Stress tolerance is influenced by artificial light at night during development and life-history strategy

Artificial light at night (ALAN), also known as ecological light pollution, is increasingly prevalent in terrestrial environments. Nearly one-quarter of land surfaces worldwide are exposed to ALAN (Falchi et al., 2016), and global ALAN levels have increased >2% each year (Kyba et al., 2017). Exposure to ALAN impacts several features of animal biology, including physiology, phenology, life-history traits, and daily activity (Fyie et al., 2021; reviewed in Sanders et al., 2021). Although ALAN tends to have negative effects, animals may shift their life histories to mitigate costs because life-history strategy can influence their stress responses (Sanders et al., 2021; Stark et al., 2004; Schultner et al., 2013; Padda et al., 2021; but see Tüzün et al., 2020). Furthermore, different strategies may exhibit different sensitivities to ALAN. In which case, ALAN may favor one strategy over another. However, it is unknown if the expression of life-history strategies is affected by ALAN exposure (e.g. whether ALAN increases use of a given strategy) or if ALAN has strategy-specific effects on fitness-related traits (e.g. whether a given strategy proves more costly in ALAN).

In addition to ALAN, stressors associated with climate change (e.g. heat and drought) are on the rise in many global regions (Sarhadi et al., 2018; IPCC, 2021). Although the effects of ALAN on the stress response have been examined (e.g. McLay et al., 2018; Grunst et al., 2020; Miner et al., 2021; Czarnecka et al., 2022), its role in heat or desiccation tolerance is unknown. Field crickets (Gryllus spp.) offer unique insight into the relationships among ALAN, life-history strategies and stress tolerance. Gryllus exhibit a wing dimorphism that mediates two distinct life-history strategies during early adulthood: flight-capable females have low reproductive investment while flight-incapable females have high reproductive investment (Roff, 1984; Zera and Mole, 1994; Zera, 2005; Stahlschmidt, 2022). This wing dimorphism in Gryllus is a genetic polymorphism and an environmental polyphenism (Zera, 2005). It is genetically determined and responsive to selection (reviewed in Roff and Fairbairn, 2001; Roff et al., 2002; Zera, 2005) but it can also be influenced by environmental factors (Roff and Fairbairn, 2007; Glass and Stahlschmidt, 2019; reviewed in Wu et al., 2014). In other crickets, a similar wing dimorphism is altered by light pulses in the scotophase (night-time) during development (Masaki and Yasuhiko, 1989) and exposure to ALAN affects mating behavior (Botha et al., 2017; Levy et al., 2021). The variable field cricket (Gryllus lineaticeps) has a geographic range encompassing areas where ALAN, heat and drought conditions are increasingly prevalent (Bennie et al., 2015; Sarhadi et al., 2018; Weissman and Gray, 2019; IPCC, 2021). Thus, we manipulated ALAN exposure during development in G. lineaticeps to test three hypotheses. (1) ALAN exposure is costly to adult phenotype. This hypothesis is based on the tendency for ALAN to impact the daily activity patterns and life-history traits of many animals, including crickets (Levy et al., 2021; reviewed in Sanders et al., 2021). Further, ALAN reduces melatonin levels in crickets (Jones et al., 2015) and melatonin is important for stress tolerance in insects (Subala et al., 2017; Fan et al., 2021). Therefore, we predict that developmental ALAN exposure is costly and thereby increases fluctuating asymmetry (a stress indicator: reviewed in Graham et al., 2010), and reduces body size, reproductive investment and stress (heat and desiccation) tolerance. (2) ALAN exposure promotes expression of the flight-capable life-history strategy, which would allow crickets to disperse to less stressful environments. Gryllus maintain or increase flight capacity at the expense of survival and/or reproduction during several types of stress (food: Glass and Stahlschmidt, 2019; heat and water: Padda et al., 2021; food and water: Padda and Stahlschmidt, 2022). Thus, we predict a greater frequency of flight-capable crickets in the ALAN-exposed treatment group. (3) However, ALAN exposure is more costly to individuals expressing the flight-capable life-history strategy because flight-capable Gryllus exhibit greater sensitivities to other stressors (e.g. oxidative and immune challenges: Stahlschmidt et al., 2020; our unpublished results), potentially due to greater energetic requirements of the flight-capable strategy (Zera and Mole, 1994; Zera et al., 1997; Nespolo et al., 2008; Sun et al., 2020; Stahlschmidt, 2022). Here, we predict interactive effects between light treatment and flight morphology whereby ALAN exposure will have disproportionate costs to body size, reproductive investment and stress tolerance for flight-capable crickets.

