According to the heat dissipation limit (HDL) theory, reproductive performance is limited by the capacity to dissipate excess heat. We tested the novel hypotheses that (1) the age-related decline in reproductive performance is due to an age-related decrease of heat dissipation capacity and (2) the limiting mechanism is more severe in animals with high metabolic rates. We used bank voles (Myodes glareolus) from lines selected for high swim-induced aerobic metabolic rate, which have also increased basal metabolic rate, and unselected control lines. Adult females from three age classes – young (4 months), middle-aged (9 months) and old (16 months) – were maintained at room temperature (20°C), and half of the lactating females were shaved to increase heat dissipation capacity. Old females from both selection lines had a decreased litter size, mass and growth rate. The peak-lactation average daily metabolic rate was higher in shaved than in unshaved mothers, and this difference was more profound among old than young and middle-aged voles (P=0.02). In females with large litters, milk production tended to be higher in shaved (least squares mean, LSM±s.e.: 73.0±4.74 kJ day−1) than in unshaved voles (61.8±4.78 kJ day−1; P=0.05), but there was no significan"t effect of fur removal on the growth rate [4.47±2.29 g (4 days−1); P=0.45]. The results provide mixed support of the HDL theory and no support for the hypotheses linking the differences in reproductive aging with either a deterioration in thermoregulatory capability or genetically based differences in metabolic rate.

The reproductive period of endotherms is extremely energy demanding, because of the costs of both increased foraging and the metabolic processes involved in provisioning the young. The lactation-induced elevation of metabolic rate in mammalian species has been described (Duarte et al., 2010) and can be considered as a by-product of selection for intense parental care. Sustained aerobic scope (the ratio between sustained metabolic rate and basic metabolic rate, BMR) describes the potential capacity of an animal to utilize energy over a prolonged period of time; for example, during its reproductive phase, relative to its basal maintenance requirements (Peterson et al., 1990; Hammond and Diamond, 1997). Sustained aerobic scope is limited to a range from 1.3 to 7 (Hammond and Diamond, 1997), but the underlying cause for this metabolic ceiling is still not fully understood.

List of symbols and abbreviations

     
  • ADEpeak

    peak-lactation apparent digestive efficiency

  •  
  • ADMRpeak

    peak-lactation average daily metabolic rate

  •  
  • A-lines

    lines selected for high swim-induced aerobic metabolism

  •  
  • BMR

    basal metabolic rate

  •  
  • C-lines

    lines unselected, control

  •  
  • DLW

    doubly labelled water

  •  
  • FCpeak

    peak-lactation food consumption

  •  
  • GRpeak

    peak-lactation litter growth rate

  •  
  • HDL

    heat dissipation limit

  •  
  • LSpeak

    litter size on the 10th day of lactation

  •  
  • LSM

    adjusted least square mean

  •  
  • Mb,peak

    body mass in lactating females on the 10th day of lactation

  •  
  • MEIpeak

    peak-lactation metabolizable energy intake

  •  
  • MEOpeak

    peak-lactation milk energy output

  •  
  • Ml,peak

    litter mass on the 10th day of lactation

The hypothetical limiting mechanisms can be considered from the perspective of the ‘energy cascade’, a process by which the flux of energy flows from the environment to offspring (Lovegrove, 2006). Two types of the limiting mechanisms have been discussed extensively: central, through limitation of energy assimilation, and peripheral, through limitation of energy expenditure of particular peripheral organs and tissues (Drent and Daan, 1980; Weiner, 1989; 1992; reviewed in Bacigalupe and Bozinovic, 2002). Several experiments designed to resolve which of the two mechanisms limits females' energy budgets at peak lactation have provided contradictory results (Hammond and Diamond, 1992, 1994; Koteja, 1996; Rogowitz, 1998; Johnson and Speakman, 2001; Johnson et al., 2001; Wu et al., 2009). Since 2003, the discussion about parental investment and energy budget has been revitalized by introducing an alternative, heat dissipation limitation (HDL) theory (Król and Speakman, 2003a). According to the HDL theory, a female's ability to dissipate excess heat limits her peak-lactation sustained metabolic rate. The excess heat is a consequence of the elevated metabolic activity resulting from the grossly increased rates of food digestion, nutrient absorption and transport, and milk synthesis (Król and Speakman, 2003a). Thus, in the strict sense, relevant from purely physiological perspective, the HDL theory proposes a mechanism limiting the energy budget of lactating females, and consequently the capacity to transfer energy as milk. As offspring growth can be limited by the rate of energy supply through milk, in a broader sense, relevant from ecological and evolutionary perspectives, HDL theory also concerns factors limiting reproductive output, which can be measured as the number or mass of the weaned offspring or their growth rate (Speakman and Król, 2010a). Thus, the HDL theory provides insights into the physiological mechanisms and constraints that influence maternal investment and offspring development during the lactation period, which potentially have an impact on reproductive output, and hence both the Darwinian fitness of individuals and population dynamics.

The HDL theory has been tested intensively by attempting to manipulate the heat dissipation constraint via manipulation of ambient temperature or removing the mother's fur during the lactation period. Several studies found support for the HDL theory, as fur removal or reduced ambient temperature increased the mothers' food consumption and milk production, along with higher pup growth rates (Johnson and Speakman, 2001; Król and Speakman, 2003a; Król et al., 2007; Wu et al., 2009; Simons et al., 2011; Yang et al., 2013; Sadowska et al., 2016). However, some studies reported findings contradictory to the predictions of the HDL theory (Zhao and Cao, 2009; Valencak et al., 2010; Zhao, 2011), or supporting them only partially. For example, shaved common vole females had an increased pup growth rate, but not increased food intake or milk production (Simons et al., 2011). Thus, heat dissipation did appear to be a factor limiting reproductive output, but not through limiting the females' reproductive energy effort. In shaved bank vole mothers, several reproductive performance traits were increased, characterizing both their effort (e.g. metabolizable energy intake and milk energy output) and reproductive output (the litter growth rate), which supports HDL theory, but a decreased digestive efficiency indicated that the ‘central’ limitation could be also involved (Sadowska et al., 2016). Despite the inconsistent results, which only partially support the HDL theory, most studies are in accordance with its prediction that the female's ability to efficiently dissipate heat excess limits their reproductive performance. If this limiting mechanism is indeed common, a question arises: what are the broader implications of the HDL theory? Several authors have already pointed out its importance in the context of global climate change, and especially global warming (Speakman and Król, 2010a; Grémillet et al., 2012; Nilsson and Nord, 2018; Bao et al., 2020). As it is known that thermoregulatory capabilities change with age (e.g. Kenney and Hodgson, 1987; Kenney and Munce, 2003; DeGroot and Kenney, 2007; Hanna and Tait, 2015), another important yet not extensively explored question is how the HDL mechanism affects traits related to age (Speakman and Król, 2010a).

A general decline in reproductive capacity with age is a commonly described phenomenon in most animals (Rose, 1991; Gaillard et al., 1994; Ericsson et al., 2001; Descamps et al., 2008; Hammers et al., 2012; Delbaere et al., 2020). The senescence hypothesis predicts that old females have a lower level of resources to allocate to reproduction because of physiological deterioration with aging (Rose, 1991; Martin and Festa-Bianchet, 2011). The reproductive performance decrease can be measured as the progressive decrease in fertility (delivery success) and fecundity (weaning success and litter size or mass) and occurs at different ages in different species (Edwards et al., 1998). For example, a study on reindeer showed that calf birth mass firstly increased and then decreased with maternal age (Weladji et al., 2010). Females of golden hamsters more frequently undergo pregnancy loss and show failure of maternal care with increasing age (Edwards et al., 1998). Descamps et al. (2008) found that old female red squirrels were less efficient at raising young, and the survival of their weaned offspring was also lower. These and other examples show that various aspects of aging affect physiological and behavioral processes directly associated with reproduction success, such as ovulatory cycle, implantation, parturition processes or hypothalamic–pituitary–ovarian function (Edwards et al., 1998). Thus, reproductive aging is explained as parallel to the general physical decline and accumulation of damage in multiple kinds of tissue with increasing age. However, despite several explanations of reproductive aging, understanding of this mechanism is still limited.

One of the possible research directions for closing the knowledge gap concerning reproductive aging is an exploration of the links between thermoregulatory properties, BMR and the capacity for sustained energy expenditure. BMR may be linked to the capacity for a sustained rate of energy expenditure over long time periods. This can lead to an association between high BMR and performance during energy-demanding periods, e.g. lactation (Johnston et al., 2007). A previous study found that high-BMR female mice have higher parental investment capacity, measured as the offspring growth rate (Sadowska et al., 2013). These females have also higher food consumption and heavier visceral organs, and the results suggest that they have higher energy acquisition abilities. Results from an earlier study on bank voles from lines selected for high aerobic metabolism demonstrated that the individual variation in metabolic rate affects the thermoregulatory properties and suggested that the heat dissipation limit can act differently on animals at different ages and with a different level of metabolic rate (Grosiak et al., 2020).

