Effect of age and mating on body condition, fecundity and metabolic rate in female sand crickets, Gryllus ﬁ rmus

Female fecundity is dependent on age and mated status. Young female insects accumulate considerable fat stores to fuel energetically expensive ovary development and egg production. Consequently, younger females are expected to have more stored fat than older females while the latter should have larger eggloads than the former. Mating is expected to increase fecundity because male ejaculates contain fecundity-enhancing substances that stimulate egg development, ovulation and oviposition. We experimentally tested the effect of female age (young versus old) and mated status (virgin versus mated) on fat and eggloads in female Gryllus ﬁ rmus ﬁ eld crickets (Orthoptera: Gryllidae), in addition to testing the prediction that young or mated females have higher resting metabolic rates (RMR) than old or virgin females because the former should be signi ﬁ cantly more engaged in the energetically expensive activity of egg production than the latter. As predicted, we found that young females had more fat and fewer eggs than old females and that fat loads negatively correlated with fecundity across all females. Young females also exhibited higher RMR, as expected if egg production is energetically costly. Contrary to expectation, however, mated status had little effect on fat load, egg production or RMR. That mating had little effect on total egg production (i.e. stored and oviposited eggs) challenges the hypothesis that male fecundity-enhancing substances in the ejaculate stimulate egg production. Our experiment also permitted us to examine the validity of two popular indices of body condition, the scaled mass index (SMI) and residual body mass ( R i ). Neither index

Female fecundity is dependent on age and mated status.Young female insects accumulate considerable fat stores to fuel energetically expensive ovary development and egg production.Consequently, younger females are expected to have more stored fat than older females while the latter should have larger eggloads than the former.Mating is expected to increase fecundity because male ejaculates contain fecundity-enhancing substances that stimulate egg development, ovulation and oviposition.We experimentally tested the effect of female age (young versus old) and mated status (virgin versus mated) on fat and eggloads in female Gryllus firmus field crickets (Orthoptera: Gryllidae), in addition to testing the prediction that young or mated females have higher resting metabolic rates (RMR) than old or virgin females because the former should be significantly more engaged in the energetically expensive activity of egg production than the latter.As predicted, we found that young females had more fat and fewer eggs than old females and that fat loads negatively correlated with fecundity across all females.Young females also exhibited higher RMR, as expected if egg production is energetically costly.Contrary to expectation, however, mated status had little effect on fat load, egg production or RMR.That mating had little effect on total egg production (i.e.stored and oviposited eggs) challenges the hypothesis that male fecundityenhancing substances in the ejaculate stimulate egg production.Our experiment also permitted us to examine the validity of two popular indices of body condition, the scaled mass index (SMI) and residual body mass (R i ).Neither index accurately represented the true treatment effects of age and mated status on fat load; however, the SMI reflected true fat content in young, but not old, females.
© 2024 The Author(s).Published by Elsevier Ltd on behalf of The Association for the Study of Animal Behaviour.This is an open access article under the CC BY license (http://creativecommons.org/licenses/ by/4.0/).
Body condition generally refers to the energetic state of an animal and assumes that individuals with greater fat reserves are in better condition because they have more resources to fuel various fitness-related functions.In insects, the fat body is responsible for the synthesis and storage of energy-rich molecules such as glycogen and lipid, among others (Arrese & Soulages, 2010;Canavoso et al., 2001;Chapman, 1998;Downer & Matthews, 1976;Skowronek et al., 2021).Fat body components are mobilized for a wide variety of conditions and processes such as flight, immunity, starvation and embryogenesis.During embryogenesis, the fat body produces considerable amounts of the major egg protein vitellogenin, which is taken up by the developing oocytes (Downer & Matthews, 1976;Hoffmann, 1995;Jacome et al., 1995;Kawooya & Law, 1988;Ziegler & Ibrahim, 2001) and supplies the major portion of lipid in the oocyte (Canavoso et al., 2001;Downer & Matthews, 1976;Jacome et al., 1995;Jouni et al., 2003).