Our study will be the first to examine the interconnectedness of the responses of animals to three increasingly prevalent features of environmental change: heat, drought and light pollution. Specifically, we will test whether ALAN influences sensitivity to climate-related stressors and determine the role of life-history strategy in mitigating the effects of ALAN.

Gryllus lineaticeps Stål 1861 has a geographic range that extends from southwestern Oregon, through California, and into Baja California, Mexico (Weissman and Gray, 2019). Like several other Gryllus spp., G. lineaticeps expresses two different wing morphs, short-winged (SW) and long-winged (LW) (Weissman and Gray, 2019). At the onset of adulthood, flight-incapable SW females invest heavily into fecundity (Zera, 2005). During early adulthood, LW crickets can be categorized into the LW(h) (histolyzed, non-functional dorso-longitudinal flight muscle (DLM)] and LW(f) (functional DLM) morphs. Thus, the morphs can be categorized into two different life-history strategies, flight-incapable SW or LW(h) and flight-capable LW(f) morphs, during early adulthood. The LW(h) morph in Gryllus has been found to be more physiologically similar to the SW morph than to the LW(f) morph (Zera et al., 1997; Zera and Larsen, 2001; Stahlschmidt, 2022). Yet, LW (both h and f morphs) and SW Gryllus generally differ in adult body mass, DLM status, rates of development and growth, and in their sensitivities to oxidative and desiccation challenges (Glass and Stahlschmidt, 2019; Stahlschmidt et al., 2020; Stahlschmidt and Glass, 2020; Padda and Stahlschmidt, 2022). Therefore, the LW morph can be used as a less precise proxy for flight capacity if DLM status is unknown.

Crickets used in this study were from a long-term colony sourced from the progeny of individuals from a natural population at Sedgwick Reserve, Santa Ynez, CA, USA. Every 1–2 years, the colony was interbred with offspring of females captured from Sedgwick Reserve. Breeding adults in the colony were maintained at an even morph (SW:LW) and sex ratio. Cohorts of 20 newly hatched crickets (<2 days post-hatching) were reared at 28°C in 15 liter translucent plastic bins. Each bin was covered with a clear sheet of acrylic plastic and contained ad libitum access to food (dry cat food) and water. Crickets spend a significant amount of time under shelter because predation risk strongly influences their decision-making (even in the absence of predators: Hedrick and Dill, 1993; Hedrick, 2000; Stahlschmidt and Adamo, 2013; Stahlschmidt et al., 2014; Atwell and Wagner, 2015). Thus, to mimic natural conditions, each bin also contained shelters (two egg cartons) that could be accessed freely throughout development

Crickets experienced 14 h of simulated sunlight (5300 Kelvin, a similar color temperature or warmth as natural sunlight; 1–500 lx, depending on shelter use) each day and one of two different night light treatments: ALAN or control. Crickets in the ALAN group (n=18 cohorts; n=360 total crickets) were exposed to 10 h of ALAN (0.3–40 lx, depending on shelter use) that was warmer than day light (2750 Kelvin) to simulate the color temperature of street lighting recommended by the American Medical Association (Kraus, 2016). Crickets in the dark night (control) group (n=18 cohorts; n=360 total crickets) were exposed to darkness at night. The illuminance and color temperature of light bulbs were measured using a light meter (range: 0.15–250,000 lx; accuracy: ±3%; Lumu Power 2, LUMULABS, Ljubljana, Slovenia). Night light in our study was warmer (lower Kelvin) and dimmer (fewer lx) than daylight, and light was dramatically limited inside the shelters that crickets regularly accessed throughout the study. Thus, although the night light source represented a relatively high level of ALAN (40 lx), our ALAN-treated animals were likely exposed to considerably lower levels of light thereby blunting the potential effects of ALAN. However, including shelter represents a more ecologically relevant exposure to light than a design in which crickets are forced to continuously experience ambient light conditions (i.e. 500 lx during the photophase for all crickets and 40 lx during the scotophase for ALAN-treated crickets).

Within 1 day of reaching adulthood, each cricket was housed individually in a small translucent deli cup (473 ml) containing shelter (overturned 30 ml opaque container with access holes) and ad libitum access to food and water. All crickets were weighed after 5 days, which approximates when the trade-off between investment into flight versus reproduction peaks in Gryllus (Gryllus firmus: Zera, 2005; G. lineaticeps: Z.S., unpublished results). After being weighed, males were euthanized (see below) and females were typically tested for desiccation or heat tolerance. To estimate desiccation tolerance at 28°C, the water source was removed from each cricket's container. Each cricket was then checked daily until death, and the time to death (days) was used as an estimate of desiccation tolerance. Although only LW individuals are capable of flight, all crickets exhibit histolyzed flight muscle after approx. 7 days, which prevented post-mortem determination of flight morphology for desiccation trials.

There are two general experimental protocols to assess heat tolerance: static (duration to knockdown upon exposure to a constant, high temperature) and dynamic (temperature at which knockdown occurs after exposure to a constantly ramping temperature) (reviewed in Jørgensen et al., 2021). These two protocols typically yield consistent results (i.e. a positive correlation between static and dynamic estimates of heat tolerance), but values from the static protocol can be more heritable (Mitchell and Hoffmann, 2010; Jørgensen et al., 2019). Therefore, heat knockdown trials were performed at a field-relevant temperature (45±1°C: Sun et al., 2020; Stahlschmidt et al., 2022a). Each cricket was acclimated in a 50 ml centrifuge tube with access to water (cotton ball soaked with 10 ml of water) for 1 h at room temperature. Then, each cricket was transferred to a 473 ml glass jar submerged in 45°C water bath between 16:00 h and 16:30 h (i.e. 10–10.5 h into the 14 h photophase). Every minute, each cricket was flipped over, and it was considered knocked down when it could no longer right itself. The time to knockdown (min) was used as an estimate of heat tolerance, after which each female was euthanized (see below)

Females not assessed for desiccation tolerance and males were euthanized and stored at −20°C. After storage, the length of hind femurs from each cricket were measured and mean femur length was used as an index of body size (Nguyen and Stahlschmidt, 2019). Functional asymmetry, a stress indicator (Graham et al., 2010), was assessed as the coefficient of variation (mean/s.d.) of femur length for each cricket. To assess reproductive and flight investment, flight muscle score and ovary dry mass were determined. Here, DLMs were scored as 0 (absence of DLM), 1 (histolyzed, non-functional and white DLM) and 2 (functional, pink DLM) (Crnokrak and Roff, 2002; King et al., 2011; Glass and Stahlschmidt, 2019). Therefore, crickets were determined to be SW (score of 0 or 1), LW(h) (score of 0 or 1), or LW(f) (score of 2). To estimate reproductive investment, gonads from each cricket were removed, dried to a constant mass at 55°C and weighed to determine dry mass. Ovary mass is proportional to fecundity (egg number) in Gryllus (G. firmus: Roff, 1994; G. lineaticeps: Z.S., unpublished results). During dissections, mating status (i.e. presence of sperm-filled spermatheca) was assessed, and all females were ‘not mated’.

