Latitudinal Variation in Life History Reveals a Reproductive Advantage in the Texas Horned Lizard (Phrynosoma cornutum)

Full text

Latitudinal Variation in Life History Reveals a Reproductive Advantage in the
Texas Horned Lizard (Phrynosoma cornutum)
Daniel F. Hughes
1
, Walter E. Meshaka Jr.
2
,CarlS.Lieb
3
, and Joseph H. K. Pechmann
4
Geographically widespread species that occupy varied thermal environments provide testable models for understand-
ing the evolution of life-history responses to latitude. Studies that draw range-wide conclusions using descriptive data
from populations in the core of a species’ distribution can overlook meaningful inter-population variation. The Texas
Horned Lizard (Phrynosomatidae: Phrynosoma cornutum) spans an extensive latitudinal distribution in North America
and has been well studied in connection with life-history evolution, yet populations occupying the most northern and
coldest areas within its range were absent from previous examinations. We tested genus-wide patterns and challenged
species-specific findings on the evolution of the life-history strategy of P. cornutum using populations at the northern
edge of its geographic range and comparative material from farther south. Traits in populations at the highest
latitudes corroborated several patterns for the genus and species, including delayed reproduction and a trend towards
smaller adult body size with increasing latitude. Novel in our study, however, was the finding of a comparative increase
in clutch size among size-adjusted females in Kansas, indicating a reproductive response for greater fecundity at the
northern edge of its geographic range. Furthermore, analyses adjusted for body size revealed that egg dimensions were
constant across variation in clutch size, suggesting that there is not a strong relationship between egg size and egg
number across latitude. We discuss the selective pressures that may have resulted in the diminution of adult body size
coupled with greater fecundity that is unique to the northernmost populations of P. cornutum. Our findings highlight
the type of insights into the study of life-history evolution that can be gained across Phrynosomatidae from the
inclusion of populations representing latitudinal endpoints.
NEARLY all facets of animal lives are affected by
latitudinal variation in climate, including seasonal
activity (Sperry et al., 2010), metabolism (Tsuji,
1988), and development (Laugen et al., 2003). These
phenomena are intensified in ectotherms because tempera-
ture strongly influences their physiology (Huey, 1982), and
many aspects of physiology have implications for morphol-
ogy and demography (Huey and Berrigan, 2001). Latitudinal
variation in ectotherm body size, growth, and mortality has
been well studied (Adolph and Porter, 1993, 1996); however,
we know far less about such variation from direct measures of
reproduction. Variation in fecundity can have significant
demographic consequences that impact population size and
persistence. For example, smaller ectothermic lizards gener-
ally have fewer young (Fitch, 1985), and if lizards are smaller
at higher latitudes (e.g., Ashton and Feldman, 2003;
Pincheira-Donoso et al., 2008), then clutch size is expected
to be smaller. Nevertheless, there is a tradeoff whereby
changes in the offspring number are associated with changes
in egg size, thus impacting offspring size and ultimately their
survival (Sinervo, 1990). Herein we examine variation in
reproductive traits of Texas Horned Lizards (Phrynosoma
cornutum) from three populations that span most of the
species’ latitudinal distribution in the United States (Fig. 1).
Body size in lizards is highly variable (Olalla-Ta
´rraga et al.,
2006), and patterns often do not conform to traditional
latitudinal explanations for vertebrates (e.g., Bergmann’s
Rule [Angilletta et al., 2004]). Body size in lizards is
constrained by seasonal resource acquisition (e.g., Ballinger,
1977; Van Noordwijk and de Jong, 1986), and the amount of
time each year that foraging is feasible will decline with
increasing latitude. Latitudinal patterns in lizard body size
covary with several life-history variables, one of which is
clutch size (King, 2000; Angilletta et al., 2004). At higher
latitudes, growth is generally slower as lower temperatures
shorten activity seasons and increase dormancy (Blouin-
Demers et al., 2002); however, a relatively larger body size is
achieved in some species by prolonged growth and delayed
maturation (Atkinson, 1994). Latitudinal variation in clutch
size should mirror the variation in body size, and any
deviations from this pattern would suggest differential
selective pressures within species populations (Tinkle et al.,
1970).
North American horned lizards (Phrynosomatidae: Phryno-
soma) have previously been the subject of many ecological
studies in connection to life-history theory (e.g., Ballinger,
1974; Howard, 1974). Pianka and Parker (1975) conducted a
genus-wide examination of horned lizard species and
described their life-history strategy as belonging to the
‘‘delayed reproduction, large clutch and single brooded’’
category per Tinkle et al. (1970). Further, Pianka and Parker
(1975) sought to reconcile the heterogeneous life-history
strategy in horned lizards (i.e., r-selected for high fecundity
and K-selected for delayed maturity and long lives) with the
expectation of higher juvenile mortality relative to the more
homogenous strategies for other phrynosomatid lizards (e.g.,
Adolph and Porter, 1993; Pianka, 1970).
Studies of the Texas Horned Lizard (P. cornutum) have
corroborated several genus-wide patterns and hinted at some
geographic trends to its life-history strategy. Across a
latitudinal gradient from Me
´xico to Colorado, Texas Horned
Lizards exhibit variation in body size with smaller individuals
at northern latitudes (Montgomery et al., 2003), yet not all
horned lizards exhibit this trend (Pianka and Parker, 1975).
1
Department of Animal Sciences, University of Illinois at Urbana-Champaign, 1207 West Gregory Drive, Urbana, Illinois 61801; Email:
dfhughes@illinois.edu. Send reprint requests to this address.
2
Section of Zoology and Botany, State Museum of Pennsylvania, 300 North Street, Harrisburg, Pennsylvania 17120; Email: wmeshaka@pa.gov.
3
UTEP Biodiversity Collections, Department of Biological Sciences, University of Texas at El Paso, 500 West University Avenue, El Paso, Texas
79968; Email: clieb@utep.edu.
4
Department of Biology, Western Carolina University, 130 Natural Science Building, Cullowhee, North Carolina 28723; Email: jpechmann@
email.wcu.edu.
Submitted: 10 July 2019. Accepted: 1 October 2019. Associate Editor: C. Bevier.
Ó2019 by the American Society of Ichthyologists and Herpetologists DOI: 10.1643/CH-19-266 Published online: 2 December 2019
Copeia 107, No. 4, 2019, 736–747

The large clutch sizes and delayed maturity among horned
lizards is associated with the production of a single clutch per
season (Pianka and Parker, 1975). Clutch frequency in the
Texas Horned Lizard, however, is geographically variable
among populations without demonstrating a regional geo-
graphic bias (Ballinger, 1974; Howard, 1974; Vitt, 1977). A
seasonal increase in testis size from May–August is conserved
across horned lizards, and this increase is consistent across
longitude in the Texas Horned Lizard (Ballinger, 1974;
Howard, 1974). Delayed reproduction is a genus-wide
phenomenon and occurs in Texas Horned Lizards from the
southwestern United States (Howard, 1974) to Oklahoma
(Endriss et al., 2007). Texas Horned Lizards exhibit latitudinal
variation in clutch size from Texas to Oklahoma with smaller
clutches observed in populations at higher latitudes (Howard,
1974; Vitt, 1977; Endriss et al., 2007). However, this latitudinal
pattern has been interpreted without accounting for estab-
lished clinal changes in adult body size (i.e., Montgomery et
al., 2003).