Ovary development and egg production are energetically expensive activities in insects (e.g.Terblanche et al., 2004), which might partially explain why younger females often have higher metabolic rates than older females (Piiroinen et al., 2010).It is therefore critical for females to accrue resources in early adult life before sexual maturity to meet the energetic demands of embryogenesis (Piiroinen et al., 2010; see also : Anand & Lorenz, 2008;Arrese & Soulages, 2010;Bommarco, 1998).One hypothesis suggests that the delay of days to weeks between the onset of adulthood (eclosion) and full sexual maturity in insects (e.g.Orthoptera: Cade & Wyatt, 1984;Kaufmann, 2017;Robson & Gwynne, 2010;Diptera: Boyce, 1934;Butterworth et al., 2020;Teskey, 1969;Lepidoptera: Scott, 1972;Odonata: Corbet, 1980) evolved to facilitate the acquisition of sufficient energy reserves to fuel the costly development of ovaries and eggs (e.g.Clifford & Woodring, 1986;Lorenz & Anand, 2004;Piiroinen et al., 2010).Indeed, empirical evidence across a wide variety of insect taxa shows that the female fat body increases in size immediately after eclosion to adulthood (e.g.Lorenz, 2007;Lorenz & Anand, 2004) and then decreases with age after sexual maturation (Casas et al., 2005;Lorenz & Anand, 2004;Nestel et al., 2005;Stahlschmidt & Chang, 2021).The depletion of the fat body near the time of sexual maturity tends to be coincident with a gain in ovary mass or egg production, suggesting that energy reserves in the fat body are transferred to oocytes in preparation for mating (Ellers, 1996;Lorenz & Anand, 2004;Ziegler & Ibrahim, 2001).The long-term effects of age on eggload (i.e.stored eggs) or total egg production (i.e.stored and oviposited eggs), however, are less clear as reproductive output can increase with adult age and plateau midlife (e.g.Bommarco, 1998;Lorenz, 2007;Sisterson, 2008), reach a peak at midlife and then decrease thereafter (e.g.Aluja et al., 2001;Lorenz & Anand, 2004;Nestel et al., 2005;Pervez et al., 2004), or decline with age (e.g.Morais et al., 2012;Tasnin et al., 2021).Predicting the egg and fat loads of older females is therefore difficult.
A female's relative fat load versus eggload likely depends on female age, particularly the time relative to sexual maturity, and mating status.Few studies have simultaneously quantified fat load and eggload of females varying in age (i.e.young versus old) (e.g.Lorenz, 2007;Lorenz & Anand, 2004) and no study to date has investigated whether age and mating status act interactively on fat content or egg production.It is possible, for example, that younger mated females have greater eggloads (and less fat) than younger virgins while older mated females have smaller eggloads (and possibly the same amount of fat) than older virgins.Moreover, knowing how age and mating status affect fat reserves and eggloads has important practical implications.The concept of body condition is foundational to a plethora of hypotheses in evolutionary and behavioural ecology because many sexually selected and life-history traits are condition dependent (Andersson, 1994;Tomkins et al., 2004).Ideally, the fat stored by individuals should be quantified to accurately assess how much fuel an individual has at their disposal at any given time.Because this approach requires destructive sampling of the individuals under study (Schamber et al., 2009;Stevenson & Woods, 2006;Williams & Robertson, 2008), several nondestructive statistical techniques have been developed (e.g.Ardia, 2005;Kelly et al., 2014;Peig & Green, 2009, 2010).All these techniques quantify, in one way or another, an individual's body mass while accounting for its structural body size while assuming that relatively heavier individuals have more stored fat.These indices, however, might inaccurately reflect body condition if animals having the same relative body mass differ in their amount of stored fat because of sex, age or mating status (e.g. Kelly et al., 2014;Stahlschmidt & Chang, 2021).
Here, we used a fully factorial experimental design to test how the relative age (young versus old) and mating status (virgin versus mated) of sexually mature female sand crickets, Gryllus firmus (Orthoptera: Gryllidae), affect fat content, eggloads and resting metabolic rate.First, we predicted that younger females would have greater fat loads than older females while the opposite pattern would be expected for eggloads.Additionally, we also expected that mated females would have less stored fat and greater egg production than virgins.Second, we measured the resting metabolic rate of female crickets to test the hypothesis that the conversion of fat to ova is energetically expensive and that these costs differ according to female age and mating status.We predicted that younger females would have higher metabolic expenditures than older females because they are developing their ovaries and growing their oocytes.We also predicted that mated females would have higher metabolic rates than virgins because copulation stimulates egg production, which is energetically expensive.Finally, we examined whether two popular indices of body condition, residual mass and scaled mass index (Peig & Green, 2009, 2010), accurately reflect the actual fat loads of females within each experimental treatment combination.