Data were tested for normality, natural logarithm-transformed when necessary and analyzed using SPSS (v.26 IBM Corp., Armonk, NY). Two-tailed significance was determined at α=0.05. Several general linear mixed models were used to examine the independent and interactive effects of light treatment (control and ALAN) and flight morphology [i.e. SW, LW(h) and LW(f)] or wing morphology (i.e. SW vs. LW) on developmental duration, heat tolerance, desiccation tolerance, adult body mass and size, functional asymmetry and gonad mass. Sex was included as a fixed effect when applicable (e.g. only females were assessed for stress tolerance) and cohort identity was included as a random effect in each model. For each model, all interactions between and among fixed effects (e.g. light treatment, sex and flight morphology) were initially included; for example, to determine and account for whether females (that invest more heavily into reproductive tissue) were more sensitive to light treatment relative to males. Sex independently influenced body and gonad mass (i.e. strong female-biased sexual dimorphisms), and sex also interacted with other factors (e.g. light treatment and flight morphology) to influence body and gonad mass. Thus, sexes were analyzed separately for body and gonad mass to better understand the effects of ALAN and life-history strategy on these two traits.

A general linear model was performed on data from each cohort to determine the effects of treatment on survivorship (i.e. proportion of crickets surviving to adulthood in each cohort). An ordinal logistic generalized linear mixed model was performed on flight morphology. To determine effect size, eta squared (η²) or partial eta squared (η²_p) were determined for each fixed effect and effect-effect interaction, and effect sizes were considered small, medium, or large at η² (or η²_p)=0.01, 0.06, and 0.14, respectively (Cohen, 1988). Significant results are reported below and all results are shown in Table 1. Pairwise post hoc analyses were used to determine differences between the three flight morphs and to interpret interactive effects, and the Type I error rate associated with multiple comparisons was controlled by using the Holm–Bonferroni method.

Survival, functional asymmetry and flight morphology were not affected by any factor (Fig. S1A–C; Table 1). Females developed significantly faster than males (mean±s.e.m.: 51±1.0 days versus 54.2±0.9 days, respectively), and flight-capable crickets [LW(f)] developed faster than either the SW or the LW(h) morph (Fig. S1D; Table 1).

Crickets exposed to ALAN generally had greater body size, and females had greater body mass and reproductive investment (Fig. 1 and Fig. S1F; Table 1). Gryllus exhibit sexual size dimorphism (Stahlschmidt and Chang, 2021; Stahlschmidt et al., 2022a), and females in our study were heavier than males and exhibited greater reproductive investment (Fig. 1E,F). Flight-incapable females were heavier than flight-capable females (mean±s.e.m: 788.8±19.4 mg versus 657.1±22.8 mg, respectively) and they had greater ovary mass (23.1±2.2 mg versus 68.8±1.6 mg) (Fig. 1B and Fig. S1f; Table 1). Body size (both sexes) and body mass (females only) were greater in LW crickets relative to SW crickets (Fig. 1A and Fig. S1f; Table 1).

Stress tolerance was only determined in females, and flight-incapable and ALAN-exposed crickets had reduced heat tolerance (Fig. 2A; Table 1). We detected only one significant interactive effect of flight capacity and light treatment, specifically, desiccation tolerance was greatest in SW crickets that were exposed to ALAN, and crickets in all other treatment combinations exhibited similar desiccation tolerances (Fig. 2B; Table 1). The effect of flight morphology on desiccation tolerance was not determined because Gryllus do not exhibit functional flight muscle after early adulthood.

