ARTICLE IN PRESS 
Journal of Arid Environments xxx (2009) 1-9 
ELSEVIER 
Contents lists available at ScienceDirect 
Journal of Arid Environments 
journal homepage: www.elsevier.com/locate/jaridenv 
Longevity and growth strategies of tiie desert tortoise {Gopherus agassizii) 
in two American deserts 
AJ. Curtin^*, G.R. Zug^ J.R. Spotila^ 
'Department of Bioscience and Biotechnology, Drexel University, 3J4J Chestnut Street Philadelphia, PA 19104, USA 
^'Department of Vertebrate Zoology, National Museum of Natural History, Smithsonian Institution, PO Box 37012, Washington DC 20013-7012, USA 
ARTICLE    INFO 
Article history: 
Received 28 May 2008 
Received in revised form 
2 September 2008 
Accepted 27 November 2008 
Available online xxx 
Keywords: 
Age 
Life history 
Sexual maturity 
Skeletochronology 
Southwestern United States 
Testudines 
ABSTRACT 
The desert tortoise occurs in two strikingly different desert regimes in the southwestern United States. In 
the Mojave Desert, rainfall is more irregular and resources are more limited than in the Sonoran Desert. 
We examined the age structure of tortoise populations from these two deserts to determine whether the 
difference in resource availability has driven an evolutionary divergence in life history strategies. Age and 
growth rates strongly reflect the ecological adaptation of the two populations. The oldest Sonoran males 
reached 54 years, compared to only 43 years in females. The oldest West Mojave (WM) males reached 
56 years, compared to only 27 years in females. WM tortoises grew faster than Sonoran ones, and 
females reached sexual maturity at earlier ages (~ 17-19 years) than Sonoran females (~ 22-26 years). 
These traits and the higher rate of clutch production in the WM population are likely the evolutionary 
adaptation for low juvenile survivorship and a significantly shorter life span. Frequent droughts in the 
WM Desert and the lowest annual rainfall area within the range of the desert tortoise cause chronic 
physiological stress, likely annually, and are proposed as a major selection force producing contrasting 
life-history strategies. 
? 2008 Elsevier Ltd. All rights reserved. 
1. Introduction 
Tortoises are portrayed regularly as animals of exceedingly long 
lives. How long an animal lives fascinates scientists and laymen 
alil^e; however, for biologists, an animal's longevity has evolu- 
tionary importance in the context of an individual's reproductive 
potential and its role in population dynamics. Longevity is critical 
relative to an animal's reproductive life span, that is, the duration of 
the reproductive interval from first reproduction (attainment of 
sexual maturity) to the last reproductive act. For tortoises, an 
individual's reproductive longevity can extend for decades because 
evidence suggests that turtles lack reproductive senescence 
(Girondot and Garcia, 1999). 
Some individual tortoises do live more than 100 years (Cham- 
bers, 2004), but these individuals are captive animals living in 
protected environments. Is longevity as great in wild and particu- 
larly in stressful environments? Desert tortoises, Gopherus agassizii. 
* Corresponding author, present address: Academic Resource Center, Duke 
University, Box 90694. Durham, NC 27708, USA.. Tel.: +1 919 684 4980: fax: +1 919 
684 8934. 
E-mail addresses: amanda.curtin@duke.edu (A.J. Curtin), zugg@si.edu (G.R. Zug), 
spotiljr@drexel.edu (J.R. Spotila). 
0140-1963/$ - see front matter ? 2008 Elsevier Ltd. All rights reserved. 
doi:10.1016/j.jaridenv.2008.11.011 
of the American Southwest are ideal for examining this question. 
Germano (1992) examined this question by estimating age of four 
regional populations of desert tortoise with the scute-annulus 
technique. His sample of 574 tortoises encompassed the entire 
range of the desert tortoise and allowed him to estimate longevity 
for the four major biomes (western Mojave Desert, eastern Mojave 
Desert, Sinaloan thorn-scrub, and Sonoran Desert). With a single 
exception, he discovered only one individual (eastern Mojave) 
older than 40 years. In his cautious reporting, he defined longevity 
as the minimum age of the oldest individuals: 32 years, western 
Mojave; 35 years, Sonoran; none, Sinaloan scrub. He further noted 
that the relative portion of adults over 25 years varied greatly: 11% 
eastern Mojave; 5% western Mojave, 29% Sonoran. These data 
reflect differences in life span, resulting in differences in age 
structure of the population. He provided estimates of age of 
maturity subsequently (Germano, 1994a), although not explicitly 
but sufficient to allow estimates of reproductive life span (i.e., 
longevity minus maturity): 25 years, 15 years, and 20 years, 
respectively. Thus, both longevity and reproductive life span for 
these populations suggest life-history adaptations, possibly driven 
by differences in the climatic regime of the three desert areas. 
Genetic history might also play a role in the reported differences. 
Using restriction site analysis, Lamb et al. (1989) resolved five 
different mtDNA genotypes  among 22  populations  of desert 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts, Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
A.J. Curtin et al. /Journal of Arid Environments xxx (2009) 1-9 
tortoises and found three well-defined genetic assemblages: 
a Mojave assemblage, Sonoran assemblage and Sinaloan assem- 
blage. The genetic distances (5.1-5.6%) observed between Mojave 
and Sonoran genotypes of desert tortoises are significantly higher 
than distance values reported for any other turtles species (Lamb 
and McLuckie, 2002; Walker and Avise, 1998). Consequently, 
depending on molecular clock timing, the Mojave and Sonoran 
mtDNA lineages appear to have diverged some 5 or 6 million years 
ago (Lamb and Lydeard, 1994; Lamb and McLuckie, 2002). 
Regardless of the timing of the divergence of the Mojave and 
Sonoran Desert populations, the genetic distance (evolutionary 
history) between these two populations is great (McCord, 2002). 
The Mojave Desert is the youngest biotic province in North America 
(Van Devender, 2002), and the Mojave population is younger 
evolutionarily than Sonoran tortoises. Moreover, within the Mojave 
Desert, West Mojave tortoises are potentially younger than their 
eastern cousins (Morafka and Berry, 2002). We, therefore, wished 
to examine age and growth strategies within and between West 
Mojave and Sonoran Desert tortoise populations because they 
represent the eastern and western extremes of morphology, 
behavior, and ecology of desert tortoises (Morafl<a and Berry, 2002; 
Van Devender, 2002). Additionally, the accuracy of Germano's scute 
aging protocol has been questioned (Wilson et al., 2003), so an 
independent aging technique (skeletochronology) offers an 
opportunity to test Germano's results and interpretation. 
2. Materials and methods 
2.J. Materials 
Our sample consisted of salvaged carcasses of 72 free-living 
desert tortoises (24 juveniles, 28 adult females, 17 adult males, 3 
sex unknown) from the Sonoran Desert (Maricopa, Mohave, 
Yavapai, Pinal and Pima Co., Arizona) and 69 carcasses (13 adult 
females, 30 adult males, 22 juveniles, 4 sex unknown) from the 
West Mojave Desert (San Bernardino Co., California); hereafter, 
Mojave equals West Mojave (in Sections 2 and 3). Identification 
of sex and maturity was made by Arizona Game and Fish 
Department staff when they collected the carcasses between 
1990 and 1997. Similarly, these data were assigned by Californian 
wildlife staff when salvaging carcasses. The West Mojave Desert 
tortoise samples were collected between 1991 and 1995 near US 
highway 58 as part of a long term study on the effectiveness of 
fences and culverts for protecting desert tortoises along highways 
(see Boarman, 1992; Sazaki et al., 1995). Mojave tortoises are 
Usted as threatened (US Fish and Wildlife Service, 1994) due to 
downward trends in population size (especially in the western 
populations). The use of salvaged carcasses of wild (free-living) 
individuals allows us to estimate natural age by skeletochronol- 
ogy without disturbing or harming any individuals of this 
threatened species. 