Museum specimens were the primary source of data for
nearly all of the studies on the Texas Horned Lizard cited
above. In keeping within this well-established framework, we
examined a large series of museum specimens from the
northern edge and a southwestern portion of the Texas
Horned Lizard’s geographic range to address the following
three main questions: (1) Are there geographic patterns in
life-history characteristics after controlling for differences in
body size among populations? (2) What is the relationship
between egg size and clutch size within and among
populations? (3) Are there latitudinal patterns in multivariate
life history? Importantly, we included populations from a
higher latitude than previously studied, and we focused on
reproductive traits for which predictability has been lacking
in previous investigations. Essentially, we attempted to use
our data—that include the northernmost populations—to
resolve previously inconsistent life-history findings deduced
from narrow geographic sampling for the Texas Horned
Lizard.
MATERIALS AND METHODS
We examined 279 specimens of the Texas Horned Lizard
from four US-based herpetology collections, which included
91 from the Sternberg Museum (Hays, Kansas), 122 from the
University of Texas at El Paso Biodiversity Collections (El
Paso, Texas), 26 from the Carnegie Museum of Natural
History (Pittsburgh, Pennsylvania), and 40 from the Univer-
Fig. 1. Map showing the populations
of the Texas Horned Lizard (Phryno-
soma cornutum) we examined for
life histories: Kansas, Me
´xico, New
Mexico, and Texas. The sample size
for each location is displayed in the
lizard-head silhouette and sample
sizes by county (or Mexican state)
are in parentheses after each loca-
tion. The species’ geographic distri-
bution (shaded area) was modified
from Hammerson (2007). The photo
of a male and female in coitus was
taken on 24 June 2016 in Do ˜
na Ana
County, New Mexico, by CSL.
Hughes et al.—Reproductive variation in the Texas Horned Lizard 737

sity of Illinois Museum of Natural History (Champaign,
Illinois). The specimens originated from three primary
locations: Kansas, Texas, and New Mexico (Fig. 1). Eighty-
seven specimens were collected from 1958–2013 in seven
northern counties in Kansas (Ellis, Lincoln, Osborne, Rooks,
Russell, Saline, and Trego). Eighty-two specimens were
collected from 1930–2010 in five western counties in Texas
(Brewster, Culberson, El Paso, Hudspeth, and Reeves). Eighty
specimens were collected collected from 1939–2015 in ten
mostly southern counties in New Mexico (Chaves, Do˜
na
Ana, Eddy, Grant, Hidalgo, Luna, Otero, San Miguel, Sierra,
and Socorro). We also examined a smaller sample of thirty
specimens that were collected from 1951–1979 in three states
in northern Me
´xico (Chihuahua, Coahuila, and Durango).
We note that the classification of populations based on
political boundaries is artificial because they may not reflect
natural groups found in the wild. Our subsequent discussions
of our study populations by states and countries refer only to
the areas represented in our samples, not other parts of these
entities.
For each specimen, we first measured snout–vent length
(SVL) with calipers to the nearest 0.1 mm and then we
dissected and sexed the lizards. We classified females as
immature or non-reproductive if the largest ovarian follicle
was ,3 mm and not vitellogenic. We assigned females as
reproductive if they contained shelled eggs or vitellogenic
ovarian follicles 3 mm. We measured the diameters of the
largest yolking follicles, and length and width of shelled eggs
to the nearest 0.1 mm. We calculated egg volumes from the
equation for an ellipsoid: v ¼(4/3) pab
2
(Mayhew, 1963). We
then calculated clutch volumes as the product of clutch size
and egg volume. We used the fraction of length and of mid-
testis width/male SVL as a measure of seasonal male fertility.
We determined the ovarian cycle by plotting monthly
distribution of the largest follicle and shelled egg length.
We estimated sexual-size dimorphism (SSD) by first dividing
the mean adult SVL of males by that of females. We also
calculated SSD using the Lovich and Gibbons (1992) index
where we divided the mean adult female SVL by the mean
adult male SVL and then subtracted 1, so when the SSD ¼0
the sexes are equal in size, and SSD ,0 when males are
larger, and SSD .0 when females are larger.
Without direct observations on the number of eggs
deposited in each clutch, we estimated clutch size using
counts of shelled eggs and/or vitellogenic ovarian follicles 3
mm. Because SVL is linearly related to clutch size in Texas
Horned Lizards (Fig. 2; Ballinger, 1974; Howard, 1974) and to
avoid potential problems with the use of ratios in statistics
(Packard and Boardman, 1999), all measurements were
analyzed using multiple regression with SVL as a covariate
to adjust for effect of body size in our comparisons of life-
history variables among populations. We used a General
Linear Model (GLM) with SVL as a covariate to test if there
was a significant difference between the two clutch measures
(shelled eggs and vitellogenic follicles). We tested whether
the slopes of the regression lines between SVL and clutch size
differed among populations or by clutch measure by testing
for interaction effects among these variables using a GLM.
We normalized the mean SVL to zero for analyses of clutch
variation across populations by subtracting the global mean
SVL (88.4 mm) from all SVL measurements, so that
comparisons of categorical variables would occur at the
mean SVL rather than at SVL ¼0. This was necessary because
preliminary analyses found differences in the slopes of the
regression lines across populations, and we wanted to make
comparisons at the most biologically relevant SVL.
We also explored multivariate divergence of the three
primary populations with principal component analysis
(PCA). To avoid weighting variables by their variance, the
PCA was performed using a correlation matrix calculated
from means of five life-history characters (clutch size, egg
length, egg width, egg volume, and clutch volume). We
restricted this analysis to females with shelled eggs only. One
female specimen from New Mexico with shelled eggs was
excluded from this analysis because the eggs were rounded.
Measurements were log transformed then size corrected
using SVL as a covariate because all five measurements were
positively correlated with body size. The residuals were saved
andincludedasinputforPCA.Weusedstandardized
guidelines to determine that our data met the assumptions
(e.g., normality) of PCA (McGarigal et al., 2000).
Data were organized in Excel 2016 (Microsoft Inc., Red-
mond, WA) and all statistical analyses were conducted in
Minitab v. 17 (Minitab Statistical Software, State College, PA)
or SAS v. 9.4 (SAS Institute, Cary, NC). We note that several
specimens were collected as dead on road with various levels
of damaged structures or some specimens had internal organ
systems removed prior to our study, and thus we could not
extract all measurements from all specimens. We chose to
include data in relevant analyses and summary statistics from
damaged specimens that still possessed interpretable charac-
ters, such as SVL when internal organs were removed or
clutch size when egg sizes were unmeasurable, but we note
that these inclusions inevitably influenced sample sizes
across categories.
RESULTS
Testicular cycle.—Testis length and width varied seasonally in
all three populations but was most similar between New
Mexico and Texas (Fig. 3A–C). Among males in Kansas, testes
were largest in April and May, and smallest in June–August
(Fig. 3A). For Texas males, testis size was largest from May to
July (Fig. 3B). In New Mexico, testes reached a maximum size
in May and June, and rapidly decreased in size thereafter,
Fig. 2. Clutch size regressed against body size for three populations of
the Texas Horned Lizard (Phrynosoma cornutum). We normalized body
size by subtracting the global mean SVL (88.4 mm) from all
measurements. Dark gray rectangles denote Kansas, filled circles Texas,
and light gray diamonds New Mexico. See text for more details of these
analyses.
738 Copeia 107, No. 4, 2019

with the smallest sizes in July and August (Fig. 3C). Late-
season increases in testis size were evident by August in
Kansas, and by September in Texas and New Mexico.
