Heatwaves inflict reproductive but not survival costs to male insects

Climate change is having a profound effect on the environment as a result of rising global temperatures and extreme climatic events, such as heatwaves, becoming more frequent (Christidis et al., 2015; Domeisen et al., 2023; Frich et al., 2002). Heatwaves are generally defined as a 5°C increase of temperature for a period of five consecutive days (Frich et al., 2002; Sales et al., 2018). Their environmental effects, mostly neglected until recently, are clearly posing a threat to biodiversity (Ummenhofer and Meehl, 2017). The sudden and extreme thermal stress occurring with a heatwave can be particularly disruptive for a variety of biological functions (Dowd et al., 2015; Van de Pol et al., 2017), and even more so than gradual increases in the mean temperature (Sheldon and Dillon, 2016). In animals, including humans, heatwaves may have not only lethal effects (McKechnie and Wolf, 2010; Ratnayake et al., 2019; Strydom et al., 2020; Wild et al., 2019; Zhao et al., 2018) but also sublethal effects by impairing animal cognitive or motor performance (Danner et al., 2021), activity levels (Mameri et al., 2020), cell physiology (Jimenez and Williams, 2014) or foraging activities (du Plessis et al., 2012; Hemberger et al., 2023). Surprisingly, the effects of heatwaves on reproduction, which is a central biological function determining individual fitness and population persistence, have only been the focus of a handful of studies (Breedveld et al., 2023; Chevrier et al., 2019; Hurley et al., 2018; Martinet et al., 2021; Pilakouta et al., 2023; Sales et al., 2018; Stahlschmidt et al., 2022; Wild et al., 2019). This is unfortunate because the heat threshold at which fertility deteriorates (namely, thermal fertility limits) is generally much lower than the critical temperature limits suggested for survival (i.e. the upper temperature thresholds at which critical body functions of an organism fail, known as CT_max; Parratt et al., 2021; Walsh et al., 2019).

Heatwaves can be particularly detrimental to male fertility through direct effects on gamete production and performance (Walsh et al., 2019) or indirect effects on mating and the expression of secondary sexual traits (behaviours or morphologies; Abram et al., 2017; Candolin, 2019; García-Roa et al., 2020). Studies reporting, for example, heat-induced damage to male gametes and gamete performance are accumulating (e.g. Gasparini et al., 2018; Nguyen et al., 2013; Pérez-Crespo et al., 2008; Reinhardt et al., 2015). But very few studies have focused specifically on heatwave-like temperature conditions (Breedveld et al., 2023; Hurley et al., 2018; Leach et al., 2021; Sales et al., 2018). Moreover, although temperature can modulate behavioural interactions among mating partners (García-Roa et al., 2020), the consequences of heatwaves on male sexual signalling and mating behaviours remain poorly understood. For example, exposure to a heatwave reduces the level of courtship displays and colour ornamentation used for sexual signalling by male guppies (Breedveld et al., 2023). Likewise, exposure to a heatwave alters courtship displays (i.e. stability of wing vibrational signals) and odour profiles of male solitary bees (Conrad et al., 2017). Meanwhile, heat may affect male mating success by altering female signal recognition and preference. For example, female zebra finches that are experiencing a heatwave stop discriminating between conspecific and heterospecific male songs (Coomes et al., 2019). In contrast, females of a desert-dwelling jumping spider only show preference for male courtship when they are exposed to heat (Brandt et al., 2020). Overall, these pioneering studies clearly suggest that heatwaves can have widespread effects on male fertility. Investigating the effects of heatwaves on reproductive traits is therefore timely and necessary to fully uncover the impact of climate change on individual reproductive success and population persistence (Buckley and Huey, 2016; Wild et al., 2019).

Insects are key for advancing our understanding of the impact of heatwaves on organisms, given that they are ectotherms and generally lack physiological mechanisms for thermoregulation (González-Tokman et al., 2020), which makes them particularly dependent on environmental temperature (Deutsch et al., 2008; Kingsolver et al., 2013). As such, unpredictable heat stress associated with heatwaves can have profound impacts on individual reproduction and population persistence (Sales et al., 2021), ultimately contributing to the global decline of insect populations (Harvey et al., 2023; Martinet et al., 2021). This idea is supported by prior work revealing that heat affects a large variety of behavioural interactions and traits that are important for mate acquisition in insects (González-Tokman et al., 2020; Leith et al., 2021) via its effects on metabolic rate and activity levels (Abram et al., 2017). Under heat stress, there is a general decline in energy production, which, together with the activation of energetically costly physiological responses, such as heightened metabolic rates, leads to reduced energy availability (Chown and Nicolson, 2004). Energetic constraints may therefore mediate survival and reproductive costs under heat stress if energy is reallocated to maintain the physiological energetic balance (i.e. homeostasis; Angilletta et al., 2003; Niehaus et al., 2012). For example, increased temperatures appear to exacerbate the trade-offs between disease resistance and reproduction in field crickets (Adamo and Lovett, 2011), and between heat tolerance and fecundity in pea leaf miners (Huang et al., 2007). Therefore, maintaining homeostasis in response to heat might require insects to reallocate resources towards the soma by, for instance, reducing investment in other functions such as reproduction and/or increasing energy intake.

Here, we tested whether heat stress derived from ecologically relevant heatwaves affects the mating behaviour, fertility, mass change and survival of the Mediterranean field cricket, Gryllus bimaculatus. This species is an ideal study system to investigate the effects of heatwaves on insect reproduction and survival for three main reasons. Firstly, G. bimaculatus might frequently experience a broad range of temperatures given that the species is often found in habitats with extreme temperature fluctuations, such as arid and semi-arid areas, and has a wide distribution that includes tropical and subtropical areas of Europe, Asia and Africa (Ferreira and Ferguson, 2010; Harrison and Bogdanowicz, 1995; Ragge, 1972). Secondly, field crickets are model organisms in the study of male mating behaviour and female mate choice given that male–female interactions are easily tractable in laboratory settings and their reproductive biology has been particularly well studied (Horch et al., 2017). Males compete over breeding territories (i.e. cracks and burrows in the ground) through aggressive male–male fights (Adamo and Hoy, 1995). Fight winners sing to attract females to their territory (Simmons, 1986; Tachon et al., 1999) and court females when they are in close vicinity (Simmons, 1986). Females exert mate preferences based on male song quality (Rantala and Kortet, 2003; Verburgt et al., 2011), body size (Bateman et al., 2001; Simmons, 1986) and chemical profile (Iwasaki and Katagiri, 2008; Tregenza and Wedell, 1997). Females mount males to allow genital coupling and transfer of the male's spermatophore (i.e. sperm-filled protein capsule; Sakai et al., 1991). Thirdly, field crickets are known to be particularly sensitive to temperature change, as temperature affects a multitude of functions, including metabolism, physiology, immunity and behaviour. For example, crickets maintained at higher temperature exhibit increased metabolic rate (Nespolo et al., 2003), faster egg-to-adult development (Odhiambo et al., 2022), increased egg laying, faster egg development and increased egg mass (Adamo and Lovett, 2011), and increased ovary mass (Stahlschmidt et al., 2022). In males, higher temperatures lead to greater food consumption and body mass (Behrens et al., 1983; Hoffmann, 1974), faster sexual maturation (Gasparini et al., 2018), increased song frequency (Martin et al., 2000; Van Wyk and Ferguson, 1995) and increased exploratory behaviour (Niemelä et al., 2019). Changes in temperature also affect sperm production and quality: males at higher temperatures produce less sperm and of lower viability (Gasparini et al., 2018). To our knowledge, however, no study so far has tested whether heatwaves can also affect male mating interactions, reproductive success and survival.

In this study, we simulated heatwaves of two different magnitudes to test whether short heat exposure affects behaviour, fertility and survival in male G. bimaculatus. We exposed males to simulated heatwave conditions for five consecutive days at 33°C, 39°C or standard rearing conditions (28°C; controls). To estimate the short-term consequences of heatwaves on male fitness, we paired experimental males with control females (kept at 28°C) and measured their mating behaviour (i.e. courtship song duration, latency to attract a female), reproductive success (i.e. successful mating and number of offspring) and lifespan. We also tested for effects of heat treatments on male energy intake, which we estimated by measuring the change in body mass over heatwave exposure. Overall, we predicted that the most severe heatwave treatment would have detrimental effects on male fitness (Gasparini et al., 2018), reducing mating success, offspring number and lifespan (Angilletta et al., 2003). Assuming that the acquisition of energetic resources is essential for recovering from heat damage, males may increase their feeding rate and/or benefit from more efficient food intake at higher temperatures (Behrens et al., 1983; Hoffmann, 1974). Hence, if such a buffering mechanism takes place efficiently, we would expect heat-treated males to gain more weight than control males.

