Ecology: Vol. 87, No. 10, pp. 2447–2458.

INDIVIDUALLY VARIABLE ENERGY MANAGEMENT STRATEGIES IN RELATION TO ENERGETIC COSTS OF EGG PRODUCTION

François Vézina,^{a, 3} John R. Speakman,^b and Tony D. Williams^a

aDepartment of Biological Sciences, Simon Fraser University, 8888 University Drive, Burnaby, V5A1S6, Canada bAberdeen Center for Energy Regulation and Obesity (ACERO), School of Biological Sciences, University of Aberdeen, Aberdeen, Scotland, AB242TZ, UK

Abstract.Marked interindividual variation in metabolic rate suggests considerable complexity in energy management strategies, but attempts to further our understanding of the relationship between resting metabolic rate (RMR), daily energy expenditure (DEE), and reproductive effort have been hampered by the complexity of studying this system in the field. Here, we describe energy management strategies in a captive-breeding system, using Zebra Finches (Taeniopygia guttata), to demonstrate the high level of complexity and interindividual variability in energy expenditure, food intake, locomotor activity, and reproductive effort. In particular, we investigated whether the increase in RMR associated with egg production is additive, resulting in higher DEE and a need for elevated food intake, or whether this cost is compensated by reduced expenditure in nonreproductive components of the energy budget. We found high levels of intra-individual variation in energy expenditure associated with egg production in female Zebra Finches, e.g., comparing nonbreeding stage with the one-egg stage, change in RMR varied from 4.0% and 41.3%, and change in DEE varied from −33.3% to +46.4%. This variation was systematically related to aspects of locomotor activity and reproductive effort. Females with the largest increase in RMR during egg production decreased locomotor activity the most but still had increased DEE at the one-egg stage, and females with high DEE at the one-egg stage produced larger clutches. Our study suggests that females minimize increases in DEE during egg production through behavioral energy reallocation (reduced locomotor activity) but that individuals differ in their use of this strategy, which, in turn, is related to the absolute level of reproductive investment. This suggests a very complex, individually variable system of energy management to meet the demands of egg production.

Key words:basal metabolic rate (BMR); daily energy expenditure (DEE); egg production costs; energy reallocation; reproduction; resting metabolic rate (RMR); Taeniopygia guttata; Zebra Finch.

Manuscript received December 12, 2005; revised March 29, 2006; accepted April 3, 2006

Introduction Return to TOC

Energy is widely held to be a universal currency and biologists are increasingly considering how energy management affects the fitness of the individual (Ricklefs 1996). A common assumption is that basal metabolic rate (BMR) is directly related to an individual's capacity for sustained levels of high daily energy expenditure (DEE), i.e., that there is a generalizable empirical (and functional) relationship between BMR and DEE (e.g., Daan et al. 1990, Ricklefs 1996). Nevertheless, there is marked, but largely unexplained, interindividual variation in both BMR (or resting metabolic rate, RMR) and DEE in free-living animals (for reviews, see Bryant and Tatner 1991, Williams and Vézina 2001, Speakman et al. 2004). Although these data hint at considerable complexity in energy management strategies, attempts to further our understanding of the relationship between BMR and DEE, especially in the context of reproductive effort, have been hampered by the complexity of studying this system in the field, e.g., in obtaining comprehensive data on energy intake and locomotor activity in addition to BMR/DEE in the same individuals over multiple breeding stages (Williams and Vézina 2001).

Here, we focus on energy management strategies in a captive-breeding system, using Zebra Finches (Taeniopygia guttata; see Plate 1 ), to demonstrate the high level of complexity and interindividual variability in the relationship between RMR, DEE, and reproductive effort. We specifically consider energy management underlying costs of egg production since numerous recent studies have shown that females producing extra eggs do incur fitness costs that are apparent either within the breeding attempt (Heaney and Monaghan 1995, Monaghan et al. 1995, 1998, Nager et al. 2000) or in terms of long-term female survival (Visser and Lessells 2001). The mechanistic link between the cost of egg production and reproductive success or future survival is not clear, but increased energy investment required for the physiological process of egg formation (Williams 1998) may play a significant role. It is therefore important to understand variation in energy management (measurable through metabolic and behavioral parameters) in relation to energy intake (food consumption), and both nonreproductive and reproductive activities; something that would be very difficult in field studies. Recent empirical studies in both the field and laboratory have shown that the metabolic cost of egg production results in a 16–27% increase in resting metabolic rate (Chappell et al. 1999, Nilsson and Raberg 2001, Vézina and Williams 2002, 2005). Furthermore, the increased RMR of birds producing eggs is repeatable between breeding attempts (Vézina and Williams 2005). Such studies therefore raise an important question in relation to energy management: how does a female cope with a 16–27% increase in RMR in terms of her overall energy budget (as measured by DEE)? There are two likely possibilities: (1) the cost of egg production could be additive to other expenditures, meaning that daily energy expenditure (DEE) would also increase in response to egg production, and require increased food intake or utilization of stored energy, or (2) the additional energy demand could be compensated for by reducing or trading off some other component of the energy budget.

One proposed mechanism for energy reallocation as a management strategy involves a reduction in locomotor activity in reproducing females. This has been demonstrated in free-living gravid lizards (Cooper et al. 1990), birds (Ettinger and King 1980), and pregnant and lactating mammals (Poppitt et al. 1993, Speakman et al. 2001). Captive-breeding Zebra Finches similarly reduce their level of locomotor activity during egg production (Houston et al. 1995, Williams and Ternan 1999), suggesting they might use a behavioral strategy for energy reallocation to direct more ingested energy toward egg production. However, no studies to date have combined measures of locomotor activity and food intake with any direct measurements of energy expenditure. If energy reallocation occurs during egg production in Zebra Finches, then assuming that compensation is complete, and all individual utilize this strategy equally, overall daily energy expenditure for a given female should be relatively constant between nonbreeding and egg-producing stages, which would be consistent with the general lack of difference in average DEE among breeding stages reported for free-living birds (Williams and Vézina 2001).