Nesting season.—The nesting season was short (ca. 2–3
months) in all three populations, and some overlap was
evident among them (Fig. 3D–F). In Kansas, shelled eggs were
present in females during May and June (Fig. 3D). In Texas,
ovigerous females were present also for two months,
beginning one month later (June–July; Fig. 3E). The longest
nesting season (May–July) was detected in New Mexico,
which subsumed the combined nesting season of the other
two regions (Fig. 3F).
Clutch size.—An initial analysis found significant effects on
clutch size of population (F
2,30
¼9.22, P¼0.0008) and SVL
(F
1,30
¼35.56, P,0.0001), and significant interactions
between SVL and population (F
2,30
¼6.85, P¼0.0035) and
between SVL and clutch measure (F
1,30
¼6.12, P¼0.0193).
Clutch measure (F
1,30
¼0.75, P¼0.392), the interaction of
population and clutch measure (F
2,30
¼1.40, P¼0.261), and
the three-way interaction of population, clutch measure, and
SVL (F
2,30
¼2.10, P¼0.140) did not have significant effects.
We next tried two different follow up analyses: 1) dropping
the non-significant terms from the full model, and 2)
conducting separate analyses for the two clutch measures.
Population (F
2,41
¼3.79, P¼0.0322), SVL (F
1,41
¼17.08, P¼
0.0002), and their interaction (F
2,41
¼3.62, P¼0.0371)
remained significant in the reduced model. The interaction
of SVL and clutch measure was far from significant in the
reduced model (F
1,41
¼0.22, P¼0.638), and the regression
lines for the two different clutch measures were nearly
identical. Results were similar when data were analyzed
separately by clutch measures, for which population, SVL,
and their interaction were the only terms in the models
(results not shown). We, therefore, decided to pool data for
the two clutch measures for our final analysis.
When data for the two clutch measures were combined,
clutch size was significantly affected by population (F
2,41
¼
3.78, P¼0.0322), SVL (F
1,41
¼17.45, P¼0.0002), and their
interaction (F
2,41
¼3.60, P¼0.0377; Fig. 2). The least square
mean clutch size (clutch size at the global mean SVL [88.4
mm]) was 29.4 for Kansas, 22.7 for New Mexico, and 10.8 for
Fig. 3. Monthly distribution of three life-history traits in the Texas Horned Lizard (Phrynosoma cornutum) from three populations. Top row: Testis
length and width as a fraction of male snout–vent length (SVL) for (A) Kansas (n¼39), (B) Texas (n¼41), and (C) New Mexico (n¼30). Gray circles
represent testis length/SVL and open circles testis width/SVL. Middle row: Largest ovarian follicle diameters and shelled egg lengths for (D) Kansas(n
¼16), (E) Texas (n¼16), and (F) New Mexico (n¼15). Gray circles represent shelled eggs and open circles yolked ovarian follicles. Bottom row: Body
size (SVL) for (G) Kansas (39 males, 26 females, 22 juveniles), (H) Texas (46 males, 29 females, 3 juveniles), and (I) New Mexico (42 males, 34
females, 3 juvenile). Gray circles represent females, open triangles males, and black stars juveniles.
Hughes et al.—Reproductive variation in the Texas Horned Lizard 739

Texas, and differed significantly between Kansas and Texas (P
¼0.0255). Because the global average SVL lies between the
body size distributions of the two populations and does not
overlap with the body sizes in either population, this
significant difference applies to Texas Horned Lizards which
are slightly larger than the largest Kansas female observed
and slightly smaller than the smallest Texas female observed
(Fig. 2). Clutch size increased with body size much faster in
the Texas population than in the other populations (Fig. 2),
indicating that differences among populations in clutch size
adjusted for body size varied with SVL. For example, the
regression lines for Kansas and Texas cross around the body
size of the largest Texas female, indicating that if Kansas
females got as large as the largest females in Texas then the
clutch sizes for both would likely be similar, but we note that
no Kansas female approached such a large SVL.
Shelled egg size.—Size-corrected analyses detected no signif-
icant differences among populations in the means of egg
length (F
2,14
¼0.45, P¼0.649), egg width (F
2,14
¼0.62, P¼
0.553), or egg volume (F
2,14
¼0.62, P¼0.550). Because egg
length (R
2
¼25.9%, F
1,16
¼5.59, P¼0.03), egg width (R
2
¼
35.1%, F
1,16
¼8.68, P¼0.009), and egg volume (R
2
¼33.5%,
F
1,16
¼8.05, P¼0.01) were positively correlated with SVL, we
removed the effects of body size by calculating residual scores
from the separate regressions of each log-transformed metric
(and clutch size) on maternal SVL. Using the resulting
residuals in separate regression analyses revealed no effect
of clutch size on three egg traits (Fig. 4A–C), yet there was a
slight decrease in residual egg length and egg volume with
increasing residual clutch size, suggesting that eggs get
smaller as clutches get larger.
Clutch frequency.—We only detected luteal scars associated
with expended oviductal eggs, and concurrent ovarian
follicles that measured ,2.5 mm. We found a single female
collected in June from Saline County, Kansas, that had 29
shelled eggs and 25 yolked ovarian follicles .3mm,
suggesting that this individual had the potential to lay a
second clutch that year given acceptable energetic or
environmental conditions. We also found four females from
Texas and one from New Mexico collected in June or July
that had shelled eggs and ovarian follicles between 2 and 3
mm. These findings suggest that females from these three
sites typically produce a single clutch annually; however, we
could not rule out multiple clutch production in some
individuals across all populations.
Monthly activity.—We detected activity based on pooled
capture dates across locations (Fig. 3G–I). We found that
the activity in the Texas Horned Lizard ranged from seven to
eight months. Months of capture were longest in Texas
(March–October), followed by Kansas, New Mexico, and
Me
´xico (April–October). Captures in Kansas were far greater
from April to June (n¼65) than from July to October (n¼22;
Fig. 3G). Similarly, captures in Texas were greater from March
to June (n¼57) than from July to October (n¼21; Fig. 3H),
and captures in New Mexico were also greater from April to
June (n¼53) than from July to October (n¼26; Fig. 3I). From
a smaller sample, this pattern was reversed in Me
´xico where
captures were greater from July to October (n¼19) than from
April to June (n¼8). For all locations, captures began in April
(except a single specimen from Texas collected in March) and
ended in October. We note that the monthly activity
differences among populations may be related to the
idiosyncratic timing that specimens were collected and/or
sample size limitations, especially for Me
´xico.
Age at sexual maturity.—The smallest juvenile we examined
measured 22.6 mm SVL from Ellis County, Kansas, captured
on 14 September 1963. Body-size cohorts were apparent in
monthly distributions of body sizes for Kansas (Fig. 3G). We
presume May and June eggs to have hatched within two
months of oviposition. We do not know how much growth
would have occurred from a July–August hatching event to
our first juvenile cohort in September (22.6–29.2 mm SVL).
However, little growth appeared between the fall and
Fig. 4. Tradeoff between egg size and clutch size after statistically
removing the effects of maternal body size in three populations of the
Texas Horned Lizard (Phrynosoma cornutum). (A) Residuals for egg
length plotted against residual clutch size; (B) residuals for egg width
plotted against residual clutch size; and (C) residuals for egg volume
plotted against residual clutch size. The prediction interval is denoted by
a dark-gray tightly spaced dashed line, and the 95% confidence interval
is denoted by a light-gray loosely spaced dashed line.