Across animal taxa, ALAN tends to entail an array of costs, from reduced cognition and offspring number to increased predation (reviewed in Sanders et al., 2021). However, key features of development (developmental duration and success, and bilateral symmetry) were not affected by ALAN exposure in our study (Fig. S1A–C), and ALAN exposure actually benefited other fitness-related traits (e.g. increased body size in both sexes and greater body mass in females: Fig. 1A,F) in contradiction to our first hypothesis that ALAN exposure is costly to adult phenotype. Developmental stress also has positive effects on fitness-related traits in crickets and other taxa (Durrant et al., 2018; Crino and Breuner, 2015) and it can promote heat tolerance in insects (Willot et al., 2021). As it does for other traits (e.g. reviewed in Sanders et al., 2021), ALAN had a relatively modest effect on body size in our study (i.e. η²_p=0.051, indicative of small-to-medium effect sizes: Cohen, 1988). Yet, ALAN exposure during development had greater effects on reproductive investment in females (i.e. η²_p=0.17, indicative of a large effect size: Fig. 1B; Table 1), which is likely due, in part, to larger body size (reviewed in Peters, 1983). ALAN-related increases in body size (Fig. 1A) and ovary mass may be due to a longer daily feeding period (sensu changes in activity patterns exhibited by other animals: Sanders et al., 2021). In another cricket (Acheta domesticus), sub-adults and adults exhibit circadian patterns of feeding (Nowosielski and Patton, 1963). However, investigation is required to determine whether developmental ALAN alters such patterns in Gryllus. Exposure to ALAN does not generally promote body size across animals (Sanders et al., 2021), potentially because of opposing pressures on body size by other aspects of environmental change (e.g. warming: Gardner et al., 2011; Fryxell et al., 2020; but see Siepielski et al., 2019). However, ALAN exposure may offset or even outweigh reductions in body size caused by warming. In addition to food subsidization and non-native species, urban environments are characterized by both warmer temperatures and ALAN exposure and urbanization promotes body size in several animal taxa, including insects (McKinney, 2002; Wilby, 2008; Shochat et al., 2010; Penick et al., 2015; Beasley et al., 2018; Merckx et al., 2018; Putman and Tippie, 2020; IPCC, 2021).

Developmental ALAN exposure and flight (or wing) morphology had substantial effects on stress tolerance (demonstrated by their medium effect sizes: Fig. 2; Table 1). For example, ALAN exposure reduced heat tolerance by approx. 25% and SW crickets exhibited approx. 38% greater desiccation tolerance (i.e. could survive 3 days longer) than LW crickets after ALAN exposure (Fig. 2). In the field, G. lineaticeps adults in the founding population emerge during a July breeding season, which is characterized by hot and dry microclimates (e.g. 40°C or higher with 0.25 mm of monthly precipitation on average) (Sun et al., 2020; Stahlschmidt et al., 2022a; US Climate Data 2022). Therefore, variation in heat and desiccation tolerance is likely to be highly ecologically relevant. That said, future work is required to fully understand the implications of the effect sizes on stress tolerance observed in our study.

We offer three physiological explanations for our results related to stress tolerance. First, in support of our first hypothesis (ALAN exposure is costly to adult phenotype), ALAN-exposed females had reduced heat tolerance (Fig. 2A) because ALAN reduces melatonin levels in other crickets (Jones et al., 2015). Melatonin supplementation increases stress tolerance in insects (Subala et al., 2017; Fan et al., 2021), and it improves heat tolerance in vertebrates (Sahin et al., 2004; Abbas et al., 2007; Sharma et al., 2013). Thus, an ALAN-induced reduction in melatonin levels may have decreased heat tolerance in our study. Second, serotonin is the precursor to melatonin (Vanecek, 1998), and it regulates circadian rhythms in crickets (Saifullah and Tomioka, 2002) and increases water excretion in other insects (Mykles et al., 2010). If developmental ALAN exposure downregulates the serotonin–melatonin system, then water stores may be conserved thereby promoting desiccation tolerance (in contradiction to our first hypothesis). Third, ALAN only benefited the desiccation tolerance of flight-incapable crickets (i.e. SW crickets, because most LW crickets were likely flight-capable: Fig. 2B and Fig. S1C) because flight capacity influences circadian rhythms. For example, flight-capable Gryllus females tend to exhibit larger daily fluctuations (i.e. greater circadian rhythmicity) in juvenile hormone levels, global gene expression patterns and metabolic rate (Zhao and Zera, 2004; Zera et al., 2018; Stahlschmidt, 2022). Thus, serotonin–melatonin systems in flight-capable crickets may be less sensitive to ALAN-related perturbations, meaning that only flight-incapable crickets receive the desiccation tolerance benefits of ALAN-induced reductions in serotonin. Melatonin is an ancient antioxidant that is important to circadian rhythms and stress tolerance for plants and animals (reviewed in Zhao et al., 2019); therefore, future efforts should explore its role in mediating the effects of ALAN on stress tolerance.

Variation in life-history strategies allows animals to adaptively respond to their environments (Ricklefs and Wikelski, 2002). Yet, the expression of life-history strategy (investment into flight versus reproduction) was not developmentally plastic in response to ALAN (Fig. S1C), contradicting our second hypothesis that ALAN exposure will promote the expression of the flight-capable life-history strategy. This hypothesis characterized ALAN as a stressor that would favor the dispersal phenotype (i.e. flight-capable and/or LW morph). In contrast, ALAN appeared to be beneficial as it promoted adult body size and ovary mass (Fig. 1), in agreement with recent work demonstrating positive effects of short-term ALAN exposure on mating behavior in G. lineaticeps (Stahlschmidt et al., 2022b). Likewise, our third hypothesis (ALAN exposure will prove more costly to individuals expressing the flight-capable life-history strategy) was not supported. Life-history strategies in G. lineaticeps did influence stress tolerance, which has been shown in other taxa. For example, ‘fast-living’ organisms, characterized by early reproduction, can be less susceptible to stress relative to their ‘slow-living’ counterparts characterized by delayed reproduction (Stark et al., 2004; Palacios et al., 2012; Schultner et al., 2013; Padda et al., 2021; but see Tüzün et al., 2020). In agreement, ‘fast-living’ G. lineaticeps (i.e. SW females with greater reproductive investment during early adulthood: Fig. 1B) had greater desiccation tolerance when exposed to ALAN and greater heat tolerance in both light treatments relative to ‘slow-living’, flight-capable females that delay reproductive investment (Fig. 2). Similarly, SW or flight-incapable G. firmus are less sensitive to water limitation and oxidative challenge relative to LW crickets (Stahlschmidt et al., 2020; Padda and Stahlschmidt, 2022). Therefore, continued exposure to ALAN and climate-related stressors (e.g. heat and drought: Sarhadi et al., 2018; IPCC, 2021) may disadvantage the flight-capable morph. Accordingly, Gryllus may offer key insight into the dynamic between life-history strategy and stress tolerance in the context of environmental change.

In Gryllus, ALAN may disrupt life-history strategy-specific variation in circadian physiology. Flight-capable crickets exhibit greater circadian rhythmicity in hormone levels, patterns of gene expression and whole-animal metabolism (Zhao and Zera, 2004; Zera et al., 2018; Stahlschmidt, 2022). Therefore, future work should continue to characterize circadian biology in Gryllus, as well as the effects of ALAN on overall and strategy-specific rhythmicity. For example, ALAN may blunt circadian patterns of physiology and behavior, particularly for individuals expressing the flight-capable life-history strategy. Currently, the connections among ALAN, circadian biology, life-history strategy and stress tolerance are unknown. Furthermore, ALAN may impose other ecological costs (e.g. increased predation: reviewed in Sanders et al., 2021) that may differentially affect particular life-history strategies. For example, temperature shifts have divergent effects on the behavior of the two G. lineaticeps wing morphs: heat waves increase boldness in SW crickets and reduce it in LW crickets (Padda et al., 2021). Because ALAN may be an evolutionary trap that is driving global declines of insects (reviewed in Hopkins et al., 2018; Owens et al., 2020), our study represents an important first step in understanding the role of ALAN in shaping two of an animal's most important tools to address ongoing global change: life-history strategy and stress tolerance.

We thank Jacob Whitlock for animal care.