We used a large outbred laboratory population of Gryllus bimaculatus De Geer 1773, which originated from wild-caught animals in Seville (Spain) collected during the summer of 2021. The average summer temperatures in Seville range from 24.5°C to 28.4°C, with average minimum temperatures of 19.0–21.8°C and average maximum temperatures of 30.6–35.3°C (https://en.climate-data.org/). Animals were transported to LMU Munich (Germany) where they were raised in multiple tanks (20×37×30 cm) containing 30–40 crickets each. They were maintained in temperature- and humidity-controlled climatic rooms with a 14 h:10 h light:dark cycle at 28°C and 65% humidity. Tanks were equipped with egg cartons to provide shelter, water vials with cotton stoppers and ad libitum access to dry cat food (Ja! Knusper-Mix Rind & Gemüse) for large nymphs and adults, and fish flakes (sera^® Pond flakes) for nymphs. Small cups (diameter×height: 7×4.5 cm) with moist soil were provided in adult tanks for egg laying. Newly hatched nymphs were raised communally as described above.

The crickets used in the experiment were randomly selected among stock nymphs at the penultimate stage before adult emergence, ensuring that all individuals in our experiment remained unmated. Males were isolated individually into smaller containers (i.e. dark grey, non-transparent plastic containers of 10×10×9 cm) equipped with a small plastic shelter, a vial containing water and ad libitum food. Females were placed jointly in groups of five in small tanks (15×23×20 cm) equipped with egg cartons, two water vials and ad libitum food. We recorded adult emergence by monitoring male individual containers and female tanks every second day. This enabled us to standardize both age at testing and the absence of prior mating experience, reducing sources of variation and facilitating comparisons across treatments. After females had reached adulthood, their tanks were rearranged in groups based on their moulting dates to establish groups of same-age individuals. To ensure sexual maturation, adult males and females were given a minimum of 1 week before the experiment began. The average age at testing, measured as the number of days from moulting to adulthood, was 9.76 days for males and 16.84 days for females. We excluded any individuals with visible injuries (e.g. missing legs).

To examine the effects of heat on male reproduction and survival, we exposed sexually mature males for five consecutive days to one of two simulated heatwave conditions (33°C or 39°C) or the control temperature of 28°C; these temperatures were kept constant throughout the treatment exposure. We selected heatwave conditions of 33°C, representing a 5°C increase above the rearing temperature, and 39°C, a biologically relevant heatwave temperature commonly observed in southern European countries, including southern Spain where the laboratory population of crickets originated (Castillo-Mateo et al., 2023). We note that the critical temperature limit (CT_max) for G. bimaculatus is unknown, while the optimal temperatures for development, reproduction and survival range from 27°C to 34°C (Behrens et al., 1983).

Male body mass was first measured to the nearest 0.01 g using a digital scale (Kern & Sohn EW 600-2M GmbH). Males, housed in containers with ad libitum food and water (as described above), were randomly assigned to one of three groups: (1) males exposed to a moderate heatwave at 33°C (N=25), (2) males exposed to a 39°C heatwave (N=25), and (3) control males not exposed to any heatwave and instead maintained in normal temperature conditions at 28°C (N=26). To achieve a temperature of 39°C, we positioned individual containers on top of 20 W heat mats (Lucky reptile Thermo Mat) in a temperature-controlled room set at 33°C, where the other moderate heatwave treatment group was also maintained during heatwave exposure. After the 5 days of heatwave exposure, experimental males were returned to 28°C and used for testing after 24 h of acclimation. Control males and females used in the experiments were kept at a constant temperature of 28°C in a separate temperature-controlled room. We note that, regardless of the heatwave treatment, we maintained stable conditions to ensure negligible temperature variation, stable, constant 65% humidity, and a 14 h:10 h light: dark cycle. Temperature and humidity conditions were verified by placing a data logger (Exo Terra^® monitoring device, Canada) inside testing containers.