740 Copeia 107, No. 4, 2019

following spring cohort (26.4 mm). The fastest growth would
have resulted in sexual maturity of males as early as 12–13
months of age. Females and all other males would have
reached sexual maturity during their second spring of life at
an age of about 21–22 months of age (Fig. 3G). Too few
juvenile and hatchling data were available for the remaining
three locations to estimate growth rates from monthly
distribution of body sizes (Fig. 3H–I).
Body size at sexual maturity.—Adult body size exhibited a
negative relationship with latitude (R
2
¼51.5%, F
1,238
¼
253.01, P,0.001; Fig. 5). Body sizes of adult males followed
a latitudinal trend of increasing size from north to south with
respect to minimum, maximum, and mean values (Table 1).
Mean male SVL in Kansas was significantly smaller than that
of New Mexico, Texas, and Me
´xico (F
3,134
¼125.3, P,0.001).
Likewise, body sizes of adult females followed a geographic
trend of increasing size from north to south with respect to
minimum and mean values. Mean female SVL in Kansas was
significantly smaller than that of New Mexico, Texas, and
Me
´xico (F
3,98
¼36.75, P,0.001; Table 1).
Sexual dimorphism in adult body size.—Mean body size of
adult males was found to be significantly smaller than that of
Fig. 5. Body size regressed against latitude in four populations of the
Texas Horned Lizard (Phrynosoma cornutum). The prediction interval is
denoted by a dark-gray tightly spaced dashed line, and the 95%
confidence interval is denoted by a light-gray loosely spaced dashed
line. Filled symbols represent males and open symbols females.
Table 1. Summary of life-history characters for populations of the Texas Horned Lizard (Phrynosoma cornutum). Means 6standard deviation are
followed by ranges and sample sizes for body size, clutch size, egg length, egg width, egg volume, and clutch volume.
Variable Kansas Texas New Mexico Me
´xico
Male SVL (mm) 60.564.3 86.266.4 83.668.5 87.168.1
50.4–68 69.2–99.2 67.6–100.4 77.4–99.2
39 47 42 10
Female SVL (mm) 70.866.4 95.669.3 90.969.9 88.1612.3
61.6–86.1 76.6–113.3 73.2–116.3 71.6–108.1
26 30 34 12
Juvenile SVL (mm) 36.6611.9 42.5616.4 52.8610.4 61.367.2
22.6–59.3 23.8–59.6 39.6–62.7 46.1–69.6
22 5 4 8
Clutch size (pooled) 20.666 28.169.3 24.666.3 36
12–35 14–44 15–40 —
15 14 13 1
Yolked ovarian follicles 21.466 28.7611.8 2567.5 —
14–35 15–44 18–40
10 6 7
Shelled eggs 1966.3 27.667.8 24.265.2 36
12–29 14–41 15–31 —
58 61
Egg length (mm) 12.361.1 14.161.4 13.561.7 13.560.5
9.8–14.6 9.3–16.2 10.6–16.8 12.2–14.5
96 212 113 29
Egg width (mm) 8.260.8 1061 9.461.5 9.860.6
6.7–9.9 6.8–11.9 6.9–12 7.9–10.7
96 212 113 29
Egg volume (mm
3
) 440.36105.9 759.76184.5 669.86272.8 —
256.2–682.5 249.4–1,076.8 264.2–1,226.7
96 212 113
Clutch volume (mm
3
) 8,365.964,156.5 20,540.569,056.9 15,137.467,273.2 —
3,675.1–14,931 8,640.3–35,964.7 8,453.5–25,450.8
58 5
Female SVL with shelled eggs 73.968.9 101.364.5 99.3610.9 99.4
62.2–86.1 96.5–107.5 84.3–116.3 —
58 61
Female SVL with yolked
ovarian follicles
68.365.9 98.162.8 96.166.2 —
61.6–77.6 94.3–101.6 89.3–107.5
10 6 9
Hughes et al.—Reproductive variation in the Texas Horned Lizard 741

adult females in Kansas (t¼–7.77, df ¼63, P,0.001), in New
Mexico (t¼–3.45, df ¼74, P,0.001), and in Texas (t¼–5.24,
df ¼75, P,0.001), but not in Me
´xico (t¼–0.21, df ¼20, P¼
0.42; Table 1). Variation in the degree of female biased SSD
followed a latitudinal trend, where the SSD, expressed as the
ratio of mean body sizes of males: females, was 0.855: 1.0 in
Kansas, 0.909: 1.0 in New Mexico, 0.902: 1.0 in Texas, and
0.99: 1.0 in Me
´xico. Using the Lovich and Gibbons (1992)
index, we found additional support that SSD varied with
latitude such that adult lizards were the most dimorphic at
northern latitudes (Kansas ¼0.17) and that sexes were more
similar in size at southern latitudes (Texas ¼0.11; New
Mexico ¼0.09), especially in our southernmost population
(Me
´xico ¼0.01).
Trends in multivariate life history.—The pattern of differenti-
ation among populations and latitudinal influences are
presented in Figure 6A–B. Relationships among populations
were adequately represented in two dimensions judging by
the magnitude of their respective eigenvalues; the first two
dimensions explained 99% of the total variance (Fig. 6A;
Table 2). Two of the five variables loaded similarly on PC1,
which overall described 76% of the variation in the data
(Table 2). The first eigenvalue showed a strong positive
loading on clutch volume and second on egg volume, with
lower positive loadings on clutch size, egg width, and egg
length. For PC2, which explained 23% of the variance, the
second eigenvalue had the highest positive loading for clutch
size and relatively high negative loading for egg volume, low
positive loading for clutch volume, and low negative
loadings for egg length and egg width (Table 2).
The majority of the variation in our data was described by
PC1 and PC2 which were easily interpreted. Thus, we used
the scores for these axes to test the hypothesis that
latitudinal clines are derived from a co-adaptation of life-
history traits. The scores on the first principal component
showed a marginally negative, non-significant relationship
with latitude (R
2
¼0.21%, F
1,16
¼0.03, P¼0.856; Fig. 6B).
DISCUSSION
Much of the life-history strategy of Texas Horned Lizards (P.
cornutum) seems to have remained highly conserved as the
species expanded its range northward after the last glacia-
tion. The nesting season in Kansas is delayed by one month
compared to Texas, whereas in New Mexico the season spans
from the beginning of that in Texas to the end of that in
Kansas. The male gonadal cycle of maximal size in spring,
minimal size in late summer, and recrudescence in early fall
is generally consistent from 398Nto308N. Further, the overall
pattern of collection records during April–October is similar
from northern Me
´xico to northern Kansas, and a two-year
interval to reach maturity seems also to be maintained. The
resemblance among the three populations is even more
remarkable given the fact that drastic climatic and adult body
size differences exist across locations. We did find pro-
nounced differences in clutch characteristics and associated
metrics among P. cornutum in the United States, however. We
detected latitudinal signatures in body sizes and consequent-
ly clutch sizes but did not detect a strong influence of latitude
in multivariate life history among females with shelled eggs
across populations. We also found that egg dimensions
remained relatively constant across a wide range of clutch
sizes, despite a slightly negative relationship. We will first
discuss reasons as to why seasonal activity patterns and age at
maturity are so conserved in Texas Horned Lizards and then
we address latitudinal variation in life history and the
implications of our results.
