Nutritional Quality of Natural Foods of Juvenile Desert Tortoises (Gopherus agassizii): Energy, Nitrogen, and Fiber Digestibility

Abstract

Wild Desert Tortoises, Gopherus agassizii, are eating different foods now than they were decades ago, because exotic plant species have invaded and flourished in the Mojave Desert over the last century. Reservations about the nutritional quality of exotic vegetation compared to native vegetation led us to conduct feeding experiments with growing, juvenile Desert Tortoises. We determined the digestibility of dry matter, energy, fiber, and nitrogen in four foods: Achnatherum hymenoides (a native grass), Schismus barbatus (an exotic grass), Malacothrix glabrata (a native forb), and Erodium cicutarium (an exotic forb). The largest nutritional differences among diets were between food types (fresh forbs and dry grasses) rather than between native and exotic species. The two grass diets were higher in fiber content and they contained less digestible energy than the two forb diets. The grasses contained little protein, and tortoises actually lost mass and body nitrogen while eating them. The exotic forb yielded more energy and nitrogen per unit dry mass than did the native forb, but this may be related to differences in phenological stages and associated fiber contents of these foods when they were collected. Juvenile tortoises gained weight rapidly when eating forbs and showed no evidence of having a lower digestive capability than did adults, despite their small size and immaturity. Estimates of nitrogen requirements compared to annual nitrogen intake on these diets suggested that growth of juveniles may be limited in part by dietary nitrogen.

Full text

Nutritional Quality of Natural Foods of Juvenile Desert Tortoises
(Gopherus agassizii): Energy, Nitrogen, and Fiber Digestibility
LISA C. HAZARD,
1
DANIELLE R. SHEMANSKI,AND KENNETH A. NAGY
Department of Ecology and Evolutionary Biology, University of California, Los Angeles, 621 Young Drive South,
Los Angeles, California 90095-1606 USA
ABSTRACT.—Wild Desert Tortoises, Gopherus agassizii, are eating different foods now than they were
decades ago, because exotic plant species have invaded and flourished in the Mojave Desert over the last
century. Reservations about the nutritional quality of exotic vegetation compared to native vegetation led us
to conduct feeding experiments with growing, juvenile Desert Tortoises. We determined the digestibility of
dry matter, energy, fiber, and nitrogen in four foods: Achnatherum hymenoides (a native grass), Schismus
barbatus (an exotic grass), Malacothrix glabrata (a native forb), and Erodium cicutarium (an exotic forb). The
largest nutritional differences among diets were between food types (fresh forbs and dry grasses) rather than
between native and exotic species. The two grass diets were higher in fiber content and they contained less
digestible energy than the two forb diets. The grasses contained little protein, and tortoises actually lost mass
and body nitrogen while eating them. The exotic forb yielded more energy and nitrogen per unit dry mass
than did the native forb, but this may be related to differences in phenological stages and associated fiber
contents of these foods when they were collected. Juvenile tortoises gained weight rapidly when eating forbs
and showed no evidence of having a lower digestive capability than did adults, despite their small size and
immaturity. Estimates of nitrogen requirements compared to annual nitrogen intake on these diets suggested
that growth of juveniles may be limited in part by dietary nitrogen.
The species composition of plants is changing
in the Mojave Desert as exotic species invade
and flourish there (Brooks, 2000); nonnative
species now comprise 66%of the annual plant
biomass in wet years and 91%in dry years
(Brooks and Berry, 2006). These introduced
plants, especially the now-abundant Schismus
barbatus and Erodium cicutarium, are being eaten
by both juvenile and adult Desert Tortoises
(Gopherus agassizii) (Hansen et al., 1976; Nagy
and Medica, 1986; Barboza, 1995; Peterson,
1996b), and there is concern about the nutri-
tional quality of these plants (Nagy et al., 1998;
Oftedal, 2002). By consuming substantial
amounts of introduced vegetation, are juvenile
tortoises getting less of the nutrients, such as
nitrogen and energy, which they need for
growth? As grasses expanded in the beginning
of the Miocene, herbivorous rodent biodiversity
also expanded, indicating adaptation by these
mammals to this increasingly available, but
low-quality, food source (Stevens and Hume,
1995).
In this study, we examined the nutritional
quality of native and exotic plants eaten by
Desert Tortoises by conducting a series of
feeding experiments. We fed four separate diets
in turn to juvenile Desert Tortoises—two native
plants (one grass and one forb) and two exotic
plants (one grass and one forb). We measured
food intake, collected and analyzed the voided
feces, and estimated by difference the juvenile
tortoises’ abilities to digest and retain dry
matter, fiber, energy, and nitrogen (an index of
protein) from each food plant. We hypothesized
that, within food type, exotic plants would in
general have lower nutritional value than native
plants.
MATERIALS AND METHODS
Animals.—Twenty captive-bred neonate and
juvenile (0.5–1.5 yr old) Desert Tortoises (G.
agassizii) were donated or loaned by members of
the California Turtle and Tortoise Club. While
the housing status prior to donation was not
known for all animals, most were housed with
adult tortoises (often the parents) and, therefore,
would have had the opportunity for coproph-
agy of adult feces, with resulting inoculation of
their digestive tracts with beneficial gut mi-
crobes. The animals were allowed several weeks
to acclimate to the controlled temperature (24.5
61uC) and humidity (23 620%rh) in a
vivarium room at the University of California,
Los Angeles. Each tortoise was housed individ-
ually in a plastic bin (39 325 314 cm), which
contained a shelter (a 16-ounce cardboard soft
1
Corresponding Author. Present address: Depart-
ment of Biology and Molecular Biology, Montclair
State University, Montclair, New Jersey 07043, USA;
E-mail: hazardl@mail.montclair.edu
Journal of Herpetology, Vol. 43, No. 1, pp. 38–48, 2009
Copyright 2009 Society for the Study of Amphibians and Reptiles

drink cup cut in half vertically and laid on its
side, cut side down). A fluorescent ‘‘full
spectrum’’ lamp (Reptisun 5.0, Zoo Med Labo-
ratories, Inc.) was placed 25–30 cm above the
animals to provide the ultraviolet B light
necessary for vitamin D synthesis. Heat from a
25-watt spotlight placed at one end of each bin
provided a thermal gradient from 27.7–31.9uC
(on average) during daytime hours (0700–2100).
Diets.—Mediterranean grass (S. barbatus), an
exotic annual grass, was collected at Fort Irwin
near Barstow, California, and near Lancaster,
California. Indian rice grass (Achnatherum hyme-
noides), a native perennial grass, was collected at
Fort Irwin and near Phelan, California. Both
grass species were collected during early and
midsummer, after the grasses were brown and
dried, which is when wild tortoises eat them
(Nagy and Medica, 1986). Their digestibility
was measured during that autumn and winter.
Erodium cicutarium, or filaree, an exotic annual
forb in the geranium family, was collected near
Lancaster, and Malacothrix glabrata, or desert
dandelion, which is a native annual forb in the
sunflower family, was obtained in Yucca Valley,
California. The two forbs were collected several
times during the following springtime, when
green and succulent, and in the flowering and
early fruiting phenological stage. We collected
only those parts of food plants that juvenile
tortoises are known to eat (Oftedal et al., 2002).
Foods were sealed in plastic bags and placed on
ice during transport.
In the laboratory, the dry grasses were
chopped in a Wiley mill, sifted to bite-sized
pieces (2–4 mm in length), and stored in a
drying oven at 60uC until they were used. Forbs
were first cleaned of any remaining detritus,
chopped into bite-sized pieces with a sharp
French knife, placed into several small sealed
plastic bags, and refrigerated or frozen until
used. Periodically, we took samples from the
bag of green forbs currently in use for measure-
ment of water content (drying at 60uCto
constant mass). This allowed us to calculate
the amount of dry food matter offered to each
tortoise from measurements of the mass of fresh
matter offered each day.
Feeding Trials.—The 20 tortoises were sorted
into two groups of 10, each group having
similar means and ranges of body mass. For
the grass feeding trials, one group of 10 (mean
mass 83 644 g) was offered A. hymenoides, and
the other (mean mass 92 638 g) was given S.
barbatus. The tortoises given A. hymenoides
remained healthy during the 133-day trial.
However, two animals offered S. barbatus
became ill and died very early in the study,
and two others refused to eat and were
excluded from the experiment, resulting in a
total of six animals in that group. After the grass
trials were completed, all 16 remaining tortoises
were used for the two forb diets, which were
offered consecutively. Tortoises were first given
E. cicutarium, which was available in early
spring that year, for 90 days. Following the E.
cicutarium trial, M. glabrata was still unavailable
in the field, so the animals were given finely
chopped leaves of commercially available kale
(Brassica oleracea, a cabbage variant, in the
mustard family) for 20 days before starting the
90-day M. glabrata feeding trial. For grasses,
passage rate as determined by the throughput
times of bite-sized pieces of plastic surveyors
tape was long (mean 21 64 days, range 9–40
days). Therefore, lengthy trials were needed to
allow the tortoises to adjust to a new diet and to
ensure that the feces collected during a partic-
ular trial would be from the diet fed at that time.
In all trials, food was weighed (wet mass) and
placed in small, low-sided Petri dishes in each
animal’s bin daily. Feces were collected daily,
and uneaten food was collected every 10 days.
A tray, cut from a waxed paper cup and
designed to catch and hold excreta, was taped
to the posterior end of each tortoise’s plastron.
This facilitated collection of excreta and simpli-
fied clean separation of urinary wastes from
feces, it reduced opportunities for coprophagy
(consumption of feces) and contamination of
food with feces and urine, and it kept the bins
cleaner. Drinking water was offered to the
animals every 10 days, by placing them in a
bowl of shallow water for about 15 min, during
the last 50 days of the grass feeding trials and
during the last 40 days of the forb feeding trials.
Accurate dry matter digestibility (DMD;
grams retained/gram ingested) measurements
were critical, because digestibilities of all other
nutrients were calculated from these numbers.
Because tortoises may only defecate once every
few days, longer trial periods improve the
accuracy of the measurement of rate of feces
output. For the grasses, we determined that
trials of at least 125 days were needed to obtain
consistent and reliable results; shorter measure-
ment periods resulted in much higher variation
in DMD. For the forbs E. cicutarium and M.
glabrata, the last 40 days of the 90-day trials gave
consistent results and were used for DMD
measurements.
We initially used the ‘‘pulse’’ method to
measure digestibility. We force-fed six 1 33–
5 mm pieces of plastic tape to each tortoise
every 10 days, varying color each time, and
started new feces collection vials when the next
color tape was found in the feces. Therefore,
feces collection periods were not necessarily
congruent with ort collection periods. This
method provides a direct measure of the
DESERT TORTOISE NUTRITION 39

amount of dry matter retained by the animal
from the food eaten between markers. However,
it assumes that the markers provide a reliable
indicator of food passage (that is, they move
through the gut at the same rate as the food).
Plastic tape markers in this study did not track
food passage reliably, necessitating switching to
the ‘‘steady-state’’ method. In this method, the
amount of food eaten in a given time period is
measured, and the feces produced during that
same time period are collected. This requires the
assumptions that (1) the animals have not
become constipated or developed diarrhea
and, thus, have not produced more or less feces
than normal, and (2) the animals have not
grown substantially. An increase in size during
the trial would lead to an overestimate of DMD,
because of an increase in the total amount of
feces stored in the animal. This method also
requires that the feces being collected are from
the food type being tested. To ensure this, we
fed test diets for at least 30 days before
beginning sample collections. During the grass
trials, steady-state digestibility was measured
from the day the second or third marker was
recovered from each tortoise (start of a new
fecal vial) to the last day grass was fed, 18
January 2000. Therefore, the total length of the
period varied somewhat among tortoises (125 6
7 days).
To determine growth rate, tortoises were
periodically weighed, and shell dimensions
(carapace length, shell width between the fourth
and fifth marginal scutes, and maximal shell
height at the same distance from the anterior
end, Nagy et al., 2002) were measured. Both
mass and shell ‘‘volume’’ (length 3width 3
height, all in cm) changes were used as
measures of growth.
Sample Analyses.—Feces and duplicate sam-
ples of food were dried to constant mass at
60uC, and then homogenized in a SPEX mixer-
mill for chemical analysis. Energy contents of
subsamples of food and feces were measured at
UCLA with a Phillipson microbomb calorime-
ter, in which each food and feces sample was
analyzed in duplicate or triplicate, and using
benzoic acid as the standard. Subsamples that
differed by more than 1.5%were reanalyzed.
Fiber and nitrogen content in the food and feces
were determined at the Division of Agriculture
and Natural Resources Analytical Laboratory,
University of California, Davis. Total nitrogen
was determined via the combustion gas analyz-
er method (Dumas, 1981), and the ankom
method (Ankom TechnologyH) was used to
measure neutral detergent fiber (Van Soest,
1985). Some analyses required larger sample
mass than was available for some animals,
reducing sample sizes for energy content to
nine for M. glabrata, for nitrogen content to nine
for A. hymenoides, and fiber content to seven for
A. hymenoides, five for S. barbatus, and six each
for M. glabrata and E. cicutarium.
Data Analysis.—To examine nutrient digest-
ibility (amount of nutrient retained relative to
nutrient ingested) and availability (amount of
nutrient retained relative to amount of food
ingested), we constructed bicoordinate utiliza-
tion plots (Raubenheimer and Simpson, 1994),
and calculated ‘‘traditional’’ digestibilities for
individuals, for comparison with earlier papers
using this method. We expressed amounts of
nutrients ingested and retained as daily rates to
adjust for differences in trial duration. Slopes of
the utilization plots give values equivalent to
traditional apparent digestibility and availabil-
ity values. Slope data were statistically analyzed
using analysis of covariance, where the re-
sponse variable was nutrient retained and the
effect variables were diet, nutrient intake, and
diet 3nutrient intake interaction. P-values of
0.05 or less were considered statistically signif-
icant. A significant interaction term indicated
that diets had differing slopes (different digest-
ibilities or availabilities). The 95%confidence
intervals (standard error of regression coeffi-
cient xt
0.05[2], [n22]
) were then calculated for the
slope of each diet (Zar, 1984); diets with
nonoverlapping confidence intervals were con-
sidered to be significantly different from one
another. Statistical tests were performed using
JMP 5.0 for Mac (SAS Institute, Cary, NC) or
calculated according to Zar (1984).
Our measurements of digestibility are termed
‘‘apparent’’ digestiblities to indicate that we did
not measure ‘‘true’’ digestibilities, because feces
contain not only undigested nutrients, but also
nutrients from endogenous sources such as
mucoproteins, pancreatic and intestinal en-
zymes, sloughed epithelial cells, salivary, gas-
tric, and bile secretions, and bacterial nitrogen
and amino acids from the ileum (Caine et al.,
1999).
RESULTS
Food Composition, Intake Rates, and Tortoise
Growth Rates.—The energy contents of the four
diets were similar (Table 1), although desert
dandelion contained about 5%less, on a dry
matter basis (a significant difference), than the
other three foods. The forbs had about five
times more nitrogen in them than the grasses
(Table 1). About half of the dry matter in the
grasses was neutral detergent fiber, whereas
fiber content of the forbs was much lower, at
about 37%in desert dandelion and 19%in
filaree (Table 1).
40 L. C. HAZARD ET AL.

Food consumption rates (milligrams dry
food/day) were significantly correlated with
body mass for all four foods, and the relation-
ships were directly proportional (slopes of log-
transformed data not different from 1.0; Fig. 1).
Mass-specific intake rates were higher for forbs
than for grasses but did not differ within those
categories (Table 1).
Tortoises grew while eating both forbs; mean
growth rates for these foods did not differ from
one another (Table 1). Tortoises eating the
grasses did not grow but instead lost body
mass and shell volume at low but statistically
significant rates: mean rates of change of mass
and shell volume for each grass diet were
negative and significantly different from zero
rate of change but not different from one
another (Table 1). Tortoises were offered drink-
ing water periodically throughout all trials;
thus, decreases in body mass were not simply
the result of loss of body water while eating dry
grasses. Growth rate was positively correlated
with dry matter intake rate for the forbs, but
slopes for grasses did not differ from zero:
higher grass intake did not result in more
growth (Fig. 2).
Nutrient Digestibility and Availability.—For all
four diets, there was a close linear relationship
TABLE 1. Nutrient content (mg/g or kJ/g of dry food) of four food plants of juvenile Desert Tortoises and
rates of voluntary food intake and growth for captive juvenile Desert Tortoises consuming those foods. Native
grass: Achnatherum hymenoides. Exotic grass: Schismus barbatus. Native forb: Malacothrix glabrata. Exotic forb:
Erodium cicutarium. Within rows, significant differences between means (shown with SD in parentheses) are
indicated by different letter superscripts (ANOVA and Tukey’s t-tests). Negative growth rates for A. hymenoides
and S. barbatus were significantly different from zero growth.
Grasses Forbs
Native Exotic Native Exotic
Food composition
Energy (kJ/g dry matter) 17.2 (0.11)
a
17.4 (0.01)
a
16.6 (0.21)
b
17.5 (0.11)
a
Nitrogen (mg/g dry matter) 6.5 (0.45)
a
8.6 (0.31)
b
29.3 (1.4)
c
42.2 (0.87)
d
Fiber (mg/g dry matter) 498 (31)
a
501 (13)
a
367 (24)
b
192 (18)
c
Food Intake Rate (mg dry matter 3
[g body mass]
21
3day
21
) 1.65 (0.46)
a
1.43 (0.37)
a
4.05 (1.13)
b
5.22 (1.82)
b
Growth rate
%mass change 3day
21
20.045 (0.038)
a
20.043 (0.025)
a
0.139 (0.128)
b
0.162 (0.133)
b
%volume change 3day
21
20.027 (0.038)
a
20.040 (0.019)
a
0.126 (0.124)
b
0.164 (0.131)
b
FIG. 1. Relationships between food intake rates
(milligrams dry food/day) and body mass for juvenile
Desert Tortoises eating four foods. Slopes did not
differ from each other or from 1.0. Open circles: native
grass (Achnatherum hymenoides); open triangles: exotic
grass (Schismus barbatus); closed circles: native forb
(Malacothrix glabrata); closed triangles: exotic forb
(Erodium cicutarium).
FIG. 2. Relationship between growth rate and food
intake rate for juvenile Desert Tortoises. Open circles:
native grass (Achnatherum hymenoides); open triangles:
exotic grass (Schismus barbatus); closed circles: native
forb (Malacothrix glabrata); closed triangles: exotic forb
(Erodium cicutarium). Slopes for forbs were significant
and did not differ; slopes for grasses did not differ
from zero.
DESERT TORTOISE NUTRITION 41

between dry matter intake and dry matter
retained (Fig. 3). Slopes of the lines differed:
the digestibility of dry matter was significantly
lower in the grasses and M. glabrata than in the
exotic forb E. cicutarium (Table 2). Intercepts of
the regression lines for dry matter digestibility
and for nearly all nutrient digestibilities (with
the exception of S. barbatus nitrogen, see below)
did not differ significantly from zero.
Energy digestibility and availability were
very similar to dry matter digestibility and
were lowest in the grasses, moderate in desert
dandelion (although not significantly different
from the grasses), and highest in filaree, which
provided nearly twice as much energy per unit
dry food than did the grasses (Table 2).
The forbs were good sources of digestible
nitrogen. Nitrogen digestibility was 67%for
Malacothrix and 73%for Erodium and did not
differ significantly between the forbs (Table 2).
Nitrogen availability on a dry matter basis was
higher for Erodium because of the higher
nitrogen content of that food. When tortoises
ate the grasses, they actually lost more nitrogen
in their feces than they consumed in their food.
Therefore, nitrogen digestibility and availability
values were negative for the grasses, although
the slope for Achnatherum digestibility did not
differ significantly from zero (P50.056,
Table 2). Regression lines for Schismus nitrogen
digestibility and availability had y-axis inter-
cepts that differed from zero (y-intercept 6SE
50.478 60.125 for both lines). Nitrogen loss
increased with increasing dry matter intake
(Fig. 4), suggesting that nitrogen loss was not
simply a result of low amounts of food eaten.
Fiber digestibility tended to be higher in the
desert dandelion than in the other three diets,
but there were no statistically significant differ-
ences among diets (Table 2). Fiber availability
on a dry matter basis was significantly lower for
filaree than for desert dandelion and Schismus
because of the low fiber content of filaree;
Achnatherum did not differ from any of the
other three diets.
DISCUSSION
There are many ways to evaluate the nutri-
tional value of foods, ranging from their
proximate chemical composition to their eco-
logical importance to populations (Oftedal,
2002). We evaluated two nutritional parameters
in this study, digestibility (a relative index of the
animal’s ability to extract a nutrient from a
food) and availability (the amount of useable
nutrient an animal gets from one unit of a food)
for dietary energy, nitrogen (an index of
protein), and fiber. Each of these conveys
different information about the nutritional
value of these food plants for juvenile tortoises.
Exotic versus Native Species.—The native and
exotic grasses had similar energy, nitrogen, and
fiber contents, although nitrogen content was
slightly higher for Schismus (Table 2). Digest-
ibility and availability (dry matter basis) of
energy and fiber did not differ between the two
grasses; however, nitrogen digestibility and
availability (very low for both grasses) were
lower for Schismus than for Achnatherum (de-
spite the slightly higher nitrogen content of
Schismus) and was significantly lower than zero
only for Schismus. The overall nutritional value
of the grasses does not appear to differ. Nutrient
content of forbs was more varied, with Erodium
having more nitrogen, less fiber, and slightly
more energy per gram than Malacothrix.Erodium
had higher dry matter and energy digestibility
and availability, higher nitrogen availability,
and lower fiber availability than did Malaco-
thrix. The nutritional value of the forbs does
vary somewhat and is likely in part caused by
differences in fiber content (see below), not
necessarily because of differences in the geo-
graphic origin of the food.
Grass versus Forb Species.—Overall, the forbs
had higher nutritional value than the grasses.
The forbs provided juvenile tortoises with more
energy and nitrogen per unit dry food than did
the grasses. Desert dandelion yielded about 14–
21%more energy, and filaree provided about
46–56%more energy per gram DM ingested
than did the dry grasses. Tortoises gained
nitrogen rapidly (desert dandelion) or very
FIG. 3. Dry matter utilization plot for four foods
eaten by juvenile Desert Tortoises. Open circles: native
grass (Achnatherum hymenoides); open triangles: exotic
grass (Schismus barbatus); closed circles: native forb
(Malacothrix glabrata); closed triangles: exotic forb
(Erodium cicutarium). Slopes and statistics for regres-
sion lines are given in Table 2.
42 L. C. HAZARD ET AL.

rapidly (filaree) while consuming green forbs,
but lost nitrogen at low rates while eating dry
grass. Nitrogen loss increased with increasing
dry mass of grass consumed. In general, the
nutritional differences between grasses and
forbs are larger than any differences between
exotic and native species (Table 2; Fig. 4).
However, the grasses we used were dead and
dried, whereas the forbs were green and
growing when harvested. Thus, part of this
difference may be the result of the different
phenological stages of the grasses and forbs.
Fresh green S. barbatus contains more nitro-
gen than does dry senescent S. barbatus (Ofte-
dal et al., 2002), but in spring, tortoises often
ignore grasses and instead select green
forbs (Nagy and Medica, 1986; Henen, 1997;
Oftedal, 2002). Tortoises eating fresh green
S. barbatus were able to digest 63%of its dry
matter, 59%of its energy, and 54%of its
nitrogen and obtained 10.6 kJ of energy and
10.2 mg nitrogen per gram of dry food (calcu-
lated from values in Barboza, 1995). Thus,
grasses can yield just as much energy and about
30–55%as much nitrogen as do forbs when both
are consumed at an early phenological stage.
Tortoises in the western Mojave Desert may
select forbs in the spring to better obtain
nutrients other than energy (e.g., nitrogen,
water, phosphorus). In summer, dried grasses
provide energy but little else. In addition to
nitrogen loss, juvenile tortoises lose phosphorus
on these foods and gain other nutrients (e.g.,
calcium, magnesium) only at low rates (L. C.
Hazard, D. R. Shemanski, and K. A. Nagy,
unpubl. data).
TABLE 2. Digestibility and availability (dry matter basis) of nutrients in four diets eaten by juvenile Desert
Tortoises. Native grass: Achnatherum hymenoides. Exotic grass: Schismus barbatus. Native forb: Malacothrix glabrata.
Exotic forb: Erodium cicutarium. Slopes are the regression coefficients (digestibility or availability) for utilization
plots (Figs. 3 and 4), with standard errors (SE) and R
2
-values for the regressions. *, P,0.05. **, P,0.01. ***, P,
0.001. Ratios are the traditional ratio-based digestibility values (retained/ingested for each individual; mean
[SD]) for comparison with previous studies. Within nutrient categories, diets sharing a letter in superscript are
not significantly different (ANCOVA and nonoverlap of 95%confidence intervals for slopes; ANOVA and
Tukey’s t-tests for ratios). Data for adults are from Nagy et al. (1998) and Meienberger et al. (1993).
Nutrient Diet N
Nutrient digestibility (mg or kJ retained per mg or kJ ingested)
Slope (SE) R
2
Ratio (SD) Adult ratio (SD)
Dry matter Native grass 10 0.49 (0.05)
a
0.92*** 0.42 (0.10)
a
0.47 (0.05)
Exotic grass 6 0.50 (0.05)
a
0.96*** 0.45 (0.06)
a
0.50 (0.08)
Native forb 16 0.58 (0.02)
a
0.98*** 0.56 (0.04)
b
0.70 (0.03)
Exotic forb 16 0.70 (0.01)
b
0.99*** 0.71 (0.04)
c
0.63 (0.05)
Energy Native grass 10 0.48 (0.06)
a
0.91*** 0.39 (0.10)
a
0.46 (0.06)
Exotic grass 6 0.45 (0.05)
a
0.96*** 0.40 (0.06)
a
0.48 (0.08)
Native forb 9 0.57 (0.03)
a
0.98*** 0.51 (0.05)
b
0.73 (0.03)
Exotic forb 16 0.69 (0.02)
b
0.99*** 0.72 (0.05)
c
0.69 (0.05)
Nitrogen Native grass 9 20.23 (0.10)
b
0.43 20.31 (0.16)
a
0.07 (0.10)
Exotic grass 6 20.83 (0.10)
a
0.94** 20.32 (0.21)
a
20.07 (0.18)
Native forb 16 0.67 (0.02)
c
0.99*** 0.63 (0.05)
b
0.79 (0.01)
Exotic forb 16 0.73 (0.02)
c
0.99*** 0.75 (0.05)
c
0.72 (0.05)
Fiber Native grass 7 0.46 (0.12)
a
0.74* 0.27 (0.17)
a
Exotic grass 5 0.52 (0.03)
a
0.99*** 0.43 (0.06)
a
Native forb 6 0.76 (0.07)
a
0.97*** 0.67 (0.07)
b
Exotic forb 6 0.42 (0.11)
a
0.80* 0.52 (0.08)
b
Nutrient Diet N
Nutrient availability (mg or kJ retained per g dry matter intake)
Slope (SE) R
2
Ratio (SD)
Energy (kJ/g) Native grass 10 8.3 (1.0)
a
0.91*** 6.7 (1.8)
a
7.8 (1.0 )
Exotic grass 6 7.7 (0.8)
a
0.96*** 6.9 (1.0)
a
8.3 (0.3)
Native forb 9 9.4 (0.5)
a
0.98*** 8.4 (0.9)
b
11.4 (0.4)
Exotic forb 16 12.1 (0.3)
b
0.99*** 12.6 (0.9)
c
10.9 (0.4)
Nitrogen (mg/g) Native grass 9 21.5 (0.7)
b
0.43 22.0 (1.1)
a
0.3 (1.4)
Exotic grass 6 27.1 (0.9)
a
0.94** 22.8 (1.8)
a
20.5 (0.3)
Native forb 16 19.9 (0.6)
c
0.99*** 18.8 (1.4)
b
21.3 (0.4)
Exotic forb 16 30.8 (0.8)
d
0.99*** 31.5 (2.0)
c
18.1 (1.3)
Fiber (mg/g) Native grass 7 231 (61)
a, b
0.74* 136 (83)
b
Exotic grass 5 257 (13)
b
0.99*** 215 (28)
b
Native forb 6 278 (25)
b
0.97*** 245 (24)
b
Exotic forb 6 81 (21)
a
0.80* 99 (16)
a
DESERT TORTOISE NUTRITION 43

FIG. 4. Utilization plots for digestibility and availability of nutrients for four diets eaten by juvenile Desert
Tortoises. Open circles: native grass (Achnatherum hymenoides); open triangles: exotic grass (Schismus barbatus);
closed circles: native forb (Malacothrix glabrata); closed triangles: exotic forb (Erodium cicutarium). Graphs on left
show nutrient digestibility (nutrient retained vs. nutrient intake) and graphs on right show nutrient availability
(nutrient retained vs. dry matter intake). Inset graph for nitrogen shows digestibility data for grasses on an
expanded scale for clarity. Slopes and statistics for regression lines are given in Table 2. (A) Energy digestibility.
(B) Energy availability. (C) Nitrogen digestibility. (D) Nitrogen availability. (E) Fiber digestibility. (F)
Fiber availability.
44 L. C. HAZARD ET AL.

Fiber Effects.—Fiber in foods of herbivorous
animals provides energy directly to the intesti-
nal microbes which can ferment cellulose, and
the fermentation products also provide energy
to the host animal (Bondi, 1987; Dierick et al.,
1989; Baer et al., 1997a). In addition, the
intestinal microbes provide their host with other
nutrients, such as some vitamins. However, the
relative difficulty of digesting fiber can be
nutritionally disadvantageous when foods con-
tain relatively high levels of fiber (Swart et al.,
1993; Baer et al., 1997b; Souffrant, 2001). The dry
grasses contained much more fiber than did the
forbs (Table 1), and this probably accounts for
their relatively low energy availability values
(Table 2). The difference in fiber content of the
forbs likely reflects the different phenological
ages of the two species when they were
collected. Desert dandelion plants were difficult
to find that year, because of drought conditions,
but some were finally located in Yucca Valley,
along the southeastern margin of the Mojave
Desert. Although they were still flowering, by
then they were more mature and apparently
woodier when collected than was the filaree we
had harvested much earlier in the year. Also, to
have sufficient mass of Malacothrix to feed the
animals for the duration of the trial, we
included stems in the finely chopped food
given to the tortoises, whereas stems were
excluded for Erodium. In a previous digestibility
study on adult Desert Tortoises, desert dande-
lion and filaree had equivalent digestibilities
and availabilities of energy and nitrogen (Nagy
et al., 1998; see Discussion below). Moreover,
the dry matter digestibility of the desert
dandelion used in that study was 70%, which
is substantially higher than the 58%measured
for desert dandelion in this study. Dry matter
digestibility in the current study was negatively
correlated with dietary fiber content (R
2
50.995;
P,0.001; Fig. 5).
Juveniles versus Adults.—Digestibility and
availability of energy and nitrogen for juvenile
tortoises were comparable to values for adult
Desert Tortoises fed the same foods in previous
studies (Meienberger et al., 1993; Nagy et al.,
1998). Given that the juveniles we studied
weighed less than 10%of the adult tortoises
studied earlier, and that there are allometric
reasons to expect lower digestibilities in small
reptilian herbivores (Pough, 1973; Wilson and
Lee, 1974), it is perhaps surprising that juveniles
were able to digest dry matter, energy, and
nitrogen just as well as did adults on three of
the four diets (Table 2). The digestibility of
nutrients in Malacothrix was substantially lower
(by 12–17%) for juveniles, which is probably the
result of the greater fiber content and age of the
dandelion plants eaten by the juvenile tortoises
(see above). Food passage times in juveniles
eating dry grass averaged 21 64 days (range 9–
40 days), essentially (and surprisingly) the same
as adult passage time (about 22 days, range 15–
30, Meienberger et al., 1993). Thus, we find little
evidence that digestive physiology was influ-
enced by body mass within the range of masses
of the tortoises we studied (from the smallest
juvenile at 33 g to the largest adult at 4,440 g).
Growth.—Tortoises lost mass and shell vol-
ume when eating grasses but gained mass and
volume when eating forbs. This is partly a result
of the lower availability of nutrients in dry
grasses. However, voluntary intake rates for
grasses were far lower than rates for forbs
(Table 1). Tortoises eating grasses simply may
have not eaten enough dry matter to maintain
body mass. Meienberger et al. (1993) found that
adult tortoises eating dry S. barbatus grass ad
libitum also lost body mass and suggested that
their tortoises were unable to eat dry grass
quickly enough to maintain or gain body mass
because the physical structure of the grass filled
the gut lumen at a relatively low density of dry
matter compared to green leaves. However, our
results suggest that the dry grasses were
qualitatively different from the forbs. When
tortoises ate grasses, increased dry matter intake
resulted in greater loss of nitrogen (Fig 4) and
no mass gain (Fig. 2). Increases in dry matter
intake would also result in increased fecal dry
matter output. The endogenous secretions and
associated substances (e.g., mucous and
sloughed epithelial cells) necessary to physical-
ly eliminate the waste matter may contain
FIG. 5. Effect of dietary fiber content (mean 6SD
for duplicate samples) on dry matter digestibility
(DMD; utilization plot regression coefficient 6SE) for
four diets eaten by juvenile Desert Tortoises. Open
circle: native grass (Achnatherum hymenoides); open
triangle: exotic grass (Schismus barbatus); closed circle:
native forb (Malacothrix glabrata); closed triangle:
exotic forb (Erodium cicutarium). DMD 520.0006(mg
fiber/g dry matter) +0.808; R
2
50.995; P50.003.
DESERT TORTOISE NUTRITION 45

enough nitrogen to account for the nitrogen
loss, and this nitrogen would increase with
increasing food intake. Thus, it appears that
tortoises that eat dry grasses to obtain energy do
so at a substantial cost in terms of nitrogen and
probably water. This is consistent with other
studies that have shown that tortoises feeding in
the summer are in negative nitrogen balance
(Peterson, 1996a).
In free-living adult tortoises, body masses
change more rapidly and extensively in re-
sponse to variations in water balance than they
do to changes in energy balance (Nagy and
Medica, 1986; Henen et al., 1998); thus, a
correlation between dry matter intake and mass
change rates may be obscured in data from a
field population. However, tortoises in this
study should have been fully hydrated contin-
uously, because drinking water was offered
regularly. Our study reveals that juvenile
tortoises can grow rapidly (up to 0.5%of body
mass added per day) while they are eating
succulent green forbs.
Nitrogen Limitation.—Using published data,
we estimated the increase in total body nitrogen
during the second year of life for wild juvenile
tortoises, and compared it to estimated net
nitrogen intake when feeding on the diets
studied here. During the second year of a
tortoise’s life, total body mass increases from
35–54 g, and dry mass increases from 6.1–12.6 g
(Nagy et al., 1997). There are no available data
for nitrogen content of Desert Tortoises, but
protein content of reptiles averages 74.7%of dry
matter (Boyd and Goodyear, 1971). Assuming
that one sixth of the mass of the protein is
nitrogen, total body nitrogen would increase
from 758 mg to 1,575 mg during the second
year, a net gain of 817 mg. During that year, a
juvenile tortoise eats an estimated 96.8 g of food
(Nagy, 1972). Approximately 66%of this food is
eaten in the spring when green food is available,
and the remaining 34%is eaten during the
summer and is likely to be dried grasses.
Assuming a diet of native foods (Malacothrix
and Achnatherum), a tortoise would take in
1,274 mg of nitrogen in the spring but then lose
49 mg in the summer, for a net intake of
1,225 mg. Similarly, a tortoise feeding on exotic
species (Erodium and Schismus) would gain
1,968 mg nitrogen in the spring and lose
235 mg in the summer, for a net intake of
1,733 mg. These net intake estimates for native
and exotic diets are 1.50 and 2.12 times the
estimated requirement for growth, respectively.
This does not take into account loss of nitrog-
enous waste in urine as a part of normal protein
turnover, nor the importance of nitrogen as a
vehicle for excretion of potassium in an insol-
uble form (potassium urate salts) if high levels
of potassium are present in the food (Oftedal
and Allen, 1996). Therefore, the small margin of
safety above the estimated nitrogen require-
ment suggests that nitrogen is likely to be a
limiting nutrient for growing juvenile tortoises.
However, because the total amount of nitrogen
lost when feeding on grasses is small relative to
nitrogen gained in the spring, feeding on
nitrogen-poor dry grasses may still be beneficial
if other nutrients such as energy can be
obtained. Chuckwallas (Sauromalus obesus) are
known to store nitrogen when feeding on
succulent high-protein plants in spring, which
offsets losses when feeding on dry summer
foods (Nagy, 1975). Nitrogen is not necessarily
the only limiting nutrient for Desert Tortoises;
similar calculations for several minerals showed
that phosphorus is also obtained in amounts
just above estimated requirements (L. C. Haz-
ard, D. R. Shemanski, and K. A. Nagy, unpubl.).
There appeared to be few nutritional differ-
ences within food type (forb or grass) between
native and exotic plant foods fed to captive
tortoises. However, the broader context of the
ecology of the Desert Tortoise must also be
considered. The dominant invasive species,
Schismus spp., E. cicutarium, and Bromus rubens,
are increasing overall annual plant biomass but
with a reduction in native plant diversity,
possibly including preferred food plants of
Desert Tortoises (Brooks, 2000; Brooks and
Berry, 2006). If tortoises are forced either to
switch from native forbs to potentially less
nutritious exotic grasses or to spend more time
searching out the less available native forbs,
there may still be ecological or nutritional
consequences for animals in the wild. Addi-
tionally, the dramatically increased biomass of
invasive grasses, although potentially providing
additional food for tortoises and other herbi-
vores, also increases the frequency of wildfires
(Brooks and Esque, 2002), which may impact
tortoises directly through increased mortality or
indirectly through alterations to the Mojave
Desert ecosystem.
Acknowledgments.—We thank the California
Turtle and Tortoise Club and D. Morafka for
loaning us juvenile tortoises for this study. We
also thank the staff at the DANR Analytical
Laboratory at the University of California at
Davis for fiber and nitrogen analysis, and C.
Park and J. Nakai for their assistance with
animal husbandry. We especially thank D.
Morafka for his support and facilitation. Finan-
cial support and means were provided by the
Academic Senate and the Department of Ecol-
ogy and Evolutionary Biology of the University
of California, Los Angeles, and by the Director-
ate of Public Works (M. Quillman, Natural and
46 L. C. HAZARD ET AL.

Cultural Resources Manager) of the U.S. Army
National Training Center at Fort Irwin, Califor-
nia, via a contract with the California State
University, Dominguez Hills Foundation. The
UCLA Office for the Protection of Research
Subjects (ARC 98-160-02) approved this study.
LITERATURE CITED
BAER, D. J., O. T. OFTEDAL,W.V.RUMPLER,AND D. E.
ULLREY. 1997a. Dietary fiber influences nutrient
utilization, growth and dry matter intake of Green
Iguanas (Iguana iguana). Journal of Nutrition
127:1501–1507.
BAER, D. J., W. V. RUMPLER,C.W.MILES,AND G. C.
FAHEY JR. 1997b. Dietary fiber decreases the
metabolizable energy content and nutrient digest-
ibility of mixed diets fed to humans. Journal of
Nutrition 127:579–586.
BARBOZA, P. S. 1995. Nutrient balances and mainte-
nance requirements for nitrogen and energy in
Desert Tortoises (Xerobates agassizii) consuming
forages. Comparative Biochemistry and Physiolo-
gy 112A:537–545.
BONDI, A. A. 1987. Animal Nutrition. John Wiley and
Sons, New York.
BOYD, C. E., AND C. P. GOODYEAR. 1971. The protein
content of some common reptiles and amphibians.
Herpetologica 27:317–320.
BROOKS, M. L. 2000. Competition between alien annual
grasses and native annual plants in the Mojave
Desert. American Midland Naturalist 144:92–108.
BROOKS, M. L., AND K. H. BERRY. 2006. Dominance and
environmental correlates of alien annual plants in
the Mojave Desert, USA. Journal of Arid Environ-
ments 67:100–124.
BROOKS, M. L., AND T. C. ESQUE. 2002. Alien plants and
fire in Desert Tortoise (Gopherus agassizii) habitat of
theMojaveandColoradoDeserts.Chelonian
Conservation and Biology 4:330–340.
CAINE, W. R., S. TAMMINGA,W.C.SAUER,M.W.A.
VERTEGEN,AND H. SCHULZE. 1999. Bacterial contri-
butions to total and endogenous recoveries of
nitrogen and amino acids in ileal digesta of newly
weaned piglets fed protease-treated soybean meal.
Livestock Production Science 57:147–157.
DIERICK, N. A., I. J. VERVAEKE,D.I.DEMEYER,AND J. A.
DECUYPERE. 1989. Approach to the energetic impor-
tance of fiber digestion in pigs. I. Importance of
fermentation in the overall energy supply. Animal
Feed Science and Technology 23:141–167.
DUMAS, J. B. 1981. Sur les porcedes de l’analyse
organique. Annalide Chimie XLVII:195–213.
HANSEN, R. M., M. K. JOHNSON,AND T. R. VAN DEVENDER.
1976. Foods of the Desert Tortoise, Gopherus
agassizii, in Arizona and Utah. Herpetologica
32:247–251.
HENEN, B. T. 1997. Seasonal and annual energy
budgets of female Desert Tortoises (Gopherus
agassizii). Ecology 78:283–296.
HENEN, B. T., C. C. PETERSON,I.R.WALLIS,K.B.BERRY ,
AND K. A. NAGY. 1998. Effects of climatic variation
on field metabolism and water relations of Desert
Tortoises. Oecologia 117:365–373.
MEIENBERGER, C., I. R. WALLIS,AND K. A. NAGY. 1993.
Food intake rate and body mass influence transit
time and digestability in the Desert Tortoise
(Xerobates agassizii). Physiological Zoology
66:847–862.
NAGY, K. A. 1972. Water and electrolyte budgets of a
free-living desert lizard, Sauromalus obesus. Journal
of Comparative Physiology 79:39–62.
———. 1975. Nitrogen requirement and its relation to
dietary water and potassium content in the lizard
Sauromalus obesus. Journal of Comparative Physi-
ology 104:49–58.
NAGY, K. A., AND P. A. MEDICA. 1986. Physiological
ecology of Desert Tortoises in southern Nevada.
Herpetologica 42:73–92.
NAGY, K. A., D. J. MORAFKA,AND R. A. YATES. 1997.
Young Desert Tortoise survival: energy, water, and
food requirements in the field. Chelonian Conser-
vation and Biology 2:396–404.
NAGY, K. A., B. T. HENEN,AND D. B. VYAS. 1998.
Nutritional quality of native and introduced food
plants of wild Desert Tortoises. Journal of Herpe-
tology 32:260–267.
NAGY, K. A., B. T. HENEN,D.B.VYAS,AND I. R. WALLIS.
2002. A condition index for the Desert Tortoise
(Gopherus agassizii). Chelonian Conservation and
Biology 4:425–429.
OFTEDAL, O. T. 2002. Nutritional ecology of the Desert
Tortoise in the Mojave and Sonoran Deserts. In T.
R. van Devender (ed.), The Sonoran Desert
Tortoise: Natural History, Biology, and Conserva-
tion. University of Arizona Press, Tucson.
OFTEDAL, O. T., AND M. E. ALLEN. 1996. Nutrition as a
major facet of reptile conservation. Zoo Biology
15:491–497.
OFTEDAL, O. T., S. HILLARD,AND D. J. MORAFKA. 2002.
Selective spring foraging by juvenile Desert
Tortoises (Gopherus agassizii) in the Mojave De-
sert: evidence of an adaptive nutritional stra-
tegy. Chelonian Conservation and Biology
4:341–352.
PETERSON, C. C. 1996a. Anhomeostasis: seasonal water
and solute relations in two populations of the
Desert Tortoise (Gopherus agassizii) during chronic
drought. Physiological Zoology 69:1324–1358.
———. 1996b. Ecological energetics of the Desert
Tortoise (Gopherus agassizii): effects of rainfall and
drought. Ecology 77:1831–1844.
POUGH, F. H. 1973. Lizard energetics and diet. Ecology
54:837–844.
RAUBENHEIMER,D.,AND S. J. SIMPSON. 1994. The
analysis of nutrient budgets. Functional Ecology
8:783–791.
SOUFFRANT, W. B. 2001. Effect of dietary fibre on ileal
digestibility and endogenous nitrogen losses in the
pig. Animal Feed Science and Technology
90:93–102.
STEVENS,C.E.,AND I. D. HUME. 1995. Com-
parative Physiology of the Vertebrate Diges-
tive System. Cambridge University Press, Cam-
bridge.
SWART, D., F. K. SIEBRI TS,AND J. P. HAYES. 1993.
Utilization of metabolizable energy by ostrich
(Struthio camelus) chicks at two different concen-
trations of dietary energy and crude fibre origi-
DESERT TORTOISE NUTRITION 47

nating from Lucerne. South African Journal of
Animal Science 23:136–141.
VAN SOEST, P. J. 1985. Definition of fiber in animal
feeds. In W. Hairsign and D. J. A. Cole (eds.),
Recent Advances in Animal Nutrition, pp. 55–69.
Butterworths, London.
WILSON, K. J., AND A. K. LEE. 1974. Energy expenditure
of a large herbivorous lizard. Copeia 1974:338–348.
ZAR, J. H. 1984. Biostatistical Analysis. Prentice-Hall,
Inc., Englewood Cliffs, NJ.
Accepted: 25 May 2008.
48 L. C. HAZARD ET AL.