On the day following the end of the heatwave exposure, male body mass was measured again and used to calculate the relative change in mass over the heat treatment as the difference in mass measured before and after treatment, divided by the mass before treatment. Each male was randomly assigned to one female. The pair was placed into a small arena (20×30×15 cm) and left to interact freely for a maximum of 20 min (i.e. the assay was interrupted sooner if mating occurred within that time). After an initial approach and antennal assessment, males generally begin to court females, performing a courtship song, and, if successful, males are mounted by the female. During mounting, males transfer a spermatophore (i.e. a protein capsule containing sperm and accessory fluids) to females by genital coupling. We scored the transfer of a spermatophore to the female's genital opening as occurrence of successful mating. As our aim was to obtain fertilised females to measure male reproductive output, males that were unsuccessful in their first mating attempts were used in a second assay. In such cases, we assayed the pair again after approximately 40 min following the end of the first assay. Allowing a 40 min waiting period provides ample time for unsuccessful males to produce a spermatophore. Mating failures often result from the absence of courtship behaviour associated with the lack of a readily produced spermatophore (McMahon et al., 2021).

All mating assays were video recorded using a webcam (Logitech) connected to a laptop. We then manually scored the videos to assess (i) mating success (1=successful, 0=unsuccessful), (ii) the total duration of male courtship song calculated as the sum of all song durations, (iii) female acceptance measured as the latency until first mounting, (iv) the latency to mate measured as the time to spermatophore transfer, and lastly (v) mating effort measured as the number of mounting attempts required for successful spermatophore transfer.

We next estimated fitness as male reproductive output and post-treatment lifespan. To measure male reproductive output, we recorded the number of offspring produced by females mated with experimental males. Immediately after the mating assays, females were placed individually in small tanks (12×14×23 cm) provided with a shelter, water and food, and a plastic cup filled with moist soil for egg laying. After 1 week, we replaced the egg cups with new ones to provide a fresh substrate for egg laying. The first set of cups were individually placed in tanks of similar size, provided with fish food, water and egg cartons. All the tanks were kept in a temperature-controlled room at 28°C and monitored until hatching. After hatching had occurred, we counted the number of nymphs from each cup.

Meanwhile, males were returned to their original individual container at 28°C, supplied with ad libitum water and food, and inspected every alternate day to record the date of their natural death. Post-treatment lifespan was calculated as the number of days between the time the treatment was completed (i.e. approximately 15 days after final moult) and death.

All analyses were performed using R, version 4.1.2 (http://www.R-project.org/). Raw data (Table S1) and script are available as Supplementary Materials and Methods. We used linear regression models for traits with a Gaussian distribution (body mass change over heat treatment), Poisson regression models for count data (brood size, lifespan, song duration, latency to first mounting, latency to mating, and number of mountings), and generalised linear models for traits with a binomial distribution (mating success). To account for overdispersion, we used quasi-models to analyse data on the number of offspring, post-heatwave lifespan, song duration, latency to first mounting and latency to mating. Our main aim was to test for the effects of a heatwave treatment on male reproductive success and survival. Thus, all models included the heatwave treatment (33°C heatwave, 39°C heatwave or control) as a fixed effect. Given that male age varied between 13 and 19 days at the end of the heat treatment, we included age as an additional fixed effect in the model analysing post-treatment lifespan. Our model analysing mating success included number of mating trials (1 or 2) as an additional fixed effect to account for the fact that males that failed to mate during their first assay were presented with a second assay. We used the Anova function of the car package to obtain model outputs (Fox and Weisberg, 2018). Whenever the heat treatment had a significant effect on a trait of interest, we ran pairwise comparisons using a Tukey's test to test for differences among treatment groups.

Three males died during the 33°C treatment and one during the 39°C treatment. The final sample size for the behavioural and reproductive data was 22 males treated at 33°C, 24 males treated at 39°C, and 26 control males. We did not record data on body mass before the temperature treatment for 8 control males, yielding a final sample size for mass change of 22, 24 and 18 males for the 33°C, 39°C and control treatment groups, respectively.