To investigate individual variability in energy management during egg production and the potential for energy reallocation via behavioral mechanisms, we measured RMR, DEE, food intake, locomotor activity, and reproductive effort on the same individual female Zebra Finches. Our first aim was to compare DEE estimates of females measured at nonbreeding and breeding (laying and chick-rearing) stages in controlled captive conditions, thus reducing any confounding effects of ambient ecological conditions. Second, we evaluated how egg-producing females manage their overall energy budgets by comparing their DEE estimates with previous RMR measurements made at the same breeding stages for the same individuals. We examined interindividual variation as this represents the basis for natural selection and may allow us to better understand flexibility in how individual birds adjust to the energy cost of egg production (see Williams and Vézina 2001). We predicted that if female Zebra Finches used a behavioral strategy for energy reallocation, locomotor activity would decrease but food intake and DEE would not change from nonbreeding to egg-producing stages. Alternatively, if egg production costs were additive, then DEE should increase when the birds are forming eggs, due to the increase in RMR.

Material and Methods Return to TOC

Animal care

Zebra Finches were maintained in controlled environmental conditions (temperature 19–23°C; humidity 35–55%; constant light schedule, 14L:10D with lights on at 0800 hours) and all birds were provided with a mixed-seed diet (50:50 Panicum and white millet, approximately 12.0% protein, 4.7% lipid; Jamieson's Pet Food, Vancouver, British Columbia, Canada), water, grit, and cuttlefish bone (calcium) ad libitum. The birds also received a multivitamin supplement in the drinking water once per week. Between experiments, all birds were housed in large reserve cages (122 × 46 × 41 cm) in single-sex groups, and during experiments all pairs (nonbreeding female/female, and breeding male/female; see Experimental groups and general protocol) were housed in separate breeding cages (61 × 46 × 41 cm) provided with an external nest box (15 × 14.5 × 20 cm). For experiments with single-sex pairs, access to the box was blocked by closing the entrance with a piece of cardboard. During breeding experiments, nest boxes were checked daily between 1000 and 1200 hours, and all new eggs were weighed (±0.001 g) and numbered. A clutch was considered complete after two consecutive days with no new eggs. All experiments and animal husbandry were carried out under a Simon Fraser University Animal Care Committee permit (692B-94), following the guidelines of the Canadian Committee on Animal Care.

Experimental groups and general protocol

All birds were adult individuals (>3 months of age) with previous successful breeding experience. Females were first paired as nonbreeding female/female pairs (NB) for seven days, food consumption and locomotor activity were measured on days 5, 6, and 7 of the nonbreeding period, and DEE was measured from day 6 to day 7 using the doubly labeled water (DLW) technique. On day 8, all female/female pairs were separated and each female was paired with a male to form a male/female breeding pair (see Plate 1 ). Locomotor activity was monitored in breeding pairs starting the following day through until clutch completion. Food intake was recorded on day 1 and 2 after pairing (pre-laying: PL) and again during egg laying (LY) beginning the day prior to laying of the first egg and during the four following days. All females had their DEE measured at the one-egg stage (i.e., on the day they laid their first egg; LY-1) with DEE estimates encompassing a complete ovulation and laying cycle (second egg). The birds were then left to incubate their eggs and raise their chicks (see Plate 1 ). Ten of the females paired with males did not produce eggs. We later compared the activity data from these unsuccessful breeding pairs with activity patterns of successful breeders (see Williams and Ternan 1999).

Food intake and locomotor activity were measured again in successful breeders between days 16 and 18 of the chick-rearing period (CK; 2–4 days before fledging) and breeding female DEE was measured from day 17–18 during the chick-rearing stage. If chicks fledged before day 16 (n = 2 pairs), locomotor activity data were discarded because of the bias introduced by chicks hopping on the perches. We used a repeated-measures approach to compare DEE values of 24 females measured as nonbreeders, at the one-egg stage and during chick-rearing (in females that successfully raised chicks, sample sizes are: NB = 24; LY-1 = 24; CK = 11).

Preliminary experiments showed that we could not measure RMR and DEE in the same breeding attempt without causing breeding failure. Therefore, to compare changes in RMR and DEE we used data on RMR measured in the same individual females and for the same breeding stages, but in an earlier breeding experiment using exactly the same protocol as described above (Vézina and Williams 2005). Since RMR is repeatable over relatively long periods (months to years) in both nonbreeding and egg-laying birds (Rønning et al. 2005, Vézina and Williams 2005) we consider the individual change in RMR, measured in the first breeding attempt, a good reflection of the metabolic investment in egg production specific to a given female. Although we measured all birds as nonbreeders before their breeding stages, we do not believe that any time-dependent effects confounded this study because all birds had already bred previously. Furthermore, the time intervals between pairing and laying, as well as egg and clutch mass, are repeatable between breeding attempts in Zebra Finches (T. D. Williams, unpublished data) as is reproductive effort measured as both RMR and DEE at LY-1 (Vézina and Williams 2005; F. Vézina, unpublished data).

Locomotor activity

We monitored locomotor activity at all breeding stages using a micro-switch system connected to a cage perch as described and validated by Williams and Ternan (1999). This system does not discriminate potential differences between sexes in locomotor activity, but validation of the technique using direct observation via video recording clearly showed that activity does not differ with sex and that the two birds of a pair follow the same activity pattern (Williams and Ternan 1999). Williams and Ternan (1999) also showed that time spent in the nest box is not dependent on sexes. Because we could only monitor locomotor activity in 14 cages at a time, the complete data set was gathered over five identical experimental runs.

Food intake

To obtain a gross estimate of energy intake we measured food consumption by giving birds 25 g/d of seeds in an open 946-mL plastic food container placed on the cage floor. This avoided any spillage and allowed us to measure seed mass intake by weighing the remaining seeds and empty husks left in the container after each 24-h period. Williams and Ternan (1999) showed that, on average, females eat slightly more food (4.5%) than males, but this difference was significant only on the two days before the first egg was laid. Measuring food intake per pair is therefore a good indicator of female food intake in our experimental context, and we report the pair values as representative of female energy input. During the experiment, the birds received 6 g of egg food supplement (20.3% protein, 6.6% lipid) daily, which was always completely consumed, by both males and females, by the following day.