We based our research on the bank vole (Myodes glareolus) from an experimental evolution model system: four lines selected for high swim-induced aerobic metabolism (A-lines), which also have increased basal, average daily and maximum cold-induced metabolic rates, and four unselected control (C)-lines (Sadowska et al., 2015a; Dheyongera et al., 2016; Stawski et al., 2017). The results of a previous study on the effects of aging indicate that the maximum run-induced aerobic metabolic rate decreased with age in the A-lines, whereas it remained constant in the C-lines (Rudolf et al., 2017). The reproductive output tended to be higher in bank vole females from lines selected for high aerobic metabolism than in voles from control lines; however, shaving similarly affected the peak-lactation metabolizable energy intake, average daily metabolic rate, as well as the milk energy output and litter growth rate in voles from the A- and C-lines (Sadowska et al., 2016).

The main aim of this study was to test a novel hypothesis that an age-related decrease of the capacity to dissipate excess heat, and hence exacerbation of the limitations predicted by the HDL theory, contributes to the decline of reproductive output with age in bank voles. As the rate of metabolism is presumably linked with thermoregulation, reproductive effort and aging, we also hypothesized that the selection for an increased metabolism alters the process of age-related decline of the heat dissipation capacity and reproductive performance. We performed a multifactor experiment to investigate the effects of (i) maternal age, (ii) fur removal (a treatment which relieves the mother's energy budget from heat dissipation limitation) and (iii) selection for high metabolic rate on reproductive performance (measured as milk production and litter growth at peak lactation). Importantly, we explored the interactions between these factors. With this complex, full factorial design, we could test several specific hypotheses. The most important were the following two. (1) The age-related decrease of reproductive performance is due to an age-related decline in thermoregulatory capability. We predicted that the reproductive performance would be higher in shaved than in unshaved females, and that this difference would be greater in old than in young mothers. Such a result would confirm that the limitations predicted by HDL theory increase with age. (2) The age-related reproductive performance differences are affected by selection for an increased exercise metabolism (which has led to a generally elevated level of metabolism). We expected that an age-related decline in reproductive performance would be more severe in the A-line than in the C-line mothers. We also expected that in the old voles, the difference in reproductive performance between shaved and unshaved mothers would be more pronounced in the A-line than in the C-line mothers.

Animal model and the selection experiment

The work was performed on bank voles, Myodes glareolus (Schreber 1780), from four replicate lines selected for the maximum 1-min swim-induced rate of oxygen consumption (A-lines, high aerobic metabolism), and four replicate lines of control, random-bred voles (C-lines, control) (Sadowska et al., 2008, 2015a).

The bank vole is one of the most numerous and widespread woodland rodents in Central Europe, inhabiting practically all types of forest (Pucek et al., 1993). In the wild, the average life expectancy is about 3 months, the maximal life expectancy is about 10–15 months, but individuals can achieve 18–21 months (Bernshtein et al., 1999; Petrusewicz, 1983). Thus, most individuals do not survive more than one breeding season. In captivity, bank voles can live for over 4 years, but they are able to breed for only 2 years (Buchalczyk, 1970).

The experimental colony originates from voles captured in the Niepołomice Forest (southern Poland) in 2000 and 2001. After a few generations of random mating, the base population was divided into four independent lines for A-selection and four lines for control (C). The replicate lines for A-selection were maintained with 15–20 reproducing families per line. Within-family selection was applied, and mating between siblings and first cousins was avoided. Mating was performed after the selection trial on animals from the age of about 90 days. The selection criterion was the highest 1 min rate of oxygen consumption achieved during an 18 min swimming trial at a water temperature of 38°C, adjusted for differences in body mass and other confounding factors, such as sex and measurement date. Because heat exchange rate in water is very high, water of a temperature approximately equal to that of an animal's body provides a thermoneutral environment. Thus, the selection trial protocol was not associated with either a cold or heat stress (Sadowska et al., 2015a). Measurements were performed on young adult animals aged 75–95 days. Since the 18th generation, the mass-adjusted maximum rate of oxygen consumption during swimming (O2,swim) is more than 60% higher in A- than in C-lines (Jaromin et al., 2019; Lipowska et al., 2022).

The selection for high O2,swim also resulted in a higher mass-corrected BMR (Sadowska et al., 2015a). The thermogenic capacity, measured as the maximum cold-induced rate of oxygen consumption, was higher in the A-lines than in the C-lines (Dheyongera et al., 2016; Stawski et al., 2017), although the non-shivering thermogenic capacity was not affected by the selection (Stawski et al., 2015). The selection increased cold tolerance, but reduced the capacity to thermoregulate at high environmental temperatures (Stawski et al., 2017). Contrary to the main assumption of the oxidative stress theory of aging (Harman, 1956; reviewed in Cedikova et al., 2016), markers of oxidative damage were not increased in A-lines, and the antioxidant activity did not differ between animals from the A- and C-lines, even if the animals' energy expenditure was elevated during reproduction (Ołdakowski et al., 2012, 2015). The animals from A-lines had significantly higher run-induced maximum metabolic rate than those from C-lines (Rudolf et al., 2017; Jaromin et al., 2019). The results of a previous study on the effects of aging indicated that the maximum run-induced aerobic metabolic rate decreased with age in the A-lines, whereas it remained constant in the C-lines (Rudolf et al., 2017). The reproductive output tended to be higher in bank vole females from the A-lines (Sadowska et al., 2016). To summarize, the selection for a high aerobic metabolism resulted in changes in the metabolic, thermoregulatory and reproductive traits and affected the pattern of age-related changes in some locomotor performance traits.

Experimental protocol

The study was performed in two experimental blocks, in 2018 (block B1) and 2019 (block B2). For the first block, 280 adult females were sampled from three generations: 94 young females from the 27th generation, 92 middle-aged females from the 26th generation, and 94 old females from the 25th generation. For the second block, 225 adult females were sampled from four generations: 60 young females from the 29th generation, 56 middle-aged females from the 28th generation, and 60 old females from the 27th generation and 49 from the 26th generation. The 505 females belonged to three age classes: young (90–141 days; mean±s.d.: 116.2±12.2 days), middle-aged (205–324 days; 283.0±28.4 days) and old (413–683 days; 490.0±70.3 days). In both blocks within the generations, the animals represented both selection directions and all replicate lines within the selection directions (Table S1). Before the experiment, the voles were kept in standard cages (267×207×140 mm, polypropylene; Tecniplast, Buguggiate, Italy) in same-sex groups of n=3 individuals. Virgin females were randomly sampled and paired with unrelated, adult males from the same selection direction, replicate line and generation. The pairs were kept separately in individually ventilated cages (Techniplast, GM500) with dust-free aspen bedding (Abedd, Munich, Germany) and equipped with a coconut-shell hut, under controlled ambient temperature (20±1°C) and photoperiod (16 h:8 h light:dark). Food (standard rodent chow: 24% protein, 3% fat, 4% fiber; Labofeed H, Kcynia, Poland) and water were provided ad libitum.

Males were kept with females until the females' pregnancy was detected (females were weighed and observed every 2 days after mating). Of the 505 mated females, 349 gave birth. Birth success was similar in the A- and C-lines, but it was much lower in the old females (54%) than in the middle-aged (77%) and young ones (82%) (Table S1). However, as the number of old females mated was purposefully larger, the number of parturitions given by the old females (109) was only slightly lower than those given by the middle-aged (114) or the young ones (126; Table S1). One young A-line female died after giving birth, and two females gave birth to one, dead pup (one old C-line female and one middle-aged A-line female).

On the 4th day of lactation (since parturition day=day 0), the shaving manipulation was applied to enhance the females' capacity for heat dissipation (Sadowska et al., 2016). Shaving was repeated on the 8th day of lactation to minimize fur regrowth. We randomly assigned all females to one of the two groups: shaved (135 females: 50 young, 45 middle-aged and 40 old) and unshaved (138 females: 51 young, 44 middle-aged and 43 old; Table S1). During this procedure, all voles were anesthetized by a controlled inhalation system (Sigma Delta Vaporizer UNO BV, Abingdon, UK) via nose cone for about 8 min during the first shaving and about 4 min during the second shaving. Isoflurane (Vetpharma, Barcelona, Spain) with oxygen was provided with a flow of 200 ml min−1 (following DLAR Veterinary Medical Staff, 4-6563) at a 3% volume of breathing air (Hohlbaum et al., 2017). We shaved voles completely except for the head (Fig. 1), using a hair razor (Wella, HS 61). Sham-shaving was performed on females from the unshaved group: the animals were anesthetized and handled in the same way, and for about the same time, as the shaved individuals but without fur removal.

Fig. 1.

Shaved (left) and unshaved (right) bank voles (Myodes glareolus) with 2–3 day old pups. The shaving manipulation was applied to elevate the females' capacity for heat dissipation during the peak of lactation (females in the photos are not the ones used in the experiment, where the shaving was done on day 4 of lactation).

Fig. 1.

Shaved (left) and unshaved (right) bank voles (Myodes glareolus) with 2–3 day old pups. The shaving manipulation was applied to elevate the females' capacity for heat dissipation during the peak of lactation (females in the photos are not the ones used in the experiment, where the shaving was done on day 4 of lactation).

Mothers’ body mass (Mb, g), litter size (LS), litter mass (Ml, g) and litter mortality were recorded daily throughout the entire 15 days of lactation. The peak-lactation growth rate [GRpeak, g (4 days−1)] of the litter was calculated as the litter mass difference between the 10th and 14th day. An approximate daily food intake was estimated by subtracting the mass of food remaining in the feeder from the mass of food provided to the feeder and corrected for the dry mass content. At the beginning of the peak of lactation period (days 11–12), a 24 h, more accurate feeding trial was performed (details in the next section). The trial was followed by a 24 h measurement of the average daily metabolic rate (ADMR) quantified by the doubly labelled water (DLW) technique (days 12–13). Pups were weaned on the 21st day of lactation.

