Introduction

The seasonal timing of a wide range of species is shifting in response to climate change (Parmesan and Yohe 2003; Root et al. 2003; Thackeray et al. 2010), largely in response to increasing temperatures (Cohen et al. 2018). In many cases, interacting species are shifting their phenology at different rates, with the resulting phenological mismatches between consumer and resource leading to natural selection on phenology and possibly negative consequences for population viability (Kharouba et al. 2018; Visser and Gienapp 2019). The relative importance of different environmental cues, such as temperature and photoperiod, to time key life history events is generally thought to play a role in determining a species’ seasonal timing shift (Chmura et al. 2019; Renner and Zohner 2018). However, to accurately predict species’ responses to climate change, we need to gain a detailed mechanistic understanding of how different environmental cues interact to produce the seasonal timing response (Chmura et al. 2019; McNamara et al. 2011).

In many species, the seasonal timing of life history events is under photoperiodic control, with additional environmental cues such as temperature used to fine-tune the response (Bradshaw and Holzapfel 2007). In insects, photoperiod similarly plays a major role in determining the timing of development, particularly in regulating dormancy responses (Danks 1987; Denlinger 2002). While photoperiod is primarily involved in the induction of diapause in insects (Denlinger 2002), it can similarly act as a cue for diapause maintenance (Tauber and Tauber 1976) and diapause termination (Brunnarius and Dumortier 1984; Koštál et al. 2017). For example, in the European corn borer (Ostrinia nubilalis) and Asian corn borer (Ostrinia furnacalis), larvae that have entered diapause remain sensitive to photoperiod throughout autumn and early winter, with day length in combination with temperature regulating the duration of diapause (McLeod and Beck 1963; Yang et al. 1950; van Dis et al. 2021). Eggs become transparent only very close to hatching, which is when they might start responding to photoperiod. