.50 3.26 2.18 2.06 2.53 2.90 2.20

4. Subdigital lamellae 2.06 1.32 2.28 1.97 0.84 1.49 2.27

5. Circumorbitals 3.60 2.74 4.02 2.39 3.86 3.36 3.04

6. Lateral supraoculars 6.25 4.79 5.60 7.96 5.85 5.32 5.34

7. Right medial supraoculars 0.53 0.52 0.53 0.53 0.53 0.48 0.52

8. Parietals-frontoparietals 1.06 0.99 0.92 0.79 1.27 0.78 1.03

9. Frontals 0.53 0.88 0.79 0.94 0.42 0.95 0.94

10. Prefrontals-frontonasals 1.05 0.63 1.14 1.23 0.48 0.88 1.03

11. Right foreleg 0.21 0.23 0.16 0.17 0.41 0.20 0.19

12. Right thigh 0.29 0.36 0.26 0.25 0.53 0.47 0.35

13. Right shank 0.41 0.36 0.28 0.33 0.35 0.32 0.22

14. Right hind foot 1.70 1.26 3.06 1.48 2.75 2.08 1.64

15. Hind fourth toe 0.81 0.62 0.85 0.57 0.95 0.63 0.83

16. Auricular lobule 0.20 0.14 0.22 0.16 0.14 0.19 0.24

17. Interparietal length 0.19 0.20 0.18 0.13 0.20 0.18 0.15

18. Interparietal width 0.15 0.24 0.13 0.15 0.22 0.19 0.18

19. Head length 0.17 0.33 0.25 0.25 0.25 0.23 0.14


*Sample sizes in parentheses














for Male Sceloporus woodi


8
(10)

1.77

1.45

2.99

1.72

1.62

4.27

0.42

0.99

0.99

0.48

0.16

0.45

0.46

1.97

0.88

0.21

0.12

0.16


9
(9)

1.90

1.32

3.14

1.24

3.20

4.65

0.50

0.83

0.53

0.67

0.17

0.24

0.26

1.90

1.06

0.14

0.15

0.13


10
(11)

2.04

1.86

3.19

2.09

2.37

4.00

0.52

1.14

0.60

1.17

0.14

0.32

0.34

2.25

0.82

0.13

0.24

0.26


11
(10)

1.82

1.70

4.23

1.69

4.09

5.48

0.52

0.88

0.42

0.94

0.12

0.32

0.32

1.37

0.63

0.19

0.20

0.21


Populations*


12
(6)

1.17

1.47

2.42

1.50

2.66

3.67

0.75

0.98

0.52

0.63

0.36

0.43

0.59

2.43

1.02

0.19

0.23

0.40


13
(10)

1.25

1.34

1.50

1.17

3.17

3.68

0.42

0.48

0.67

0.92

0.32

0.52

0.45

3.27

1.18

0.14

0.20

0.16


14
(10)

1.29

1.43

2.48

1.97

1.26

5.76

0.32

0.71

0.48

0.32

0.16

0.45

0.36

1.89

0.93

0.17

0.22

0.13


15
(9)

1.81

1.56

2.24

1.86

2.29

5.21

0.67

0.73

0.33

0.71

0.19

0.48

0.30

1.39

0.84

0.19

0.09

0.14


16
(14)

1.78

1.93

2.14

1.68

2.58

3.86

0.50

1.07

0.76

0.85

0.21

0.45

0.30

2.73

1.18

0.16

0.18

0.17


17
(11)

1.42

1.81

2.02

1.97

1.85

3.56

0.50

0.79

0.67

0.63

0.19

0.50

0.37

2.44

0.83

0.14

0.14

0.19


18
(12)

1.47

1.09

2.02

1.91

3.90

5.20

0.29

0.79

0.87

0.95

0.24

0.39

0.36

2.83

1.26

0.19

0.11

0.28


19
(13)

1.35

1.82

1.81

1.99

2.44

5.58

0.52

0.95

0.86

1.04

0.28

0.39

0.26

3.04

0.89

0.11

0.10

0.24


20
(9)

1.36

1.39

1.00

1.80

2.24

6.12

0.44

0.87

0.73

0.44

0.25

0.32

0.30

2.47

0.88

0.15

0.10

0.17


0.16 0.36 0.29 0.25 0.51 0.36 0.33 0.25


0.41 0.30 0.24 0.43 0.25














TABLE 7. Standard Deviations


Character

1. Dorsal scales

2. Scales around midbody

3. Femoral pores

4. Subdigital lamellae

5. Circumorbitals

6. Lateral supraoculars

7. Right medial supraoculars

8. Parietals-frontoparietals

9. Frontals

10. Prefrontals-frontonasals

11. Right foreleg

12. Right thigh

13. Right shank

14. Right hind foot

15. Hind fourth toe

16. Auricular lobule

17. Interparietal length

18. Interparietal width

19. Head length



*Sample sizes in parentheses


1
(10)

2.23

1.71

2.20

1.94

2.60

5.12

0.52

1.05

0.67

1.33

0.20

0.44

0.25

1.33

0.94

0.21

0.19

0.08

0.28


2
(10)

1.73

1.45

4.03

1.58

3.35

8.18

0.52

0.99

0.82

1.20

0.20

0.52

0.29

1.37

1.10

0.18

0.20

0.14

0.32


Populations*
3 4 5 6
(10) (10) (10) (7)

2.78 2.22 1.45 1.72

1.81 0.97 1.71 2.16

2.51 2.36 2.26 1.41

1.40 1.19 1.37 2.34

4.31 3.31 2.75 3.86

6.89 5.30 3.41 5.19

0.52 0.48 0.52 0.53

1.32 1.20 0.84 0.95

1.27 0.48 0.32 1.07

0.74 1.08 0.84 0.76

0.23 0.25 0.38 0.16

0.42 0.30 0.52 0.50

0.39 0.30 0.35 0.21

2.42 3.58 2.27 2.29

0.99 0.70 0.55 0.69

0.23 0.20 0.17 0.30

0.15 0.13 0.09 0.08

0.16 0.14 0.06 0.13

0.26 0.30 0.22 0.38


7
(10)

2.44

1.84

1.93

1.25

2.73

5.11

0.53

1.37

0.88

1.29

0.14

0.41

0.35

2.07

0.78

0.25

0.13

0.14

0.19














for Female Sceloporus woodi



Populations*


8
(10)

2.17

1.14

2.94

1.55

3.81

4.43

0.48

0.84

0.97

0.57

0.19

0.70

0.32

3.29

1.08

0.21

0.16

0.19

0.30


9
(11)

1.90

1.37

3.63

0.93

2.53

4.83

0.52

0.90

0.81

1.10

0.26

0.23

0.36

2.29

1.08

0.17

0.12

0.25

0.17


10
(11)

2.42

2.00

2.56

2.09

2.00

4.27

0.47

1.04

0.50

0.93

0.27

0.48

0.32

2.14

0.75

0.20

0.20

0.15

0.46


11
(10)

1.23

2.10

2.74

0.94

3.58

5.28

0.42

1.35

0.85

0.99

0.17

0.53

0.43

2.95

1.17

0.16

0.20

0.18

0.38


3 1 14 15 16 17 18 19 20


(14)

2.52

1.01

2.23

1.60

2.49

8.39

0.50

1.19

0.61

0.95

0.16

0.47

0.35

1.39

0.91

0.11

0.21

0.25

0.28


(10)

0.92

1.56

3.02

1.87

2.63

2.33

0.32

0.71

0.53

1.10

0.31

0.49

0.28

3.13

0.79

0.20

0.07

0.13

0.46


(10)

1.17

1.34

2.35

2.20

2.33

2.68

0.42

0.99

0.85

0.97

0.22

0.57

0.42

2.26

1.13

0.19

0.18

0.14

0.56


(11)

1.91

2.14

2.66

1.42

3.74

5.66

0.54

0.44

0.00

0.93

0.29

0.25

0.28

1.86

0.96

0.15

0.14

0.12

0.35


(14)

1.68

1.53

1.83

1.85

1.75

4.42

0.61

0.91

0.61

1.09

0.41

0.44

0.45

2.95

1.20

0.21

0.14

0.24

0.40


(9)

1.72

1.94

3.20

2.44

2.26

2.82

0.50

0.87

0.83

0.73

0.17

0.36

0.32

1.79

0.92

0.11

0.10

0.19

0.22


(8)

2.03

1.58

1.30

1.68

2.96

5.12

0.35

1.16

0.89

0.53

0.13

0.30

0.38

1.19

0.75

0.14

0.10

0.22

0.34


(13)

1.61

2.58

2.57

2.07

2.03

6.49

0.48

0.94

0.90

1.18

0.30

0.42

0.25

3.25

0.70

0.19

0.22

0.28

0.47


(11)

1.49

1.38

2.38

1.37

3.17

4.66

0.50

0.81

0.46

1.08

0.15

0.37

0.27

3.04

0.41

0.24

0.22

0.24

0.34














TABLE 8. Character Correlation


1 2 3 4 5 6 7 8 9

1 1.000

2 0.613 1.000

3 0.322 0.358 1.000

4 0.349 0.320 0.527 1.000

5 0.258 0.330 0.289 0.371 1.000

6 0.258 0.335 0.116 0.227 0.544 1.000

7 0.337 0.353 0.291 0.496 0.423 0.381 1.000

8 0.221 0.294 0.236 0.321 0.331 0.344 0.340 1.000

9 0.180 0.053 -0.012 0.191 0.266 0.203 0.224 0.181 1.000

10 0.233 0.305 0.139 0.263 0.347 0.304 0.384 0.321 0.079

11 0.319 0.373 0.467 0.708 0.409 0.309 0.596 0.353 0.277

12 0.233 0.382 0.505 0.692 0.406 0.290 0.575 0.335 0.176

13 0.360 0.382 0.481 0.708 0.419 0.318 0.607 0.307 0.253

14 0.250 0.315 0.482 0.725 0.369 0.231 0.551 0.272 0.220

15 0.210 0.296 0.530 0.712 0.343 0.176 0.546 0.244 0.194

16 0.262 0.203 0.149 0.331 0.202 0.098 0.306 0.071 0.158

17 -0.347 -0.276 -0.276 -0.410 -0.250 -0.184 -0.266 -0.306 -0.362

18 -0.222 -0.254 -0.262 -0.407 -0.245 -0.137 -0.268 -0.198 -0.177

19 -0.392 -0.441 -0.476 -0.721 -0.450 -0.345 -0.589 -0.328 -0.239














Matrix for Male Sceloporus woodi




10 11 12 13 14 15 16 17 18 19


1.000

0.894

0.933

0.881

0.827

0.479

-0.396

-0.420

-0.898


1.000

0.895 1.000

0.855 0.882 1.000

0.816 0.838 0.913 1.000

0.434 0.520 0.404 0.307 1.000

-0.341 -0.361 -0.357 -0.387 -0.047 1.000

-0.398 -0.425 -0.444 -0.451 -0.213 0.376 1.000

-0.892 -0.883 -0.842 -0.770 -0.494 0.426 0.426 1.000


1.000

0.362

0.389

0.390

0.352

0.373

0.066

-0.312

-0.168

-0.389














TABLE 9. Character Correlation


1 2 3 4 5 6 7 8 9

1 1.000

2 0.488 1.000

3 0.348 0.447 1.000

4 -0.240 -0.442 -0.376 1.000

5 0.309 0.430 0.374 -0.363 1.000

6 0.316 0.408 0.257 -0.288 0.575 1.000

7 0.322 0.380 0.397 -0.573 0.446 0.440 1.000

8 0.164 0.244 0.277 -0.275 0.367 0.454 0.341 1.000

9 0.199 0.244 0.081 -0.216 0.273 0.349 0.266 0.247 1.000

10 0.238 0.229 0.290 -0.334 0.407 0.422 0.373 0.376 0.110

11 0.174 0.408 0.524 -0.803 0.359 0.284 0.528 0.282 0.202

12 0.169 0.422 0.507 -0.806 0.390 0.272 0.512 0.277 0.161

13 0.224 0.436 0.562 -0.798 0.388 0.292 0.524 0.315 0.162

14 0.186 0.370 0.592 -0.725 0.357 0.291 0.509 0.303 0.163

15 0.202 0.368 0.612 -0.667 0.362 0.259 0.484 0.275 0.146

16 0.169 0.202 0.164 -0.450 0.070 0.065 0.145 -0.017 0.053

17 -0.262 -0.239 -0.207 0.440 -0.238 -0.211 -0.290 -0.288 -0.353

18 -0.147 -0.278 -0.170 0.432 -0.255 -0.098 -0.215 -0.158 -0.245

19 -0.287 -0.481 -0.473 0.913 -0.428 -0.321 -0.584 -0.321 -0.247













Matrix for Female Sceloporus woodi




10 11 12 13 14 15 16 17 18 19




















1.000

0.307 1.000

0.301 0.874 1.000

0.364 0.927 0.889 1.000

0.298 0.871 0.830 0.896 1.000

0.334 0.831 0.800 0.868 0.909 1.000

0.080 0.443 0.449 0.437 0.416 0.393 1.000

-0.285 -0.409 -0.310 -0.376 -0.406 -0.401 -0.099 1.000

-0.167 -0.411 -0.405 -0.423 -0.410 -0.394 -0.312 0.356 1.000

-0.363 -0.845 -0.852 -0.847 -0.797 -0.758 -0.450 0.499 0.501 1.000









Tables 10 and 11 give the coefficients of the first six canonical

variates and the proportion of the total variance associated with each

variate. The percentage of total variance accounted for by the first

three canonical variates is very close to that for the three largest

principal components. For each sex three-dimensional plots of

population means were made on the first three (Figure 2 and Figure 3)

and second three canonical variates. The trends on the first three

canonical variates are similar to those on the principal component

axes but the clusters are tighter.

Inspection of Figure 2 indicates four clusters surrounded by

several outliers. The tightest cluster is formed of populations 1,

2, 3, and 4 of the Ocala National Forest. Populations 14, 15, 16, 17,

18, 19, and 20 from the Atlantic Coast compose a looser cluster near

the Ocala group. Another cluster comprises populations 9, 10, and 11

from the southern end of the Lake Wales Ridge. Populations 6 and 7

form a cluster from the northern portion of the Lake Wales Ridge and

are closest to population 5 from the Bartow Ridge. Finally, popula-

tions 12 and 13 from the Gulf Coast and 8 from the Lake Wales Ridge

are well separated from other populations. Linkage of each population

to that other with the smallest generalized distance between them

reveals no serious distortion by the canonical plot of the relation-

ships in character hyperspace. Further, examination of Table 12

indicates general agreement between the plot and second, third, and

fourth nearest neighbors in hyperspace.

The clustering pattern of the females (Figure 3) is essentially

similar. The main difference is the merging of the Ocala National

Forest populations (1, 2, 3, and 4) with part of the Atlantic Coast














TABLE 10. Coefficients of the First



Canonical Variate
1 2
Character Raw Std. Raw Std.

1 1.56 0.27 -2.35 -0.41

2 1.49 0.25 1.02 0.17

3 -1.02 -0.26 0.98 0.24

4 0.31 0.06 0.96 0.17

5 -0.04 -0.01 0.79 0.23

6 0.13 0.07 0.01 0.01

7 0.10 0.05 -0.70 -0.35

8 0.12 0.11 0.02 0.02

9 -0.07 -0.05 -0.34 -0.25

10 -0.39 -0.34 -0.06 -0.05

11 -8.85 -0.20 6.28 0.14

12 7.01 0.28 13.00 0.52

13 -3.33 -0.12 -20.86 -0.72

14 -1.13 -0.26 0.88 0.20

15 -5.76 -0.52 -3.49 -0.32

16 1.81 0.31 -2.23 -0.38

17 0.12 0.02 1.52 0.26

18 1.65 0.33 -0.05 -0.01

19 -5.93 -0.18 8.36 0.25


% of canonical
variation













Six Canonical Variates for Males



Canonical Variate
3 4 5 6
Raw Std. Raw Std. Raw Std. Raw Std.

-2.78 -0.48 -0.04 -0.01 2.56 0.44 -2.10 -0.36

-0.40 -0.06 0.09 0.02 -2.80 -0.48 -1.82 -0.31

-2.27 -0.56 1.37 0.34 -1.02 -0.25 0.35 0.09

0.98 0.18 2.09 0.37 0.65 0.12 0.28 0.05

1.02 0.30 -0.36 -0.10 1.54 0.45 0.48 0.14

0.03 0.02 -0.47 -0.24 -0.03 -0.02 -0.93 -0.48

-0.15 -0.08 0.21 0.10 -0.34 -0.17 0.43 0.22

-0.10 -0.09 -0.04 -0.04 -0.02 -0.02 0.03 0.03

0.46 0.34 0.01 0.01 -0.36 -0.26 -0.28 -0.21

-0.27 -0.24 -0.76 -0.66 0.30 0.26 0.09 0.08

9.85 0.22 -15.21 -0.34 -16.28 -0.37 -4.42 -0.10

-3.58 -0.14 -3.80 -0.15 4.49 0.18 7.86 0.32

-6.18 -0.21 8.21 0.28 1.32 0.04 12.95 0.45

1.75 0.40 1.61 0.37 0.58 0.13 -1.36 -0.31

-2.80 -0.25 -1.94 -0.18 1.38 0.12 -2.53 -0.23

0.55 0.10 0.66 0.11 2.28 0.39 0.71 0.12

1.76 0.30 0.67 0.11 2.16 0.37 -2.37 -0.40

-1.40 -0.28 0.58 0.12 0.21 0.04 1.64 0.33

-12.36 -0.37 -5.76 -0.17 15.21 0.46 -7.82 -0.23


9% 8%













TABLE 11. Coefficients of the First


Canonical Variate
1 2
Character Raw Std. Raw Std.

1 -0.52 -0.10 1.43 0.28

2 -1.81 -0.31 2.38 0.41

3 1.50 0.39 -0.71 -0.18

4 1.06 0.18 -2.07 -0.35

5 -0.74 -0.22 -0.92 -0.27

6 -0.09 -0.05 -0.41 -0.22

7 0.20 0.10 1.17 0.58

8 -0.05 -0.05 0.13 0.13

9 0.21 0.16 0.27 0.20

10 0.45 0.45 0.30 0.30

11 -8.78 -0.22 14.39 0.36

12 -7.59 -0.34 -4.74 -0.21

13 13.73 0.46 14.84 0.50

14 1.55 0.38 -1.22 -0.30

15 4.22 0.38 2.15 0.20

16 -0.15 -0.03 1.51 0.29

17 1.26 0.21 -1.25 -0.20

18 -0.45 -0.08 -0.13 -0.02

19 7.36 0.26 -9.28 -0.33


% of canonical
variation 26% 18%














Six Canonical Variates for Females



Canonical Variate


Raw

2.57

1.79

0.24

0.45

-0.79

-0.20

0.25

0.10

-0.41

-0.22

-5.24

-6.88

2.70

-1.17

1.35

0.66

3.06

-0.85

10.25


Std.

0.50

0.31

0.06

0.08

-0.23

-0.11

0.12

0.10

-0.31

-0.22

-0.13

-0.31

0.09

-0.29

0.12

0.13

0.50

-0.16

0.36


Raw

2.78

-3.79

-0.96

1.04

0.81

-0.41

0.09

-0.16

0.24

0.15

-14.89

-8.83

4.82

0.91

-0.01

2.00

0.60

-0.08

-9.73


Std.

0.54

-0.65

-0.25

0.18

0.24

-0.22

0.04

-0.16

0.18

0.15

-0.37

-0.40

0.16

0.22

-0.00

0.39

0.10

-0.01

-0.34


Raw

0.14

2.53

-1.13

0.51

2.08

0.21

-0.22

-0.14

0.39

-0.05

1.99

-5.01

-10.28

1.68

3.27

-0.87

2.13

-0.16

-3.68


Std.

0.03

0.43

-0.29

0.09

0.61

0.11

-0.11

-0.14

0.30

-0.05

0.05

-0.22

-0.35

0.42

0.30

-0.17

0.35

-0.03

-0.13


10% 8%


6


6
Raw

1.87

-0.66

0.80

-1.66

0.68

-0.25

-0.20

-0.07

-0.14

0.26

0.05

12.51

-20.12

-0.29

4.19

1.13

-1.95

1.89

8.25


Std.

0.36

-0.11

0.21

-0.28

0.20

-0.13

-0.10

-0.07

-0.10

0.26

0.00

0.56

-0.68

-0.07

0.38

0.22

-0.32

0.35

0.29


































00
0 0

C
.0m
=0 -H


0 4-.
to


0







H
sa

CO






.4
C-i






0
cti
a
H

0 c



0 -
-mo
O o
00

U *il
4iu





38






















LG
1---t






/ 01
I = = r___
/ ^=^-----






























4 Ul




0
tC
C

co




Od.

0


0 a) ,



O)
4 41 (


P 8


4-4 U 0
9 .


o C


(U 0





0
C
4- C
C, *H




SoC
C 'C










C
C3




40














^^^^ \Y














TABLE 12. Generalized Distances (D2) Between


4 5 6 7


1

0.0

4.24

6.82

6.46

17.85

11.18

9.73

22.82

16.74

13.30

16.31

20.72

15.33

14.86

16.39

12.03

12.61

12.18

7.90

13.90


0.0

2.24

4.38

20.26

8.28

10.38

15.83

15.17

17.80

15.96

15.99

11.59

11.85

17.34

9.38

9.44

11.96

8.04

11.28


0.0

4.30

18.21

8.12

9.00

14.52

14.58

21.49

15.50

20.73

12.16

12.12

18.37

10.45

12.55

17.05

8.25

9.12


0.0

24.02

12.96

14.57

11.17

22.54

21.98

15.36

15.86

11.42

16.76

21.27

12.74

11.23


8 9 10


0.0

19.48

24.32

25.72

13.31

11.18

20.70

25.14

21.30

22.74

30.09

20.77

24.51


0.0

8.57

7.56

19.75

13.63

9.86

17.70

8.90

8.90

18.53

7.54

17.41


0.0

13.22

7.77

7.31

16.02

13.89

17.36

13.59

20.04

12.15

8.90

11.44

9.18

10.77

11.90

8.38

6.84


2 3


0.0

7.12

16.26

15.30

17.62

19.99

11.29

11.67

18.06

13.33

22.96


0.0

11.54

9.94

39.62

26.55

23.32

15.03

41.33

33.50

22.63

22.57

15.72

22.90

26.79

25.57

15.01


0.0

3.42

21.53

17.83

20.21

16.70

23.34

22.52

14.92

15.83

10.79

16.19

21.13

16.98

12.74














Sceloporus woodi Populations (Males)



11 12 13 14 15 16 17 18 19 20






















0.0

23.13 0.0

17.29 14.24 0.0

14.08 28.30 10.63 0.0

19.54 33.17 19.63 5.37 0.0

7.67 24.74 12.00 4.65 10.63 0.0

11.01 23.47 12.21 4.56 11.13 6.22 0.0

21.17 29.71 14.98 6.14 8.30 10.30 5.02 0.0

14.35 22.54 9.60 7.45 15.16 7.80 8.51 10.58 0.0

15.94 28.94 12.72 5.66 11.27 8.46 12.28 12.32 9.49 0.0














TABLE 13. Generalized Distances (D2) Between


2 3 4 5 6 7


1

0.0

6.08

10.29

6.28

15.64

9.07

9.12

18.22

11.53

18.21

13.91

16.95

18.34

9.83

11.88

7.20

7.14

12.63

9.19

8.08


0.0

6.41

2.74

20.53

12.97

9.30

23.73

12.37

15.46

12.65

18.18

16.85

16.02

22.01

7.37

10.49

20.31

8.90

10.68


0.0

19.10

15.17

11.07

24.06

16.90

17.38

15.99

17.71

19.42

16.04

20.18

5.64

7.15

18.71

10.96

12.01


0.0

11.75

10.09

38.20

29.55

28.28

20.41

39.77

34.86

25.52

18.30

16.73

23.37

26.02

22.61

16.28


0.0

8.03

22.06

13.08

7.24

25.33

15.98

14.82

13.79

22.56

24.72

16.68

21.59

16.29

13.30

23.87

12.97

13.14


0.0

20.80

14.03

15.16

10.41

21.05

23.05

15.62

18.82

12.35

14.20

23.43

15.12

13.79


8 9 10


0.0

15.26

20.18

22.09

10.62

14.52

10.56

16.71

17.29

15.46

24.90

26.84

23.06


0.0

9.00

5.75

15.38

16.92

14.20

18.89

14.34

12.34

23.71

13.89

19.68


0.0

5.71

12.64

17.14

15.16

17.51

15.89

12.39

23.43

13.66

17.54


0.0

5.88

25.58

20.99

26.94

20.93

31.97

32.78

19.27

17.75

14.61

19.68

19.51

16.25

17.72














Sceloporus woodi Populations (Females)


11 12 13 14


0.0

18.04

19.21

13.65

18.66

14.80

12.04

23.98

11.66

14.10


0.0

13.74

16.70

24.07

15.86

14.33

29.20

23.36

22.66


0.0

13.30

21.30

12.30

14.11

24.32

17.04

12.56


0.0

6.02

10.72

6.68

8.47

15.33

8.09


15 16 17 18 19 20


0.0

12.15

11.07

7.83

17.24

12.67


0.0

6.76 0.0

12.90 11.36 0.0

10.82 11.77 12.94 0.0

11.83 10.35 13.34 10.28 0.0









group (16, 17, 19, and 20) and the partial separation of these from the

remaining Atlantic Coast populations (14, 15, and 18). The Southern

Lake Wales Ridge populations (9, 10, and 11) still form a unit. Popu-

lation 6 is farther from its neighboring population 7 and closer to

population 5. Populations 8, 12, and 13 are again outliers. Certain

discrepancies between the canonical plot and nearest neighbor by D2

are obvious. Several populations form pairs, 5 and 6, 9 and 10, 16

and 17, on the canonical plot without being nearest neighbors in

hyperspace. Examination of distance rankings (Table 13) usually shows

these pairs to be second or third nearest neighbors, so it can be

stated that this canonical plot does not seriously distort the

configuration in hyperspace.

When the populations were plotted on a combination of the fourth,

fifth, and sixth canonical variates, formless scatters of points

were obtained that displayed only traces of the correspondence

between phenetic and geographic affinities of the previous canonical

plots. Some of the discrepancies between proximity on the first three

canonical axes and generalized distance were resolved, but many more

were generated. The canonical variates after the third apparently

represent the residue of phenetic variation not related to geographic

proximity, i.e., the independent evolutionary tack of each population.

Thus geographic proximity, whether because of gene flow or similar

ecological conditions, is associated with about half the phenetic

variation observed. A multivariate analysis of variance showed the

phenetic differences among geographic regions to be statistically

significant.









Figures 4 and 5 show the influence each character would have on

the first two canonical variates if it varied alone. Comparison of

this information with the projections of the population means on the

canonical axes (Figure 2 and Figure 3) exposes the major trends of

variation. Among the males, populations with low K1 and K2 values

have longer shanks and fourth toes and to a lesser degree longer

foreleg and hind feet. High values on K2 are associated with longer

thighs. Populations with large K but small K values are characterized

by a larger number of dorsal scales and longer auricular lobules. Low

values on K correlate particularly with larger numbers of femoral pores.

Character vectors among the females are similar in general

direction but not in magnitude; the limb measurements do not so

greatly dominate the canonical variates. Populations with large values

on K1 tend to possess longer shanks, hind feet, and fourth toes, while

large values on K2 are particularly associated with a larger number

of right medial supraoculars. Small K values tend to characterize
2
populations possessing more fifth toe lamellae and longer heads, and

small values on K3 are correlated with a high dorsal scale count.

The average regional generalized distances separating populations

are given in Table 14 for four regions. The same trend is evident in

both sexes. The Ocala National Forest populations are most similar,

but little more so than populations from the Southern Lake Wales

Ridge. The Northern Lake Wales-Bartow Ridge populations are most

distinct, and the Atlantic Coast populations show an intermediate

degree of separation from each other.

Color variation in life is not extensive in S. woodi, but some

trends are noticeable. The Marco Island population (13) differs most























Figure 4. For each character the vector shows the direction and
magnitude of its influence on the first two canonical
variates of the males. See Table 2 for
character code.




48







K2

12






17

S4/ K2





-1






15
7 16








13

























Figure 5. For each character the vector shows the direction
and magnitude of its influence on the first two
canonical variates of the females. See Table 2
for character code.





50







K2

7

13


2
11
10
16


1










3
12 6 1

14
19














0 0
M 0
4X C ) 4o
z a -c






z n


00t
f)

U "o

Url O
> Ll r
Url

4-) In1
cC -cc
< co











C)
a,S
UL)

*.H i-
F0 Cd
,- nU '-1
c/i
k o


v,C
OUU)T-


rz





0
o -
O


cc- c-a
N '0

Tt \D









strongly by being notably pallid and by a tendency for the males to

retain the dorsal markings usually limited to juveniles and females.

Since the Marco Island scrub lacks the normally dense Pinus clausa

overstory and has instead only widely spaced Caribbean pine (Pinus

caribaea), the light coloration may be an adaptation to higher ambient

illumination and thus an example of the phenomena described by

Gloger's rule (H. W. Campbell, pers. com.). Generally, specimens from

the Southern Lake Wales Ridge lack the reddish cast seen in woodi to

the north and east and are grayer. The dorso-lateral dark stripe of

males is least well-defined on the Lake Wales Ridge, more prominent

in the Ocala National Forest, and very conspicuously dark and

straight-edged on the Atlantic Coast.



Discussion


Phenetic Affinities


For mean character states of the characters considered, a

pattern of geographic variation exists among S. woodi populations.

Northern and Southern Lake Wales Ridge populations form distinct

assemblages with affinities between the northern populations and the

Bartow Ridge woodi. Populations from the Ocala National Forest

represent another phenetic unit with close affinities to the Atlantic

Coast populations and more distant ones to woodi from the Northern

Lake Wales Ridge. The two Gulf Coast populations show similarities

but are not particularly close.









Adaptive Interpretation of Variation


The adaptive meaning of variation in most of the characters is

cryptic and probably comprehensible only through extensive ecological

and genetic knowledge of each population. However, two of the stronger

variational patterns are interpretable, at least qualitatively.

Elongation of the limbs is generally considered a cursorial adapta-

tion in terrestrial lizards (Kramer, 1951; Snyder, 1962). Examina-

tion of a character correlation matrix based on woodi population means

shows that mean lengths of the shank, hindfoot, fourth toe, and,to a

lesser degree, thigh and foreleg tend to vary as a unit; this suggests

that running ability of these ground-dwelling lizards is more important

in some populations. Generally, a correspondence exists between a

tendency to longer legs and my own subjective impression of the

openness of the habitat. Whether this characteristic of a locality

is consistent over long periods is unknown. The understory at

localities 5, 6, 7, and 20 is dense and continuous and oak leaves

carpet the ground, whereas localities 8, 9, 11, 12, and 13 have oaks

and other shrubs scattered as clumps among considerable expanses of

bare sand. Other localities are intermediate. Since the typical

escape behavior of S. woodi is rapid flight into a tangle of shrubs,

cursorial ability would reasonably be at a premium where cover is

minimal. Additionally, the Florida Scrub Jay (Aphelocoma c.

coerulescens), probably a major predator of S. woodi, occurs most

abundantly in open scrub (Westcott, 1970).

Although many lizards exhibit geographic variation in dorsal

scale number, the adaptive nature of the variation is not understood.

Hellmich (1951) and Soule (1966) associated increase in dorsal scale









number and the concomitant decrease in scale size in Liolaemus and Uta,

respectively, with climates subjectively assessed as progressively

cooler, but actual climatological data were not available for the

regions concerned. Existence of such data for Florida allows statis-

tical evaluation of any association between dorsal scale number and

environmental temperature in S. woodi.

The January, August, and annual temperature regimes for each woodi

population were estimated by averaging the corresponding mean tempera-

tures over the ten-year period 1961-1970 at the nearest weather station

(Climatological Data, ESSA). Table 15 gives these overall means and

indicates the assignment of weather stations for the lizard popula-

tions. Since the Ocala National Forest was represented by a single

weather station, an overall mean dorsal scale number was used for

populations 1, 2, 3, and 4; for the same reason, populations 9, 10,

and 11 were lumped. In each sex, significant negative product-

moment correlations were obtained between mean dorsal scale number and

both mean annual temperature (males: r = -.467, p = .038; females:

r = -.549, p = .012) and mean January temperature (males: r = -.435,

p = .052; females: r = -.543, p = .017). The correlations with mean

August temperature were negative but not significant (males:

r = -.264, p = .174; females: r = -.226, p = .212). Hellmich (1951)

and Soule (1966) hypothesized that variation in dorsal scales is

based on thermoregulatory adaptation and specifically that, other

things being equal, a lizard with smaller scales retains heat more

effectively. In the absence of heat-flux data, that interpretation

seems reasonable for the present case. Variation in mean August

temperature in peninsular Florida is very small and accounts for only
















TABLE 15. Assignment of Temperature Data to S. woodi Populations


Weather Station

Ocala

Winter Haven

Lake Alfred

Mountain Lake

Avon Park

Lake Placid

Naples

Everglades

Titusville

Melbourne

Vero Beach

Ft. Pierce

Stuart

West Palm Beach

Pompano Beach


Temperature (F)
January i August i

57.6 82.0

60.4 81.5

59.0 81.6

60.3 81.5

62.0 82.9

61.2 81.9

64.4 82.1

64.3 82.2

59.6 81.7

60.7 80.9

61.7 81.2

62.1 81.1

63.9 82.1

65.1 82.3

67.1 82.7


Population

1,2,3,4

5

6

7

8

9,10,11

12

13

14

15

16

17

18

19

20


Annual 7

71.1

72.0

71.1

71.7

73.3

72.4

73.6

74.1

71.5

71.8

71.9

72.2

73.4

74.2

75.2









a small percentage of the variation in mean dorsal scale number;

however, mean January temperatures are associated with approximately

25 percent of mean dorsal scale variation and range from those

forcing S. woodi into frequent quiescence in north and central Florida

to those of coastal south Florida, which are high enough for sustained

activity. Thus more numerous, smaller dorsal scales may function in

extending the temporal range of woodi in localities marginally suited

to winter activity. Obviously, the majority of dorsal scale variation

remains unaccounted for.



Dispersal History


The location of scrubs has been correlated by Laessle (1958) with

such ancient marine features as dunes, beaches, bars, and submerged

hilltops. Scrub grows on siliceous sands that were very well sorted

by wind or marine currents when sea levels were higher than at present.

Under the condition that these fossil marine features can be dated, this

correlation allows educated speculation regarding the history and

dispersal of Sceloporus in Florida. Unfortunately geologists do not

yet agree on the number or ages of the higher sea levels. Basing

their interpretations largely on topographic data, Cooke (1945) defined

seven marine terraces and MacNeil (1950) four (Table 16), but both

considered all the terraces formed during Pleistocene interglacials.

Employing physiographic and stratigraphic data, Alt and Brooks (1965)

recognized five ancient shorelines; they believed the oldest to be

Miocene and suggested that Pleistocene sea levels were never higher

than 70-80 feet above present. More recently both Alt (1967) and
















TABLE 16. Estimated Ages


Authority


Cooke (1945)


Aftonian

Yarmouthian

Yarmouthian



Sangamon

Sangamon

Sangamon

Mid-Wisconsin


MacNeil (1950)


Yarmouthian

Sangamon





Mid-Wisconsin



Post-Wisconsin


Terrace
Elevation
(in feet)


Sea


270

200-250

170

140-150

90-100

70-80

45

25-30

16-18

5-10


Okefenokee

Wicomico





Pamlico

Princess Ann

Silver Bluff
















of Florida Marine Terraces


flIAL*&tflL ~-


Alt and
Brooks (1965)


Late Miocene






Pliocene

Late Pliocene

Early Pleistocene

Pleistocene


Alt (1967)


Late Miocene






Pliocene



Aftonian

Yarmouthian


Brooks (1968)


Late Miocene



Late Pliocene

Aftonian

Yarmouthian

Yarmouthian

Sangamon

Mid-Wisconsin


Sangamon


Brooks (pers.
comm., 1971)




Late Miocene



Late Pliocene

Aftonian

Yarmouthian

Yarmouthian

Yarmouthian

Sangamon

Sangamon









Brooks (1968; pers. comm., 1971) have separately modified their earlier

scheme; the important difference between the modifications is the age

of the 90-100 terrace. Alt dates it as Pliocene, while Brooks believes

it marks the first interglacial period, the Aftonian. Of course, later

interglacials would be correlated with higher sea levels by Brooks than

by Alt. Because the interpretations of Alt and Brooks are grounded on

more extensive evidence than those of Cooke and MacNeil, they will be

followed below, specifically in the form of Brooks (pers. comm., 1971).

There is no doubt on morphological grounds that S. woodi is a

member of the undulatus group of Smith (1938) and, as will be shown

later, woodi hybridizes sparingly with undulatus where their distribu-

tions meet now. However, a distributional peculiarity suggests

that woodi might not have been derived directly from S. undulatus

undulatus in Florida; this is the absence of undulatus in the seven

islands of sandhill vegetation that are surrounded by scrub in the

Ocala National Forest. The north-south oriented ridge that runs

through the northern portion of the Ocala National Forest is mostly

covered by scrub, but Hughes, Salt Springs, Kerr, and Riverside

Islands are extensive stands of sandhill vegetation largely above 90

feet and with some elevations near 150 feet. The three eastern sand-

hill islands, Pats, Syracuse, and Norwalk, are between 90 and 50 feet

in elevation. Laessle (1958) suggested the scrub bordering the western

edges of Salt Springs, Kerr, and Riverside Islands grows on sand

sorted by marine currents when sea levels were 150 and 90 feet above

present and that these sandhill islands and their lower neighbors to

the east were completely surrounded by scrub after extensive dunes

formed on the shore of the sea 25 feet above present (near the present









west shore of Lake George) and migrated westward. Brooks (pers. comm.,

1971), however, believes these paleodunes, presently vegetated by

scrub, developed as aeolian deposits under more xeric soil conditions

when the water table was lower because of lowered glacial sea levels;

he considers them Kansan, Illinoian, and Wisconsin in age.

Lack of undulatus in the eastern sandhill islands is not conclusive

since Laessle (1958) believes dunes could have covered them, killing

the vegetation, and then moved on, so that sandhill vegetation may not

have been continuously present in these islands. However, the western

four islands were presumably never overblown and probably have been

vegetated by open forests since the earliest Pliocene, when the sea

which covered the area during the Miocene retreated. The western

sandhill islands were not definitely connected with the mainland until

the Nebraskan glaciation, but they would have joined it during proposed

low sea levels during the Pliocene (Webb and Tessman, 1967). Certainly

during all the glacial periods and during the late Yarmouthian inter-

glacial these islands would have joined the emergent peninsula and

would doubtless have been colonized by undulatus if it were present in

Florida then. During the Yarmouthian, following Laessle, or perhaps

earlier, following Brooks, the sandhill islands were surrounded by

scrub-covered dunes, which have prevented entrance by undulatus both

by offering unsuitable habitat and by harboring woodi from which

undulatus is not reproductively isolated. The only fossil undulatus

from Florida is of uncertain age but is no older than Sangamon

(Brattstrom, 1953; Auffenberg, 1956).

In coloration, proportions, and in ground-dwelling tendencies,

woodi is nearer the southwestern members of the undulatus group than









to S. u. undulatus. Consequently, and in view of the possible late

arrival of S. u. undulatus in Florida, woodi may have been derived

from a form close to S. virgatus or S. u. consobrinus that invaded

Florida from the southwestern United States or northern Mexico.

Considerable biogeographic and paleontological evidence indicates

biotic exchange between this region and Florida from Oligocene through

Pleistocene (Neill, 1957; Auffenberg and Milstead, 1965). Pitelka

(1951) suggested that the Florida Scrub Jay (Aphelocoma c. coerulescens),

the other vertebrate most closely associated with scrub, reached Florida

from the southwest in the Pliocene during a maximum expansion of

sclerophyll woodland. The progenitor of woodi probably entered the

peninsula at the same time.

Laessle (1958) linked no scrub to the 200-250-foot terrace, but

that on the slopes of Red Hill at the Archbold Biological Station may

be on sands sorted during the Miocene (H. K. Brooks, pers. comm., 1971).

Recession of the Late Pliocene 140-160-foot sea left considerable

expanses of well-sorted sand along the Southern Lake Wales Ridge and

smaller deposits in the northern portion of the Ridge and in the

present Orange County area (scrubs 71-74). Development of the scrub

as a plant community probably took place on the Lake Wales Ridge in

the Pliocene; endemism among scrub plants is highest there today

(John D. Beckner, pers. comm., 1971). Presumably woodi arose there

at the same time in response to the ecological opportunity offered by

the scrub to a ground-dwelling sit-and-wait feeder.

The enigmatic present distribution of woodi, with central and

peripheral habitats colonized but suitable areas in between unoccupied,

could conceivably have arisen in two ways. The distribution of scrub









could have formerly been much more extensive allowing expansion of

the lizard's range followed by extinction in many relict scrubs. Though

it is known that more xeric climates have existed in Florida, there is

no evidence for assuming that most scrubs were parts of a single

continuous unit. Further, the hypothetical extinctions would have

occurred non-randomly. An alternative and much more probable

hypothesis for generating the present distribution requires dispersal

by woodi to some scattered scrubs through other vegetation types. This

dispersal has been impeded by the lizard's habitat requirement of

ground largely free of herbaceous vegetation. S. woodi is never found

in hammocks or low flatwoods and in this study has been seen in only

a few locations with sandhill or scrubby flatwoods vegetation

(Laessle, 1942). These were exceptional in having very sparse

wiregrass (Aristida; Sporobolus) growth and were closely adjacent

to scrub. Besides scrub, the only extant vegetation in which woodi

could have maintained populations even temporarily would have been

sandhills or scrubby flatwoods so sterile, dry, or subject to frequent

ground fires that little wiregrass existed.

Due to the short distances separating scrubs on the Lake Wales

Ridge, woodi spread to most of them, including some that developed

on sands of the 90-foot Wicomico sea. Slight morphological differences

developed between Southern Lake Wales Ridge populations and those

toward the northern end of the Ridge, while more isolated populations

diverged uniquely. By an unknown route, woodi reached several small

scrubs (43, 44; Figure 1) between Lake Dora and Lakes Harris and

Eustis and probably from there colonized the extensive scrub of the

present Ocala National Forest. This large population became slightly

differentiated from those of the Lake Wales Ridge.









The lack of woodi in a scrub is not proof the lizard has not been

there. Extinction of small populations is a possibility because

stands of Pinus clausa, which has largely serotinous cones (Laessle,

1968), must burn by crown fires each several decades to be maintained.

Nevertheless, in view of the existence of numerous scrubs without

woodi and of its presence in some very small scrubs, poor dispersal,

stemming from habitat restriction, must in many cases be responsible

for absence of the species. The absence of woodi in certain scrubs

is particularly astonishing because of their great age, large size,

or proximity to woodi populations. Scrubs 71, 72, 73, and 90 were

submerged hilltops in the Okeefenokee Sea and so have been exposed

for over 2 million years. Scrubs 64, 88, and 91 seemingly afford

substantial targets, since each is several square miles in area.

Though less than 5 miles from two woodi populations (44 and 45),

scrub 89 is separated from them by cypress sloughs. It lacks the

lizard. Only a narrow strip of sandhills and Alexander Springs

Creek with its half-mile-wide hammock prevent the colonization of

scrub 67. And scrub 65 is less than two miles, through sandhills,

from the route (now followed by SR-19) by which woodi probably reached

the Ocala National Forest scrub.

Derivation of the Atlantic Coast populations from Ocala National

Forest woodi is suggested by their phenetic similarities. Given the

lizard's apparent ineptitude at overland dispersal, and the want of

any intervening relict populations, dispersal may have been by rafting.

When the Pamlico Sea stood at full height, the eastern edge of the

Ocala National Forest scrub was very near sea level and across a lagoon

that occupied the present St. Johns River valley a chain of barrier









islands consisting mainly of dunes, undoubtedly covered with scrub,

began some 50 miles to the southeast and extended south for about

125 miles (MacNeil, 1950). If individuals from the Ocala National

Forest population had been set adrift, south-flowing currents

(Laessle, 1958) would have carried them through a lagoon bordered

on the east by over 100 miles of suitable habitat. During those

interglacials when the sea did not cover the Atlantic scrubs, the

Ocala National Forest scrub was the only woodi-inhabitated scrub

at seaside. Once a beachhead was made, the present range on the

Atlantic Coast would have been readily achieved because of the almost

continuous distribution of scrub there. Dispersal north of Titusville

has apparently been prevented by low flatwoods.

Spread of woodi to the Gulf Coast is less understandable. It

must have been post-Silver Bluff, for the colonized scrubs are merely

10 feet above sea level. The present flatwoods between these scrubs

and the scrub at the south tip of the Lake Wales Ridge frequently

approach the physiognomy of scrubby flatwoods; survival of woodi

in these may have been possible during more xeric times. Inasmuch

as Marco Island is a Wisconsin dune (H. K. Brooks, pers. comm., 1971),

the woodi there are very recently derived from those near Naples.

The reason for their rapid divergence is not apparent, but may

relate to the Marco scrub being floristically atypical, lacking Pinus

clausa and containing a number of Antillean plant species.









Comparison with Other Species


Patterns of distribution and differentiation of several other

animals confined to the Florida xeric plant associations have been

described and may be compared with those of the present study. First,

if presence of woodi on the coasts is regarded as fortuitous, dispersal

overland to available xeric habitats has been poor by woodi and

Neoseps reynoldsi (Telford, 1959; 1962) but good by the scarabaeid

genus Mycotrupes (Hubbell, 1954), the lycosid species-pair Geolycosa

patellonigra-G. xera (McCrone, 1963), Stilosoma extenuatum (Highton,

1956), Eumeces egregius (Mount, 1965), Tantilla relicta (Telford, 1966),

and Rhineura floridana (Zug, 1968) since these forms occur between

the central highlands and the coasts. A geographic range identical

to that of woodi in the central highlands implies that Neoseps finds

poorly drained plant communities as great a barrier as does woodi, but

why other fossorial reptiles do not is unknown. Though the direction

of causality is unclear, effective dispersal may be related to a

second observation: that only woodi and Neoseps show no evidence of

fragmentation into Lake Wales Ridge and north Florida populations

during the higher (>100 feet) stands of sea level. Possibly they never

reached north Florida because of poor dispersal ability or perhaps

some of the other species mentioned invaded the regions centered on

Orange and Hernando Counties more readily from north Florida than from

the Lake Wales Ridge.

Quantitative techniques for comparing the degree of intraspecific

differentiation between species scrutinized for different characters

are yet to be developed, and qualitative assessments are very subjective.

Nevertheless, though most investigators of variation in the xeric








Florida species cited above have not been reluctant to apply formal

nomenclature to infraspecific taxa, it appears that such taxa are fairly

well-defined in their organisms. This is not the case for S. woodi. A

powerful multivariate technique and adequate sample sizes have permitted

detection of differentiation, but differences among the means (Tables 4

and 5) are generally minuscule and most populations show nearly the whole

range of variation in each character. Even on the canonical axes,

individuals are not clustered closely about their population mean but

overlap greatly with individuals of other populations. The weak develop-

ment of morphological differences, even under conditions of isolation,

is probably a consequence of both the uniformity of scrub and the

restriction of woodi to it. The extensive radiation of insular Uta

populations, isolated in the Gulf of California for periods equivalent

to those of woodi populations, occurred on islands differing considerably

in ecological conditions (Soule, 1966). Secondly, isolated populations

of species with broader habitat tolerance would be more likely to

diverge by making dissimilar adaptive compromises with geographic regions

that differ in proportional representation of habitats. Differentia-

tion in Rhineura may have resulted from such an effect (Zug, 1968). The

only other Florida reptile examined and found to lack strong geographic

variation is Rhadinaea flavilata (Myers, 1967), which is limited to a

single habitat, pine flatwoods.



Gene Flow and Differentiation


Ehrlich and Raven (1969) began what surely will be a fruitful

polemic by suggesting that gene flow, unlike stabilizing selection,

is usually unimportant in preventing divergence among populations.









One line of their evidence was the small degree of divergence in many

organisms between colonies that have lacked gene flow for great periods.

However, these observations are without controls that would allow

divergence among populations connected by gene flow to be compared

with that among isolated populations. Further, degree of divergence as

assessed subjectively by systematists with varying taxonomic philosophies

seems unnecessarily qualitative given the availability of multivariate

methods which measure distances in character space. Metter and Pauken

(1969) showed that differentiation, expressed as D2, generally corresponds

with paleoecological changes that have restricted gene flow in Ascaphus

true.

Although it is clear from the small absolute differences between

isolated woodi populations that strong stabilizing selection must be

operative, the degree of divergence among the populations of several

regions (Table 14) is noteworthy since it can be correlated with the

likelihood of gene flow. Populations 1-4 are parts of one continuous

population extending throughout the scrub of the Ocala National Forest;

no physical or ecological barriers separate these populations, which

average 10 miles apart. Populations 9-11 were similarly connected

prior to the interposition of citrus groves in recent decades. Their

average physical separation is 7 miles. Gene flow among populations

in both these sets is restricted only by distance. Kerster (1964)

estimated neighborhood area in Sceloporus olivaceus to be approximately

425 m in diameter. Populations 14-20 are from a linear series of

scrubs partly separated by ecological and physical barriers presently,

but probably more nearly continuous during the Wisconsin glaciation.

They average 27 miles apart. Populations 5-8 are from scrubs completely

isolated by sandhills and flatwoods and about 15 miles apart.









Inferentially, the four regions can be ranked by amount of gene flow

between the populations examined: Ocala National Forest and Southern

Lake Wales Ridge, high; Atlantic Coast, moderate; Northern Lake

Wales Ridge, extremely low. The ordination is the inverse of that

for differentiation of the populations and implies that gene flow

retards differentiation.

Two objections to that implication can be raised. First, although

the regions seem equal in the degree of similarity of their scrubs,

there are no data to prove the range of selection regimes does not

differ between the regions. Second, the regions vary in the time that

has been available for differentiation. Populations 5-8 are probably

Kansan in age (1.2 million years) though 5 could be Nebraskan (1.5

million years). Populations 1-4 and 14-20 probably date from the

Yarmouthian (0.5 million years) with the latter set somewhat the

younger. Thus, populations 5-8, having had more than twice as long

to diverge, might be expected to be less similar. On the other hand,

populations 14-20, in spite of being younger than populations 1-4, are

more widely separated phenetically. And populations 9-11, being

Pliocene in age, are older than populations 5-8; nevertheless, they

are more alike. Finally, it is the populations most isolated from

genetic exchange with other populations (5, 12, and 13) that are the

most phenetically distinct; however, here the founder effect cannot

be discounted as a cause. Obviously data from a single species do

not substantiate the conservative effect of gene flow; numerous

other studies are needed and material for them might be most readily

found among reptiles of montane western North America.


















CHAPTER II


THE PHENETICS AND ECOLOGY OF AN EXTRAORDINARILY NARROW HYBRID ZONE



Introduction


Hybrid zones between races and semi-species hold considerable

interest for evolutionary biology. They may be examined from the point

of view of the selection pressures that adapt each population to its

geographic range (Dice and Blossom, 1937; Blair, 1943). More importantly

though, the width and age of the hybrid zone and the mechanisms which

restrict gene flow across it are of particular relevance to under-

standing the origin and evolution of ethological isolating mechanisms.

In this chapter three very narrow hybrid zones between the

iguanid lizards Sceloporus woodi and S. u. undulatus are considered.

Sceloporus woodi is normally restricted to the sand-pine scrub

association of the Florida peninsula. S. u. undulatus inhabits dry,

open forests in the southeastern United States; in Florida it is most

abundant in the longleaf-pine/turkey-oak association (hereafter termed

sandhill vegetation). Laessle (1942) defined these associations, and

the striking differences in physiognomy between them have been detailed

(Laessle, 1958; 1968). The overstory of sand-pine scrub consists of









only Pinus clausa, normally in even-aged populations. The understory

is a dense thicket of sclerophyllous shrubs and is dominated by oaks

which range in height from one-half to two meters. The ground surface

is a mosaic of bare sand and thin, but well-compacted, leaf litter over

which herbaceous growth is extremely sparse. Pinus australis was the

dominant tree in primeval sandhill vegetation, and Quercus laevis was

both less common and lower in height. However, logging and fire

control have favored Quercus, and it is presently the more abundant

tree in most stands of sandhill vegetation. In contrast to sand-pine

scrub, shrubs are relatively few and scattered while herbaceous ground

cover is well-developed. The herbaceous layer, composed mainly of

wiregrasses (Aristida; Sporobolus) but with many species of forbs,

combines with fallen pine needles and oak leaves and forms a loose

cushion that varies in depth depending on the recency of ground fire.

Both associations occur on well-drained soils and often are adjacent.

The transition between them usually takes place over a distance of

only a few meters; in some places, one can place a foot in both

associations simultaneously.

The three hybrid zones studied are along ecotones in the Ocala

National Forest. The Lake Eaton ecotone is near the intersection of

FS-79A and FS-96, about .5 mile south of Lake Eaton (Section 26, T. 14

S., R. 24 E.). The Forts Bear Hole ecotone is south of FS-95 about

1.5 miles southwest of Forts Bear Hole (Sections 17 and 18, T. 16 S.,

R. 25 E.). The Alexander Springs ecotone parallels SR-445 .3 mile

to its west and about 2.0 miles southwest of Alexander Springs.









Methods


The identification of an individual as belonging to one of two

groups is a problem best solved by the use of a linear discriminant

function. The function is generated using data taken from individuals

which have been classified into groups a priori. It is then used to

classify unknown individuals. In order to distinguish unambiguously

between individuals of the two species that were from populations

unexposed to hybridization, a linear discriminant function based on

ten characters was computed using the BMD 04M program (Dixon, 1968).

For each sex of S. undulatus, twenty-five specimens from Alachua and

Marion Counties, Florida, comprised the parental sample. The parental

sample for each sex of S. woodi was twenty-five specimens from the

central portion of the scrub of the Ocala National Forest. The

discriminant function allows a value, Z, to be obtained for each

individual by summing the products of each measurement and the

corresponding discriminant coefficient. If the discriminant function

is effective, there is no overlap in the distributions of Z values of

the two species. Calculation of Z values for individuals from the

ecotones permits evaluation of their morphological affinities.

Table 17 lists the characters included in the discriminant function.

All were counts except for the last, which is the sum of values for

five qualitatively scored characters; these characters were the

degree of contact between the first canthal and the lorilabials on

either side, the degree of reduction to one row of scales between the

subocular and the supralabials, and the size of the auricular lobules.

Possible values for each character ranged from 0-2, with the low values

assigned to the condition typical in undulatus and the high values to













TABLE 17. Description of Characters in the Discriminant Function







1. Femoral pores: sum of both sides

2. Subdigital lamellae of hind fifth toe: sum of both sides

3. Circumorbitals: sum of both sides

4. Lateral supraoculars: sum of both sides

5. Subdigital lamellae of right hind fourth toe

6. Supradigital lamellae of right front fourth toe

7. Infratarsals on right foot between base of second toe and
base of fifth toe

8. Infratibials on right leg between base of fifth toe and knee

9. Mid-dorsal scales between interparietal and hind margin of thigh

10. Qualitative head scalation: value from 0-10









that in woodi. Several metric characters were examined but could not

be used in the discriminant function, since adjustment for differences

in body size would require a priori a choice between the regressions

of the two species. Additionally, the color pattern of each specimen

was scored subjectively on a five-point scale in regard to its approach

to a pattern typical of one of the species. A pattern typical of

woodi was scored as 5 and one like that of undulatus was scored as 1.

In order to identify the factors that minimize gene flow across

the ecotone, several behavioral and ecological characteristics of each

species were studied. For each species, male choice tests (Ferguson,

1969) were made during the middle of the breeding season in a non-

hybridizing population and in a population adjacent to a hybrid zone.

These populations were near Gainesville, Florida, and near the Alexander

Springs ecotone, respectively, for undulatus. They were located west

of FS-88, about 1.5 miles south of FS-75, in the Ocala National Forest

(Section 32, T. 12 S., R. 25 E.) and near the Lake Eaton ecotone,

respectively, for woodi. An adult male encountered in the field was

simultaneously offered an adult female of each species. Each female

was tethered at the end of a thin pole nine feet long. The females

were of equal size and were positioned so that they were equally

exposed to the view of the male. Any test in which movement by a

female caused the male to court her was discarded. A test was

completed when the male courted (and frequently attempted mating with)

one of the females. The male was then collected so that he would not

be retested.

Direct female choice tests are not feasible in lizards because of

agonistic interactions between males and variations in female

receptivity. Experiments by Hunsaker (1962) indicate that female









Sceloporus, when given a choice of compartments in a cage, choose

most frequently that in which there is a conspecific male. In nature,

such a tendency would function to reduce interspecific matings. To

test such a possibility, females were placed in cages divided into

three 15-inch square compartments; a male woodi was tethered in one

end compartment and a male undulatus in the other. Each compartment

was lighted by a 60-watt bulb positioned so that the temperature of

the cage floor below it was 300 C. Every half hour the compartment

occupied by the female was noted. Finally, interspecific agonistic

behavior between males was investigated by placing a male of one

species, tethered to the end of a pole, near a male of the other in

the field.

The tendency of each species to remain restricted to the usual

plant community, even in the absence of the other species across the

ecotone, was investigated by observation at appropriate localities.

More than fifty sand-pine scrub localities that lacked woodi were

searched for undulatus. The seven islands of sandhill vegetation in

the Ocala National Forest unoccupied by undulatus were examined for

woodi. In one of these (Riverside Island; Sections 32 and 29,

T. 12 S., R. 25 E.) the amount of bare ground and the density of grass

blades were measured by line intercept; from each central point

two-meter lines were run in the four cardinal directions. Three sets

of twenty-five points were obtained. The points of one set were

chosen by use of a random number table in an area where woodi had never

been observed. The points of the second set were taken in an area

where large expanses of open sand had been created by logging operations;

each point was determined as the position of an undisturbed individual









of woodi when first seen by the investigator. Locations of points of

the last set were chosen by the random number table in the same area

where the second set had been taken.

Foraging behavior was investigated by direct observation of

undisturbed individuals in the field; forty hours' observation was

logged for each species in addition to many hours of more casual field

observation. Since Sceloporus normally sits and waits for prey move-

ment instead of actively foraging, the time spent by an individual

on perches of varying heights above the ground and the orientation

of the individual's head were recorded together with information on

actual prey capture. Stomach contents of seventy adults of each species

collected between May and August were examined; each species sample

had an equal sex ratio.



Results


Phenetics on the Ecotones


Character means for the two species and the discriminant coeffi-

cients for each character are given in Table 18. The discriminant

functions distinguish the non-hybridizing populations without mis-

classification of any individuals (Figure 6 and Figure 7). Of the

twenty-seven male Sceloporus collected near the Alexander Springs

ecotone, eight had Z values within the range observed for undulatus

and ten had values within the range for woodi. The Z values of the

remaining nine individuals fell between the ranges of the allopatric

undulatus and woodi samples. These individuals are interpreted as

hybrids. Thirty-eight females from the Alexander Springs ecotone












4 -1
do N OS C N O a N I 00N
C0 l OS 0 0 D OS a .h m 0 I $0 rO
r- O 0 0 N I M n '0 i 0 I CLn C p
*H*H O 0 0 0 0 0 0 0 0 0 0 I 0 N
S *+ * * * a I 0-H N *-
u0q4 0 0 0 0 0 0 0 0 0 0 1 > 0
40 I I I I 000I 0 0
ad m0c a) 0




i I N
00 I00 C .t
O 0 0 I C 0 m
E 0 OS OS C I N N C C I o 0
S3 On . .. .4
a0 N N N N N 0 N
H rI 0




5 HS 00 10 -:t tn m MD M Ed :1 0)
0-H -4 N N N N O 0 N N I N aNO
S -i N 0 I cd 0 H


0 m N



44



0 -H0 ) N N, OS '0 ,t Lo m '0 I COH H Cm
-H 0C 0 N C C OS 4 0 O I 4 O .4

0 0- a a C a C a C 0 a C I 4> C C

o ) C
+.)

'I
O 0
0 I NC N
p 4 Oo ) 0 )N N '0 .4 o '0 I 0 N N
44 00 'U 0 N N N O N oN '0 C I 0 ; N
N N O . . . . . . . . . 00 0 M
Cd oH oZ N C a L i i)
0-i 0 0 0 0 0 0 0 0 0 0 C O
0 2O N N N N N r4 14 N 0I I 0 In N





H 0 H OS O N 0 0
*H O N N O N N I N
(0 00 N OS .4 O I N r- I 0 H N











H a M I (
O *H 0 0




00 0 0 O




0 N N H 0 0 I
Ni 0 NM 0 T- 0 Ic r
N-1 0 0 i 0 4 I *









0 CU $ N N U 0 0 I O
0 -H 0 M 0 4 0 0 N 0 I t




S0 0 40 0 00 $4 0 I 44


N N IU'N OS 0

.44 I































> 4 in






, 01 0


P-O 4-H



.4- 4- Cd
0 0 C) c*

C) .0 I
i 3 i p








S)0 4N





U 0
rq 0 n0


41) 4 a
*H il H








NO C
S4 4-f





i 00







0 l C


H0 l 0(

*H d L
4-' C "-4



4- 0 a) H"



L0
























E-

ca




-------- q E-------




E 0 1 r"

C 0
. r


-0
-oL 0


0 L( 0
o -
o 0








,







-v
*^ *T -^ *C
































tOCH
C4 > 4
r0 0 '4-

1H N*







0
H '-H' 3/





H 3 V)0





H 0
O C C3
.0 0 4-1 I-'



,C 3





.C -O -






000 0
-a p' o






0 M




V4- ) -H*
0 0








V) 0

3 tO 0




0 -I) '0 F



C r-d 1.
0 dl ru











r-
F. I -




























E





o


-- ---

F
L-
S E
C



'o E
m-l----


[
I












rL

c














[
C
-----I


n m-i


C






-
C
r
E C


. c
C
C
'1


------


*
C









"0
r


03









were examined. The discriminant function classified fourteen as

undulatus and thirteen as woodi; eleven were intermediate. Figure 8

maps the location of each specimen from the Alexander Springs ecotone

and approximately indicates its position on the phenetic continuum.

All the undulatus were restricted to the sandhill side of the ecotone

and most were found in the area between SR-445 and FS-38 that was

minimally disturbed by logging. The woodi were collected in the main

body of the scrub, in the small scrub patches within the sandhills,

and in the parts of the sandhills that have been severely modified

by logging. Most of the hybrids occurred in these disturbed areas,

but some were found outside them in the sandhills.

Twenty-five males and twenty-two females were collected near the

Forts Bear Hole ecotone. Two males and a single female were classified

as undulatus. Eighteen males and fifteen females were classified as

woodi. Five males and six females were intermediate. The phenetic

position and location of each specimen is shown in Figure 9. The

woodi were in scrub and in disturbed portions of the sandhills,

particularly along sand roads; inexplicably, large areas of undisturbed

sandhills were without undulatus, although those collected were in such

an area. The hybrids occurred primarily near sand roads through the

sandhills.

Fifty-three males were available from the Lake Eaton ecotone.

Four had Z values within the undulatus range and thirty-five had Z

values within the range for woodi. Fourteen individuals with inter-

mediate Z values are judged to be hybrids. Of the sixty-nine females

collected near the Lake Eaton ecotone, nine were classified as undulatus,


























-a






OUU
C a)






0rnt *





*H rl T 0
4C U I-,











cd V)C



Ot 4 r
O -lH H

C
m U 0 F -I




C) H .C)







( '0 -to





to PLC 3C)





C I3 Q3 m





83





4

o0
0 0
rc) 0




0 0
0 0O

0
0 0





S.-. -



S.0.







-. C. .











"" ,:" '
: : ..** ,
. ., ,:., '.U





























V tu
03

c0 0 w n
*H 03*H


0d V d 0
nH V H
0 -H 0
03 VI



0 >,






Ln H H

0 cd- t o
s* .H 30)
0 -0 -
0-H 00*0
4-' *H 0 S
0 P 03 H
0 0 0 > 01

0 0 *-H 0
> > 03 0
co cnl-H1
SPa tCo



14 H 03 0
03 V) to 3 0
03 03 *H
to to in V>

0 -H .--

0C.H -U



to ,. 1-









*H
LI,






85
















S* *




***














'



)'-"o 0









forty-one as woodi, and nineteen as hybrids. Figure 10 shows that

most of the woodi were located in the scrub or near sand roads adjacent

to it, but that woodi was also present in the sandhill vegetation in

a band that extended perpendicularly from the ecotone across FS-96 to

FS-79A. The band contained considerable bare sand within it. The

undulatus were only collected in the sandhill vegetation; they were

concentrated in an area with heavy wiregrass growth parallel to FS-79A.

Almost all the hybrids were located in the sandhill vegetation.

When the mean Z value of the non-hybridizing woodi was compared

with such values calculated for individuals from each ecotone which

were classified as woodi, the mean Z values of the ecotonal woodi were

in every case shifted in the direction of undulatus (Figures 6 and 7).

Those of the females from the Forts Bear Hole and the Lake Eaton

ecotones were significantly shifted (p < .05; one-tailed test). The

equivalent comparison for undulatus reveals a shift of all mean ecotonal

Z values toward those of woodi. Here, however, only the means from

the Forts Bear Hole ecotone were not significantly different from the

mean of the non-hybridizing population, and this exception is probably

the result of the small sample sizes.



Interspecific Behavior


The results of the male choice tests are presented in Table 19.

Both away from and near the ecotones, males showed no tendency to

distinguish between females of the two species. Ferguson (1969) showed

in similar tests that free-living Uta males in Colorado discriminate

against Texas female Uta in favor of Colorado female Uta; he suggested






























a 00

*H ,a a)
0 ) 0 r*




0 C r-






>4
*d > 7C

C) 0 r- r-
--4Cfl-H 0
o O CC m




0 u dn M
0 > 4 o




I CC 04*' 1

0 tHOC
CH 0








C) H- C c ) a3
0 4J L C
4-'P4 1a l i











'n *FH 0 C
0 0
c< U IH0








bo
4r a)
i-i b0 ^ v
(u t a
J3 -1 h
mc' < ^
&< V] -i-





0=E
oC r 1 a
cgab~Jl
M~~a
*r-1C
[_ ~ aa





88










,. .. ..
;. ',--. :. . .









S" :: S




."'.. 'o"

OS
'L' ~S
~S


















TABLE 19. Results of Male Choice Tests


Males Tested



Lake Eaton ecotone woodi


Alexander Springs ecotone undulatus


non-hybridizing woodi


non-hybridizing undulatus


Female Courted*
woodi undulatus


7 6


*p > .95 in all cases that males do not discriminate









this happens because the Texas Uta resembles Sceloporus graciosus, which

is sympatric with the Colorado Uta.

Results of the female association tests are more ambiguous (Table 20).

Female woodi and female undulatus from non-hybridizing populations both

tended to associate more frequently with a male woodi than with a male

undulatus or with neither. Female woodi from ecotonal populations were

more often with a male than expected by chance, but did not discriminate

between the species. The female undulatus from ecotonal populations

were unassociated with a male more often than expected, but they tended

to be in the compartment with the male undulatus more frequently than

in that of the male woodi. Although undulatus females from the ecotonal

populations tend to discriminate in favor of conspecific males more than

undulatus females from non-hybridizing populations, woodi females show

the opposite trend. Consequently, the data are perhaps best inter-

preted as failing to clearly show character displacement of association

behavior on the ecotones.

When males of either species were presented to males of the other

species in the field, territorial defense behavior was elicited; this

included head bobbing, lateral compression, lateral presentation, and

biting.



Habitat Selection


Sceloporus undulatus was found in only two of the fifty-three scrub

communities that lack woodi, and both these localities are atypical.

One is a scrub one mile northwest of Sunnyside, Bay County, portions of

which contain very large and widely separated sand-pine and have a




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs