The use of stored resources to fuel reproduction, growth and maintenance to balance variation in nutrient availability is common to many organisms. The degree to which organisms rely upon stored resources in response to varied nutrients, however, is not well quantified. Through stable isotope methods we quantified the use of stored versus incoming nutrients to fuel growth, egg and fat body development in lizards under differing nutrient regimes. We found that the degree of capital breeding is a function of an individual’s body condition. Furthermore, given sufficient income lizards in poor condition can allocate simultaneously to storage, growth, and reproduction, which allowed them to catch up to better conditioned animals. In a parallel, inter-specific survey of wild lizards we found that the degree of capital breeding varied widely across a diverse community. These findings demonstrate that capital breeding in lizards is not simply a one-way flow of endogenous stores to eggs, but is a function of the condition state of individuals and the availability of nutrients during both breeding and non-breeding seasons. Here we explore the implications of these findings for our understanding of capital breeding in lizards and the utility and value of the capital-income concept in general.
Lizard Capture and Maintenance:
Thirty two female prairie lizards (Sceloporus undulatus) were caught on Bureau of Land Management reserves near Albuquerque, NM during the last two weeks of July, 2007 and maintained in a room at the biology department of the University of New Mexico under the approval of the UNM-IACUC (#07UNM007). Two lizards were housed per 20 gallon glass terrarium and were provided a sand substrate, as well as perch and shelter spaces built by stacked pieces of plywood and rock. Lizards were kept on a 12 hour (light:dark) photoperiod and a temperature gradient was provided by a 100 watt heat lamp placed at one end of the terrarium and focused on the wood perch, which provided a stable heat gradient that ranged from 39 ± 1.7ºC at the perch to 26 ± 0.8ºC at the cool end of the tank. Resulting mean daytime body temperatures were 36.3 ± 6.2ºC (n = 18). An ultraviolet-B fluorescent light (ZooMed® UVB 10.0 fluorescent) was also provided for vitamin D synthesis.
Experimental dietary treatments and reproduction:
S. undulatus were captured from a cottonwood woodland field site and paired by Snout Vent Length (SVL) and then randomly split into either a high (n = 16) or low nutrient treatment (n = 16). The high nutrient diet consisted of seven crickets and two mealworms per week; similar to an ad libitum diet found by Angilletta (2001). We estimated that a low nutrient diet reduced by ~30% of ad libitum (five crickets/week and one mealworm every other week) would reduce body condition and reflect the poor conditions experienced by lizards in the wild (see Ballinger 1977, 1979, Ballinger and Congdon 1980, Sinervo and Adolph 1994). These low diet lizards were switched to the high diet after hibernation, referred to hereafter as the LH treatment (n = 16). The high treatment prior to hibernation was split into a high diet post-hibernation (HH, n = 8) and a low treatment (HL, n = 8). We did not have an LL treatment because we assumed that they would be in such low body condition that they would not reproduce.
The lizards were prepared for hibernation during November 2007 by gradually reducing the photoperiod to 7 hours per day, and were fasted for two weeks. Lizards were then placed in 27 liter plastic containers with a sandy substrate and wood shavings for burrowing on November 17, 2007 and maintained at 10.2 ± 3.1ºC. The lizards were removed from hibernation on February 2, 2008. To induce reproduction, male prairie lizards that were maintained for a separate study were introduced for two weeks to the female terrarium in mid-February of 2008. Reproduction was observed in numerous tanks (mounting and copulation), and signs of reproduction (bite marks) were apparent on all females. The female lizards were then palpated weekly to monitor egg development. When eggs appeared to be nearly shelled or shelled, the lizards were euthanized via an intraperitoneal injection of sodium pentobarbital (using a dose of 60 mg/kg). Two lizards were euthanized in late March following rapid development of shelled eggs. All other lizards were euthanized during the last week of May 2008, at which time most were found to have either large follicles or shelled eggs.
Stable isotope treatments:
After the lizards were euthanized liver, fat body, and thigh muscle samples were harvested, freeze dried and a 0.5 mg sample was placed into a pre-cleaned tin capsule (Costech, #041074, Valencia, CA) for stable isotope analysis. Eggs and follicles were also harvested, their length and width measured and freeze dried. All lipids were extracted from freeze dried and ground muscle and eggs/follicles by a 2:1 chloroform and methanol bath; lizard muscle had undetectable amounts of lipids. The suspended lipids from eggs were pipetted into separate storage vials and air dried. Lipids and lipid-free egg tissues were then loaded into tin capsules. We measured the δ13C of each egg and follicle greater than 6mm in length (½ the length of shelled eggs and assumed to reflect reproductive allocation). Our stable isotope methodology follows standard methods and our protocol is described in detail in Warne et al. (2010a, 2010b). We report all isotope values in the standard delta notation (δX = (Rsample /Rstandard – 1) x 1000) in parts per thousand (‰) relative to the international carbon standard VPDB (Vienna Pee Dee Belemnite). Measurements were conducted on a continuous flow isotope ratio mass spectrometer in the UNM Earth and Planetary Sciences Mass Spectrometry lab. The precision of these analyses was ± 0.1‰ SD for δ13C based on long-term variation of the working laboratory standard (valine δ13C = -26.3‰ VPDB), samples of which were included on each run in order to make corrections to raw values obtained from the mass spectrometer.
Essential to this study is the observation that differences in photosynthetic biochemistry inherent to C3- and C4-plants produces distinct differences in the d13C of their tissues, which can be used to trace the movement of nutrients through consumers (Hobson et al. 1997, O'Brien et al. 2000). The lizards were collected from cottonwood woodlands in which their diet was largely composed of C3-plant derived carbon, as evidenced by a baseline muscle δ13C of -25.1± 0.1‰ VPDB, near that of the mean value for C3 plants of -27.3 ± 0.04‰. Prior to hibernation lizards were maintained on a diet composed of crickets (mean δ13C ± SEM for lipids = -22.5 ± 0.1‰ and lipid free carbon = -21.7 ± 0.1‰ VPDB, n = 16) raised on C3-plant derived dog food (Nutro® Natural Choice® large breed puppy lamb and rice formula) and mealworms (lipids = -26.3 ± 0.1‰ and lipid free = -24.35 ± 0.1‰ VPDB, n = 8) raised on bran meal. Here we used the mathematical normalization model of Post et al. (2007) to determine the lipid-free δ13C values for these insects, assuming reported lipid contents of 13.8% for crickets and 32.8% for mealworms (Bernard and Allen 1997). Maintaining lizards on a C3 diet prior to hibernation insured that their capital stores of fat bodies and muscle would have consistent carbon isotope values. After hibernation the lizards were switched to a C4 based insect diet of crickets (lipids = -16.3 ± 0.1‰ and lipid free = -15.53 ± 0.1‰VPDB, n = 35) raised on a C4 - corn based dog food (Iams® Smart Puppy large breed formulaTM) and mealworms (lipids = -13.0 ± 0.3‰ and lipid free = -10.2 ± 0.2‰ VPDB, n = 9) raised on coarse ground cornmeal; which provided an ‘income’ diet with δ13C values distinct from the pre-hibernation C3 diet.
We used tissue d13C values and a standard two-end-point mixing model to estimate the proportion of endogenous fat or muscle (capital) and incoming insect-dietary sources used to provision eggs. Because crickets and mealworms had different δ13C values and the dietary treatments imposed on the lizards also consisted of differing quantities of feeder insects (high = 7 crickets + 2 mealworms/week; low = 5 + 0.5/week) we used a weighted mean to calculate the insect δ13C for each treatment in this model. The weighted δ13C value for the high dietary treatment for C4 insect lipids was -15.6‰ and -14.3 for lipid free carbon; for the low treatment C4 insect lipids was -16‰ and -15‰ for lipid-free carbon. The discrimination (Δ13C) values used in this model for muscle (-1.9‰) and fat bodies (0‰) were experimentally determined for S. undulatus (Warne et al. 2010a).
The effect of dietary treatment on the body condition of lizards was analyzed by repeated measures ANCOVA with treatment and stage of the experiment as fixed effects, individuals as random effects nested within treatment, and snout-vent length (SVL) as a covariate. Body condition was estimated as the least squares (LS) mean of body weight (minus eggs) adjusted for SVL in this ANCOVA model; a method argued to be more statistically sound than other condition indices (Packard and Boardman 1988, García-Berthou 2001). Treatment effects on SVL were similarly analyzed by repeated measures ANOVA. Mauchly’s test was used to confirm that the assumption of sphericity for repeated measures analysis was valid, and epsilon corrections to the degrees of freedom were applied when necessary. Dietary treatment effects on growth were measured by the specific growth rate of SVL (ln(SVL2/SVL1)/Δdays) for the pre- and post hibernation periods, and analyzed by one-way ANOVA. Dietary treatment effects on reproductive effort, measured as relative clutch mass (RCM = clutch mass/body mass with no eggs) and clutch size were analyzed by one-way ANOVA. The effect of dietary treatment on tissue δ13C values were similarly analyzed by one-way ANOVA. Post-hoc comparisons of treatment effects during the four experimental stages were conducted using Tukey-Kramer’s HSD test. Prior to all analyses the data were tested for homogeneity of variance and confirmed to meet model assumptions. These analyses were performed in JMP® 8.0 (SAS Institute Inc., Cary, NC, 1989-2007). All values are reported as mean ± SEM.