Daily energy expenditure (DEE)

We measured DEE in females that successfully laid eggs, using the doubly labeled water technique (Lifson and McClintock 1966, Speakman 1997). The injection solution was made of 4.2182 g H₂¹⁸O (96.1 atom%), 3.9372 g H₂¹⁸O (95 atom%), and 3.8825 g ²H₂O (99.8 atom%). Using a high-precision gas tight syringe (Hamilton, Reno, Nevada, USA), all individuals received 76 μL of labeled water intramuscularly in the right pectoral muscle. To determine the exact amount of water injected, the syringe was weighed (±0.0001 g) before and after each injection. The birds were then returned to their cages for an hour to allow for isotope equilibration with the body water pool (Williams and Nagy 1984). After the equilibration period, the birds were weighed (±0.1 g) and blood was sampled by puncturing the brachial vein of one wing. About 100 μL of blood was collected per individual into micro-hematocrit tubes that were immediately flame sealed after releasing the birds back in their cages. Twenty-four hours later, a second blood sample was taken from the other wing following the same procedure. The birds were then reweighed and released into their cages. Background isotope enrichments can vary significantly over time, and this may result in significant errors in CO₂ production measurements (reviewed in Williams and Vézina 2001). Because background levels are mainly affected by the enrichment of the input sources: water, food, and atmospheric oxygen (Speakman and Racey 1987, Tatner 1988, 1990, Thomas et al. 1994), it is unlikely that our birds would exhibit marked background enrichment variation in controlled conditions, as the only variable input source was drinking tap water. However, we were concerned about changes in body composition between breeding stages and a possible time effect (Williams and Vézina 2001). We therefore took blood samples for background measurements at all breeding stages and all experimental runs (total background sample size NB = 10, LY-1 = 5, CK = 3). All samples were stored at 4°C until distillation and analysis.

Samples of blood were vacuum distilled into glass Pasteur pipettes (Nagy 1983) and the water obtained was used for isotope-ratio mass spectrometric analysis of ²H and ¹⁸O. The ²H analysis was performed on hydrogen gas, produced by on-line chromium reduction of water (Morrison et al. 2001). For analysis of ¹⁸O enrichment in blood samples, water distilled from blood was equilibrated with CO₂ gas using the small-sample equilibration technique (Speakman et al. 1990). For analysis of ¹⁸O:¹⁶O ratios, equilibrated water samples were admitted to an ISOCHROM μGAS system (Micromass UK, Waters Corporation, Milford, Massachusetts, USA), which uses a gas chromatograph column to separate nitrogen and CO₂ in a stream of helium gas before analysis by isotope ratio mass spectrometry (IRMS). Total body water estimates were based on the plateau technique (Speakman 1997) and DEE was calculated from CO₂ production using Speakman (1997) equation 7.17, taking evaporative water loss into account and assuming a RQ value of 0.8. All sample analyses and calculations were performed blind of the experimental manipulations.

Resting metabolic rate (RMR)

RMR (metabolic rate of birds in post-absorptive state, measured during the inactive phase of the circadian cycle at a temperature within the thermoneutral zone) was measured as described in Vézina and Williams (2005) using a flow-through respirometry system (Sable Systems International, Las Vegas, Nevada, USA). Since our DEE data are presented in kJ/d, to obtain RMR values in comparable units we converted VO₂ (mL O₂/h) to kJ/d. Average nighttime respiratory quotient (RQ) was 0.82 and showed a large degree of variability between individuals (maximal range 0.69 to 0.99), but no differences between breeding stages (repeated-measures ANOVA, F_2,28 = 0.4, P = 0.7; NB = 0.81 ± 0.01, LY-1 = 0.83 ± 0.01, CK = 0.84 ± 0.03 [mean ± SE]). Daytime RQ was 0.85 (range 0.73 to 1.01, n = 14; F. Vézina, unpublished data). We cannot consider that individual RQ remained constant within a 24-h period because RQ values are not stable through time (Powers 1991, Walsberg and Wolf 1995, Walsberg and Hoffman 2005). Therefore, we used an energy equivalent of 20.38 kJ/L O₂, reflecting an averaged RQ value of 0.835 (Weir 1949). Gessaman and Nagy (1988) showed that using a RQ of 0.8 to convert VO₂ values into kilojoules results in less than ±3% error regardless of the actual protein, fat, or carbohydrate mixture being catabolized.

Statistical analysis

All data were tested to ensure normality (Shapiro-Wilk test; Zar 1996). Metabolic rate varies allometrically with body mass (Schmidt-Nielsen 1984), but using log₁₀-transformed values did not change any of the statistical results compared to untransformed data. We therefore present results of analysis on untransformed values (body mass in grams and RMR and DEE in kJ/d). To compare within-individual changes from the nonbreeding to one-egg and chick-rearing stages, we used repeated-measures ANOVA and ANCOVA with body mass as a covariate when appropriate (for RMR and DEE). This was performed using mixed general linear models that included individual bird as random factor, breeding stage as the main fixed effect, and body mass as covariate if necessary. We removed the interaction term body mass × breeding stage from the ANCOVA when not significant (i.e., indicating no difference in slopes). Post hoc comparisons between stages were performed using Tukey hsd and multiple contrasts (when using a covariate) corrected with the Bonferroni procedure (Rice 1989). Within breeding stage, relationships between continuous variables were investigated using Pearson's correlation analyses. Data are reported as means ± SE.

Results Return to TOC

Locomotor activity

We first investigated locomotor activity to confirm the potential for behaviorally based energy reallocation (sensu Williams and Ternan 1999). Locomotor activity (number of perch hops per day) was independent of male or female body mass at all breeding stages (P 0.2 in all cases) but was highly dependent on breeding stage (F_3,48 = 10.5, P < 0.0001; Fig. 1 A). Mean activity was higher at the nonbreeding stage compared to all other stages, and was 42.1%, 61.1%, and 52.9% lower during the pre-laying, laying, and chick-rearing stages, respectively. Mean locomotor activity did not differ significantly among these three stages (Fig. 1 A) and, during the laying stage, it was independent of mean egg mass, clutch mass, and clutch size (P > 0.5 in all cases). Absolute locomotor activity levels varied considerably between breeding pairs, so we standardized these data by calculating the percentage of deviation in daily activity from mean activity for the whole laying period for each individual pair (following Williams and Ternan 1999). Locomotor activity in breeding pairs decreased rapidly between eight and four days before the onset of laying (Fig. 2 A) but thereafter the activity level remained low and constant until clutch completion (overall F_15,203 = 6.0, P < 0.0001; independent contrast day −8 to day −4 F_1,203 = 15.2, P < 0.0001; day −4 to day 8 F_1,203 =1.8, P = 0.2). This pattern of change in locomotor activity was not apparent in the unsuccessful male/female pairs that did not produce eggs (F_15,114 = 1.2, P = 0.3; Fig. 2 B).

Food intake

Food intake was independent of male, female, or combined (pair) body mass at all breeding stages (P 0.1 in all cases) but was related to breeding stage (F_3,52 = 52.3, P < 0.0001; Fig. 1 B). However, mean food intake did not differ among nonbreeding, pre-laying, and laying stages (average intake; NB = 5.0 ± 0.3 g/d, PL = 4.8 ± 0.3 g/d, LY = 5.4 ± 0.3 g/d) and only increased markedly during chick-rearing (10.6 ± 0.4 g/d) due to food being consumed by the chicks. None of our measures of female reproductive effort (clutch size, clutch mass, mean egg mass) were related to either mean (four days) food intake during laying (P 0.4 in all cases) or to food intake the day prior to laying of the first egg (P 0.2 in all cases).

DEE in relation to locomotor activity, food intake, and reproductive effort

DEE was independent of female body mass for all breeding stage (P 0.2 in all cases). At the nonbreeding stage, DEE was also independent of locomotor activity recorded between the two blood samples (P = 0.5), and mean food intake over the three days (5, 6, and 7) prior to sampling (P = 0.1). However, DEE was positively correlated with food intake at day 6 (the day of DEE measurement; r = 0.51, n = 24, F_1,22 = 7.7, P < 0.05; Fig. 3 ). In contrast, DEE at the one-egg stage was positively correlated with locomotor activity recorded between the blood samples (r = 0.48, n = 23, F_1,21 = 6.1, P < 0.05; Fig. 4 A), i.e., during egg production more active birds spent more energy on a daily basis. Furthermore, average food intake per pair, measured either over four days or on the day of DEE measurement, was significantly correlated with DEE at the one-egg stage (mean food intake, r = 0.51, n = 24, F_1,22 = 7.7, P < 0.05; day of LY-1, r = 0.52, n = 24, F_1,22 = 8.3, P < 0.01; Fig. 4 B). Finally, during chick-rearing, DEE was independent of locomotor activity and food intake (P > 0.5 in both cases).

On average, females laid 6.0 ± 0.4 eggs (range: 4–10 eggs) and mean egg mass was 1.040 ± 0.017 g (range: 0.841–1.201 g). DEE at the one-egg stage was positively correlated with both clutch size (r = 0.49, n = 24, F_1,22 = 6.9, P < 0.05) and total clutch mass (r = 0.46, n = 24, F_1,22 = 6.1, P < 0.05; Fig. 5 A and B); i.e., during egg-laying overall energy expenditure was positively related to reproductive effort. In contrast, during chick-rearing DEE was independent of brood size and total brood mass at 21 days (P 0.3 in both cases).

Individual variation in RMR and DEE and relationship between these traits

For the subset of RMR birds for which we measured DEE, mean resting metabolic rate varied with breeding stage, as shown by Vézina and Williams (2005; repeated-measures ANCOVA with body mass as a covariate; F_2,27 = 17.0, P < 0.0001, no significant interaction; least-square mean NB RMR = 19.8 ± 0.4 kJ/d, LY-1 RMR = 23.3 ± 0.5 kJ/d, CK RMR = 23.1 ± 0.9 kJ/d). RMR increased by 3.5 kJ/d, or 17.6% from the nonbreeding to the one-egg stage and remained high during chick-rearing. However, there was marked interindividual variation in the change in RMR (ΔRMR) from the nonbreeding to the one-egg stage ranging from 4.0% to 41.3%. This individual variability in ΔRMR was independent of both nonbreeding and one-egg stage body mass and was independent of the change in mass between the two stages (P 0.3 in all cases). Considering the change in RMR in relation to locomotor activity, only three out of 23 pairs showed an increase in activity comparing nonbreeding stage with one-egg stage. One of these pairs represented a clear outlier with a 97% increase in activity (5.9 times the standard deviation of the mean difference in activity). Excluding this pair, there was a significant negative relationship between the relative change in locomotor activity (percent difference compared to nonbreeding level) and RMR at the one-egg stage (r = −0.57, n = 22, P < 0.01; Fig. 6 ). Thus, individuals that had the highest one-egg stage RMR were also the ones showing the largest reduction in locomotor activity from nonbreeding to one-egg stage, suggesting that the level of behavioral energy compensation (reduced locomotor activity) for the cost of producing eggs is a function of the individual's initial metabolic investment.

In contrast to the within-stage analysis, there was a significant overall effect of body mass on DEE (F_1,30 = 15.3, P < 0.001) as well as a significant breeding-stage effect (F_2,30 = 6.6, P < 0.01) and interaction term, body mass × breeding stage (F_2,30 = 6.1, P < 0.01). The mass × stage interaction was due to a slightly higher DEE at the chick-rearing stage (5.6%) compared with the one-egg stage (no significant difference between nonbreeding and one-egg stages, independent contrast P = 0.9; mean NB DEE = 53.6 ± 1.0 kJ/d, LY-1 DEE = 51.3 ± 1.3 kJ/d, CK DEE = 54.2 ± 1.8 kJ/d). Excluding the chick-rearing stage from the analysis resulted in no significant breeding-stage effect (F_1,21 = 0.04, P = 0.8). Again there was marked individual variation in the change in DEE (ΔDEE) comparing nonbreeding and one-egg stage values, with ΔDEE varying between −33.3% and +46.4%. As a consequence of this individual variation mean DEE did not change significantly within females between nonbreeding and one-egg stage (controlling for body mass). Individual variability in ΔDEE between nonbreeding and one-egg stages was independent of the relative change in locomotor activity for that period (P = 0.7), and was independent of body mass at both the nonbreeding and one-egg stages (P 0.1). However, ΔDEE was significantly related to the gain in mass between the two stages (r = 0.63, n = 24, F_1,22= 14.8, P < 0.001), which partly reflects the mass of the reproductive machinery at the one-egg stage (T. D. Williams and C. E. Ames, unpublished data).

Residual RMR (correcting for body mass) and DEE were positively correlated at the nonbreeding stage (r = 0.59, n = 24, F_1,22 = 11.6, P < 0.005; Fig. 7 A) but not at the one-egg stage (P = 0.7; Fig. 7 B). Furthermore, ΔRMR from the nonbreeding to one-egg stage was positively correlated with ΔDEE for the same period (r = 0.42, n = 24, F_1,22 = 4.7, P < 0.05; Fig. 8 ). When corrected for the effect of the gain in body mass on the change in DEE, the relationship remained significant (residual ΔDEE vs. ΔRMR; r = 0.41, n = 24, F_1,22 = 4.6, P < 0.05).

Discussion Return to TOC

Although we used a captive-breeding species, the Zebra Finches, in controlled and presumably relatively benign breeding conditions, we have demonstrated a high level of complexity and interindividual variability in energy management strategies, specifically in the relationship between RMR, DEE, and reproductive effort. We believe that these results are relevant for free-living birds and although a comprehensive study of this nature has so far proved elusive in field studies we hope our data will inform future experimental fieldwork. Our study suggests that the metabolic cost of egg production requires individuals to make behavioral adjustments that minimize large variations in overall energy expenditure (DEE): female Zebra Finches reduced energy expenditure devoted to locomotor activity during egg production but did not change food intake. This is in agreement with previous studies (Williams and Ternan 1999) and supports the hypothesis of behavioral energy reallocation. On average, RMR increased by 18% in response to egg production but this mean increase masked a high level of individual variability (+4–40%). Similarly, we did not detect a significant change in average DEE during the same period, and this was due to even greater individual variability in the change in DEE (−33% to +46%). Importantly however, this variability in RMR and DEE was not random but was systematically related to aspects of locomotor activity and reproductive effort. For example, birds that had high levels of DEE at the one-egg stage produced large clutches, and individual variation in the change in RMR associated with egg production was positively correlated with the change in DEE. Increased RMR associated with egg production is consistent with previous observations of elevated RMR during egg production in Zebra Finches and other passerines (Chappell et al. 1999, Nilsson and Raberg 2001, Vézina and Williams 2002, 2005). In our study, we interpret the individual variability in the change in RMR associated with egg production as a measure of the metabolic demands of investment in reproductive tissues. We base this interpretation on the fact that positive relationships between RMR and oviduct mass have been demonstrated in three species including Zebra Finches (Chappell et al. 1999, Vézina and Williams 2003, 2005). The size of the oviduct has been shown to be related to egg size in European Starlings Sturnus vulgaris (Ricklefs 1976, Vézina and Williams 2003) and egg quality may also be related to oviduct size. Indeed, Christians and Williams (2001) have shown a positive relationship between egg albumen protein content and oviduct size in European Starlings. Thus, generating a larger oviduct potentially allows for the production of larger, better eggs, but this incurs a higher metabolic cost visible in elevated RMR. Therefore the question remains: how do females adjust their energy budget relative to their costs of physiological reproductive investment?

The marked interindividual variation in RMR and DEE that we have documented clearly confounds the interpretation of average, population-level changes in energy expenditure because mean values mask the particular direction of changes within individuals. Thus, although mean RMR increased from nonbreeding to one-egg and chick-rearing stages, mean DEE remained constant in our study. This was because, at the one-egg stage, mean DEE was based on measurements of individuals that had either increased or decreased their overall energy expenditure, and this variation in change in DEE cancelled out such that there was no statistical difference in the mean values. We suggest that females “manage” their energy budget during egg production in order to minimize potential increases in DEE by adjusting their behavior, i.e., decreasing locomotor activity, to compensate for the increase in RMR. In fact, the individuals showing the largest reproductive investment in terms of RMR at the one-egg stage were also members of pair that decreased locomotor activity the most (down to −86%). Individual variation in measures of energy expenditure has been reported in earlier studies (Moreno 1989, Williams 1987) but the basis of this variation has not been investigated (see Williams and Vézina 2001). Our results clearly demonstrate that more attention needs to be given to this level of variability as it may help us to understand adjustments that individual animals have to make to balance their energy budget.

Reduction of activity as a strategy for energy reallocation

Birds show decreased activity when facing thermoregulatory challenges (Cherel et al. 1988) or the cost of molting into new feathers (Austin and Fredrickson 1987, Robin et al. 1989). In Zebra Finches, food limitation also resulted in reduced locomotor activity (Meijer et al. 1996, Dall and Witter 1998), and egg production has now been shown in three independent studies to be associated with reduced activity (Houston et al. 1995, Williams and Ternan 1999; this study). Decreases in activity have also been documented in free-living Willow Flycatchers (Empiodonax traillii) during egg production (Ettinger and King 1980) and in mammals such as pregnant female common shrew (Sorex araneus; Poppitt et al. 1993) and rats (Slonaker 1924), lactating mice (Mus musculus; Speakman et al. 2001) and rats (Slonaker 1924, Wang 1924), as well as pregnant women (Butte et al. 2004). This suggests that reducing locomotor activity is a widespread behavioral mechanism that helps to compensate for temporarily increased energetic demands from a range of other physiological processes. In contrast to marked changes in behavior associated with reproductive investment, Zebra Finches in our study did not change food (energy) intake. Energy expenditure in Zebra Finches does not appear to be limited by their food processing capabilities. Wild Australian Zebra Finches eat 28–76% more seeds per day than their domesticated counterpart (Zann 1996) and captive breeding pairs kept in cold temperatures (7°C) consume 72% more seeds per day than pairs kept in normal conditions (21°C; K. G. Salvante, F. Vézina, and T. D. Williams, unpublished data). Nevertheless, our experiment demonstrated that even if these birds are potentially capable of sustaining higher levels of DEE by increasing food intake during egg production, egg-laying females appeared to avoid this strategy.

Additive vs. compensated costs of egg production

At the individual level, it appeared that energy investment in egg production (increased RMR) generated a wide spectrum of effects on DEE, from overcompensation (decrease in DEE) to additive effects (increase in DEE). This was shown by the positive relationship between the change in RMR from the nonbreeding to the one-egg stage and the change in DEE during the same period (see Fig. 8 ). Individuals with relatively high reproductive effort (large increase in RMR) appeared to not be able to fully compensate for this increased energy demand. Even though they showed the largest decrease in locomotor activity, they still had an elevation in DEE. Thus, in these birds, there was an additive effect of the cost of egg production on DEE. In contrast, individuals with low reproductive effort (small increase in RMR) were much better at avoiding an increase in DEE and, in some cases, even overcompensated for the elevated RMR (reduction in DEE during egg production; see Fig. 8 ). Therefore, behavioral adjustments allowed all birds to decrease the total energy expended, but the level of energy reallocation appeared to depend on the actual level of reproductive investment. Ultimately, the females with the highest DEE at the one-egg stage produced larger clutches, which suggests that they did benefit from their higher metabolic investment.

Relationship between RMR and DEE

We found a significant positive correlation between RMR and DEE at the nonbreeding stage, but this relationship was lost when the birds were actively producing eggs. A general interpretation for such a correlation is that increased sustained energy expenditure is supported by enlarged organs which results in higher maintenance costs and thus elevated RMR (Peterson et al. 1990, Hammond and Diamond 1997, Piersma and Lindström 1997, Piersma 2002; but see Ricklefs et al. 1996, Speakman 2000, Speakman et al. 2003, 2004). Our nonbreeding birds showed the highest levels of locomotor activity. However, DEE was not related to activity at this stage, which does not provide strong support for this interpretation. If the relationship between RMR and DEE in nonbreeding individuals was due to a common effect of the underlying physiological machinery, then its influence on energy expenditure seems unrelated to locomotor activity as measured by hopping counts. Speakman et al. (2003) suggested that RMR and DEE follow independent trajectories that may or may not correlate in response to direct ecological conditions that the organism faces and this may be the case in our egg-laying birds. The lack of relationship between RMR and DEE in females producing eggs may have been because these two variables were responding differently to the demands of egg production. Mean RMR increased in response to egg production, but mean DEE remained relatively unchanged because females individually adjusted their behavior to maintain a constant overall energy expenditure. The birds that showed the largest increase in RMR were also the ones compensating the most. Therefore, they minimized the impact of an increased RMR on DEE, which resulted in a correlation between these two variables showing a slope not significantly different from zero. In other words, we suggest that reallocation of energy during egg formation uncoupled the relationship between RMR and DEE.

DEE and egg production costs

We are aware of only two studies on birds specifically investigating energy expenditure, using the doubly labeled water technique, during egg production. Stevenson and Bryant (2000) demonstrated a positive correlation between DEE and egg mass in one of two years in free-living Great Tits (Parus major), but did not report nonbreeding or chick-rearing DEE or potential relationships between DEE and clutch size. Consistent with the present data, Ward (1996) found no significant differences between DEE of wild Barn Swallows (Hirundo rustica) measured during egg-formation, incubation, or chick-rearing. Ward (1996) suggested that egg production costs are small relative to routine energy requirements because DEE was not significantly elevated or correlated with egg energy content. We disagree somewhat with Ward's (1996) interpretation. The energy expended in egg formation results from several physiological processes that have their own associated energetic costs (reproductive organ development, function and maintenance, hepatic yolk precursor production and uptake by the ovary, albumen and shell deposition in the oviduct). The energy spent in these physiological activities is not directly comparable to the chemical energy content of the eggs (see Vézina and Williams 2002). In fact, our data suggest that egg-producing females compensate for the extra demand associated with egg formation by decreasing the energy expended in locomotion in relation to their change in RMR. Furthermore, this energy compensation strategy appears to be closely related to the period of follicular growth as the reduction in activity occurred just before the onset of rapid yolk development (rapid yolk development begins four days prior to laying of the first egg in Zebra Finches [Haywood 1993:Fig. 2 a]). Although a 16–27% increase in RMR (Nilsson and Raberg 2001, Vézina and Williams 2002, 2005) may represent a relatively small fraction of DEE, the fact that average DEE does not increase significantly during egg production may not mean that the amount of energy invested in egg production is trivial, but rather that this cost is demanding enough for the animal to adopt behavioral strategies to minimize the increase in overall energy expenditure. That we find no significant differences in DEE between egg-producing and other life-history stages, some of which are believed to be energetically costly (i.e., chick-rearing [Drent and Daan 1980]), suggests that other expensive activities are also subject to compensatory adjustments. In agreement with this idea, Williams and Vézina (2001) compiled absolute values of DEE (kJ/d) in studies where measurements were obtained at more than one physiological stage within species; they found a general tendency for a constant level of DEE during all phases of reproduction and perhaps even throughout the year.

In conclusion, our study suggests that females minimize increases in DEE during egg production through behavioral energy reallocation (reduced locomotor activity) but that individuals differ in their use of this strategy which, in turn, is related to the absolute level of reproductive investment. This suggests a very complex, individually variable system of energy management to meet the demands of egg production. We suggest that this complexity might help explain why so few studies of free-living birds have found support for positive relationships between energy expenditure and putative correlated ecological variables (e.g., temperature, food availability, parental provisioning effort; Williams and Vézina 2001), and why some studies report contradictory results in different years (e.g., Stevenson and Bryant 2000).

Acknowledgments

We are grateful to Oliver P. Love, Katrina G. Salvante, Julian K. Christians, François Fournier, Piet van den Hout, Mark A. Chappell, Joseph B. Williams, and Theunis Piersma for stimulating discussions and comments on earlier versions of this article. We also thank Dr. Paula Redman and Peter Thomson for DLW analysis as well as Andréa and Jean-Baptiste Vin for their help with bird care and data collection and processing. A special thanks to Ray Holland, who designed and built the automated micro-switch monitoring system that allowed to record hopping activity in Zebra Finches. This research was funded by an operating NSERC grant to T. D. Williams, post-graduate funding to F. Vézina from NSERC and FCAR, and student research grants from SICB, Sigma Xi and AOU to F. Vézina.

Literature Cited Return to TOC

Austin, J. E., and L. H. Fredrickson. 1987. Body and organ mass and body composition of postbreeding female Lesser Scaup. Auk 104: 694–699.

Bryant, D. M., and P. Tatner. 1991. Intraspecies variation in avian energy expenditure. Ibis 133: 236–245.

Butte, N. F., W. W. Wong, M. S. Treuth, K. J. Ellis, and E. O. Smith. 2004. Energy requirements during pregnancy based on total energy expenditure and energy deposition. American Journal of Clinical Nutrition 79: 1078–1087.

Chappell, M. A., C. Bech, and W. A. Buttemer. 1999. The relationship of central and peripheral organ masses to aerobic performance variation in house sparrows. Journal of Experimental Biology 202: 2269–2279.

Cherel, Y., J. P. Robin, and Y. Le Maho. 1988. Physiology and biochemistry of long term fasting in birds. Canadian Journal of Zoology 66: 159–166.

Christians, J. K., and T. D. Williams. 2001. Intraspecific variation in reproductive physiology and egg quality in the European Starling Sturnus vulgaris. Journal of Avian Biology 32: 31–37.

Cooper, W. E. Jr., L. J. Vitt, R. Hedges, and R. B. Huey. 1990. Locomotor impairment and defense in gravid lizards (Eumeces laticeps): behavioral shift in activity may offset costs of reproduction in an active forager. Behavioral Ecology and Sociobiology 27: 153–157.

Daan, S., D. Masman, and A. Groenewold. 1990. Avian basal metabolic rates: their association with body composition and energy expenditure in nature. American Journal of Physiology 259: R333–R340.

Dall, S. R. X., and M. S Witter. 1998. Feeding interruptions, diurnal mass changes and daily routines of behaviour in Zebra Finch. Animal Behaviour 55: 715–725.

Drent, R. H., and S. Daan. 1980. The prudent parent: energetic adjustments in avian breeding. Ardea 68: 225–252.

Ettinger, A. O., and J. R. King. 1980. Time and energy budgets of the Willow Flycatcher (Empidonax traillii) during the breeding season. Auk 97: 533–546.

Gessaman, J. A., and K. A. Nagy. 1988. Energy metabolism: errors in gas-exchange conversion factors. Physiological Zoology 61: 507–513.

Hammond, K. A., and J. Diamond. 1997. Maximal sustained energy budgets in humans and animals. Nature 386: 457–462.

Haywood, S. 1993. Sensory control of clutch size in the Zebra Finch (Taeniopygia guttata). Auk 110: 778–786.

Heaney, V., and P. Monaghan. 1995. A within-clutch trade-off between egg production and rearing in birds. Proceedings of the Royal Society of London B 261: 361–365.

Houston, D. C., D. Donnan, and P. J. Jones. 1995. The source of the nutrients required for egg production in Zebra Finches Poephila guttata. Journal of Zoology London 235: 469–483.

Lifson, N., and R. McClintock. 1966. Theory of use of the turnover rates of body water for measuring energy and material balance. Journal of Theoretical Biology 12: 46–74.

Meijer, T., J. Rozman, M. Schulte, and C. Stach-Dreesmann. 1996. New findings in body mass regulation in Zebra Finches (Taeniopygia guttata) in response to photoperiod and temperature. Journal of Zoology, London 240: 717–734.

Monaghan, P., M. Bolton, and D. C. Houston. 1995. Egg production constraints and the evolution of avian clutch size. Proceedings of the Royal Society of London B 259: 189–191.

Monaghan, P., R. G. Nager, and D. C. Houston. 1998. The price of eggs: increased investment in egg production reduces the offspring rearing capacity of parents. Proceedings of the Royal Society of London B 265: 1731–1735.

Moreno, J. 1989. Variation in daily energy expenditure in nesting northern wheatears Oenanthe oenanthe. Auk 106: 18–25.

Morrison, J., T. Brockwell, T. Merren, F. Fourel, and A. M. Phillips. 2001. On-line high-precision stable hydrogen isotopic analyses on nanoliter water samples. Analytical Chemistry 73: 3570–3575.

Nager, R. G., P. Monaghan, and D. C. Houston. 2000. Within-clutch trade-offs between the number and quality of eggs: experimental manipulations in gulls. Ecology 81: 1339–1350.

Nagy, K. A. 1983. The doubly labeled water (³HH¹⁸O) method: a guide to its use. UCLA Publication no. 12-1417 University of California, Los Angeles, California, USA.

Nilsson, J-A., and L. Raberg. 2001. The resting metabolic cost of egg laying and nestling feeding in Great Tits. Oecologia 128: 187–192.

Peterson, C. C., K. A. Nagy, and J. Diamond. 1990. Sustained metabolic scope. Proceedings of the National Academy of Sciences (USA) 87: 2324–2328.

Piersma, T. 2002. Energetic bottlenecks and other design constraints in avian annual cycles. Integrative and Comparative Biology 42: 51–67.

Piersma, T., and Å. Lindström. 1997. Rapid reversible changes in organ size as a component of adaptive behaviour. Trends in Ecology and Evolution 12: 134–138.

Poppitt, S. D., J. R. Speakman, and P. A. Racey. 1993. The energetics of reproduction in the common shrew (Sorex araneus): a comparison of indirect calorimetry and the doubly labeled water method. Physiological Zoology 66: 964–982.

Powers, D. R. 1991. Diurnal variation in mass, metabolic rate, and respiratory quotient in Anna's and Costa's hummingbirds. Physiological Zoology 64: 850–870.

Rice, W. R. 1989. Analyzing tables of statistical tests. Evolution 43: 223–225.

Ricklefs, R. E. 1976. The chemical composition of the ovary, oviduct and follicles of the starling. Auk 93: 184–187.

Ricklefs, R. E. 1996. Avian energetics, ecology and evolution. Pages. 1–30. in C. Carey, editor. Avian energetics and nutritional ecology Chapman and Hall. New York, New York, USA.

Ricklefs, R. E., M. Konarzewski, and S. Daan. 1996. The relationship between basal metabolic rate and daily energy expenditure in birds and mammals. American Naturalist 147: 1047–1071.

Robin, J.-P., Y. Handrich, Y. Cherel, and Y. Le Maho. 1989. Energy saving during breeding and molt in birds. Pages. 293–304. in C. Bech and R. E. Reinertsen, editors. Physiology of cold adaptation in birds Plenum Press. New York, New York, USA.

Rønning, B., B. Moe, and C. Bech. 2005. Long-term repeatability makes basal metabolic rate a likely heritable trait in the zebra finch Taeniopygia guttata. Journal of Experimental Biology 208: 4663–4669.

Schmidt-Nielsen, K. 1984. Scaling: why is animal size so important? Cambridge University Press, New York, New York, USA.

Slonaker, J. R. 1924. The effect of copulation, pregnancy, pseudo-pregnancy and lactation in the voluntary activity and food consumption of the albino rat. American Journal of Physiology 71: 362–394.

Speakman, J. R. 1997. Doubly-labelled water: theory and practice Chapman and Hall, London, UK.

Speakman, J. R. 2000. The cost of living: field metabolic rates of small mammals. Advances in Ecological Research 30: 177–297.

Speakman, J. R., T. Ergon, R. Cavanagh, K. Reid, D. M. Scantlebury, and X. Lambin. 2003. Resting and daily energy expenditures of free-living field voles are positively correlated but reflect extrinsic rather than intrinsic effects. Proceedings of the National Academy of Sciences (USA) 100: 14057–14062.

Speakman, J. R., A. Gidney, J. Bett, I. P. Mitchell, and M. S. Johnson. 2001. Limits to sustained energy intake IV. Effect of variation in food quality on lactating mice Mus musculus. Journal of Experimental Biology 204: 1957–1965.

Speakman, J. R., E. Krol, and M. S. Johnson. 2004. The functional significance of individual variation in basal metabolic rate. Physiological and Biochemical Zoology 77: 900–915.

Speakman, J. R., K. A. Nagy, D. Masman, W. G. Mook, S. D. Poppitt, G. E. Strathearn, and P. A. Racey. 1990. Interlaboratory comparison of different analytical techniques for the determination of oxygen-18 abundance. Analytical Chemistry 62: 703–708.

Speakman, J. R., and P. A. Racey. 1987. The equilibrium concentration of oxygen-18 in body water: implication for the accuracy of the doubly labeled water technique and a potential new method of measuring RQ in free-living animals. Journal of Theoretical Biology 127: 79–95.

Stevenson, I. R., and D. M. Bryant. 2000. Climate change and constraints on breeding. Nature 406: 366–367.

Tatner, P. 1988. A model of the natural abundance of oxygen-18 and deuterium in the body water of animals. Journal of Theoretical Biology 133: 267–280.

Tatner, P. 1990. Deuterium and oxygen-18 abundance in birds: implication for DLW energetic studies. American Journal of Physiology 258: R804–R812.

Thomas, D. W., K. Martin, and H. Lapierre. 1994. Doubly labeled water measurements of field metabolic rate in White-tailed Ptarmigan: variation in background isotope abundance and effects on CO₂ production. Canadian Journal of Zoology 72: 1967–1972.

Vézina, F., and T. D. Williams. 2002. Metabolic costs of egg production in the European Starling (Sturnus vulgaris). Physiological and Biochemical Zoology 75: 377–385.

Vézina, F., and T. D. Williams. 2003. Plasticity in body composition in breeding birds: what drives the metabolic costs of egg production?. Physiological and Biochemical Zoology 76: 716–730.

Vézina, F., and T. D. Williams. 2005. The metabolic cost of egg production is repeatable. Journal of Experimental Biology 208: 2533–2538.

Visser, M. E., and C. M. Lessells. 2001. The cost of egg production and incubation in Great Tits (Parus major). Proceedings of the Royal Society of London B 268: 1271–1277.

Walsberg, G. E., and T. C. M. Hoffman. 2005. Direct calorimetry reveals large errors in respirometric estimates of energy expenditure. Journal of Experimental Biology 208: 1035–1043.

Walsberg, G. E., and B. O. Wolf. 1995. Variation in the respiratory quotient of birds and implications for indirect calorimetry using measurements of carbon dioxide production. Journal of Experimental Biology 198: 213–219.

Wang, G. H. 1924. The changes in the amount of daily food intake of the albino rat during pregnancy and lactation. American Journal of Physiology 71: 735–741.

Ward, S. 1996. Energy expenditure of female Barn Swallows Hirundo rustica during egg formation. Physiological Zoology 69: 930–951.

Weir, J. B. V. 1949. New methods for calculating metabolic rate with special reference to protein metabolism. Journal of Physiology 109: 1–9.

Williams, J. B. 1987. Field metabolism and food consumption of savannah sparrows during the breeding season. Auk 104: 277–289.

Williams, J. B., and K. A. Nagy. 1984. Validation of the doubly labelled water technique for measuring energy metabolism in savannah sparrows. Physiological Zoology 57: 325–328.

Williams, T. D. 1998. Avian reproduction, overview. Pages. 325–336. in E. Knobil and J. D. Neil, editors. Encyclopedia of reproduction. Volume I Academic Press. New York, New York, USA.

Williams, T. D., and S. P. Ternan. 1999. Food intake, locomotor activity, and egg laying in Zebra Finches: contributions to reproductive energy demand?. Physiological Zoology 72: 19–27.

Williams, T. D., and F. Vézina. 2001. Reproductive energy expenditure, intraspecific variation and fitness. Current Ornithology 16: 355–405.

Zann, R. A. 1996. The Zebra Finch: a synthesis of field and laboratory studies Oxford University Press, Oxford, UK.

Zar, J. H. 1996. Biostatistical analysis Prentice Hall, London, UK.

---

Figures Return to TOC

³Present address: Department of Marine Ecology and Evolution, Royal Netherlands Institute for Sea Research (NIOZ), P.O. Box 59, 1790 AB Den Burg, Texel, The Netherlands. E-mail: fvezina@nioz.nl

Corresponding Editor: S. J. Simpson.