All the animal care and measurement procedures were approved by the II Local Bioethical Committee in Kraków, Poland (no. 292/2018 and 168/2019).

Measurement of mothers’ metabolic rate and reproductive performance traits

At the beginning of the feeding trial (day 11 of lactation), the females with pups were placed in new cages with fresh sawdust and a known mass of food (about 30±0.001 g per cage). The amount of sawdust was much lower than the standard amount to allow collection of the remaining food and feces. The dry mass content of the given food estimation was based on three samples of fresh food weighed on each day when the feeding trials were started. After 24 h, the remaining food and feces were collected, separated and dried at 60°C to constant mass (±0.001 g; Radwag PS 200/2000.R2 Precision Balance, Radom, Poland). The large pellet pieces left in the feeder or in cages and food orts (i.e. the food chewed by the animals to powder) were collected separately. Samples of fresh food were dried at the same time.

The energy content of dry food (5 samples from each block) and of dry feces (from all females) was measured by bomb calorimetry (Model 6100, Parr Instrument Company, Moline, IL, USA). Mean (±s.d.) energy content of dry food was 17.72±0.04 kJ g−1 and that of dry feces was 15.52±1.75 kJ g−1.

Food consumption rate (FC, g day−1) and apparent digestive efficiency (ADE, %) were calculated from data obtained during the feeding trial according to the following equations:
(1)
(2)
where Efeces and Efood are the energy content of feces and food and Mfeces is the dry mass of feces.
The proportion of orts (PO, %) relative to the entire dry mass food intake (i.e. the amount that disappeared from feeders) was calculated following equations used in Sadowska et al. (2016):
(3)
The mean daily food consumption in the peak-lactation period was based on the daily measurements of food intake recorded on days 11–14 (FCpeak, g day−1), and was corrected for PO:
(4)
The metabolizable energy intake at peak lactation (MEIpeak, kJ day−1) was based on FCpeak:
(5)
The 0.97 multiplier accounts for an assumed 3% urinary energy loss (Drożdż, 1968).

The average daily metabolic rate was measured on days 12–13 days of lactation using the doubly labelled water (DLW) technique (Speakman, 1998; Butler et al., 2004). The procedure was performed as in Sadowska et al. (2016). The protocol was performed on 61 young, 62 middle-aged and 56 old mothers chosen from each selection and shaving group. Females were weighed (±0.1 g) and injected interperitoneally with a known mass of DLW (representing approximately of 0.62% of a lactating animal's body mass and about 0.41% of a non-breeding animal's body mass) containing enriched 18O (30 atom%) and 2H (18.4 atom%). Syringes were weighed before and after injection (±0.0001 g) to calculate the precise mass of the exact dose (Wu et al., 2009; Sadowska et al., 2016). The injection was followed by two rounds of blood sampling: the first one was taken 1 h after the injection to estimate initial isotope enrichment, and the final one was taken 24 h after the first to estimate isotope elimination rates. At each sampling round, two 50 µl samples were taken by retro-orbital puncture and were stored in heat–sealed glass capillaries in ambient room temperature. Blood samples were also taken from non-breeding and non-injected voles from each line to estimate the background of isotope enrichment in an animal's water mass. Samples were vacuum distilled, and the distilled water was used to produce CO2 and H2 (Speakman and Król, 2005; Wu et al., 2009). Gas source isotope ratio mass spectrometry (ISOCHROM μGAS system and IsoPrime IRMS, Micromass, Manchester, UK) were used to analyze the isotope ratios 18O:16O and 2H:1H (Speakman and Król, 2005). The isotope ratios were converted to ADMR using a model recommended for small rodents (equation 7.17 in Speakman, 1997). Peak-lactation milk energy output (MEOpeak, kJ day−1) was calculated as the difference between MEIpeak and ADMRpeak (Sadowska et al., 2016).

Statistical analyses

Overall, records from 255 mothers were analyzed (Tables S1, S5). Among them, records from 7 animals were excluded from the peak-lactation analyses because of death of the mothers (3 shaved and 4 unshaved), and 21 were excluded because of death of the litters (13 shaved and 8 unshaved mothers).

All statistical analyses were performed in SAS v.9.4 (SAS Institute, Inc. Cary, NC, USA). For analyses of the relationship between peak-lactation growth rate and litter size, we used a mixed non-linear procedure (NLMIXED) to apply a stage-regression model. The aim was to check whether the GRpeak increases linearly with litter size or has an upper limit (‘break point’, the litter size at which the slope of regression line changes). The full model had four estimable parameters: an intercept, the slope of the regression line below a break point, the value of the break point and the slope above the break point. Body mass measured on the 10th day of lactation (before the peak lactation) was included in the model as a covariate. The residual error was assumed to have a normal distribution (verified by inspection of diagnostic graphs). The full model was then reduced to a simpler model, with the slope above the break point fixed to zero (i.e. with a hard upper limit assumed for a trait value). Finally, the model was reduced to simple regression line, without a break point. We compared all the models using likelihood ratio tests and AIC criterion to check which model fitted the data better.

For the main analyses of Mb,peak, LSpeak, Ml,peak, GRpeak, FCpeak, ADEpeak, ADMRpeak, MEIpeak and MEOpeak, we used a cross-nested mixed ANCOVA model (MIXED procedure with REML method and variance components constrained to non-negative values). All models included mothers’ age class (young, middle-age and old), selection (A- versus C-lines) and shaving treatment (shaved versus unshaved) as the main fixed categorical factors, and interactions between these factors. All models also included experimental block (B1 versus B2) as a categorical factor, and the random effect of replicated lines nested within selection groups as well as its interactions with the main categorical factors. The main analyses were divided into three parts. In the first part, we aimed to answer the question whether the main factors affect absolute values of reproductive and metabolic traits. In the second part, body mass of the mother was included as a fixed covariate in the models, to compare the performance trait values adjusted for the variation in the body mass of mothers. In the last part, focused on testing the HDL effect, the analyses were limited to females with large litters, with a litter size of 4 or more (based on results of the stage regression, Table S4, and results of Sadowska et al., 2016). In the last part, female body mass was also included as a fixed covariate in models, as in the second part. The statistical models had a similar structure to those described above.

To test the assumption of the homogeneity of body mass slopes, all initial full ANCOVA models included the first and the second order interactions between body mass and the main factors. If these interactions were not significant, we removed them. If an interaction with body mass was significant, it was retained in the final model and we tested the differences between groups across the entire range of body mass (at a minimum body mass of 24 g, at an average body mass of 29 g, and at a maximum body mass of 32 g). We used the ‘at’ option in the LSMEANS function, which allows determination of the covariate value for computing the least squares means. For the analyses of simple effects in cases of a significant interaction (P<0.05), we used the SLICE function of the SAS Mixed procedure, which allows comparison of groups of one factor separately from groups of the other factor (involved in the interaction). The Satterthwaite approximation for non-orthogonal models was applied to calculate the denominator degrees of freedom (Satterthwaite option in the Model command of SAS MIXED procedure). In all results, descriptive statistics are provided (mean±s.d.), and the trait values estimated with the ANOVA or ANCOVA models are expressed as adjusted least squares means (LSM)±s.e. All presented LSMs (when adjusted for body mass) were calculated for a fixed body mass of 29 g in lactating females. In cases with heterogenic slopes, the estimates and significance tests were also obtained for body mass near the minimum (24 g) and maximum (32 g) of the usual range of lactating females. Before estimating the final models, we analyzed studentized residuals and removed from the final analysis observations with residuals higher than 3.5 and lower than −3.5 as the outliers (Table S2).

Comparison of reproductive performance and metabolic traits

Mother and litter characteristics at the onset of peak lactation

This study was focused on reproductive performance in the energetically critical peak lactation period, rather than a comprehensive description of the whole reproductive cycle. Therefore, to avoid overloading the report with results not directly related to the focal question, we begin the description by presenting the state on day 10 of lactation, i.e. at the onset of the critical period. In the first section, we present results concerning the question whether and how the main factors in this experiment affect both reproductive performance and metabolic traits. These analyses were based on data from all available litters. The analyses aimed specifically at testing the HDL hypothesis, based on large litters only, are presented below (‘Testing the HDL hypothesis: performance traits of females with large litters’).

The body mass of lactating mothers measured on the 10th day of lactation (Mb,peak) ranged from 19.9 to 44.4 g (mean±s.d.: 29.3±4.37 g), and was on average 19% higher in the A-lines than in the C-lines (F1,6.2=35.7, P=0.0009; Tables S2 and S3; Fig. 2). The mean Mb,peak was affected by age (F2,234=3.86, P=0.023), but the pattern was complicated by a significant selection×age interaction (F2,234=4.40, P=0.013): in A-lines, Mb,peak was the highest in the middle-aged group, while in C-lines it increased with increasing age (Fig. 2). The effect of shaving was not significant (F1,233=0.41, P=0.52). Although in A-lines, but not in C-lines, Mb,peak appeared to be slightly lower in shaved than in unshaved voles, the interaction was not significant (F1,233=3.29, P=0.071). Other interactions between the main effects were also not significant.

Fig. 2.

Peak-lactation body mass of females (Mb,peak).Mb,peak was measured on the 10th day of lactation in voles from the control (C) and selected (A) lines, from young (Y, red), middle­-aged (M, green) and old (O, blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the least squares means (LSMs) and whiskers represent the standard error (s.e.).

Fig. 2.

Peak-lactation body mass of females (Mb,peak).Mb,peak was measured on the 10th day of lactation in voles from the control (C) and selected (A) lines, from young (Y, red), middle­-aged (M, green) and old (O, blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the least squares means (LSMs) and whiskers represent the standard error (s.e.).

The litter size on the 10th day of lactation (LSpeak) ranged from 1 to 9 (mean±s.d.: 4.00±1.63) and the litter mass (Ml,peak) ranged from 3.02 to 58.94 g (23.7±10.57 g). Litter size and mass were higher in voles from the A-lines than in those from the C-lines (P<0.001), but the pattern was opposite in the analysis with Mb,peak as covariate (Tables S2 and S3). The absolute values of LSpeak and Ml,peak in both selection directions were much higher in young and middle-aged than in old voles (Fig. 3A,C). The absolute LSpeak and Ml,peak did not differ between shaving groups (P>0.3). In the analysis of LSpeak, the interaction between selection and shaving manipulation was significant (P=0.04), but shaving effect was not significant in both selection lines (P>0.2). The analysis of LSpeak and Ml,peak with Mb,peak added as covariate gave similar results.

Fig. 3.

Peak-lactation litter size (LSpeak), litter mass (Ml,peak) and growth rate (GRpeak). LSpeak (A,B), Ml,peak (C,D) and GRpeak (E,F) absolute values (left) and mother mass-adjusted values (right). The voles were from the control (C) and selected (A) lines, from the young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs (A,C,E) or LSMs adjusted for a body mass of 29 g (B,D,F) and whiskers represent the s.e.

Fig. 3.

Peak-lactation litter size (LSpeak), litter mass (Ml,peak) and growth rate (GRpeak). LSpeak (A,B), Ml,peak (C,D) and GRpeak (E,F) absolute values (left) and mother mass-adjusted values (right). The voles were from the control (C) and selected (A) lines, from the young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs (A,C,E) or LSMs adjusted for a body mass of 29 g (B,D,F) and whiskers represent the s.e.

Peak-lactation growth rate

Of the 255 litters alive on day 10 of lactation, 21 (8%) died during the peak lactation period (days 10–14). The mean (±s.d.) litter size of the lost litters was 3.90±1.58, i.e. it was similar to the mean litter size. The peak-lactation litter survival rate was generally high (about 90%), and we did not observe effects of the mothers’ age, selection type or shaving manipulation on their mortality rate (Table S1).

The peak-lactation growth rate of the litters (GRpeak) ranged from 0.31 to 12.0 g (4 days−1) [mean±s.d: 4.47±2.29 g (4 days−1)]. Litters from A-lines generally grew faster [LSM±s.e: 5.18±0.35 g (4 days−1)] than those from C-lines [3.71±0.35 g (4 days−1); F1,6.24=8.79, P=0.024; Tables S2 and S3; Fig. 3E]. GRpeak was lower in the young [4.48±0.31 g (4 days−1)] and old [3.65±0.31 g (4 days−1)] than in the middle-aged class [5.21±0.31 g (4 days−1); F2,185=9.99, P<0.0001; Fig. 3E]. GRpeak did not differ between shaving groups (F1,185=0.22, P=0.64). The interactions between these three factors were not significant (P>0.08). GRpeak did not differ between the two experimental blocks (F1,186=0.77, P=0.38).

GRpeak increased with maternal body mass [slope±s.e.: 0.14±0.080 g (4 days−1) g−1; F1,179=32.2, P<0.0001]. However, the analysis showed significant interactions of body mass with age (F2,175=5.28, P=0.006; Tables S2 and S3; Fig. 3F). At the minimum body mass of 24 g, the mass-adjusted GRpeak was 39% higher in young than in old mothers, and in middle-aged mothers it was 32% higher than in old mothers. At a mean mass of 29 g, mass-adjusted GRpeak was about 30% higher in young than in old mothers, and in middle-aged mothers it was 5% lower than in old mothers. At the upper end of the mass range of 32 g, it was 24% higher in young than in old mothers, and 47% higher in middle-aged than in old mothers. GRpeak did not differ between voles from A- and C-lines after correction for body mass (F1,7.99=2.07, P=0.19). The interaction between selection direction and age was significant (F2,68.6=3.54, P=0.034): in A-line voles, GRpeak did not differ significantly between age classes (F2,57.9=2.48, P=0.093), while in C-line voles, GRpeak was lower in the old than in the young and middle-aged class (F2,46.3=8.32, P=0.0008; Fig. 3F).

Peak-lactation metabolizable energy intake

The metabolizable energy intake at peak lactation (MEIpeak) was based on the calculation of the mean daily food consumption in the peak-lactation period (FCpeak; Tables S2 and S3) and the apparent digestive efficiency (ADEpeak; Tables S2 and S3).

MEIpeak ranged from 70.87 to 253.73 kJ day−1 (mean±s.d: 153.5±38.65 kJ day−1). MEIpeak was 22% higher in voles from A-lines than in those from C-lines (F1,14.1=31.0, P<0.0001), and was on average 6% and 10% higher in middle-aged than in young and old age classes, respectively (F2,131=2.93, P=0.057; Tables S2 and S3; Fig. 4A). MEIpeak was also 14% higher in the shaved than in the unshaved group (F1,14.1=13.0, P=0.003; Fig. 4A). The interactions between these three factors were not significant (P>0.38). MEIpeak was 22% higher in block B2 than in block B1 (F1,131=13.0, P=0.0004).

Fig. 4.

Peak-lactation metabolizable energy intake (MEIpeak), average daily metabolic rate (ADMRpeak) and milk energy output (MEOpeak). MEIpeak (A,B), ADMRpeak (C,D) and MEOpeak (E,F) absolute values (left) and mother mass-adjusted values (right). The voles were from the control (C) and selected (A) lines, from the young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs (A,C,E) or LSMs adjusted for a body mass of 29 g (B,D,F) and whiskers represent the s.e.

Fig. 4.

Peak-lactation metabolizable energy intake (MEIpeak), average daily metabolic rate (ADMRpeak) and milk energy output (MEOpeak). MEIpeak (A,B), ADMRpeak (C,D) and MEOpeak (E,F) absolute values (left) and mother mass-adjusted values (right). The voles were from the control (C) and selected (A) lines, from the young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs (A,C,E) or LSMs adjusted for a body mass of 29 g (B,D,F) and whiskers represent the s.e.

MEIpeak increased with body mass (slope±s.e: 1.33±0.68 kJ day−1 g−1; F1,126=3.77, P=0.055). After adjusting for Mb,peak, the differences between selection directions and between age classes alone were not significant (P>0.12; Tables S2 and S3; Fig. 4B). Other effects were similar to those in the analysis without covariate.

Peak lactation average daily metabolic rate

The average daily metabolic rate (ADMRpeak) ranged from 54.43 to 162.66 kJ day−1 (mean±s.d: 102.8±22.36 kJ day−1) and was 26% higher in A-lines than in C-lines (F1,5.94=51.0, P=0.0004; Tables S2 and S3; Fig. 4C). ADMRpeak was on average 7% higher in middle-aged than in young and old age classes (F2,159=4.21, P=0.017; Fig. 4C), and 14% higher in the shaved than in the unshaved group (F1,5.54=30.6, P=0.002). However, the interaction between age and shaving was also significant (F2,159=3.83, P=0.024): in the old age class, ADMRpeak was much lower in the unshaved than in the shaved group (F1,44.9=28.8, P<0.0001), while in the middle-aged class, the shaving effect was smaller (F1,35.1=5.60, P=0.024) and in the young age class it was not significant (F1,39.9=3.60, P=0.065; Fig. 4C). Other interactions between these three factors were not significant (P>0.20). ADMRpeak did not differ between the two experimental blocks (P=0.81).

ADMRpeak increased with increasing body mass (slope±s.e.: 2.91±0.33 kJ day−1 g−1; F1,145=76.9, P<0.001). Other effects were similar to those in the analysis without covariates, but after adjusting for covariates, the difference between age classes alone was not significant (P=0.26; Tables S2 and S3; Fig. 4D).

Peak-lactation milk energy output

The peak-lactation milk energy output (MEOpeak) ranged from 30.25 to 123.43 kJ day−1 (mean±s.d.: 62.4±20.28 kJ day−1) and was not affected by selection (F1,5.44=0.01, P=0.94), age (F2,18.5=0.12, P=0.89), or shaving manipulation (F1,18.4=1.43, P=0.25; Tables S2 and S3;Fig. 4E). The interactions between these three factors were not significant (P>0.55). In block B2, MEOpeak was 13% higher than in block B1 (F1,79.8=5.24, P=0.025).

MEOpeak was not affected by body mass (slope±s.e.: 0.93±0.57 kJ day−1 g−1; P=0.11). Other effects were similar to those found in the analyses for non-adjusted MEOpeak (Tables S2 and S3; Fig. 4F).

Testing the HDL hypothesis: performance traits of females with large litters

First, we used the non-linear stage regression models to analyze the relationship between the litter growth rate and litter size to determine the litter size at which the females approached their physiological limit. The mass-adjusted litter growth rate increased with litter size [slope±s.e.: 0.95±0.13 g (4 days−1) per pup; P<0.001] up to a litter size of 4.75±0.35 pups (Table S4, Fig. S1). The model with a constant level above this break point was better than a simple linear regression model (χ2=12.8, d.f.=1, P=0.0003), and allowing a free slope above the break point did not improve the model (likelihood ratio test, χ2=0.1, d.f.=1, P=0.75; Table S4). Based on these results and on results from our previous study, where a litter size of 4 was found as the break point (Sadowska et al., 2016), we decided to include in the analyses aimed at testing HDL theory females with a litter size of 4 or larger.

The reproductive performance traits (GRpeak, ADMRpeak and MEIpeak) of females with the large litters increased significantly with increasing body mass of mothers (P<0.004), but MEOpeak did not correlate with body mass (P=0.18; Table S3). Neither the selection, age or shaving manipulation, or their interactions affect the mass-adjusted MEIpeak (Tables S2 and S3; Fig. 5B).

Fig. 5.

Peak-lactation litter GRpeak and metabolic traits in lactating females with large litters. Mass-adjusted GRpeak (A), MEIpeak (B), ADMRpeak (C) and MEOpeak (D). The voles were from the control (C) and selected (A) lines, from young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs adjusted for a body mass of 29 g and whiskers represent the s.e.

Fig. 5.

Peak-lactation litter GRpeak and metabolic traits in lactating females with large litters. Mass-adjusted GRpeak (A), MEIpeak (B), ADMRpeak (C) and MEOpeak (D). The voles were from the control (C) and selected (A) lines, from young (red), middle-aged (green) and old (blue) age classes, and from shaved (open symbols) and unshaved groups (filled symbols). The points represent the LSMs adjusted for a body mass of 29 g and whiskers represent the s.e.

GRpeak was not affected by selection (F1,7.31=1.76, P=0.22), age (F2,25.2=2.47, P=0.11) or shaving manipulation (F1,7.5=0.64, P=0.45; Tables S2 and S3; Fig. 5A). The interactions between these three factors were not significant (P>0.22).

ADMRpeak was 12% higher in A-lines than in C-lines (F1,9.86=8.34, P=0.016; Tables S2 and S3; Fig. 5C). It did not differ between age classes (F2,97.7=0.93, P=0.40), but was 15% higher in the shaved than in the unshaved group (F1,98.9=29.3, P<0.0001). The interaction between age and shaving was also significant (F2,95.2=4.28, P=0.017): in the old age class, ADMRpeak was much higher in the shaved than in the unshaved group, while in the middle-aged class the shaving effect was lower and in the young age class it was not significant (Fig. 5C). Other interactions between these three factors were not significant (P>0.11).

The mass-adjusted MEOpeak was not affected by selection (F1,8.97=0.03, P=0.86) and age (F2,21.7=0.01, P=0.99), but tended to be higher in shaved (LSM±s.e.: 73.0±4.74 kJ day−1) than in unshaved females (61.8±4.78 kJ day−1; F1,25=4.10, P=0.054; Tables S2 and S3; Fig. 5D). The interactions between these three factors were not significant (P>0.25).

The results of this study show that old bank vole females had generally lower reproductive performance measured as the absolute values of litter size and litter mass, or the peak-lactation litter growth rate, than females from younger age classes, which is consistent with the previous studies describing the age-related decrease in reproductive output in mammals (i.e. Descamps et al., 2008; Hammers et al., 2012; Delbaere et al., 2020). Both the analyses performed for all mothers and those limited to mothers with large litters revealed that the peak-lactation average daily metabolic rate (ADMRpeak) was higher in shaved than in unshaved mothers. The effect of shaving on ADMRpeak was more profound among old than young and middle-aged mothers. In females with large litters, the milk production (MEOpeak) tended to be higher in shaved than in unshaved voles (P=0.05; Fig. 5D), which provided support (although only weak) for the HDL theory in its strict sense, i.e. concerning the mechanism limiting the energy budget of lactating females. However, fur removal had no effect on the litter growth rate (GRpeak). Thus, the results did not support the HDL theory in a broader sense, concerning factors limiting reproductive output in general. As the interaction between the effects of age and shaving did not affect either MEOpeak or GRpeak, the results did not support the hypothesis that the age–related decline of reproductive performance is related to an age–related decline of the capacity to dissipate heat. The analyses of females rearing litters of any size revealed that voles from lines selected for high aerobic capacity (A-lines) had higher metabolic traits in comparison to control (C-line) voles. However, interactions between the effects of selection and age or shaving did not affect any of the reproductive traits. Thus, the results did not support the hypothesis linking reproductive aging with the selection for an increased exercise metabolism.

In endotherms, the age-specific patterns of reproduction are usually described as a bell–shaped curve (Bouwhuis et al., 2012; Hammers et al., 2012), and it is known that the offspring survival and litter mass of very young and elderly mothers can be decreased (Broussard et al., 2003). Bank vole females aged 6–14 months have the largest litters and from the 15th month of life the average litter size gradually decreases (Buchalczyk, 1970). As expected, the majority of the unsuccessful reproductive attempts were noted in old vole pairs, and the reproductive performance traits, i.e. the absolute values of litter size and litter mass, and the peak-lactation litter growth rate decreased in the old mothers (Tables S2 and S3; Figs 3 and 5A). Surprisingly, MEOpeak did not differ between age classes (Figs 4E,F and 5D), which means that the old females that successfully gave birth were able to retain the peak-lactation milk production as high as that of younger ones. In our study, females from all age classes only bred once. Thus, unlike old females living under natural conditions or as typical breeders in lab mouse colonies, a cumulative cost of previous reproductive burden did not contribute to their aging. This potentially allowed the females to maintain a high reproductive performance at an older age. Moreover, in the most energy-demanding period of lactation, younger females might invest less in milk production and feeding their current offspring compared with older females, to ensure their own survival and future reproduction (e.g. Stearns, 1976; Clutton-Brock, 1984; Ericsson et al., 2001). Aging female mammals experience various reproductive function declines, and life history factors, e.g. the number of reproductive attempts, should be considered in studies on reproductive aging (Lemaitre and Gaillard, 2017).

Lactation is the period of highest maternal energy expenditure in most mammals (Hammond and Diamond, 1994; Król and Speakman, 2003b; Johnston et al., 2007; Wu et al., 2009; Schulte, 2015), and the age–related decline in lactation performance can considerably affect the age-related decrease in reproductive performance. The milk glands are subject to senescence (Daniel, 1977), which could contribute significantly to the decreased ability of older females to meet the offspring energy requirements during peak lactation. The peak-lactation litter growth rate was generally lower in younger than in old mothers, but depending on mothers' body mass, litters of young or middle-aged mothers had the highest GRpeak (Fig. 3). Our results revealed also that litters from A-lines generally grew faster than those from C-lines (Fig. 3E,F). However, the MEOpeak was not significantly affected by age or selection (Figs 4E,F and 5D). A possible explanation for this result is that the milk biochemical composition could differ between age classes and selection directions. For example, milk of high-BMR laboratory mice females had lower protein concentration and higher lactose concentration than milk of low-BMR females (Sadowska et al., 2015b). Age-related and/or between-line differences of the primary milk carbohydrate, as well as differences in pups’ suckling abilities and milk digestion efficiency may significantly affect the offspring peak-lactation growth rate and quality (Bateman, 1957; Fiorotto et al., 1991; Schroeder et al., 2007).

Our main objective was to test the hypothesis that the age-related decline in reproductive performance is related to an age-related decrease of heat dissipation capacity. As in our previous report (Sadowska et al., 2016), we have found evidence of a ceiling limiting reproductive performance (Table S4 and Fig. S1), and performed all analyses aimed to test the HDL theory only on large litters. Within mothers of such litters, shaving increased ADMRpeak and MEIpeak, indicating increased metabolic heat production (Fig. 5B,C). The increase in ADMRpeak due to shaving could imply that the shaved female has been relieved of the heat dissipation limitation and can increase food consumption and milk production (according to the HDL theory), or that shaved females had higher metabolism to maintain body temperature. Consistent with our prediction, the ADMRpeak difference between shaved and unshaved mothers was also more profound among old than young and middle-aged mothers. This result suggests that the limitations predicted by HDL theory could increase with age. With no correction for body mass, MEIpeak was also higher in young and middle-aged than in old mothers, but a shaving by age interaction was not revealed by both analyses. When analysis was restricted to mothers with large litters, consistent with the previous results on bank voles supporting the HDL theory (Sadowska et al., 2016), the shaved voles tended to have higher MEOpeak than unshaved ones (Fig. 5D). These results suggest that mothers with large litters approached a physiological ceiling limiting their milk energy output, and that the limitation can be removed by fur shaving, which led to increased heat loss. Thus, the results provided support for the HDL theory in its strict sense (concerning the mechanisms limiting energy budget of lactating females). However, the observed milk production increase did not translate into increased litter growth rate (Fig. 5A), which leads to the conclusion that these results did not support the HDL theory in a broader sense (concerning factors limiting the reproductive output). However, just the litter growth, size or mass at weaning may not adequately reflect the reproductive success of females. Possibly, even if the offspring of shaved and unshaved mothers had the same growth rate and final body mass, the quality of the offspring of mothers with lower milk production could be worse than that of mothers with higher reproductive investment. In addition, potentially differential impacts on the mother's future reproductive success or survival are unknown yet also relevant to the overall fitness. This hypothesis may provide direction for future investigation.

Several previous studies reported that at cold ambient temperature during lactation (a condition analogous to mother fur removal) milk production was not increased (e.g. Wu et al., 2009; Zhao and Cao, 2009) and such results may support the peripheral limitation hypothesis (limitation by the capacity of mammary glands; but see Speakman and Król, 2010b). Wu et al. (2009) in their study on Brandt's voles, suggested that the peripheral limitation hypothesis can be more relevant for mothers raising smaller litters. Our study was performed on the first litters, in contrast to the previous study with similar experimental approaches and different generations of the same model species, but performed on the second litters (Sadowska et al., 2016). Possibly, this difference was due to a generally smaller number of raised pups in the first reproductive attempt. The limitations described in the central, peripheral and heat dissipation limitation theories represent different aspects of the complex interplay of physiological processes involved in energy allocation and utilization, and these results highlight that multiple factors could interact and contribute to shaping an organism's reproductive capacity.

According to the ‘increased-intake hypothesis’ (Nilsson, 2002), high metabolic rate may be associated with maintenance of larger internal organs, higher rate of biosynthesis and rate of energy intake, and these differences may lead to higher milk production rates and consequently higher reproductive output (Blackmer et al., 2005; Boratyński and Koteja, 2010; Książek et al., 2004; Sadowska et al., 2015b). In contrast, according to the alternative ‘compensation hypothesis’ (Nilsson, 2002), mammals with a high metabolic rate have high costs of self-maintenance and may transfer less energy for reproduction (Boratyński et al., 2013). A previous study on the same animal model reported the elevated litter mass at the end of lactation in voles from A-lines (Sadowska et al., 2016). Similarly, mice selected for high BMR raised heavier litters than low-BMR mothers (Sadowska et al., 2013, 2015b). Here, as expected, voles from the selected lines had higher absolute values of reproduction and peak-lactation metabolic traits in comparison to control voles: litter size, litter mass, GRpeak, FCpeak, MEIpeak and AMDRpeak (Tables S2 and S3; Figs 3, 4 and 5). However, the mean Mb was much lower in C-line than in A-line voles and the differences in all reproductive output traits were attributable to differences in maternal body mass. Thus, our results did not provide support for the increased-intake hypothesis, and, consequently, for the hypothesis of the first phase of the triphasic model that the increased metabolism and body temperature in endotherms evolved in response to natural selection favoring investment in parental care and high locomotor activity (Lovegrove, 2017).

Fur removal resulted in a large increase in thermal conductance in lactating females (Sadowska et al., 2016), and previously reported results showed that the thermal conductance was higher in the A-than in the C-lines and tended to be higher in the old than in the young voles (Grosiak et al., 2020). The previous study also suggested that the selection for high aerobic exercise performance resulted in a compromised ability to cope with the hot environment in older voles, even though aging had no such adverse effect in the control lines. Thus, we expected a different response to shaving in old animals from the A- and C-lines. However, neither interaction of selection direction with age or selection with shaving manipulation was found in the peak-lactation metabolic traits and litter growth rate analyses restricted to mothers with large litters. Therefore, we did not find that the heat dissipation limitation manipulation effects are more pronounced in old mothers from A-lines than in old mothers from C-lines.

To conclude, we found an evidence of a age-related reproductive performance decrease, but it was not linked with genetically based differences in metabolic rate. The results provided support for the HDL theory in its strict sense, concerning the mechanisms limiting the females' energy budget, but not the HDL theory in a broader sense, concerning factors limiting reproductive output in general. The results together do not support the novel hypothesis linking reproductive aging with the age-related decline of thermoregulatory capabilities.

We would like to acknowledge Sylwester Kunysz for providing invaluable technical assistance during the experiment. We are very grateful to the many technicians and students who helped with animal maintenance, especially Barbara Bober-Sowa and Katarzyna Baliga–Klimczyk. We thank members of our research team for their comments on the manuscript. We also would like to convey our thanks to the reviewers for valuable comments on our manuscript. The work would not be possible without free access to the animal maintenance facility and laboratories of the Institute of Environmental Sciences (Faculty of Biology, Jagiellonian University), the work of technical staff maintaining the facilities, and administration support from both the Institute secretary and the University Research Support Centre. Some results and excerpts from the Discussion in this paper are reproduced from the PhD thesis of M.G. (Grosiak, 2023).

Author contributions

Methodology: M.G., E.T.S.; Formal analysis: M.G., P.K., J.R.S., C.H.; Investigation: M.G.; Data curation: M.G.; Writing - original draft: M.G.; Writing - review & editing: P.K., J.R.S., C.H., E.T.S.; Visualization: M.G.; Supervision: P.K., J.R.S., E.T.S.; Project administration: E.T.S.; Funding acquisition: E.T.S., P.K.

Funding

This project was funded by National Science Center in Poland (Narodowe Centrum Nauki, grant no. 2016/22/E/NZ8/00416 to E.T.S.). Narodowe Centrum Nauki (grant no. 2016/23/B/NZ8/00888 to P.K.) and Jagiellonian University (Uniwersytet Jagielloński w Krakowie, grant no. N18/DBS/000003) provided funds for maintaining the long-term selection experiment and the animal facility.

Data availability

All relevant data can be found within the article and its supplementary information.

Bacigalupe
,
L. D.
and
Bozinovic
,
F.
(
2002
).
Design, limitations and sustained metabolic rate: lessons from small mammals
.
J. Exp. Biol.
205
,
2963
-
2970
.
Bao
,
M. H.
,
Chen
,
L. B.
,
Hambly
,
C.
,
Speakman
,
J. R.
and
Zhao
,
Z. J.
(
2020
).
Exposure to hot temperatures during lactation in Swiss mice stunts offspring growth and decreases future reproductive performance of female offspring
.
J. Exp. Biol.
223
,
jeb223560
.
Bateman
,
N.
(
1957
).
Some physiological aspects of lactation in mice
.
J. Agric. Sci.
49
,
60
-
77
.
Bernshtein
,
A. D.
,
Apekina
,
N. S.
,
Mikhailova
,
T. V.
,
Myasnikov
,
Y. A.
,
Khlyap
,
L. A.
,
Korotkov
,
Y. S.
and
Gavrilovskaya
,
I. N.
(
1999
).
Dynamics of Puumala hantavirus infection in naturally infected bank voles (Clethrinomys glareolus)
.
Arch. Virol.
144
,
2415
-
2428
.
Blackmer
,
A. L.
,
Mauck
,
R. A.
,
Ackerman
,
J. T.
,
Huntington
,
C. E.
,
Nevitt
,
G. A.
and
Williams
,
J. B.
(
2005
).
Exploring individual quality: basal metabolic rate and reproductive performance in storm-petrels
.
Behav. Ecol.
16
,
906
-
913
.
Boratyński
,
Z.
and
Koteja
,
P.
(
2010
).
Sexual and natural selection on body mass and metabolic rates in free–living bank voles
.
Funct. Ecol.
24
,
1252
-
1261
.
Boratyński
,
Z.
,
Koskela
,
E.
,
Mappes
,
T.
and
Schroderus
,
E.
(
2013
).
Quantitative genetics and fitness effects of basal metabolism
.
Evol. Ecol.
27
,
301
-
314
.
Bouwhuis
,
S.
,
Choquet
,
R.
,
Sheldon
,
B. C.
and
Verhulst
,
S.
(
2012
).
The forms and fitness cost of senescence: age-specific recapture, survival, reproduction, and reproductive value in a wild bird population
.
Am. Nat.
179
,
E15
-
E27
.
Broussard
,
D. R.
,
Risch
,
T. S.
,
Dobson
,
F. S.
and
Murie
,
J. O.
(
2003
).
Senescence and age-related reproduction of female Columbian ground squirrels
.
J. Anim. Ecol.
72
,
212
-
219
.
Buchalczyk
,
A.
(
1970
).
Reproduction, mortality and longevity oi the bank vole under laboratory conditions
.
Acta Theriol.
15
,
153
-
176
.
Butler
,
P. J.
,
Green
,
J. A.
,
Boyd
,
I. L.
and
Speakman
,
J. R.
(
2004
).
Measuring metabolic rate in the field: the pros and cons of the doubly labelled water and heart rate methods
.
Funct. Ecol.
18
,
168
-
183
.
Cedikova
,
M.
,
Pitule
,
P.
,
Kripnerova
,
M.
,
Markova
,
M.
and
Kuncova
,
J.
(
2016
).
Multiple roles of mitochondria in aging processes
.
Physiol. Res.
65
,
S519
-
S531
.
Clutton-Brock
,
T. H.
(
1984
).
Reproductive effort and terminal investment in iteroparous animals
.
Am. Nat.
123
,
247
-
268
.
Daniel
,
C. W.
(
1977
).
Cell longevity in vivo
. In
Handbook of the Biology of Aging
, 1st edn (
ed.
C. E.
Finch
and
L.
Hayflick
), pp.
122
-
158
.
New York
,
USA
:
Van Nostrand Reinhold
.
DeGroot
,
D. W.
and
Kenney
,
W. L.
(
2007
).
Impaired defense of core temperature in aged humans during mild cold stress
.
Am. J. Physiol. Reg.
292
,
R103
-
R108
.
Delbaere
,
K.
,
Close
,
J. C.
,
Brodaty
,
H.
,
Sachdev
,
P.
and
Lord
,
S. R.
(
2020
).
Fall risk and fear of falling in older people: the vigorous, the anxious, the stoic, and the aware
.
BMJ
125
,
167
-
174
.
Descamps
,
S.
,
Boutin
,
S.
,
Berteaux
,
D.
and
Gaillard
,
J. M.
(
2008
).
Age–specific variation in survival, reproductive success and offspring quality in red squirrels: evidence of senescence
.
Oikos.
117
,
1406
-
1416
.
Dheyongera
,
G.
,
Grzebyk
,
K.
,
Rudolf
,
A. M.
,
Sadowska
,
E. T.
and
Koteja
,
P.
(
2016
).
The effect of chlorpyrifos on thermogenic capacity of bank voles selected for increased aerobic exercise metabolism
.
Chemosphere
149
,
383
-
390
.
Drent
,
R. H.
and
Daan
,
S.
(
1980
).
The prudent parent: energetic adjustments in avian breeding 1
.
Ardea
55
,
225
-
252
.
Drożdż
,
A.
(
1968
).
Digestibility and assimilation of natural foods in small rodents
.
Acta Theriol.
13
,
367
-
389
.
Duarte
,
L. C.
,
Vaanholt
,
L. M.
,
Sinclair
,
R. E.
,
Gamo
,
Y.
and
Speakman
,
J. R.
(
2010
).
Limits to sustained energy intake XII: is the poor relation between resting metabolic rate and reproductive performance because resting metabolism is not a repeatable trait?
J. Exp. Biol.
213
,
278
-
287
.
Edwards
,
H. E.
,
Tweedie
,
C. J.
,
Terranova
,
P. F.
,
Lisk
,
R. D.
and
Wynne-Edwards
,
K. E.
(
1998
).
Reproductive aging in the Djungarian hamster, Phodopus campbelli
.
Biol. Reprod.
58
,
842
-
848
.
Ericsson
,
G.
,
Wallin
,
K.
,
Ball
,
J. P.
and
Broberg
,
M.
(
2001
).
Age–related reproductive effort and senescence in free–ranging moose, Alces alces
.
Ecology
82
,
1613
-
1620
.
Fiorotto
,
M. L.
,
Burrin
,
D. G.
,
Perez
, and
Reeds
,
P. J.
(
1991
).
Intake and use of milk nutrients by rat pups suckled in small, medium, or large litters
.
Am. J. Physiol. Regul. Integr. Comp. Physiol.
260
,
R1104
-
R1113
.
Gaillard
,
J. M.
,
Allaine
,
D.
,
Pontier
,
D.
,
Yoccoz
,
N. G.
and
Promislow
,
D. E.
(
1994
).
Senescence in natural populations of mammals: a reanalysis
.
Evolution
48
,
509
-
516
.
Grémillet
,
D.
,
Meslin
,
L.
and
Lescroël
,
A.
(
2012
).
Heat dissipation limit theory and the evolution of avian functional traits in a warming world
.
Funct. Ecol.
26
,
1001
-
1006
.
Grosiak
,
M.
(
2023
).
Can heat dissipation limit (HDL) theory explain reproductive ageing? Insights from experimental evolution in bank voles. PhD thesis, Jagiellonian University, Kraków. Repository of the Jagiellonian University
. ruj.uj.edu.pl/xmlui/bitstream/handle/item/320603
Grosiak
,
M.
,
Koteja
,
P.
,
Bauchinger
,
U.
and
Sadowska
,
E. T.
(
2020
).
Age-related changes in the thermoregulatory properties in bank voles from a selection experiment
.
Front. Physiol.
11
,
576304
.
Hammers
,
M.
,
Richardson
,
D. S.
,
Burke
,
T.
and
Komdeur
,
J.
(
2012
).
Age-dependent terminal declines in reproductive output in a wild bird
.
PLoS One
7
,
e40413
.
Hammond
,
K. A.
and
Diamond
,
J.
(
1992
).
An experimental test for a ceiling on sustained metabolic rate in lactating mice
.
Physiol. Zool.
65
,
952
-
977
.
Hammond
,
K.
and
Diamond
,
J.
(
1994
).
Limits to dietary nutrient intake and intestinal nutrient uptake in lactating mice
.
Physiol. Zool.
67
,
282
-
303
.
Hammond
,
K. A.
and
Diamond
,
J.
(
1997
).
Maximal sustained energy budgets in humans and animals
.
Nature
386
,
457
-
462
.
Hanna
,
E. G.
and
Tait
,
P. W.
(
2015
).
Limitations to thermoregulation and acclimatization challenge human adaptation to global warming
.
Int J. Environ. Res. Public Health
12
,
8034
-
8074
.
Harman
,
D.
(
1956
).
Aging: a theory based on free radical and radiation chemistry
.
J. Gerontol.
11
,
298
-
300
.
Hohlbaum
,
K.
,
Bert
,
B.
,
Dietze
,
S.
,
Palme
,
R.
,
Fink
,
H.
and
Thöne-Reineke
,
C.
(
2017
).
Severity classification of repeated isoflurane anesthesia in C57BL/6JRj mice - Assessing the degree of distress
.
PLoS One
12
,
e0179588
.
Jaromin
,
E.
,
Sadowska
,
E. T.
and
Koteja
,
P.
(
2019
).
Is experimental evolution of an increased aerobic exercise performance in bank voles mediated by endocannabinoid signaling pathway?
Front. Physiol.
10
,
640
.
Johnson
,
M. S.
and
Speakman
,
J. R.
(
2001
).
Limits to sustained energy intake: V. Effect of cold-exposure during lactation in MUS MUSCULUS
.
J. Exp. Biol.
204
,
1967
-
1977
.
Johnson
,
M. S.
,
Thomson
,
S. C.
and
Speakman
,
J. R.
(
2001
).
Limits to sustained energy intake: III. Effects of concurrent pregnancy and lactation in MUS MUSCULUS
.
J. Exp. Biol.
204
,
1947
-
1956
.
Johnston
,
S. L.
,
Souter
,
D. M.
,
Erwin
,
S. S.
,
Tolkamp
,
B. J. O.
,
Yearsley
,
J. M.
,
Gordon
,
I. J.
,
Illius
,
A. W.
,
Kyriazakis
,
I.
and
Speakman
,
J. R.
(
2007
).
Associations between basal metabolic rate and reproductive performance in C57BL/6J mice
.
J. Exp. Biol.
210
,
65
-
74
.
Kenney
,
W. L.
and
Hodgson
,
J. L.
(
1987
).
Heat tolerance, thermoregulation and ageing
.
Sports Med.
4
,
446
-
456
.
Kenney
,
W. L.
and
Munce
,
T. A.
(
2003
).
Invited review: aging and human temperature regulation
.
J. Appl. Physiol.
95
,
2598
-
2603
.
Koteja
,
P.
(
1996
).
Limits to the energy budget in a rodent, Peromyscus maniculatus: the central limitation hypothesis
.
Physiol. Zool.
69
,
981
-
993
.
Król
,
E.
and
Speakman
,
J. R.
(
2003a
).
Limits to sustained energy intake VI. Energetics of lactation in laboratory mice at thermoneutrality
.
J. Exp. Biol.
206
,
4255
-
4266
.
Król
,
E.
and
Speakman
,
J. R.
(
2003b
).
Limits to sustained energy intake VII. Milk energy output in laboratory mice at thermoneutrality
.
J. Exp. Biol.
206
,
4267
-
4281
.
Król
,
E.
,
Murphy
,
M.
and
Speakman
,
J. R.
(
2007
).
Limits to sustained energy intake. X. Effects of fur removal on reproductive performance in laboratory mice
.
J. Exp. Biol.
210
,
4233
-
4243
.
Książek
,
A.
,
Konarzewski
,
M.
and
Łapo
,
I. B.
(
2004
).
Anatomic and energetic correlates of divergent selection for basal metabolic rate in laboratory mice
.
Physiol. Biochem. Zool.
77
,
890
-
899
.
Lemaître
,
J. F.
and
Gaillard
,
J. M.
(
2017
).
Reproductive senescence: new perspectives in the wild
.
Biol. Rev. Biol. Proc.
92
,
2182
-
2199
.
Lipowska
,
M. M.
,
Sadowska
,
E. T.
,
Palme
,
R.
and
Koteja
,
P.
(
2022
).
Evolution of an increased performance under acute challenge does not exacerbate vulnerability to chronic stress
.
Sci. Rep.
12
,
1
-
16
.
Lovegrove
,
B. G.
(
2006
).
The power of fitness in mammals: perceptions from the African slipstream
.
Physiol. Biochem. Zool.
79
,
224
-
236
.
Lovegrove
,
B. G.
(
2017
).
A phenology of the evolution of endothermy in birds and mammals
.
Biol. Rev.
92
,
1213
-
1240
.
Martin
,
J. G.
and
Festa–Bianchet
,
M.
(
2011
).
Age–independent and age–dependent decreases in reproduction of females
.
Ecol. Lett.
14
,
576
-
581
.
Nilsson
,
J. Å.
(
2002
).
Metabolic consequences of hard work
.
Proc. R. Soc. B: Biol.
269
,
1735
-
1739
.
Nilsson
,
J. Å.
and
Nord
,
A.
(
2018
).
Testing the heat dissipation limit theory in a breeding passerine
.
Proc. R. Soc. B: Biol.
285
,
20180652
.
Ołdakowski
,
Ł.
,
Piotrowska
,
Ż.
,
Chrząścik
,
K. M.
,
Sadowska
,
E. T.
,
Koteja
,
P.
and
Taylor
,
J. R.
(
2012
).
Is reproduction costly? No increase of oxidative damage in breeding bank voles
.
J. Exp. Biol.
215
,
1799
-
1805
.
Ołdakowski
,
Ł.
,
Wasiluk
,
A.
,
Sadowska
,
E. T.
,
Koteja
,
P.
and
Taylor
,
J. R.
(
2015
).
Reproduction is not costly in terms of oxidative stress
.
J. Exp. Biol.
218
,
3901
-
3910
.
Peterson
,
C. C.
,
Nagy
,
K. A.
and
Diamond
,
J.
(
1990
).
Sustained metabolic scope
.
Proc. Natl. Acad. Sci. USA
87
,
2324
-
2328
.
Petrusewicz
,
K.
(
1983
).
Ecology of the bank vole
.
Acta Theriol.
28
,
1
-
242
.
Pucek
,
Z.
,
Jędrzejewski
,
W.
,
Jędrzejewska
,
B.
and
Pucek
,
M.
(
1993
).
Rodent population dynamics in a primeval deciduous forest (Białowieża National Park) in relation to weather, seed crop, and predation
.
Acta Theriol.
38
,
199
-
232
.
Rogowitz
,
G. L.
(
1998
).
Limits to milk flow and energy allocation during lactation of the hispid cotton rat (Sigmodon hispidus)
.
Physiol. Zool.
71
,
312
-
320
.
Rose
,
M. R.
(
1991
).
Evolutionary Biology of Aging
.
Oxford University Press
.
Rudolf
,
A. M.
,
Dańko
,
M. J.
,
Sadowska
,
E. T.
,
Dheyongera
,
G.
and
Koteja
,
P.
(
2017
).
Age-related changes of physiological performance and survivorship of bank voles selected for high aerobic capacity
.
Exp. Gerontol.
98
,
70
-
79
.
Sadowska
,
E. T.
,
Baliga–Klimczyk
,
K.
,
Chrząścik
,
K. M.
and
Koteja
,
P.
(
2008
).
Laboratory model of adaptive radiation: a selection experiment in the bank vole
.
Physiol. Biochem. Zool.
81
,
627
-
640
.
Sadowska
,
J.
,
Gębczyński
,
A. K.
and
Konarzewski
,
M.
(
2013
).
Basal metabolic rate is positively correlated with parental investment in laboratory mice
.
Proc. R. Soc. B: Biol.
280
,
20122576
.
Sadowska
,
E. T.
,
Stawski
,
C.
,
Rudolf
,
A.
,
Dheyongera
,
G.
,
Chrząścik
,
K. M.
,
Baliga-Klimczyk
,
K.
and
Koteja
,
P.
(
2015a
).
Evolution of basal metabolic rate in bank voles from a multidirectional selection experiment
.
Proc. R. Soc. B.
282
,
20150025
.
Sadowska
,
J.
,
Gębczyński
,
A. K.
,
Paszko
,
K.
and
Konarzewski
,
M.
(
2015b
).
Milk output and composition in mice divergently selected for basal metabolic rate
.
J. Exp. Biol.
218
,
249
-
254
.
Sadowska
,
E. T.
,
Król
,
E.
,
Chrzascik
,
K. M.
,
Rudolf
,
A. M.
,
Speakman
,
J. R.
and
Koteja
,
P.
(
2016
).
Limits to sustained energy intake. XXIII. Does heat dissipation capacity limit the energy budget of lactating bank voles?
J. Exp. Biol.
219
,
805
-
815
.
Schroeder
,
M.
,
Lavi–Avnon
,
Y.
,
Zagoory–Sharon
,
O.
,
Weller
,
A.
and
Moran
,
T. H.
(
2007
).
Preobesity in the infant OLETF rat: the role of suckling
.
Dev. Psychobiol.
49
,
685
-
691
.
Schulte
,
P. M.
(
2015
).
The effects of temperature on aerobic metabolism: towards a mechanistic understanding of the responses of ectotherms to a changing environment
.
The J. Exp. Biol.
218
,
1856
-
1866
.
Simons
,
M. J.
,
Reimert
,
I.
,
van der Vinne
,
V.
,
Hambly
,
C.
,
Vaanholt
,
L. M.
,
Speakman
,
J. R.
and
Gerkema
,
M. P.
(
2011
).
Ambient temperature shapes reproductive output during pregnancy and lactation in the common vole (Microtus arvalis): a test of the heat dissipation limit theory
.
J. Exp. Biol.
214
,
38
-
49
.
Speakman
,
J.
(
1997
).
Factors influencing the daily energy expenditure of small mammals
.
Proc. Nutr. Soc.
56
,
1119
-
1136
.
Speakman
,
J. R.
(
1998
).
The history and theory of the doubly labeled water technique
.
Am. J. Clin. Nutr.
68
,
932S
-
938S
.
Speakman
,
J. R.
and
Król
,
E.
(
2005
).
Limits to sustained energy intake IX: a review of hypotheses
.
J. Comp. Physiol.
175
,
375
-
394
.
Speakman
,
J. R.
and
Król
,
E.
(
2010a
).
The heat dissipation limit theory and evolution of life histories in endotherms-time to dispose of the disposable soma theory?
Integr. Comp. Biol.
50
,
793
-
807
.
Speakman
,
J. R.
and
Król
,
E.
(
2010b
).
Limits to sustained energy intake. XIII. Recent progress and future perspectives
.
J. Exp. Biol.
214
,
230
-
241
.
Stawski
,
C.
,
Valencak
,
T. G.
,
Ruf
,
T.
,
Sadowska
,
E. T.
,
Dheyongera
,
G.
,
Rudolf
,
A.
and
Koteja
,
P.
and
Koteja
,
P. Å.
(
2015
).
Effect of selection for high activity-related metabolism on membrane phospholipid fatty acid composition in bank voles
.
Physiol. Biochem. Zool.
88
,
668
-
679
.
Stawski
,
C.
,
Koteja
,
P.
and
Sadowska
,
E. T.
(
2017
).
A shift in the thermoregulatory curve as a result of selection for high activity-related aerobic metabolism
.
Front. Physiol.
8
,
1070
.
Stearns
,
S. C.
(
1976
).
Life-history tactics: a review of the ideas
.
Q. Rev. Biol.
51
,
3
-
47
.
Valencak
,
T. G.
,
Hackländer
,
K.
and
Ruf
,
T.
(
2010
).
Peak energy turnover in lactating European hares: a test of the heat dissipation limitation hypothesis
.
J. Exp. Biol.
213
,
2832
-
2839
.
Weiner
,
J.
(
1989
).
Metabolic constraints to mammalian energy budgets
.
Acta Theriol.
34
,
3
-
35
.
Weiner
,
J.
(
1992
).
Physiological limits to sustainable energy budgets in birds and mammals: ecological implications
.
Trends Ecol. Evol.
7
,
384
-
388
.
Weladji
,
R. B.
,
Holand
,
Ø.
,
Gaillard
,
J. M.
,
Yoccoz
,
N. G.
,
Mysterud
,
A.
,
Nieminen
,
M.
and
Stenseth
,
N. C.
(
2010
).
Age-specific changes in different components of reproductive output in female reindeer: terminal allocation or senescence?
.
Oecologia
162
,
261
-
271
.
Wu
,
S. H.
,
Zhang
,
L. N.
,
Speakman
,
J. R.
and
Wang
,
D. H.
(
2009
).
Limits to sustained energy intake. XI. A test of the heat dissipation limitation hypothesis in lactating Brandt's voles (Lasiopodomys brandtii)
.
J. Exp. Biol.
212
,
3455
-
3465
.
Yang
,
D. B.
,
Li
,
L.
,
Wang
,
L. P.
,
Chi
,
Q. S.
,
Hambly
,
C.
,
Wang
,
D. H.
and
Speakman
,
J. R.
(
2013
).
Limits to sustained energy intake. XIX. A test of the heat dissipation limitation hypothesis in Mongolian gerbils (Meriones unguiculatus)
.
J. Exp. Biol.
216
,
3358
-
3368
.
Zhao
,
Z. J.
(
2011
).
Energy budget during lactation in striped hamsters at different ambient temperatures
.
J. Exp. Biol.
214
,
988
-
995
.
Zhao
,
Z. J.
and
Cao
,
J.
(
2009
).
Effect of fur removal on the thermal conductance and energy budget in lactating Swiss mice
.
J. Exp. Biol.
212
,
2541
-
2549
.

Competing interests

The authors declare no competing or financial interests.