There was a significant effect of heat treatment on body mass change over the 5 day treatment period (F_2,61=6.77, P=0.002; Fig. 1A), indicating that males exposed to a 39°C heatwave gained 1.86% of their initial body weight, which was significantly greater compared with control males, which lost 8.35% of their initial mass (39°C versus 28°C: estimate=0.102, s.e.=0.027, t=3.66, P=0.001). Males exposed to 33°C lost 3.42% of their initial mass, which did not significantly differ from either control males (33°C versus 28°C: estimate=0.049, s.e.=0.028, t=1.73, P=0.201) or 39°C-treated males (39°C versus 33°C: estimate=0.052, s.e.=0.026, t=2.00, P=0.120).

There was no significant effect of heat treatment on male mating success (χ²=0.655, d.f.=2, P=0.720). However, there was a significant negative association between the number of mating assays and mating success (χ²=37.2, d.f.=1, P<0.0001), indicating that males that were not successful in mating during the first assay were also less likely to succeed during the second assay.

We found that heat treatment significantly affected the duration of male courtship song (χ²=10.1, d.f.=2, P=0.006; Fig. 1B), with males treated at 39°C singing 96 s longer than control males (39°C versus 28°C: estimate=0.868, s.e.=0.327, z=2.66, P=0.021) and 94 s longer than 33°C-treated males (39°C versus 33°C: estimate=0.845, s.e.=0.353, z=2.39, P=0.043).

There was, however, no effect of heat treatment on the latency to the first mounting attempt (χ²=4.46, d.f.=2, P=0.107), the latency to successful mating (i.e. with spermatophore transfer; χ²=5.91, d.f.=2, P=0.052), or the total number of mountings during the mating assay (χ²=0.176, d.f.=2, P=0.915). The effect of heat treatment on the latency to successful mating, although only marginally significant, was relatively large as it took on average 46.8% and 84.6% more time for 33°C- and 39°C-treated males, respectively, to initiate mating relative to control males.

There was a significant effect of heat treatment on the number of offspring produced by experimental males (χ²=12.1, d.f.=2, P=0.002; Fig. 2A), indicating that males exposed to a 39°C heatwave produced 69% fewer offspring relative to males exposed to 33°C (39°C versus 33°C: estimate=−1.17, s.e.=0.366, z=−3.202, P=0.003) but a similar number of offspring to control males (39°C versus 28°C: estimate=−0.806, s.e.=0.374, z=−2.15, P=0.077). Likewise, 33°C-treated males produced a similar number of offspring to control males (33°C versus 28°C: estimate=0.367, s.e.=0.263, z=1.39, P=0.340).

There was no effect of heat treatment on male post-heatwave lifespan (χ²=2.17, d.f.=2, P=0.337; Fig. 2B). There was also no effect of age at the time of treatment on survival following the heat treatment (χ²=0.044, d.f.=1, P=0.833).

Through experimental simulation of two ecologically relevant heatwave conditions, our study revealed that male G. bimaculatus exposed to the most severe heatwave at 39°C invested more time in courtship singing to achieve successful mating. However, they experienced reduced fertility, resulting in the siring of the fewest offspring. These individuals also gained relatively more weight following heatwave exposure, suggesting a potential shift in resource allocation towards somatic maintenance at the expense of reproduction. In contrast, a milder heatwave (33°C) did not lead to changes in any of the measured traits. Interestingly, we found no differences in lifetime survival across treatments, which highlights that heatwaves have the potential to affect population growth and persistence through non-lethal effects that negatively impact reproduction.

We found that heatwaves can greatly alter male mating behaviour: males that underwent the most severe heat treatment performed longer courtship singing. On the one hand, this result may indicate that heat induces an increase in signing, potentially by elevating metabolic rate and activity levels (Schulte, 2015). Such effects could result from cascading systemic changes due to greater energy availability with higher temperatures (i.e. from biochemical to whole-organism behavioural processes; Abram et al., 2017). Male crickets sing by scraping their forewings against each other in repeated pulses of sounds grouped in chirps (Koch et al., 1988). Greater temperatures enable males to increase the speed of wing opening and closure, which, ultimately, leads to a linear increase in pulse rates up to optimal temperatures (Doherty, 1985; Martin et al., 2000; Pires and Hoy, 1992). On the other hand, our results may reflect plastic responses of males seeking to increase investment in courtship in response to more intense female choosiness. Males that experienced the most severe heatwave appeared to be less-preferred mating partners given that they needed on average 84.6% longer to initiate mating when compared with control males (a result that was only statistically marginally significant). These males may ultimately signal low quality to prospective partners, if heat stress alters chemical profiles used to identify potential mates such as cuticular hydrocarbons (CHCs; Groot and Zizzari, 2019) or specific song parameters (Martin et al., 2000). Both male CHCs and song structure are in fact known to affect female mate choice in field crickets (Iwasaki and Katagiri, 2008; Tregenza and Wedell, 1997; Rantala and Kortet, 2003; Verburgt et al., 2011). Regardless of the specific reason why male field crickets have to court females for longer to achieve successful mating, our study establishes that heatwaves can alter male mating success by altering male–female behavioural interactions.

Although males exposed to a 39°C heatwave spent more time courting females, they sired fewer offspring relative to control or 33°C-treated males. These results are in line with our prediction that heatwaves should reduce male mating success and reproductive output. These fertility costs caused by heat stress resemble those documented in other arthropods, such as flour beetles (Sales et al., 2018, 2021), burying beetles (Pilakouta et al., 2023) and tidepool copepods (Siegle et al., 2022). Hyperthermia, the elevation of the body temperature above its normal range, can affect fertilisation ability in males by damaging sperm as the result of, for example, a reduction in sperm cell motility or number, or an increase in the frequency of cell abnormalities (Hurley et al., 2018; Rahman et al., 2018; Rao et al., 2016). Detrimental effects of heat stress on sperm performance and morphology have been reported in various insect species (e.g. Chevrier et al., 2019; Martinet et al., 2021; Sales et al., 2021), including our study system G. bimaculatus. In this species, males kept at higher temperature for a maximum of 20 consecutive days during adulthood produced less sperm (i.e. fewer sperm cells in a spermatophore) and sperm of lower quality (i.e. lower cell viability) compared with males kept at 24°C (Gasparini et al., 2018). In contrast, when such temperature treatment was applied throughout development, there was no effect on sperm number or quality at adulthood (Gasparini et al., 2018). These findings suggest that elevated temperatures are most detrimental to fully mature gametes when exposure is relatively short.

The reduced fertility of thermally stressed males could alternatively be caused by a failure in completing sperm transfer. Sperm and accessory fluids of field crickets are packed in a hardened protein capsule, the spermatophore, which is transferred to the female upon genital coupling during female mounting. Sperm are then ejected from the spermatophore by osmotic pressure through the evacuating fluid that is released inside the female reproductive tract (i.e. the spermathecal duct; Khalifa, 1949), whereby heat may affect transfer by, for example, altering the osmotic pressure. The detrimental effects of heat on male fertility may also be driven by the female through mate choice and control of fertilisation. Once released, sperm migrate to the female sperm storage organ (i.e. the spermatheca) and females can actively facilitate or inhibit sperm storage (Bretman et al., 2009; Hall et al., 2010; Tuni et al., 2013). The control of sperm storage is indeed considered one of the major mechanisms for females to exert post-copulatory mate choice, favouring sperm of preferred males (Bretman et al., 2009; Tuni et al., 2013). For example, females of the cricket Teleogryllus oceanicus modulate sperm storage based on pre-copulatory chemicals and their assessment of the behaviour of males (Tuni et al., 2013). If, as suggested by our findings on mating latency, heat-stressed males are the least attractive to females, selective sperm storage and the resulting higher siring success of control males should reflect female mate preference. Reduced sexual attractiveness of heat-stressed males was found in the springtail Orchesella cincta, where females are more likely to selectively choose the spermatophores deposited by control males compared with those deposited by heat-treated males (Zizzari and Ellers, 2011). Overall, uncovering the mechanistic explanations, either male or female driven, for the reduced number of offspring of males that experienced heat stress would provide an important step in elucidating the consequences of heatwaves on insect fertility.

Although males that experienced a 39°C heatwave produced fewer offspring, contrary to our expectation the heatwave treatments had no effect on male lifespan. Hence, elevated temperatures can have detrimental effects on male fitness without increasing mortality. These results are in line with the theoretical predictions that heatwaves differentially impact reproduction and survival. The threshold temperature at which heat begins to harm the reproductive function (i.e. the thermal fertility limit; Iossa, 2019; Walsh et al., 2019) might often be lower than the threshold temperature at which heat becomes lethal (i.e. the critical thermal limit; Angilletta et al., 2002; Overgaard et al., 2012). Thus, with increasing temperature, male crickets may first suffer a reduction in fertility before heat has any detrimental effects on survival. An alternative explanation is that males facing heat stress may shift their investment towards somatic maintenance at the expense of reproduction. There is a well-documented trade-off between survival and reproduction (Reznick, 1992; Roff, 2002; Stearns, 1992; Williams, 1966), and adverse conditions, such as extreme temperatures, often lead animals to favour survival and future reproduction at the expense of current fertility (e.g. Marshall and Sinclair, 2010). Flexible resource allocation can enable organisms to cope with unfavourable conditions by shifting their investment towards more essential functions, such as maintaining the physiological balance and the integrity of the soma. There is good evidence that responses to environmental stress might often come at a cost in terms of reduced reproduction, as documented by research on Drosophila flies (e.g. Chippindale et al., 1993; Salmon et al., 2001; Wang et al., 2001). We encourage future studies to examine the link between current fertility and survival, and how environmental stressors such as heatwaves can alter this link.

Our final key result indicating that males exposed to a 39°C heatwave increased body mass suggests that crickets facing heat stress may prioritise survival over current reproduction by reallocating resources towards somatic maintenance. We anticipated that males exposed to a heatwave would gain more mass, assuming that recovering from heat damage incurs a cost in terms of heightened energy expenditure and resource use. Gains in body mass following a heatwave have commonly been documented in insects and, although they may be due to increasing digestion efficiency with increasing temperatures, they generally reflect an increased food consumption at higher temperatures (Chown and Nicolson, 2004). This is the case in field crickets, where food consumption increases with increasing temperature over a range spanning 20°C to 38°C (Hoffmann, 1974) and can be associated with a gain in body mass (Adamo and Lovett, 2011). Taken together, these findings and the fact that our heatwave treatment did not affect survival may suggest that males under heat stress prioritise homeostasis maintenance by acquiring additional food resources that can be allocated to the soma. Acquisition of energetic resources may be essential for recovering from heat damage. For example, growth rate in damselflies increases following exposure to heatwaves, presumably as a result of greater food consumption and energy storage in the form of fat content (Van Dievel et al., 2017). Thus, increasing food consumption might be a common response in insects to cope with the adverse effects of heatwaves.

With the ever-increasing threat of climate change leading to intense heatwaves, there is a pressing need to understand organisms' physiological and behavioural responses to heat stress. Our study can be added to the growing list of results showcasing the consequences of heatwaves on reproductive success in animals. The findings stemming from our study reveal a crucial aspect: despite an increase in mass gain and no impact on survival, heatwaves can have detrimental effects on male reproduction. Through their negative effects on fitness, non-lethal temperatures can have potential overarching effects on population growth, jeopardising population persistence. There are many additional interesting aspects that warrant further advances in our understanding of how individuals and populations cope with thermal heat stress, including understanding sex-specific differences in the sensitivity of fertility to temperature variation (Iossa, 2019), life stage-dependent consequences of heatwaves (Pilakouta et al., 2023; Sales et al., 2021), and the plausible interaction effects of heatwaves and other environmental stressors (Adamo et al., 2012). Thermal stress is likely to be one of the biggest drivers of biodiversity loss (Smale et al., 2019; Vasseur et al., 2014). Hence, understanding how extreme temperatures affect organisms will be crucial in mediating and managing biodiversity loss in the face of global changes.

We thank Magdalena Matzke for collecting animals and Pim Edelaar (Universidad Pablo de Olavide, Seville, Spain) for providing the necessary support for animal collection in Spain, Yvonne Cammerer for assistance in animal maintenance and Niels Dingemanse (Ludwig Maximilians University of Munich, Germany) for providing logistics.