Study Animals
Experiments were conducted in the laboratory from July to September 2019 at Universit e du Qu ebec a Montr eal and used crickets that were laboratory-raised descendants of individuals collected in Gainesville, Florida, U.S.A. Crickets were maintained communally in 70-litre bins in a room at constant temperature (28 C) and relative humidity (60%) on a reversed 12:12 h light:dark cycle and were provided with cotton-plugged water tubes, Iams Proactive Health cat food ad libitum and layers of cardboard egg cartons for shelter.

Experimental Design
We haphazardly collected penultimate-instar females from the colony and placed them individually in 250 ml plastic deli containers with holes in the lid for aeration.Each cricket was provided with ad libitum water, a piece of cardboard egg carton for shelter and a piece of cat food once per week.Crickets were checked daily and the date of their eclosion to adulthood was recorded.Only long-winged morphs were used in this study.
We used a 2 Â 2 fully factorial experimental design to test the effect of age and mating status (and their interaction) on the fat load and eggload of sexually mature females.On the day of eclosion to adulthood, females were haphazardly assigned to one of four treatment groups: young (7e10 days posteclosion) or old (23e26 days posteclosion) and either mated or virgin (unmated).Females assigned to the mated treatment were individually placed in an empty 250 ml deli cup and paired a with sexually mature stimulus male that was randomly selected from the cricket colony.The pair was observed under red light until a spermatophore was transferred.If a spermatophore was not transferred within 30 min, then the female was discarded.Stimulus males were reused on different days.Unmated (virgin) females were similarly treated but were paired for 30 min with a randomly selected penultimate-instar male (i.e.nonreproductive juvenile).Females were weighed to the nearest 0.01 mg using a Mettler Toledo XP26 analytical balance before the mating treatment.After the mating treatment, females were replaced in their original container with food, water and shelter and given 48 h to oviposit.Females oviposited in the cotton plug of their water tubes; all oviposited eggs were counted.The final sample sizes for the four treatment groups were: young virgin (N ¼ 47) and mated (N ¼ 51) and old virgin (N ¼ 39) and mated (N ¼ 52).Body size (pronotum length) did not significantly differ among treatment groups (age: estimate ¼ 0.08 ± 0.05, t ¼ 1.51, P ¼ 0.133; mated status: estimate ¼ 0.10 ± 0.06, t ¼ 1.81, P ¼ 0.072).

Respirometry
All food (but not the water/oviposition vial) was removed 24 h before the respirometry measurements to standardize fasting across individuals.We measured the resting metabolic rate (RMR) of each female over 23 h, 2 days after the experimental mating treatments: 9e12 days posteclosion for young females and 25e28 days for old females.Measurements were recorded from approximately 1600 to 1500 hours the following day.Females were weighed to the nearest 0.01 mg immediately before placement in a 21.99 cm 3 respirometry chamber.Respirometry chambers were made of glass and wrapped in black electrical tape to block out light and to prevent females from seeing each other.We measured H 2 O and CO 2 production as proxies for metabolic rate using a Sable Systems stop-flow respirometry system (Sable Systems International (SSI), Las Vegas, NV, U.S.A.).
Dry, CO 2 -free air was pumped with a MFS-2 mass flow system (SSI) at 50 ml/min through eight chambers connected to an RM-8 multiplexer (SSI), allowing sequential measurements in the chambers.A combination of Pharmed (Sigma-Aldrich, St Louis, MO, U.S.A.) and Bev-A-Line (Cole-Parmer, Vernon Hills, IL, U.S.A.) tubing was used to connect the different components of the respirometry system.H 2 O and CO 2 were scrubbed via a purge gas generator (Parker Balston, Haverhill, MA, U.S.A.: Model 75-45) and an inline scrubbing column (Drierite, Xenia, OH, U.S.A.: Stock no.26800) filled with 30% Ascarite and 70% Drierite.H 2 O was measured with a RH-300 relative humidity/dewpoint meter (SSI) and CO 2 with a CA-10 carbon dioxide analyser (SSI).A scrubbing column containing magnesium perchlorate (Mg(ClO 4 ) 2 ) was placed in line between the CO 2 and H 2 O analysers, and similar column filled with Mg(ClO 4 ) 2 and Ascarite was placed between the CO 2 and O 2 analysers.
Data were acquired at 1 Hz with a UI-2 interface using the software Expedata (SSI).Experiments were conducted in a temperature-controlled room (range 22.2e23.7 C) and temperature was monitored using a thermistor temperature sensor (SSI) connected to the UI-2 interface.Baseline values were taken for 3 min by analysing gas in an empty chamber at the beginning and end of each 27 min respirometry cycle.Baselines were used to provide accurate zero values and to correct instrument drift and sensor lag.Mean _ VCO 2 was calculated from the areas under the curves.
Females were weighed to the nearest 0.01 mg immediately after removal from the chambers to account for any fat stores being mobilized during respiration measurement.Females were euthanized after weighing by freezing at À20 C. All three gas analysers were calibrated (zero and span) every 2 weeks.

Fat Load and Egg Production
Females were removed from the freezer and their pronotum length (mm), defined as the distance between the anterior and posterior edges of the pronotum, was measured to the nearest 0.01 mm under a stereomicroscope using Leica LAS image analysis software (Leica Microsystems Inc., Buffalo Grove, IL, U.S.A.).
Thawed females were weighed to the nearest 0.01 mg (fresh mass) and their mature eggs in the ovary were removed by dissection of the abdomen and counted.Total egg production was calculated as the sum of the number of mature eggs in the ovary and the number of eggs oviposited (e.g.Bentur et al., 1977).
All the dissected material (except the eggs) was placed back into the female's abdominal cavity and the female was then dried at 60 C for 24 h and reweighed to obtain dry mass.Body fat was then extracted using petroleum ether (Fisher Scientific, Hanover Park, IL, U.S.A.) reflux in a Soxhlet apparatus for 24 h.Individuals were again dried at 60 C for 24 h and then reweighed to obtain their lean dry mass.Body fat content (mg) was obtained by subtracting lean dry mass from dry mass.

Scaled Fat and Body Mass Indices
We calculated standardized fat load as the scaled fat index (SFI) and body condition as the scaled mass index (SMI).Both measures account for the covariation between body size and fat load or body mass, respectively (Peig & Green, 2009).
The predicted fat load or body mass (hereafter adjusted trait) for female i is: where T i is the trait to adjust (measured after respirometry) and L i is the pronotum length of female i.L o is the mean pronotum length of the population, which in our case is 3.96 mm.b SMA is the scaling exponent calculated by the standardized major axis (SMA) regression of ln(T) on ln(L), which in our case was 8.28 (fat load) and 2.85 (body mass).We calculated residual mass (R i ) by entering the dependent variable log body mass (mass after respirometry) into an ordinary least squares (OLS) regression model with log pronotum length as the independent variable.The standardized residual was then extracted for each cricket.

Ethical Note
Our research followed the ASAB/ABS Guidelines for the use of animals in research.Gryllus firmus is not listed as a threatened species and requires no licence to be studied in Canada.Crickets were maintained under optimal laboratory conditions and were freeze-killed prior to dissection and fat extraction.

Statistical Analysis
Statistical analyses were performed within the R statistical environment (R Core Team, 2022).We used a linear model (LM) to estimate treatment effects on scaled fat load (SFI).Our model comprised SFI as the response variable and female age (young versus old), mated status (virgin versus mated) and their interaction as explanatory factors.Body size was not included as a covariate in this model because fat load is already standardized for body size.We used the 'caret' R package (Kuhn, 2008) to perform a YeoeJohnson transformation (Yeo & Johnson, 2000) of SFI because model residuals failed to meet the assumption of normality.
A negative binomial model was used to test treatment effects on the number of mature eggs in the ovary.A negative binomial error distribution was required because the data were overdispersed (i.e. the variance exceeds the mean).Our model comprised egg number as the response variable, female age, mated status and their interaction as explanatory factors and female body size (i.e.log pronotum length) as a covariate.Female body size was entered as a covariate because fecundity is generally positively correlated with female body size in gryllid field crickets (e.g.Worthington & Kelly, 2016).
Treatment effects on the number of eggs laid were tested using a hurdle negative binomial model because egg counts were overdispersed and had an excess of zeroes (78% of females did not lay eggs).A hurdle model is a two-part model comprising a binomial model and a zero-truncated negative binomial model in which the number of eggs laid was the response variable, female age, mated status and their interaction were explanatory factors and female body size (i.e.log pronotum length) was a covariate.This analysis resulted in a two-stage analysis, in which we first determined whether treatments significantly affected a female's probability of laying eggs (i.e.compare the proportion of individuals that did not lay eggs versus those that did), and then whether treatments affected the number of eggs laid for only those individuals that laid eggs.We used 'hurdle' from the 'pscl' R package (Zeileis et al., 2008) to perform this analysis.
We used a negative binomial model to test treatment effects on total egg production (ovarian eggs plus oviposited eggs).A negative binomial was required because data were overdispersed.Our model comprised total egg number as the response variable, female age, mated status and their interaction as explanatory factors and female body size (i.e.log pronotum length) as a covariate.Pearson product-moment correlation was used to quantify the strength and direction of the relationship between fat load (SFI) and residual total egg production (i.e.fecundity controlling for body size) for each of the four treatment groups.Residual total egg production was calculated by regressing total egg count (ovarian eggs plus oviposited eggs) on log pronotum length and then extracting the standardized residual for each cricket.
We used a LM to test the effect of our treatments on two indices of body condition, the scaled mass index (SMI) and residual mass (R i ).Our models comprised either SMI or SFI as the response variable and female age and mated status as explanatory factors.Pearson product-moment correlation was used to quantify the strength and direction of the relationship between fat load (SFI) and SMI or R i for each of the four treatment groups.
A general linear mixed model was used to test the effect of female age, mated status and their interaction on resting metabolic rate.We entered VCO 2 as the response variable, female age, mated status and their interaction as explanatory factors, female body mass (log-transformed mass before respirometry trial) as a covariate and individual identity (ID) as a random effect.Because each female contributed N ¼ 49 VCO 2 measures, a mixed model allowed us to control for repeated measurements.
Model assumptions were checked using the R package 'performance' (Lüdecke et al., 2021).All continuous variables were scaled (mean ¼ 0, SD ¼ 1) prior to analysis and raw (untransformed) means are presented ± 1 SD unless otherwise noted.All statistical tests were conducted at the a ¼ 0.05 level.

Body Condition Indices
If body condition indices accurately reflect body condition (i.e.fat load), then female age and mated status should have the same effect on them as they do on fat load.We therefore expected that female age and mated status would have the same effect on SMI and residual mass as on SFI (see Fig. 1a).
An interaction between female age and mated status on SMI (estimate ¼ 0.58 ± 0.24, t ¼ 2.39, P ¼ 0.018) indicates that young virgin and young mated females did not differ in SMI whereas old mated females had a larger SMI than old virgin females (Fig. 3a).

DISCUSSION
As predicted, young female G. firmus field crickets had significantly greater fat loads (i.e.SFI) than old females, and old females had significantly greater eggloads (mature eggs in the ovary) and total egg production (ovarian and oviposited eggs) than young females (Fig. 1).Moreover, we found that fat load negatively correlated with total egg production when all females were pooled.These results provide strong support for the hypothesis that fat stores are converted to eggs in G. firmus field crickets.Unfortunately, empirical studies quantifying fat stores and egg number simultaneously with female age in insects are rare.One such study, however, showed that as the fat stores in female Gryllus bimaculatus field begin to decline 2e3 days posteclosion, ovary mass begins to increase (Lorenz & Anand, 2004), suggesting that the fat stores are fuelling ovary development and egg production.Another study by Clifford and Woodring (1986) showed that the fat body of Acheta domesticus house cricket females steadily declined for 12 days after eclosion while egg production concomitantly increased.The bulk of studies investigating these topics have examined the effect of age on fat load and eggload separately.For example, Stahlschmidt and Chang (2021) showed that, as in our study, the fat content of female G. firmus that recently sexually matured (7e11 days old) was significantly larger than that of old females (21e25 days old).Ellers (1996) found that the fat load of female Asobara tabida parasitoids linearly decreases with age, whereas Nestel et al. (2005) found that, although the total amount of lipid in adult female Mediterranean fruit flies, Ceratitis capitata, fluctuated with age, it declined overall.These latter two studies assumed that female fat load was converted to eggs over time.However, this assumption must be directly tested because egg production can have different relationships with age: egg production can decline linearly with age (e.g.Morais et al., 2012;Tasnin et al., 2021) or reach a peak at midlife and decrease thereafter (e.g.Aluja et al., 2001;Lorenz & Anand, 2004;Nestel et al., 2005;Pervez et al., 2004).Our data suggest that egg production does not decline with age in G. firmus since ovarian eggload and total egg production were both larger in older females than in younger females.Regardless, our two-point assessment of egg production (young versus old) cannot reveal egg production over time at a finer scale.For example, we do not know whether egg production plateaus at midlife as in female Teleogryllus commodus (Rence et al., 1987) field crickets and Pterostichus cupreus ground beetles (Bommarco, 1998), whether there are multiple peaks and valleys across time as in Homalodisca vitripennis leafhoppers (Sisterson, 2008), or whether it increases linearly from first mating.
We expected to find more mature eggs in the ovary of mated females because male gland products in the spermatophore should increase egg production at the expense of stored fat (e.g.Gillott, 2003).However, contrary to prediction, mating did not affect either fat load or egg production (eggload, total production).An alternative prediction is that, because male gland products are known to also stimulate oviposition behaviour in crickets (Strambi et al., 1997;Worthington & Kelly, 2016), which we observed in the current study, perhaps we might expect fewer mature eggs in the ovaries of mated females.For example, Clifford and Woodring (1986) showed that a high oviposition rate in mated female A. domesticus kept the number of eggs in the ovary low at any given time.Again, our data do not support this alternative prediction despite mated females ovipositing at a higher rate than virgin females.One explanation for our results is that 72 h (i.e.48 h oviposition time plus 24 h during respirometry) might not be sufficient to permit the male gland products to stimulate egg production.This seems unlikely since ejaculate products can act quickly on egg-laying behaviour in crickets (Destephano et al., 1982;Loher & Edson, 1973;Stanley-Samuelson et al., 1986;Zhao & Zhu, 2011).That said, however, some studies that have performed daily egg counts after mating have found that female crickets begin to lay eggs 5 days after copulation (A.domesticus: Murtaugh & Denlinger, 1985).
Contrary to prediction, neither old nor mated female G. firmus were more likely to initiate oviposition.However, as expected, of those females that did lay eggs, older females laid more eggs than younger females and mated females laid more eggs than virgins.This finding suggests that older females have more eggs ready to oviposit (as also suggested by their larger ovarian eggloads) and that male ejaculates contain fecundity-enhancing substances that stimulate oviposition (but apparently not egg production).In line with our findings, Rence et al. (1987) found that older female T. commodus crickets laid more eggs than younger females.Our results contrast with Limberger et al. (2021), who found no difference between young mated and old mated female Gryllus assimilis, and Wilson and Walker (2019) and Lorenz (2007), who found that younger A. domesticus and G. bimaculatus females, respectively, lay more eggs than older females.Our results are also in line with studies demonstrating that mated female crickets lay more eggs than virgins (e.g.Gryllus texensis: Worthington & Kelly, 2016;G. assimilis: Limberger et al., 2021;A. domesticus: Clifford & Woodring, 1986), although Bentur et al. (1977) did not support this hypothesis in G. bimaculatus.Regardless of the proximate mechanism underlying egg-laying behaviour, it is apparent that males might prefer to mate with older females because these females not only have more mature eggs that are ready for oviposition but also lay more eggs, particularly if they are terminally investing due to declining residual reproductive value (Clutton-Brock, 1984).This advantage, however, could be counterbalanced by the increased likelihood that older females are also storing more sperm of rivals and thus likely represent a greater risk of sperm competition (e.g.Herdman et al., 2004;Martin & Hosken, 2002).As expected, young female G. firmus crickets in our study had significantly higher RMR than old females.This is a general pattern across humans (Fukagawa et al., 1990;Roberts & Rosenberg, 2006) and other animals (e.g.Andreasson et al., 2020;Briga & Verhulst, 2021;Elliott et al., 2015;McCarter & Palmer, 1992), including insects (e.g.Gray & Bradley, 2003;Hack, 1997), and is generally due to a decrease in the volume of metabolically intense tissue or a decline in average tissue metabolic intensity with age (see Elliott et al., 2015).Young crickets in our study were likely engaged in the metabolically intense (and energetically expensive) activity of mobilizing fat stores to fuel ovary development and egg production.
We expected copulation to further stimulate the energetically expensive activity of egg production.Contrary to this expectation, however, mated females did not have significantly greater eggloads and did not exhibit a significant increase in RMR compared with virgins.Gray and Bradley (2003) also showed no effect of mating on RMR in Culex tarsalis mosquitoes.In contrast to our findings, Giesel et al. (1989) showed that mated female Drosophila simulans have higher metabolic activity than virgins, but intriguingly they found that even after controlling for fecundity (i.e.egg production), mated females still had significantly higher metabolic rates.This suggests that something beyond egg production is imposing an energetic cost on females.A possible explanation for this result lies in their observation that continual presence and prolonged physical contact with a male addition to an ejaculate) increased RMR compared to insemination and then isolation (Giesel et al., 1989).They suggested that mated females that are in continual physical presence of males might expend additional energy producing antiaphrodisiac substances.Perhaps something similar occurs in crickets and consequently, we might have observed higher RMR in mated females if their mates were permitted to remain with them postcopula.Terblanche et al. (2004) also found that mating increased the metabolic rate of female G. pallidipes flies.This result, however, is confounded by female age as mated females were older than virgins.
Our experiment provided an opportunity to assess the validity of two body condition indices that are commonly used by evolutionary and behavioural ecologists: the scaled mass index and residual body mass (Peig & Green, 2009, 2010).In short, neither of the two indices accurately reflected actual body condition, assuming that body condition is based on fat load or energy stores.We found that SFI was greater in young females but was unaffected by a female's mated status.Neither of the two indices exhibited this pattern, meaning that their use would lead to false conclusions of the effect of female age and mated status on body condition.Residual mass more closely matched the effect of female age and mated status on SFI: there was no interaction between female age and mated status, and young females had significantly greater residual mass than old females.However, residual mass was also greater in virgins whereas there was no effect of female mated status on fat load.SMI poorly matched body condition since an interaction between female age and mated status indicated that old virgin females were in better condition than old mated females and that young females exhibited similar condition regardless of whether they were mated or virgin.What is more, the use of SMI suggested that young females were generally in poorer condition than older females, which is opposite to reality.We also found that residual mass negatively correlated with SFI within each treatment group, as well as when females were pooled across treatment groups, which suggests that relatively heavier females had smaller fat loads than lighter females.Therefore, it would not be advisable to correlate residual mass with, for example, putatively condition-dependent behaviours.In contrast, SFI correlated with SMI in young but not in old females, probably because the body mass of older females is composed of eggs and not fat.Therefore, correlating SMI with, for example, putatively condition-dependent behaviours in young females would be valid.These latter results highlight the complex relationship between body condition indices, fat load and other condition-dependent traits.As such, they reinforce the need to identify how fat load relates to body mass in experimental subjects before deciding whether to use an index and, if so, which one.Of course, this is not possible for those working on animals in which destructively quantifying fat load is not possible (e.g. Kelly & Gwynne, 2023;Stevenson & Woods, 2006).
In conclusion, our experimental study supports the hypothesis that young female G. firmus crickets mobilize fat stores for egg production and that this process is energetically expensive.Our study also suggests, somewhat surprisingly, that copulation has little effect on egg production.We also show that two popular indices of body condition have little general relevance to actual fat load but that one (i.e.SMI) could be valid under very strict circumstances (i.e. using young females).Finally, we argue that fat load should be directly quantified if hypotheses of condition dependence are to be tested.

Figure 1 .
Figure 1.Effect of age and mated status on (a) scaled fat index (SFI) and (b) ovarian eggload in female G. firmus sand crickets.The box represents the lower (25%) and upper (75%) quartiles, the solid dark horizontal line is the median, and the whiskers indicate 1.5 times the size of the hinge, i.e. the 75% minus 25% quartiles.Sample sizes are given below the boxes.*P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3 .
Figure 3.Effect of age and mated status on (a) scaled mass index (SMI) and (b) residual total egg production in female G. firmus sand crickets.See Fig. 1 for explanation of boxes and whiskers.Sample sizes are given below the boxes.*P < 0.05; **P < 0.01; ***P < 0.001.