The nesting season of P. cornutum, for the most part, is
restricted to a few summer months in the United States:
May–June in Texas (Pianka and Parker, 1975) and Kansas
(this study); May–July in southern Kansas (Givler, 1922),
central Texas (Ballinger, 1974), and southern New Mexico
(this study); June–July in western Texas (this study); and
June–August in the southwestern United States (Howard,
1974). Our results corroborate the evidence for a relatively
Fig. 6. Multivariate life-history divergence for three populations of the
Texas Horned Lizard (Phrynosoma cornutum). (A) Scatter plot of the
first two principal components extracted from five life-history variables
that were corrected for maternal body size. (B) Scores of principal
components 1 plotted against latitude (see Table 2 for loadings).
Table 2. Principal components analysis comparing populations of the
Texas Horned Lizard (Phrynosoma cornutum) with five life-history
characters regressed against snout–vent length (SVL). Loadings,
eigenvalues, variance, and cumulative variance are shown for the first
two principal components. Values in bold represent the variables with
the highest loadings.
Variable PC1 PC2
1 Clutch size/SVL 0.237 0.737
2 Egg length/SVL 0.147 –0.197
3 Egg width/SVL 0.192 –0.186
4 Egg volume/SVL 0.531 –0.568
5 Clutch volume/SVL 0.777 0.247
Eigenvalue 0.052 0.016
Proportion (variance) 0.76 0.23
Cumulative (variance) 0.76 0.99
742 Copeia 107, No. 4, 2019

short nesting season throughout its range in the United
States, and, as did Givler (1922) in southern Kansas, we
found that the northernmost population exhibits a strikingly
similar range. Our expectation that nesting activity patterns
would vary latitudinally was based on temperature being of
significant importance to ectotherm physiology (Huey,
1982). However, the nesting season seemed to exhibit only
minor shifts in relation to temperature changes. Nesting
activity does not appear to simply be explained by temper-
ature differences; perhaps some combination of factors that
vary geographically or seasonally is responsible. Lizards
behaviorally thermoregulate to maintain activity levels
necessary to meet seasonal needs, such as reproduction
(e.g., Adolph, 1990). Mating, based on testicular cycles and
nesting seasons, in the Texas Horned Lizard occurs largely in
the spring and can extend into early summer: April–May in
Kansas, April–June in Texas, and June–July in New Mexico.
Increased activity during the mating season suggests that,
following mating, the lizards become less active than thermal
conditions would allow. This pattern presumably represents a
tradeoff between activity and survival.
Characterized as delayed breeders, several species of
horned lizards reach sexual maturity at two years of age
(Pianka and Parker, 1975), including the Texas Horned Lizard
(Ballinger, 1974; Howard, 1974; Endriss et al., 2007). Using
our smallest individual (22.6 mm SVL), a two-year estimate of
age at sexual maturity in northern Kansas does not differ
from that of other studies, which is even more remarkable in
light of marked differences in body sizes among populations.
Juveniles and hatchling Texas Horned Lizards typically are
found in fall across its range: August in central Texas
(Ballinger, 1974), June–October from several sites along the
US–Me
´xico border (Howard, 1974), late July–August in
central Oklahoma (Endriss et al., 2007), and late August–
September in Colorado (Montgomery and Mackessy, 2003).
We do not know if all of the cohort of young individuals
taken in September from northern Kansas represent hatch-
lings or a mix of both hatchlings and individuals hatched
earlier in the season, yet the latter seems likely. Based on the
most probable hatching months after the nesting season, the
earliest maturing individuals in northern Kansas would still
not breed for the first time until nine months later.
Phryonosoma cornutum does not appear to follow Adolph
and Porter’s (1996) model of populations from colder
environments maturing at a later age and at a larger size
than populations in warmer areas (e.g., Wapstra et al., 2001).
Alternatively, Texas Horned Lizards occupying colder, north-
ern habitats seem to reach sexual maturity at nearly the same
age as their counterparts from warmer, southern areas but at
much smaller sizes. Several other phrynosomatid lizards
exhibit this same general trend (e.g., Parker and Pianka, 1975;
Mathies and Andrews, 1995), as does an Australian skink
(Forsman and Shine, 1995).
From Colorado to Me
´xico, adult body sizes in Texas
Horned Lizards of both sexes decreased from south to north
along a latitudinal cline (Montgomery et al., 2003). We found
a similar pattern of decreasing adult body size with increasing
latitude. A latitudinal trend in body size is often attributed to
differences in climate decreasing activity (Angilletta et al.,
2004), environmental differences in food availability (Ash-
mole, 1963), or day length (Lack, 1954). Our results suggest
that to maintain the highly conserved two-year constraint to
reach sexual maturity, adult body sizes exhibit marked
latitudinal differences dependent upon environmental vari-
ables. Indeed, we found that the most northern population
and those in the middle of its range exhibit the same pattern
of delayed maturity, but maturity was reached at a smaller
size at higher latitudes to compensate for reduced activity
and likely reduced growth. Furthermore, the more extreme
degree of female-biased SSD in Kansas compared to popula-
tions at lower latitudes suggests that the intensity of selection
driving dimorphism is greater at the highest latitudes,
perhaps to compensate for greater seasonality of the shorter
warm periods (Moreau, 1944; Lack, 1954). Interestingly,
Zamudio (1998) found phylogenetic evidence that female-
biased SSD in three species of viviparous Phrynosoma was
actually driven by negative selection on male body size for
earlier maturation, and not fecundity selection.
Egg size in Texas Horned Lizards has not been examined
previously with respect to clutch size. With female SVL held
constant, we found that there was not a strong relationship
between clutch size and three measures of egg size in the
United States. The tradeoff between offspring number and
egg size is mediated by a fecundity advantage of producing
many small offspring balanced against the survival advan-
tage of a few large offspring. Although not measured for P.
cornutum, egg size is generally correlated with offspring size
in lizards (Sinervo, 1990; Forsman and Shine, 1995). We
found some support for the prediction that clutch size is
inversely related to egg size (Stewart, 1979) based on slightly
negative relationships detected for egg length and egg
volume with increasing clutch size. The amount of energy
allocated to reproduction by females is divided between
clutch size and egg size, and thus as clutch size increases, egg
size tends to decrease, a tradeoff that has been observed in
many oviparous squamates (Ford and Seigel, 1989). None-
theless, we found that for Texas Horned Lizard populations
in the United States, egg size did not change much as clutch
size increased, after controlling for maternal body size. Thus,
egg metrics of females that attain smaller adult body sizes in
northern populations are comparable to those of much larger
females from more southern localities.
Although a single clutch per year is produced in central
Texas (Ballinger, 1974) and eastern Arizona (Vitt, 1977),
multiple annual clutch production appears to be more
common than previously thought in P. cornutum. We could
not entirely rule out the possibility of multiple clutches from
our Texas and New Mexico samples, and we found only one
of 15 gravid females in Kansas that had shelled eggs and
yolked ovarian follicles (.3 mm) concurrently. Howard
(1974) found evidence of double clutching in a sample of
21 females from four populations ranging longitudinally
along the US–Me
´xico border. Wolf (2012) found that six of
nine radio-tracked females from central Oklahoma deposited
two clutches in 2011. Further, Wolf et al. (2014) showed that
nearly half of the females from central Oklahoma and about
a third from south Texas produced two clutches annually.
There remains the possibility that triple-clutching occurs in
southern Texas populations (Burrow, 2000) based on obser-
vations of continual nesting throughout the active season
(Wolf, 2012; Wolf et al., 2014). Texas Horned Lizards exhibit
a slightly longer season of activity in south Texas (February–
December with most activity April–August [Moeller et al.,
2005]), which may contribute to additional egg-laying bouts.
Sufficient time for the production of multiple clutches is
feasible in south Texas. However, this interval does not
explain multiple clutch production in Oklahoma. Mecha-
nisms underlying multiple clutch production in northern
populations of the Texas Horned Lizard are unclear but may
be related to a combination of favorable environmental
Hughes et al.—Reproductive variation in the Texas Horned Lizard 743

triggers that may co-occur in some years. The effects of these
variables on horned lizard fecundity are worth exploring.
A significant departure in the life history of the Texas
Horned Lizard was found in the clutch sizes among our
samples. Across its geographic range, the Texas Horned Lizard
produces large clutches (Givler, 1922; Ballinger, 1974;
Howard, 1974; Pianka and Parker, 1975; Vitt, 1977; Endriss
et al., 2007), and our data generally corroborate that finding.
A significant difference, however, exists among our sites,
such that females from the northernmost site predictably
produced the smallest clutches in association with their
smaller body sizes. When adjusted for female body size, we
found that clutch size at the average SVL was actually larger
in Kansas than in our southern populations. Thus, selection
seems to favor relatively larger clutches in P. cornutum at the
northern edge of their range, placing them at a unique
advantage in terms of fecundity. This is a somewhat
counterintuitive response for northern populations, whereby
reproductive output increases even as selection enforces
diminution of minimum, mean, and maximum adult body
size of both sexes. That both sexes remain small and the SSD
in Kansas is greater than the SSD measured elsewhere lead us
to suspect that while a highly seasonal environment results
in increased fecundity in what is probably always a single
clutch per year (and likely once per lifetime), it is a short
adult life that maintains the small adult body sizes and
disproportionately high reproductive investment per unit
clutch in northern Kansas. The direction of fecundity
selection operating on P. cornutum in Kansas may maximize
fitness via an adaptive increase in clutch size, despite
persistent ecological costs associated with smaller body sizes
driven by living in a colder environment (Pincheira-Donoso
and Hunt, 2017). The greater increase in clutch size with
body size in the Texas population than in the other
populations suggests that fitness is maximized by increasing
clutch size more than body size where females are already the
largest. Historically, Ballinger (1974) found no evidence of
geographic variation in clutch size, either latitudinally in
Texas or longitudinally along the US–Me
´xico borderlands.
The absence of a latitudinal trend in his study is understand-
able. We suspect that Ballinger’s (1974) finding is best
explained by his sampling of the species within the center
of its geographic range, where clutch sizes are all similarly
large.
From their respective study sites, Ballinger (1974) consid-
ered the Texas Horned Lizard to be K-selected for most traits
(e.g., delayed maturity), yet r-selected for clutch size and
juvenile survivorship, and Howard (1974) concluded that the
species was also K-selected for late maturity and r-selected to
produce multiple large clutches annually. Our study included
specimens from sites much farther north than examined by
Howard (1974) and Ballinger (1974). Those specimens
provided the material necessary to test the extent to which
earlier findings concerning life history traits were applicable
to this species and the genus. Our analyses show extreme
conservatism in many life-history characteristics across its
range, yet we detected a considerable extent of geographic
variation in several direct measures of fecundity.
Concluding remarks, implications, and future directions.—The
Texas Horned Lizard is a geographically widespread species
(Price, 1990) whose categorization in life-history theory has
been based on populations largely exclusive of the most
northern latitudes. Interestingly, even with samples from the
northern edge of its range included, we found that the
species generally conformed to Ballinger’s (1974) findings in
Texas populations and Pianka and Parker’s (1975) genus-wide
model. However, we found the smallest adults and clutch
sizes from populations in the northernmost locations and,
interestingly, clutch size was larger in Kansas when we
controlled for maternal body size variation among popula-
tions. Geographic variation in body size varies among many
species of reptiles, and directions or absences of latitudinal
signatures in this trait vary among them (e.g., Meshaka and
Layne, 2015). In Texas Horned Lizards, the costs associated
with living in northern Kansas are exemplified by small body
sizes and likely reduced survivorship (Shine, 1992). The Texas
populations, on the other hand, are generally comprised of
very large females that produce a single clutch or multiple
clutches each year. Despite the smaller body size of females in
Kansas, they produce a significantly larger size-adjusted
clutch than Texas females, with a potentially shorter lifespan
and reduced likelihood of multiple clutch production. It
remains to be tested as to what the latitudinal extent is for
this ostensibly advantageous life-history response in P.
cornutum. Furthermore, ectotherms are sensitive to climatic
changes (Paaijmans et al., 2013), and modeling the response
of life-history traits of Texas Horned Lizards to climate
fluctuations and then to future climate predictions would
provide an understanding as to how they might respond
reproductively to climate change (e.g., McCallum et al., 2009;
McCallum, 2010; Milanovich et al., 2010).
We wonder what environmental context may have driven
the greater fecundity in the northern edge of its range,
despite the likelihood that this species at that latitude may be
subjected to reduced opportunities for not only foraging on a
bereft prey base and but also for photosynthesizing vitamin
D
3
, a hormone necessary for calcium metabolism (Holick,
2010). Whitford and Bryant (1979) demonstrated that Texas
Horned Lizards are food limited and exhibit a pronounced
preference for harvester ants (Pogonomyrmex spp.). Although
the extent of this dietary dependence is subject to variation
(e.g., Milne and Milne, 1950; Ramakrishnan et al., 2018),
harvester ant densities and colony sizes are geographically
variable (MacMahon et al., 2000; Warburg et al., 2017). In
fact, harvester ant densities were found to be far greater at a
central Texas site compared to a southern New Mexico site
(Whiting et al., 1993). Also, the diversity of New World ants
follows a latitudinal cline with fewer species at higher
latitudes (Kaspari et al., 2003), and this pattern is consistent
for seed-eating ants across the American Southwest (David-
son, 1977). At northern latitudes, lizard life histories are
sensitive to not only reductions in nutritional energy intake
but also the availability of suitable thermal microclimates
(Ballinger, 1977). The endogenous production of vitamin D
3
by exposure to ultraviolet-B (290–315 nm) is crucial for lizard
reproduction (Ferguson et al., 1996) and is a major
component of nutritional quality for P. cornutum (Ferguson
et al., 2015). Consequently, the response in fecundity we
observed in northern populations of the Texas Horned Lizard
persists despite potentially compromised opportunities for
vitamin D
3
production from reduced aboveground activity
and presumed reductions in the extent of its preferred prey
base, perhaps both diversity and abundance.
We are particularly intrigued by the increase in clutch size
apart from female body size in northern Kansas because these
populations seem to persist at a distinct advantage with
respect to fecundity. The most likely explanation is that the
observed increase in clutch size is in fact an adaptive
response from positive fecundity selection for P. cornutum at
744 Copeia 107, No. 4, 2019

northern latitudes (Pincheira-Donoso and Hunt, 2017). The
colder climate of northern Kansas, relative to the climate
experienced by individuals in the more southern popula-
tions, is likely the major environmental selection pressure
driving smaller body sizes in both sexes, yet the strength of
positive fecundity selection may have acted as a counterbal-
ance to what should have been be a diminishing capacity to
reproduce when the Texas Horned Lizard expanded its range
northward. If the larger size-adjusted clutches are indeed an
adaptive response to occupying northern latitudes, then
selection for larger eggs, which might be expected as a
reproductive response to smaller body sizes (e.g., Bidgood,
1974), must have been less favorable for increasing offspring
survival under the particular selective pressures of Kansas. It
seems that larger clutch sizes outweighed the benefits of
larger offspring (i.e., via larger eggs) in the evolution of the
life-history strategy of the Texas Horned Lizard at northern
latitudes. With these factors in mind, we suggest that
common-garden experiments where P. cornutum from differ-
ent latitudes are reared under controlled conditions (e.g.,
Berven and Gill, 1983) would be a worthwhile endeavor to
identify whether these life-history patterns in fecundity are
innate to populations or an example of reproductive
plasticity in response to environmental variation.
ACKNOWLEDGMENTS
First and foremost, we dedicate this work to Royce E.
Ballinger and the late C. Wayne Howard in recognition for
their excellent contributions to this topic over 40 years ago.
This study would not have been possible without the
personal and professional kindnesses of Curtis J. Schmidt,
Curator of vertebrate collections of the Sternberg Museum,
and we extend our most sincere gratitude to him. We thank
Stephen P. Rogers and Jennifer A. Sheridan (Carnegie
Museum of Natural History), and Daniel B. Wylie and
Christopher A. Phillips (University of Illinois Museum of
Natural History) for access to specimens.
LITERATURE CITED
Adolph, S. C. 1990. Influence of behavioral thermoregula-
tion on microhabitat use by two Sceloporus lizards. Ecology
71:315–327.
Adolph, S. C., and W. P. Porter. 1993. Temperature, activity,
and lizard life histories. The American Naturalist 142:273–
295.
Adolph, S. C., and W. P. Porter. 1996. Growth, seasonality,
and lizard life histories: age and size at maturity. Oikos 77:
267–278.
Angilletta, Jr., M. J., P. H. Niewiarowski, A. E. Dunham, A.
D. Leache
´, and W. P. Porter. 2004. Bergmann’s clines in
ectotherms: illustrating a life-history perspective with
sceloporine lizards. The American Naturalist 164:E168–
E183.
Ashmole, N. P. 1963. The regulation of numbers of tropical
oceanic birds. Ibis 103b:458–473.
Ashton, K. G., and C. R. Feldman. 2003. Bergmann’s rule in
nonavian reptiles: turtles follow it, lizards and snakes
reverse it. Evolution 57:1151–1163.
Atkinson, D. 1994. Temperature and organism size: a
biological law for ectotherms? Advances in Ecological
Research 25:1–58.
Ballinger, R. E. 1974. Reproduction of the Texas horned
lizard, Phrynosoma cornutum. Herpetologica 30:321–327.
Ballinger, R. E. 1977. Reproductive strategies: food availabil-
ity as a source of proximal variation in a lizard. Ecology 58:
628–635.
Berven, K. A., and D. E. Gill. 1983. Interpreting geographic
variation in life-history traits. American Zoologist 23:85–
97.
Bidgood, B. F. 1974. Reproductive potential of two lake
whitefish (Coregonus clupeaformis) populations. Journal of
the Fisheries Board of Canada 31:1631–1639.
Blouin-Demers, G., K. A. Prior, and P. J. Weatherhead.
2002. Comparative demography of black rat snakes (Elaphe
obsoleta) in Ontario and Maryland. Journal of Zoology 256:
1–10.
Burrow, A. L. 2000. The effect of prescribed burning and
grazing on the threatened Texas horned lizard (Phrynosoma
cornutum) in the western Rio Grande plains. Unpubl. M.S.
thesis, Oklahoma State University, Stillwater, Oklahoma.
Davidson, D. W. 1977. Species diversity and community
organization in desert seed-eating ants. Ecology 58:711–
724.
Endriss, D. A., E. C. Hellgren, S. F. Fox, and R. W. Moody.
2007. Demography of an urban population of the Texas
horned lizard (Phrynosoma cornutum) in central Oklahoma.
Herpetologica 63:320–331.
Ferguson, G. W., W. H. Gehrmann, A. M. Brinker, G. C.
Kroh, and D. C. Ruthven III. 2015. Natural ultraviolet-b
exposure of the Texas horned lizard (Phrynosoma cornutum)
at a North Texas Wildlife Refuge. The Southwestern
Naturalist 60:231–239.
Ferguson, G. W., J. R. Jones, W. H. Gehrmann, S. H.
Hammack, L. G. Talent, R. D. Hudson, E. S. Dierenfeld,
M. P. Fitzpatrick, F. L. Frye, M. F. Holick, and T. C. Chen.
1996. Indoor husbandry of the panther chameleon
Chamaeleo [Furcifer]pardalis: effects of dietary vitamins A
and D and ultraviolet irradiation on pathology and life-
history traits. Zoo Biology 15:279–299.
Fitch, H. S. 1985. Variation in clutch and litter size in New
World reptiles. University of Kansas Museum of Natural
History Miscellaneous Publications 76:1–76.
Ford, N. B., and R. A. Seigel. 1989. Relationships among
body size, clutch size, and egg size in three species of
oviparous snakes. Herpetologica 45:75–83.
Forsman,A.,andR.Shine.1995. Parallel geographic
variation in body shape and reproductive life history
within the Australian scincid lizard Lampropholis delicata.
Functional Ecology 9:818–828.
Givler, J. P. 1922. Notes on the oecology and life-history of
the Texas horned lizard, Phrynosoma cornutum. Journal of
the Elisha Mitchell Scientific Society 37:130–137.
Hammerson, G. A. 2007. Phrynosoma cornutum. The IUCN
Red List of Threatened Species 2007:e.T64072A12741535.
https://www.iucnredlist.org/details/64072/0 (accessed 4
October 2017).
Holick, M. F. (Ed.). 2010. Vitamin D: Physiology, Molecular
Biology, and Clinical Application. Second edition. Humana
Press, New York.
Howard, C. W. 1974. Comparative reproductive ecology of
horned lizards (genus Phrynosoma) in southwestern United
States and northern Mexico. Journal of the Arizona
Academy of Science 9:108–116.
Huey, R. B. 1982. Temperature, physiology, and the ecology
of reptiles, p. 25–91. In: Biology of the Reptilia. Volume 12.
C. Gans and F. H. Pough (eds.). Academic Press, London.
Hughes et al.—Reproductive variation in the Texas Horned Lizard 745

Huey, R. B., and D. Berrigan. 2001. Temperature, demogra-
phy, and ectotherm fitness. The American Naturalist 158:
204–210.
Kaspari, M., M. Yuan, and L. Alonso. 2003. Spatial grain
and the causes of regional diversity gradients in ants. The
American Naturalist 161:459–477.
King, R. B. 2000. Analyzing the relationship between clutch
size and female body size in reptiles. Journal of Herpetol-
ogy 34:148–150.
Lack, D. 1954. The Natural Regulation of Animal Numbers.
Clarendon Press, Oxford, UK.
Laugen, A. T., A. Laurila, K. Ra
¨sa
¨nen, and J. Merila
¨.2003.
Latitudinal countergradient variation in the common frog
(Rana temporaria) development rates—evidence for local
adaptation. Journal of Evolutionary Biology 16:996–1005.
Lovich, J. E., and J. W. Gibbons. 1992. A review of
techniques for quantifying sexual size dimorphism.
Growth, Development and Aging 56:269–281.
MacMahon, J. A., J. F. Mull, and T. O. Crist. 2000. Harvester
ants (Pogonomyrmex spp.): their community and ecosystem
influences. Annual Review of Ecology and Systematics 31:
265–291.
Mathies, T., and R. M. Andrews. 1995. Thermal and
reproductive biology of high and low elevation popula-
tions of the lizard Sceloporus scalaris: implications for the
evolution of viviparity. Oecologia 104:101–111.
Mayhew, W. W. 1963. Reproduction in the granite spiny
lizard, Sceloporus orcutti. Copeia 1963:144–152.
McCallum,M.L.2010. Future climate change spells
catastrophe for Blanchard’s cricket frog, Acris blanchardi
(Amphibia: Anura: Hylidae). Acta Herpetologica 5:119–
130.
McCallum, M. L., J. L. McCallum, and S. E. Trauth. 2009.
Predicted climate change may spark box turtle declines.
Amphibia–Reptilia 30:259–264.
McGarigal,K.,S.Cushman,andS.Stafford.2000.
Multivariate Statistics for Wildlife and Ecology Research.
Springer, New York.
Meshaka, W. E., Jr., and J. N. Layne. 2015. The Herpetology
of southern Florida. Herpetological Conservation and
Biology 10 (Monograph 5):1–353.
Milanovich, J. R., W. E. Peterman, N. P. Nibbelink, and J.
C. Maerz. 2010. Projected loss of a salamander diversity
hotspot as a consequence of projected global climate
change. PLoS ONE 5:e12189.
Milne, L. J., and M. J. Milne. 1950. Notes on the behavior of
horned toads. The American Midland Naturalist 44:720–
741.
Moeller, B. A., E. C. Hellgren, D. C. Ruthven III, R. T.
Kazmaier, and D. R. Synatzske. 2005. Temporal differ-
ences in activity patterns of male and female Texas horned
lizards (Phrynosoma cornutum) in southern Texas. Journal of
Herpetology 39:336–339.
Montgomery, C. E., and S. P. Mackessy. 2003. Natural
history of the Texas horned lizard, Phrynosoma cornutum
(Phrynosomatidae), in southeastern Colorado. The South-
western Naturalist 48:111–118.
Montgomery, C. E., S. P. Mackessy, and J. C. Moore. 2003.
Body size variation in the Texas horned lizard, Phrynosoma
cornutum, from central Mexico to Colorado. Journal of
Herpetology 37:550–553.
Moreau, R. E. 1944. Clutch-size: a comparative study, with
special reference to African birds. Ibis 86:286–347.
Olalla-Ta
´rraga, M. A
´., M. A
´. Rodr´
ıguez, and B. A. Hawkins.
2006. Broad-scale patterns of body size in squamate reptiles
of Europe and North America. Journal of Biogeography 33:
781–793.
Paaijmans, K. P., R. L. Heinig, R. A. Seliga, J. I. Blanford, S.
Blanford, C. C. Murdock, and M. B. Thomas. 2013.
Temperature variation makes ectotherms more sensitive to
climate change. Global Change Biology 19:2373–2380.
Packard, G. C., and T. J. Boardman. 1999. The use of
percentages and size-specific indices to normalize physio-
logical data for variation in body size: wasted time, wasted
effort? Comparative Biochemistry and Physiology Part A:
Molecular Integrative Physiology 122:37–44.
Parker, W. S., and E. R. Pianka. 1975. Comparative ecology
of populations of the lizard Uta stansburiana. Copeia 1975:
615–632.
Pianka, E. R. 1970. On r- and K-selection. The American
Naturalist 104:592–597.
Pianka, E. R., and W. S. Parker. 1975. Ecology of horned
lizards: a review with special reference to Phrynosoma
platyrhinos. Copeia 1975:141–162.
Pincheira-Donoso, D., D. J. Hodgson, and T. Tregenza.
2008. The evolution of body size under environmental
gradients in ectotherms: Why should Bergmann’s rule
apply to lizards? BMC Evolutionary Biology 8:68.
Pincheira-Donoso, D., and J. Hunt. 2017. Fecundity
selection theory: concepts and evidence. Biological Re-
views 92:341–356.
Price, A. H. 1990. Phrynosoma cornutum.Catalogueof
American Amphibians and Reptiles 469:1–7.
Ramakrishnan, S., A. J. Wolf, E. C. Hellgren, R. W. Moody,
and V. Bogosian, III. 2018. Diet selection by a lizard ant-
specialist in an urban system bereft of preferred prey.
Journal of Herpetology 52:79–85.
Shine, R. 1992. ‘‘Costs’’ of reproduction in reptiles. Oecolo-
gia 46:92–100.
Sinervo, B. 1990. The evolution of maternal investment in
lizards: an experimental and comparative analysis of egg
size and its effects on offspring performance. Evolution 44:
279–294.
Sperry, J. H., G. Blouin-Demers, G. L. Carfagno, and P. J.
Weatherhead. 2010. Latitudinal variation in seasonal
activity and mortality in ratsnakes (Elaphe obsoleta).
Ecology 91:1860–1866.
Stewart, J. R. 1979. The balance between number and size of
young in the live-bearing lizard Gerrhonotus coeruleus.
Herpetologica 35:342–350.
Tinkle, D. W., H. M. Wilbur, and S. G. Tilley. 1970.
Evolutionary strategies in lizard reproduction. Evolution
24:55–74.
Tsuji, J. S. 1988. Thermal acclimation of metabolism in
Sceloporus lizards from different latitudes. Physiological
Zoology 61:241–253.
Van Noordwijk, A. J., and G. de Jong. 1986. Acquisition and
allocation of resources: their influence on variation in life
history tactics. The American Naturalist 128:137–142.
Vitt, L. J. 1977. Observations on clutch and egg size and
evidence for multiple clutches in some lizards of south-
western United States. Herpetologica 33:333–338.
Wapstra, E., R. Swain, and J. M. O’Reilly. 2001. Geographic
variation in age and size at maturity in a small Australian
viviparous skink. Copeia 2001:646–655.
Warburg, I., W. G. Whitford, and Y. Steinberger. 2017.
Colony size and foraging strategies in desert seed harvester
ants. Journal of Arid Environments 145:18–23.
Whitford, W. G., and M. Bryant. 1979. Behavior of a
predator and its prey: the horned lizard (Phrynosoma
746 Copeia 107, No. 4, 2019

cornutum) and harvester ants (Pogonomyrmex spp.). Ecology
60:686–694.
Whiting, M. J., J. R. Dixon, and R. C. Murray. 1993. Spatial
distribution of a population of Texas horned lizards
(Phrynosoma cornutum: Phrynosomatidae) relative to hab-
itat and prey. The Southwestern Naturalist 38:150–154.
Wolf, A. J. 2012. Spatial and demographic ecology of Texas
horned lizards within a conservation framework. Unpubl.
M.S. thesis, Southern Illinois University at Carbondale,
Carbondale, Illinois.
Wolf, A. J., E. C. Hellgren, E. M. Schauber, V. Bogosian III,
R. T. Kazmaier, D. C. Ruthven III, and R. W. Moody. 2014.
Variation in vital-rate sensitivity between populations of
Texas horned lizards. Population Ecology 56:619–631.
Zamudio, K. R. 1998. The evolution of female-biased sexual
size dimorphism: a population-level comparative study in
horned lizards (Phrynosoma). Evolution 52:1821–1833.
Hughes et al.—Reproductive variation in the Texas Horned Lizard 747