2.2. Skeletochronology 
Tortoises, like other ectothermic tetrapods, display a cyclical 
pattern of growth. Cyclic skeletal growth produces successive bony 
layers in association with an internal (genetically based) rhythm 
synchronized and reinforced by seasonal cycles (Castanet et al., 
1993; Meunier et al., 1979), hence the number of skeletal growth 
layers yields an estimate of an individual's age. Unlike tree rings, 
the bony growth layers do not persist throughout a turtle's life. As 
the turtle grows, endosteal resorption at the medullary cavity 
margin removes the earliest growth layers. A simple count of the 
visible growth layers, thus, yields fewer layers than were actually 
deposited and an underestimate of the actual number of growth 
cycles of an individual. We assume that each bony growth layer 
(cycle) equals one year and, elsewhere (Curtin et al., 2008) we 
demonstrated the reliability of this assumption with known-aged 
desert tortoises. 
Skeletochronology in turtles must incorporate resorption in the 
estimation of each individual's age. Either the use of known-age 
individuals or a back-calculation method (Castanet and Smirina, 
1990) allows a determination of the rate of periosteal bone 
resorption. In an initial study, we sectioned all appendicular skel- 
etal elements, together with vertebrae and shell elements (gular, 
marginal, neural and plastral elements) to determine which 
elements showed the greatest number of growth marks (GMs). The 
humerus, femur, ilium, and scapula contain the most "resident" 
GMs and were the bones we used for our analyses. Additionally, we 
recorded longitudinal length (LL), proximal width (PW), midshaft 
width (MW) and distal width (DW) of each bone used for age 
estimation, and if the carcass was intact, we recorded also carapace 
length (CL), carapace width (CW) and plastron length (PL). 
We fixed, decalcified, and processed bones by standard histo- 
logical protocol. All bones were embedded in Paraplast Plus?, cut in 
20 \im cross sections through the midshaft area of each bone using 
a rotary microtome, and stained with hematoxylin and eosin. We 
measured all growth layer diameters, outside diameters, and 
resorption cores along the longest axis of each element with a stage 
micrometer and transmitting light microscope, and counted the 
number of entire and partial growth layers. 
Comparisons of skeletochronology age estimates (determined 
by two protocols) to a sample of known-aged desert tortoises from 
Rock Valley, Nevada (Curtin et al., 2008) indicated that the 
correction factor method (CF) provided the most accurate juvenile 
age estimates and the growth-layer ranking protocol (Rank) the 
most accurate for estimating the age of adults. Both sets of age 
estimates are examined below. 
Both protocols rest on the basic assumption that each successive 
growth layer represents one year in the life of a desert tortoise. The 
ranking protocol (Zug, 1990) uses three data items (resorption core 
diameter, the set of sequential diameters of the complete growth 
layers (GLs), external diameter of bone at death) to obtain an age 
estimate of each individual. The first step is to order individuals by 
the increasing diameters of their individual resorption cores. The 
rows created by this ordering consist of sequential growth layer 
diameters from smallest complete layer to the outside diameter. 
These sequential GL diameters yield columns of successive 
increasing diameters, and the innermost diameter for each turtle is 
assigned to the column to which its diameter most closely matches 
those diameters of preceding turtles. The assignment of innermost 
diameter fixes the sequential assignment of column position of all 
succeeding diameters; no column is skipped in the sequence of 
diameters. This ranking protocol yields a table with a stepwise 
appearance due to the increasing size of the innermost diameter of 
successive individuals. Each column represents one year, and the 
position of the outer diameter denotes the age at death of an 
individual. 
In the correction factor protocol (Parham and Zug, 1998), an 
estimate of a tortoise's age derives from the number of growth 
layers (diameters) observed in the outer region of the bone section 
plus an estimate of the number of growth layers lost by resorption 
as the center of bone is remodeled. The latter, unobservable 
component is estimated as (D - DH)/CF, where D is the diameter of 
the resorption core, DH is the diameter of a hatchling's bone (before 
the beginning of increment formation), and CF is the correction 
factor. The correction factor is a constant "ageing rate" (years 
mm 1) assumed to apply to the resorption core and calculated as 
the mean slope of the successive bone-growth diameters in 
a sample of the smaller individual turtles. 
Please cite this article in press as: Curtin, A.J., et al., Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts. Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
A.]. Curtin et al. /journal of Arid Environments xxx (2009) 1-9 
2.3. Statistical analysis 
Because the Arizona sample consisted of multiple populations, 
we compared all size measurements and age estimations to 
determine whether differences existed among the populations. We 
used one-way analysis of variance (ANOVA) tests with a Tukey post 
hoc test and analysis of covariance (ANCOVA) to examine differ- 
ences within and between samples of the two desert populations. 
We conducted least squares regression analyses and ANOVAs to 
compare shell and bone measurements between the Sonoran and 
West Mojave population samples. We compared the various shell 
and bone measurements and age estimates from within and 
between both groups to discern sexual dimorphism using ANOVA. 
All statistical analyses were performed using the SYSTAT (Version 
9.0) statistical package; we accepted statistical significance as 
p < 0.05. 
3. Results 
3.1. Body size 
A comparison of carapace length (CL), carapace width (CW) and 
plastron length (PL) between the various Sonoran Desert pop- 
ulations shows no significant differences between localities for any 
group (adult males, adult females and juveniles). In both the Sonoran 
and Mojave sample, these three metrics are strongly correlated, for 
example, Sonoran female CL and CW r^ = 0.74 (n = 28), male CL and 
PL r^ = 0.90 (n = 17). Sonoran adult females and males have no 
dimorphism in the three shell metrics (Fig. 1; CL: r^ = 0.06, 
fi,33 = 1.95, p = 0.17; CW: r^ = 0.76, F2,28 = 0.23, p = 0.64; PL: 
r^ = 0.90, F2.23 = 0.63, p = 0.44). In contrast, the Mojave population 
shows a sexual dimorphism in CL (r^ = 0.20, Fi 33 = 8.24, p = 0.007; 
Fig. 1), but not in CW (r2 = 0.90, F2,23 = 0.56, p = 0.46) or PL 
(r^ = 0.92, F2.22 = 0.10, p = 0.76) when the effect of size is removed. 
Sonoran females are significantly larger than Mojave females in 
CL (means 232.12 ? 19.50, 214.50 ? 20.70 mm CL, respectively; 
r^ = 0.16, Fi 33 = 6.46, p = 0.02; Fig. 1). Mojave females, however, are 
significantly wider than Sonoran females (r^ = 0.73, F 2,25 = 6.05, 
p = 0.02; Fig. 1). There is no difference in PL between Sonoran and 
Mojave females {r^ = 0.79. F2.26 = 0.02, p = 0.88). Mojave and 
Sonoran males have similar body length (means 244.69 ? 35.28, 
243.43 ? 28.29 mm CL, respectively; r^ = 0.01, Fi,33 = 0.01, p = 0.19; 
PL F131 =3.46, p = 0.07). As in females, Mojave males are signifi- 
cantly wider than Sonoran males (r^ = 0.89, F2,26 = 11.26, p = 0.002; 
Fig. 1). There is no significant difference in male PL between the two 
areas (r^ = 0.88, F2.26 = 3.64, p = 0.07). Juveniles have similar shell 
metrics between the two deserts, lacking significance in all tests (CL: 
r^ = 0.01, Fi,42 = 0.44, p = 0.51; CW: r^ = 0.97, F2.32 = 0.09, p = 0.77; 
PL: r^ = 0.99, F2.30 = 0.56, p = 0.46). 
3.2. Resorption core diameters 
Resorption core diameters (RCDs) are crucial for the estimation 
of an individual's age. In Sonoran tortoises, the RCDs for adult males, 
females or juveniles at all localities are nearly identical. There is no 
significant difference in RCDs, irrespective of bone, between 
Sonoran males and females (humerus: r^ = 0.06, Fi 21 = 1.43, 
p = 0.24; ilium: r^ = 0.05, Fi,33 = 1.76, p = 0.19; scapula: P = 0.02, 
Fiio = 0.19, p = 0.67). Similarly, Mojave males and females show no 
sexual dimorphism in RCDs (humerus: r^ = 0.06, Fij = 0.42, 
p = 0.54; ilium: r^ = 0.04, Fi,25 = 0.94, p = 0.34; scapula: r^ = 0.22, 
fi.i3 = 2.74, p = 0.13). Comparison between the two deserts reveals 
that Sonoran females have significantly larger ilia RCDs (Fig. 2A) 
than Mojave females (r^ = 0.58, F2.19 = 14.46, p = 0.001). Sonoran 
males have significantly larger humeri (r^ = 0.66, F2,8 = 8.14, 
p = 0.02) and ilia RCDs (r^ = 0.20, F2.29 = 6.41,p = 0.02) than Mojave 
males (Fig. 3B). The juveniles from the two desert populations show 
no significant difference in RCDs (humerus: r^ = 0.12, Fi.26 = 3.37, 
p = 0.08; ilium: r^ = 0.02, Fi,2i = 0.33, p = 0.57; scapula: r^ = 0.09, 
Fi.i5 = 1.46, p = 0.25). 
We combined adults and juveniles to gain insight into pop- 
ulation trends. All three bones in both deserts display a close 
relationship between body size and RCDs (Table 1). 
3.3. Age and longevity 
The ilium and humerus consistently have the most periosteal 
growth   layers   and   usually   yield   similar   age   estimates.   In 
275 
S   225 
s 
I   200 
125 
{     t 
*     " 
Adult Sonoran Desert tortoises Adult West Mojave Desert Tortoises 
Carapace Carapace Plastron Carapace Carapace Plastron 
length width length length width length 
Fig. 1. Mean shell measurements of adult male and female Copherus agassizii from the Sonoran Desert, Arizona and the West Mojave Desert, California. Error bars represent 95% 
confidence intervals; m, males; /, females. 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts. Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
A.J. Curtin et al. /Journal of Arid Environments xxx (2009) 1-9 
4.5 
4 
B  3.5 
t   2.5 
1.5 
0.5- 
4.5- 
^     4 
r 3.5- 
? 
.2     3 - 
?c 
g  2.5 
B O 
Q     2 ? u 
o 
4^ 
t   1.5 
0.5- 
0 
SD WM SD WM 
B 
SD WM 
Hiimerus Ilium Scapula 
Fig. 2. Mean resorption core diameters for adult (A) female and (B) male Coplierus agassizii from the Sonoran Desert (SD), Arizona and the West Mojave (WM) Desert, California. 
Error bars represent 95% confidence intervals. 
juveniles, Rank and CF give similar age estimates (? 1 year) 
for the Sonoran (30-120 mm CL; r^ = 0.01, F1.34 = 0.56, p = 0.46) 
and Mojave (40-110 mm CL; r^ = 0.003, Fi.2i=0.07, p = 0.70) 
samples. Beyond this body size, variation between the two 
methods increases significantly, especially beyond the reported 
minimum size at sexual maturity (around 180 mm CL in both 
Sonoran and Mojave tortoises (Medica et al., 1975; Germano, 
1994b). The Rank age estimates are the most accurate ones for 
adults, i.e., >180 mm CL, based on the known-age validation 
study (Curtin et al., 2008). 
In adults, regional variation is considerable in the relationship 
between age and size (Fig. 3). The oldest Sonoran males, estimated 
at 47-54 years old, range from 241 to 266 mm CL. The oldest 
Sonoran females, 42-43 years, are 223-239 mm CL. The oldest 
Mojave male is 56 years old and 262 mm CL. The second oldest 
male is significantly larger (280 mm CL) but only 36 years. The 
oldest Mojave female is only 27 years, yet at 235 mm CL, is at the 
upper end of the size range of the oldest Sonoran females. A much 
smaller female (198 mm CL) is estimated as 26 years. 
In the Sonoran sample, the age estimates of the different 
localities are similar for adult males, females, and juveniles. The 
absence of statistical difference in adult males and females from 
different localities, allow the combination of each sex into a single 
Sonoran sample and to test for dimorphism in the average age of 
the two sexes. No dimorphism exists within the Sonoran adults 
(r^ = 0.39, F2.33 = 1.25, p = 0.27; Fig. 4). Mojave adults derive from 
the same general area, and they similarly display no dimorphism of 
female-male age estimates (r^ = 0.33, F2.30 = 0.42, p = 0.52; Fig. 4). 
Sonoran tortoises reach significantly older ages than Mojave 
tortoises (r^ = 0.55, Fa.ee = 38.56, p < 0.001; Fig. 4). Sonoran 
females are significantly older ages than Mojave females, when the 
effect of size is removed (means 32.82 it 1.4, 20.05 it 1.86 years, 
respectively; r^ = 0.61, F2.30 = 22.38, p < 0.001; Fig. 4). Similarly, 
Sonoran males reach significantly older ages than Mojave males 
with size as the covariate (means 37.03 ? 2.47, 24.41 it 1.77 years, 
respectively; r^ = 0.53, F2.33 = 18.26, p < 0.001; Fig. 4). 
Mojave tortoises grow faster than Sonoran tortoises and attain 
adult size (>180mm CL) at younger ages (Figs. 3 and 4). Hence, 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts, Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
Aj. Curtin et al. /journal of Arid Environments xxx (2009) 1-9 
a. 
U 
? Sonoran Desert tortoises /? 
350 0 West Mojave tortoises 
^ 
/ 
^ / ^^ 
JOU 
? %     ^-i r^ ^  ^  ? 
250 
? 
? 
o 
200 
O         ^-^0%^^ 
a   -^?     5U^ 
150 
100 
? ? 
50 r ? 
0 
20 30 -10 
Age (years) 
50 60 
Fig. 3. Relationship of sl<eletochironology age estimates to carapace lengtli for 
Go^herus agasiizii from the Sonoran Desert, Arizona (solid circles) and the West Mojave 
Desert. California (open circles). 
Sonoran tortoises of similar sizes to Mojave ones are older in age 
and are also older at death (Fig. 3). A comparison of the size/age 
relationships between Sonoran and Mojave populations, where the 
effect of age is removed, shows that Mojave tortoises grow signif- 
icantly faster than Sonoran tortoises (r^ = 0.72, F2,io5 = 16.04, 
p < 0.001). The adjusted least squares mean for Mojave sample is 
200.6 mm CL (? 6.0 SE) and for Sonoran tortoises 168.2 mm CL 
(zb 5.8 SE) at the mean age of 21.8 years. 
Males not only live longer but are also larger at similar ages than 
females. There is a significant relationship between age and size for 
all adults (Sonoran females: r^ = 0.23, Fi 20 = 6.00, p = 0.02; 
Sonoran males: r^ = 0.46, Fi,i2 = 10.01, p = 0.008; Mojave females: 
r^ = 0.51, Fi,9 = 9.36, p = O.oi; Mojave males: r^ = 0.25, Fi,2o = 6.74, 
p = 0.02). Comparison of CL between adults, absent the effect of 
age, shows no significant difference between the sexes for either 
Sonoran (r^ = 0.37, F233 = 0.13, p = 0.72) or Mojave tortoises 
(r^ = 0.34, F2.30 = 0.84, p = 0.37). 
The distribution of ages among the Sonoran and Mojave samples 
is highlighted by size classes. To obtain a better idea of size/age 
relationships, we compared variance in age between the two 
groups using ANOVA (Fig. 5). We excluded a 125 mm CL Mojave 
tortoise from the analysis; although classified as a juvenile based on 
size, its estimated age is 40 years and represents an extreme outlier 
in its size class (101-150 mm CL). Sonoran tortoises are significantly 
older than Mojave tortoises within the 51-100 mm CL (1^ = 0.53, 
Table 1 
Regression statistics of carapace length (CL) versus resorption core diameter (RCD) 
for Sonoran Desert and West Mojave Desert tortoises. 
CL vs RCD F P r" 
Humerus 
Sonoran Desert 92.553 <0.001 0.77 
West Mojave Desert 50.503 <0.001 0.71 
Ilium 
Sonoran Desert 135.045 <0.001 0.77 
West Mojave Desert 10.682 0.003 0.27 
Femur 
Sonoran Desert 128.824 <0.001 0.90 
West Mojave Desert 
- - - 
Scapula 
Sonoran Desert 34.421 <0.001 0.62 
West Mojave Desert 27.848 <0.001 0.70 
Fi.15 = 16.68, p = 0.001), 201-250 mm CL (r^ = 0.72, Fi,35 = 90.63, 
p < 0.001) and 251-300 (r^ = 0.25, F1.22 = 7.40, p = 0.01) mm CL 
size classes (Fig. 5). 
3.4. Age at sexual maturity 
The minimum size at sexual maturity allows us to associate our 
age estimates to age at sexual maturity. The minimum carapace 
length of a sexually mature individual is 176 mm CL for Mojave 
tortoises (Germano, 1994b) and 180 mm CL for Sonoran tortoises 
(Germano, 1994b; Averill-Murray and Averill-Murray, 2005). In the 
Sonoran sample, three individuals of 175, 180 and 195.5 mm CL, 
show ages of 26, 27 and 25 years, respectively (mean age 26 years). 
In comparison, we extrapolated age at maturity (~ 180 mm CL) 
from the age versus size regression line, and obtained an estimate 
of around 25 years (Fig. 3), similar to that obtained from the actual 
data. The Mojave (WM) sample has four individuals (168,176,177 
and 178 mm CL) around the minimum sexual maturity size and 
their estimated ages are 17, 22, 19 and 11 years (mean 17 years). 
When we extrapolated age at maturity from the age/size regression 
line, we obtained a similar estimate of around 17 years (Fig. 3). Even 
with Mojave individuals of similar size to Sonoran tortoises, the 
estimated ages of the Mojave tortoises average lower (mean 
18.7 years for 176-195 mm CL tortoises). 
4. Discussion 
4.J. Body size 
We found no significant difference in adult or juvenile body size 
or bone metrics between the various Sonoran Desert localities. Our 
small sample sizes for the different populations might prevent the 
recognition of populational size differentiation. In previous studies 
at two of the localities from which we received samples, namely 
Little Shipp Wash (LSW, Yavapai) and Eagletail Mountains (EM, 
Maricopa), those studies found sexual dimorphism of body size. 
Averill-Murray (2002) and Murray and Klug (1996) noted that 
males reached larger maximum sizes than females (LSW: 299 mm, 
267 mm CL; EM: 288, 268 mm CL; males, females, respectively). At 
the Granite Hill sites (Final County), they reported that the two 
sexes attained a similar maximum size (around 244 mm CL) and 
were significantly smaller than the preceding two populations 
(Murray and Klug, 1996). Averill-Murray (2002) found that females 
reached the same or larger CL than males at three plots south of the 
Gila River (of which we had carcasses from San Pedro Valley and 
West Silver Bell Mountains). These conflicting data on sexual 
dimorphism in the different populations suggest either sampling 
problems or populations of different age structure. Our age and size 
data cannot resolve the conflicting results. 
Our West Mojave sample shows males larger than females (CL; 
Fig. 1), matching the results of Germano (1994a). In both sexes, CL 
and bone lengths are highly correlated. Sonoran females reached 
significantly larger sizes (CL) than Mojave females, but this differ- 
ence was not observed in male body size comparisons. Whereas the 
mean CLs of our samples match those of (Germano, 1994a, Table 2) 
for West Mojave males and Sonoran females, our Mojave females 
have a strikingly smaller mean and the Sonoran males a larger 
mean. We interpret these differences as sampling error from 
smaller samples. 
The striking size dimorphism between Mojave females and 
males possibly results from females channeling all or most of their 
non-maintenance energy into reproductive effort, a physiological 
trait directly influenced by food (quantity and quality) availability 
(Peterson, 1996). Mean annual rainfall in the Arizona Upland 
subdivision of the Sonoran Desert (where most of our sample 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts. Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.n.0n 
ARTICLE IN PRESS 
A.J. Curtin et al. /Journal of Arid Environments xxx (2009) 1-9 
13 
40- 
35- T                     *' (1 
30 
L 
? 
c   20 
1 
?
 
1 
1 
?
 
1 
15 
10 
5- 
n. 
SD females SD males W'M Teniales WM males 
Fig. 4. Mean ranking protocol age estimates for adult Goplierus agassizii from the Sonoran Desert (SD), Arizona and tlie West Mojave (WM) Desert, California. Error bars represent 
95% confidence intervals. 
derived) is muchi more reliable tlian rainfall in thie Mojave Desert 
(Averill-Murray, 2002). Tlie Sonoran Desert receives about 85- 
300 mm more annual rainfall than the Mojave Desert, and the 
western portion of the latter receives even less annual rain (~ SO- 
TS mm) than the eastern portion (Wallis et al., 1999). Because 
rainfall effects the availability of food and its digestion (Oftedal, 
2002), rainfall will affect growth and reproductive effort. This 
association is demonstrated by the correlation of geographic 
variation in rainfall with the variation in reproductive traits of 
G. agassizii (Wallis et al., 1999). These researchers observed that 
East Mojave females lay eggs at smaller body sizes, lay propor- 
tionally smaller eggs, and lay more eggs than West Mojave females. 
Sonoran tortoises often invest their entire annual reproductive 
output in a single clutch laid prior to more predictable rainfall 
(Averill-Murray et al., 2002a). Rainfall seems to influence mean 
clutch size and the proportion of females reproducing each year 
45 
40 
35 
30 
~   25 
? 
-t-   20 
on 
? 
?   15 
10 
5 
0 
-5 
f 
W 
w 
w 
\ 
w 
w 
Size class (iiini CL) 
Fig. 5. Mean age within size classes of Gopherus agassizii from the Sonoran Desert (SD), Arizona and the West Mojave Desert (WM), California. Error bars represent 95% confidence 
intervals; s, Sonoran Desert; w, West Mojave Desert. 
Please cite this article in press as: Curtin, A.J., et al., Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts, Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
Aj. Curtin et al. /journal of Arid Environments xxx (2009) 1-9 
(Averill-Murray et al., 2002b). In dry years, smaller females are less 
likely to lay eggs than larger ones. Sonoran females lay on average 
only one clutch per year (Averill-Murray et al., 2002b), and in 
drought periods, they often do not reproduce. Wirt and Holm 
(1997) reported that only two of the six females they studied in the 
Maricopa Mountains in 1994 laid eggs after almost 10 years of 
drought, whereas all seven female tortoises under observation laid 
eggs at a nearby site, which apparently was less influenced by 
drought. The average annual egg-laying date does not appear to be 
directly related to recent rainfall in Sonoran tortoises (Averill- 
Murray et al., 2002a). Rainfall, however, greatly influences Mojave 
tortoise reproduction, and Mojave females can lay as many as three 
clutches in a year (although the average is 1-2 clutches; Wallis 
et al., 1999; Averill-Murray, 2002). The energy investment of 
a single clutch may allow Sonoran females to invest more energy 
into body growth than West Mojave females (Wallis et al., 1999). 
Additionally, annual resource availability appears to be more 
consistent in the Sonoran Desert (Van Devender et al., 2002). In 
contrast. West Mojave females generally produce more than one 
clutch each year, even during periods of drought (Turner et al., 
1986; Wallis et al., 1999). The physiological stress of low food 
resources constrains the amount of energy available for growth. In 
addition, resources available during egg development would 
generally affect the health and energy input available to female 
desert tortoises. Annual field metabolic rate (FMR) of female desert 
tortoises has been positively correlated with the number of eggs 
laid (Henen, 1997), which indicates that energy expenditures 
associated with reproduction (e.g. digging nests, producing eggs 
and increasing foraging effort to gather extra nutrients for 
producing eggs) are substantial (Henen et al., 1998). Droughts of 
18 months or longer occur regularly in the Mojave Desert (Oftedal, 
2002). Absence of rain precludes germination of annuals and re- 
growth of perennials. At these times, the desert is nearly devoid of 
food for tortoises, except for some of the smaller, less armored cacti 
that tortoises can eat and whatever non-woody senescent material 
that has not disintegrated or blown away, like dried grasses 
(Oftedal, 2002). In years of low winter rainfall, foraging choices are 
limited by the small number of plant species that germinate and 
grow, not to mention that many of them are introduced weedy 
species (Oftedal, 2002). 
4.2. Resorption core diameters 
Resorption core diameter (RCD) gives an indication of bone 
remodeling, which involves the removal of primary (first bone 
deposited) bone layers and, thus, the early growth layers. Sexual 
dimorphism in RCD is not evident in either sample. A comparison of 
adults between the two samples reveals that Sonoran females had 
larger RCDs than West Mojave females (Fig. 2), and Sonoran males 
had larger humeral and ilial RCDs than West Mojave males (Fig. 2). 
There was an association between body size and RCD for all bones 
in both groups. Sonoran tortoises displayed larger RCDs at similar 
carapace lengths than West Mojave tortoises. Resorption core 
diameters include not only the medullary cavity area, but also any 
secondary (endosteal) bone deposited in the area of resorption. 
Larger RCDs indicates higher rates of bone removal and, in some 
instances, higher rates of secondary bone deposition, usually 
associated with intracortical remodeling (Castanet and Smirina, 
1990). Sonoran tortoises generally had a greater rate of endosteal 
remodeling than West Mojave tortoises, especially since juveniles 
start off with similar RCDs at similar sizes. 
Bone was originally seen as a temporary store of calcium and 
phosphates that remodeling then released back into circulation. 
However, as Currey (2003) states, bone is being resorbed and 
deposited at the same time, almost contiguously, so any advantage 
to the body as a whole must be small. It is now apparent that other 
processes, like mechanical competence, changing the grain of bone 
as an adaptation to muscle insertion, or the replacement of dead 
cells, are more important (Currey, 2003). Non-mechanical factors 
also need to be taken into account. For example in Nile monitors 
{Varanus niloticus), bone in juvenile males and females share the 
same mechanical properties and density, but upon sexual maturity 
the endosteal cavity of females enlarges with each subsequent egg- 
laying cycle (de Buffrenil and Francillon-Viellot, 2001). Bone lost 
while producing eggs is not fully regained before the next egg- 
laying period. Such differential bone resorption for egg-shell 
development would explain the sexual dimorphism in RCDs in 
desert tortoises due to a female's need for extra calcium and 
phosphates for egg-shell development, but such an explanation is 
not necessary. A similar explanation for the differences in RCD size 
between the Mojave and Sonoran samples is contraindicated 
because with the greater number of clutches produced by the 
Mojave sample, they should have the larger RCDs. Further, if 
increased remodeling promoted ion release due to nutrient defi- 
ciencies (resulting from a diet of less nutritious exotic species), we 
would again expect the Mojave sample to have larger RCDs. For the 
moment without more details on calcium physiology, average 
annual calcium use in egg-shell production, and related items, this 
general explanation is unworkable. 
Another possible explanation of differential remodeling is as 
a biomechanical adaptation to different locomotor behaviors. The 
highest tortoise densities in the Mojave Desert are associated with 
occurrence in intermountain valleys and flat open land, where 
friable soils allow for the construction of deep burrows (Germano 
et al., 1994; McLuckie et al., 1999). In contrast, Sonoran tortoises 
reach their highest densities on steep, rocky hills and desert 
mountain slopes and are generally absent from or occur at low 
densities in intermountain valleys and washes (Averill-Murray and 
Averill-Murray, 2005; Averill-Murray et al., 2002a; Riedle et al., 
2002). The biomechanical constraints of locomotion on mountain 
slopes and rocky hills may demand greater remodeling in Sonoran 
tortoise bone, thus producing larger RCDs than in West Mojave 
tortoises. In addition, the potential dual foraging period and 
frequent lack of hibernation in Sonoran tortoises (Averill-Murray 
et al., 2002a; Van Devender, 2002) would cause them to be more 
active throughout the year than Mojave tortoises, where drought 
and winter freezes generally promote hibernation. Burrows serve 
as thermal refugia (Spotila et al, 1994) and during these stressful 
periods, Mojave tortoises sustain a much reduced metabolic rate, 
rarely emerging to feed (Nagy and Medica, 1986; Peterson, 1996). 
4.3. Longevity and age at maturity 
The number of visible growth layers varies among the various 
limb bones. Our preliminary skeletal screening showed that the 
tibia, fibula, ulna and radius of an individual had fewer visible 
growth layers (i.e. thinner cortical bone) than the humerus, scapula, 
femur and ilium. While these results were expected for the 
humerus and femur, the "high" number of layers for the two girdle 
bones was not, and the ability to use these two bones has been 
especially helpful owing to our reliance on salvaged specimens, 
which often lack parts of the skeleton. We recommend skeletal 
screening as a first step in any future skeletochronological study 
involving a new (unscreened) species. Screening can be performed 
by x-ray, as in de Buffrenil and Castanet (2000), or preferably by 
histological sectioning. 
Previous studies reported that male and female desert tortoises 
generally attain similar ages (Germano, 1994a; Germano et al., 
2002; Turner et al., 1987). Our results bolster that conclusion; our 
samples show no sexual dimorphism in age (Fig. 4). Sonoran 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts, Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
A.J. Curtin et al. /Journal of Arid Environments xxx (2009) 1-9 
tortoises, however, reach significantly older ages than West Mojave 
tortoises and, in general, are older at similar sizes than West Mojave 
tortoises (Figs. 3 and 4). Even though Sonoran males are signifi- 
cantly older than West Mojave males, they are similar in body size. 
The similarity in maximum size, but the difference in the age 
structure within the equivalent size classes shows that Sonoran 
tortoise are growing at a slower rate, and under this differential 
growth regime, the two populations attain sexual maturity at 
strikingly different ages. West Mojave tortoises ~ 17 years, Sonoran 
tortoises ~ 26 years. 
4.4. Growth strategies, ecological implications and conclusions 
The most surprising result of our study is that West Mojave 
tortoises grow faster and reach sexual maturity earlier than 
Sonoran tortoises (Figs. 3 and 5, also see Germano, 1992, Fig. 4; 
Germano, 1994a, Table 4). Because an equitable environment and 
adequate food promote rapid growth rate and early sexual maturity 
in captive desert tortoises (Jackson et al., 1976, 1978), Sonoran 
tortoises would be expected to have the fastest growth rates and 
earliest maturity of the two desert populations. The Sonoran Desert 
seems to have an environment more favorable or at least less harsh 
than the West Mojave Desert. The longer average life span of 
Sonoran tortoises supports this hypothesis, but their delayed sexual 
maturity and slower growth rates do not. If the West Mojave Desert 
is then the more productive of the two deserts, why do West 
Mojave females average significantly smaller than Sonoran females, 
and why do West Mojave tortoises have shorter life spans than 
Sonoran tortoises? 
Lagarde et al. (2001) observed that in steppe tortoises {Testudo 
horfieldii), which live in a similar habitat and climate to the West 
Mojave tortoises, fast-growing individuals mature early at small 
body sizes, although females are significantly larger than males 
(Lagarde et al., 2001). Wikelski et al. (1997) in his study on marine 
iguanas {Amblyrhynchus cristatus), proposed that rapid growth and 
early maturity at small sizes might be beneficial owing to the lower 
maintenance requirements of small size. This energy saving can 
then be invested into reproduction and survival during periods of 
low resources (Wikelski and Thom, 2000). Lagarde et al. (2001) 
proposed that early maturity and small size are especially beneficial 
to male steppe tortoises because annual activity (3.5 months) is 
strongly constrained by the harsh environment. Aim (1959) 
proposed that the early maturity in stunted perch {Perca fluviatilis) 
derives from either a genetically fixed maturation age or a change 
in environmental conditions (i.e., from good growth conditions 
during early life to poor ones after maturity). Jansen (1996) argued 
that the wide range of maturation age within and between various 
populations of perch species, as well as the known response of age 
at first reproduction to environmental factors (e.g., temperature 
and nutrient levels) made a strictly genetic basis for the onset of 
maturity unlikely. Optimal life history strategies allocate resources 
to maintenance, growth, and reproduction in ways that maximize 
individual fitness, i.e., age specific fecundity and survivorship 
(Congdon and Gibbons, 1990). The early reproduction of stunted 
perch decreases their risk of death before reproducing (Jansen, 
1996), and as a consequence of this strategy, stunted perch expe- 
rienced poor somatic conditions and a reduced life expectancy 
(Jansen, 1996). 
West Mojave females might have adopted this life history 
strategy of early reproduction, thereby producing more clutches 
than Sonoran females. This strategy would have a selective 
advantage for populations with high juvenile mortality and 
comparatively shorter life spans (i.e., reduced reproductive life) of 
surviving adults. In iteroparous species with moderate to long 
potential life span, the benefit of early maturity usually associates 
with increased adult survivorship augmented by increasing the 
number of reproductive episodes each year (reproductive period) 
(Congdon et al., 1982; Hamilton, 1966; Lagarde et al., 2001). The 
trade-off, however, is that Mojave females appear tied to annual 
reproduction, laying at least one, but usually more, clutches per 
year, and they do this consistently even during the chronic and 
frequent droughts of the West Mojave Desert. These frequent 
droughts, coupled with the West Mojave Desert having the lowest 
annual rainfall of the entire desert tortoise range, even if it may be 
more productive following winter rains than either the Sonoran or 
East Mojave Desert (Wallis et al., 1999), likely results in chronic 
physiological stress, greatest in the females. Peterson (1996) 
concluded from his studies in the West and East Mojave Desert of 
California that the desert tortoise is not physiologically adapted to 
live in the desert, but is a tenuous relic of a less rigorous climate. 
This conclusion is consistent with the hypothesis proposed by Van 
Devender (2002) that the Mojave tortoises evolved from Sonoran or 
Sinaloan tortoises, a population adapted to a summer rainfall cycle. 
The skeletochronological estimate of 15-17 years as age at 
maturity lies within the published data of 9-26 years in Mojave 
tortoises (Medica et al., 1975; Germano, 1994a,b), but the skel- 
etochronological estimate of 22-26 years is distinctly greater than 
previously published maturity estimates of 10-20 years for Sonoran 
tortoises. The delayed maturity in Sonoran tortoises could be 
related to the more extreme summer temperatures (regularly over 
38 ?C; Van Devender et al., 2002) experienced in this desert. Ber- 
rigan and Charnov (1994) suggested that low food quality and/or 
availability could result in delayed maturity and small body size at 
maturity but that a reduction or increase of environmental 
temperatures (causing a reduced period of annual activity) yielded 
delayed maturity and a larger body size at maturity. For example, in 
populations of one perch species {Perca flavescens) experiencing 
elevated water temperatures simultaneously with poor feeding 
conditions, the perch adopted an energy-saving strategy. The perch 
reduced clutch size or stopped egg production entirely in the year(s) 
following the year of first reproduction; this strategy reduced the 
high mortality risk associated with the considerable energy 
investment into gamete production (Jansen, 1996; Sandstrom et al., 
1995). Delayed sexual maturity could be an energy-saving strategy 
in Sonoran females, together with the tendency to produce only one 
clutch or even no clutches as a response to unfavorable environ- 
mental conditions (Averill-Murray et al., 2002b). 
The estimated longevity in both groups is higher (noticeably so 
in Sonoran tortoises) than those proposed by (Germano 1992, 
1994a,b) for West Mojave (32 years) and Sonoran Desert (35 years) 
populations. Maximum age estimates are 47-54 years in Sonoran 
males and 40-43 years in Sonoran females. In West Mojave males, 
the oldest age estimate is 56 years. The oldest males, however, were 
typically 27-37 years old. In West Mojave females, maximum age 
estimates were only 24-27 years. It is significant to note that, 
irrespective of desert, females had shorter life spans than males, 
especially West Mojave females. This age differential must be 
considered in future conservation strategies. 
Acknowledgments 
We gratefully acknowledge funding provided by a Panaphil 
Foundation Fellowship to A.J. Curtin. We are also grateful to Dr 
William Boarman and the US Geological Society for donating the 
West Mojave Desert tortoise carcasses, and Roy Averill-Murray and 
the Arizona Game and Fish Department for donating the Sonoran 
Desert tortoise carcasses used in this study. We also thank Dr Mike 
O'Connor and especially Dr Hal Avery for comments on earlier 
drafts of this manuscript and for valuable discussions and insights 
on desert tortoise ecology. 
Please cite this article in press as: Curtin, A.J., et al.. Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts. Journal of Arid Environments (2009), doi;10.1016/j.jaridenv.2008.11.011 
ARTICLE IN PRESS 
Aj. Curtin et al. /journal of Arid Environments xxx (2009) 1-9 
References 
Aim, G., 1959. Connection between maturity, size and age in fisiies. Report of tlie 
Institute of Fresiiwater Researcfi. Drottningfiolm 25,1-146. 
Averill-Murray, R.C., 2002. Reproduction of Gopherus agassizii in tiie Sonoran 
Desert, Arizona. Ciielonian Conservation and Biology 4, 295-301. 
Averill-Murray, R.C., Martin, B.T., Bailey S.J., Wirt, E.B., 2002a. Activity and behavior 
of the Sonoran Desert tortoise in Arizona. In: Van Devender, T.R. (Ed.), The 
Sonoran Desert Tortoise:  Natural History, Biology, and Conservation. The 
University of Arizona Press and the Arizona-Sonora Desert Museum, Tucson, 
AZ, pp. 135-158. 
Averill-Murray, R.C., Woodman, A.P., Howland, J.M., 2002b. Population ecology of 
the Sonoran Desert tortoise in Arizona. In: Van Devender, T.R. (Ed.), The 
Sonoran Desert Tortoise:  Natural History, Biology, and Conservation. The 
University of Arizona Press and the Arizona-Sonora Desert Museum, Tucson, 
AZ, pp. 109-134. 
Averill-Murray, R.C., Averill-Murray, A., 2005. Regional-scale estimation of density 
and habitat use of the desert tortoise (Gopherus agassizii) in Arizona. Journal of 
Herpetology 39, 65-72. 
Berrigan, D., Charnov, E.L., 1994. Reaction norms for age and size at maturity in 
response to temperature: a puzzle for life historians. Oikos 70, 474-478. 
Boatman, W.I., 1992. Effectiveness of fences and culverts for protecting desert 
tortoises along California state highway 58: summary of initial field season. 
Report to California Energy Commission. Bureau of Land Management. River- 
side, CA, 31 pp. 
de Buffrenil, V., Castanet, J., 2000. Age estimation by skeletochronology in the Nile 
monitor (Varanus niloticus), a highly exploited species. Journal of Herpetology 
34, 414-424. 
de   Buffrenil,   V.,   Francillon-Viellot,   H.,   2001.   Ontogenetic  changes   in   bone 
compactness in male and female Nile monitors (Varanus niloticus). Journal of 
Zoology, London 254. 539-546. 
Castanet, J., Smirina, E., 1990. Introduction to the skeletochronological method in 
amphibians and reptiles. Annales Science Naturelle. Zoologie 13 (11), 191-196. 
Castanet, J., Francillon-Vieillot, H.J., Meunier, F.J., De Ricqies, A., 1993. Bone and indi- 
vidual ageing. In: Hall, B.K. (Ed.), Bone, Vol. 7 CRC Press, London, pp. 245-283. 
Chambers, P., 2004. A Sheltered Life: The Unexpected History of the Giant Tortoise. 
John Murray, London, 306 pp. 
Congdon, J.D., Dunham, A.E., Tinkle, D.W., 1982. Energy budgets and life histories of 
reptiles. In: Cans, C, Pough, F.H. (Eds.), Biology of the Reptilia, Vol. 13. Academic 
Press, London, pp. 233-271. 
Congdon, J.D., Gibbons, J.W., 1990. Evolution of turtle life histories. In: Gibbons, J.W. 
(Ed.), Life History and Ecology of the Slider Turtle. Smithsonian Institution 
Press, Washington DC, pp. 45-56. 
Currey, J.D., 2003. The many adaptations of bone. Journal of Biomechanics 36,1487-1495. 
Curtin, A.J., Zug, G.R., Medica, PA., Spotila, J.R. 2008. Aging desert tortoises 
(Gopherus agassizii): testing skeletochronology with known-age individuals. 
Endangered Species Research 5, 21-27. 
Germano, D.J., 1992. Longevity and age-size relationships of populations of desert 
tortoises. Copeia 1992, 367-374. 
Germano, D.J., 1994a. Growth and age at maturity of North American tortoises in 
relation to regional climates. Canadian Journal of Zoology 72, 918-931. 
Germano, D.J., 1994b. Comparative life histories of North American Tortoises. In: 
Bury, R.B., Germano, D.J. (Eds.), Biology of North American Tortoises. National 
Biological Survey, Fish and Wildlife Research 13, pp. 174-185. 
Germano, D.J., Bury, R.B., Esque, T.C., Fritts, T.H., Medica, P.A., 1994. Range and 
habitats of the desert tortoise. In: Bury, R.B., Germano, D.J. (Eds.), Biology of 
North American Tortoises.  National  Biological  Survey,  Fish  and  Wildlife 
Research 13, pp. 73-84. 
Germano, D.J., Pough, F.H., Morafka, D.J., Smith, E.M., Demlong, M.J., 2002. Growth 
of desert tortoises: implications for conservation and management. In: Van 
Devender, T.R. (Ed.), The Sonoran Desert Tortoise Natural History, Biology and 
Conservation. The University of Arizona Press and the Arizona-Sonora Desert 
Museum, Tucson, AZ, pp. 265-288. 
Girondot, M., Garcia, J., 1999. Senescence and longevity in turtles: what elomeres 
tell us. In: Miaud, CL, Guyetant, R. (Eds.), Current Studies in Herpetology. 
Societas Europaea Herpetologica, Le Bourget du Lac, pp. 133-137. 
Hamilton, W.D., 1966. The moulding of senescence by natural selection. Journal of 
Theoretical Biology 28,12-45. 
Henen, B.T., 1997. Seasonal and annual energy budgets of female desert tortoises 
(Gopherus agassizii). Ecology 78, 283-296. 
Henen, B.T., Peterson, C.C, Wallis, I.R., Berry, K.H., Nagy, K.A., 1998. Effects of climate 
variation on field metabolism and water relations of desert tortoises. Oecologia 
117 365-373. 
Jackson Jr, C.G., Trotter, J.A., Trotter, T.H., Trotter, M.W., 1976. Accelerated growth 
rate and early maturity in Gopherus agassizii (Reptilia: Testudines). Herpeto- 
logica 32,139-145. 
Jackson Jr., C.G., Trotter, J.A., Trotter, T.H., Trotter, M.W., 1978. Further observations of 
growth and sexual maturity in captive desert tortoises (Reptilia: Testudines). 
Herpetologica 34, 225-227 
Jansen, W.A., 1996. Plasticity in maturity and fecundity of yellow perch, Perca fla- 
vescens (Mitchill): comparisons of stunted and normal-growing populations. 
Annales Zoologici Fennici 33, 403-415. 
Lagarde, R, Bonnet, X., Henen, B.T., Corbin, J., Nagy, K.A., Naulleau, G., 2001. 
Sexual  size  dimorphism  in  steppe  tortoises  (Testudo  horsfieldi):  growth. 
maturity, and individual variation. Canadian Journal of Zoology 79, 
1433-1441. 
Lamb, T, Lydeard, C, 1994. A molecular phytogeny of the gopher tortoise, with 
comments on familial relationships within the Testudinoidea. Molecular 
Phylogenetics and Evolution 3, 283-291. 
Lamb, T, McLuckie, A.M., 2002. Genetic differences among geographic races of the 
desert tortoise. In: Van Devender, T.R. (Ed.), The Sonoran Desert Tortoise: 
Natural History, Biology, and Conservation. The University of Arizona Press and 
the Arizona-Sonora Desert Museum, Tucson, pp. 67-85. 
Lamb, T., Avise, J.C, Gibbons, J.W., 1989. Phylogeographic patterns in mitochondrial 
DNA of the desert tortoise (Xerobates agassizii), and evolutionary relationships 
among North American gopher tortoises. Evolution 43, 76-87. 
McCord, R.D., 2002. Fossil history and evolution of the gopher tortoises (genus 
Gopherus). In: Van Devender, T.R. (Ed.), The Sonoran Desert Tortoise: Natural 
History, Biology and Conservation. The University of Arizona Press and the 
Arizona Sonora Desert Museum, Tucson, Arizona, pp. 52-66. 
McLuckie, A.M., Lamb, T, Schwalbe, C.R., McCord, R.D., 1999. Genetic and 
morphometric assessment of an unusual tortoise (Gopherus agassizii) pop- 
ulation in the Black Mountains of Arizona. Journal of Herpetology 33, 36-44. 
Medica, P.A., Bury, R.B., Turner, F.B., 1975. Growth of the desert tortoise (Gopherus 
agassizii) in Nevada. Copeia 1975, 639-643. 
Meunier, F.J., Pascal, M., Loubens, G., 1979. Comparaison de methods squelletto- 
chronologiques et considerations fonctionelles sur le tissu osseux acellulaire 
d'un Osteichthyen du Lagon NeoCaledonien. Aquaculture 17,137-157. 
Morafka, D.J., Berry, K.H., 2002. Is Gopherus agassizii a desert-adapted tortoise, or an 
exaptive opportunist? Implications for tortoise conservation. Chelonian 
Conservation and Biology 4, 263-287. 
Murray, R.C., Klug, CM., 1996. Preliminary data analysis from three desert tortoise 
long-term monitoring plots in Arizona: sheltersite use and growth. Proceedings 
of the Desert Tortoise Council Symposium 1996,10-17. 
Nagy, K.A., Medica, PA., 1986. Physiological ecology of desert tortoises in southern 
Nevada. Herpetologica 42, 73-92. 
Oftedal, O.T., 2002. Nutritional ecology of the desert tortoise in the Mojave and 
Sonoran Deserts. In: Van Devender, T.R. (Ed.), The Sonoran Desert Tortoise 
Natural History, Biology and Conservation. The University of Arizona Press and 
the Arizona-Sonora Desert Museum, Tucson, AZ, pp. 194-241. 
Parham, J.F., Zug, G.R., 1998. Age and growth of loggerhead sea turtles (Caretta 
caretta) of coastal Georgia: an assessment of skeletochronological age-esti- 
mates. Bulletin of Marine Science 61, 287-304. 
Peterson, C.C, 1996. Ecological energetics of the desert tortoise (Gopherus agassizii): 
effects of rainfall and drought. Ecology 77,1831-1844. 
Riedle, J.D., Averill-Murray, R.C., Bolen, D.K., 2002. Desert tortoise habitat use and 
home range size on the Florence Military Reservation: progress report. 
Nongame Endangered Wildlife Program, Technical Report 194, pp. 21. 
Sandstrom, 0., Neuman, E., Thoresson, G., 1995. Effects of temperature on life 
history variables in perch, Perca fluviatilis. Journal of Fish Biology 47, 652-670. 
Sazaki, M., Boarman, W.I., Goodlett, G., Okamoto, T, 1995. Risk associated with long- 
distance movements by desert tortoises. Proceedings of the Desert Tortoise 
Council Symposium 1994, 33-48. 
Spotila, J.R., O'Connor, M.R, Zimmerman, L.C., Ruby, D.E., 1994. Introduction: 
conservation biology of the desert tortoise, Gopherus agassizii. Herpetological 
Monographs 8,1-4. 
Turner, F.B., Hayden, P., Burge, B.L., Roberson, J.B., 1986. Egg production by the 
desert tortoise (Gopherus agassizii) in California. Herpetologica 42, 93-104. 
Turner, F.B., Medica, PA., Bury, R.B., 1987. Age-size relationships of desert tortoises 
(Gopherus agassizii) in southern Nevada. Copeia 1987, 974-977. 
US Fish and Wildlife Service, 1994. Desert Tortoise (Mojave Population) Recovery 
Plan. US Fish and Wildlife Service, Portland, pp. 1-73. 
Van Devender, T.R., 2002. Cenozoic environments and the evolution of the gopher 
tortoises (genus Gopherus). In: Van Devender, T.R. (Ed.), The Sonoran Desert 
Tortoise Natural History, Biology and Conservation. The University of Arizona 
Press and the Arizona-Sonora Desert Museum, Tucson, AZ, pp. 29-50. 
Van Devender, T.R., Averill-Murray, R.C., Esque, T.C, Holm, P.A., Dickson, V.M., 
Schwalbe, C.R., Wirt, E.B., Barrett, S.L., 2002. Grasses, mallows, desert vine, and 
more: Diet of the desert tortoise in Arizona and Sonora. In: Van Devender, T.R. 
(Ed.), The Sonoran Desert Tortoise Natural History, Biology and Conservation. 
The University of Arizona Press and the Arizona-Sonora Desert Museum, 
Tucson, AZ, pp. 159-193. 
Walker, D., Avise, J.C, 1998. Principles of phylogeography as illustrated by fresh- 
water and terrestrial turtles in the southeastern United States. Annual Review 
of Ecology and Systematics 29, 23-58. 
Wallis, I.R., Henen, B.T., Nagy, K.A., 1999. Egg size and annual egg reproduction by 
female desert tortoises (Gopherus agassizii): the importance of food abundance, 
body size, and date of egg shelling. Journal of Herpetology 33, 394-408. 
Wikelski, M., Carrillo, V., Trillmich, K, 1997. Energy limits to body size in a grazing 
reptile, the Galapagos marine iguana. Ecology 78, 2204-2217. 
Wikelski, M., Thorn, C, 2000. Marine iguanas shrink to survive El Nino. Nature 403,37. 
Wilson, D.D., Tracy, C.R., Tracy, C.R., 2003. Estimating age of turtles from growth 
rings: a critical evaluation of the technique. Herpetologica 59,178-194. 
Wirt, E.B., Holm, PA., 1997. Climatic effects on survival and reproduction of the 
desert tortoise (Gopherus agassizii) in the Maricopa Mountains, Arizona. 
Unpublished Report to Arizona Game and Fish Department, Phoenix. 
Zug, G.R., 1990. Age determination of long-lived reptiles: some techniques for sea- 
turtles. Annales des Sciences Naturelles, Zoologie, Paris, 13"^ Serie 11,219-222. 
Please cite this article in press as: Curtin, A.J., et al., Longevity and growth strategies of the desert tortoise {Gopherus agassizii) in two American 
deserts, Journal of Arid Environments (2009), doi:10.1016/j.jaridenv.2008.n.0n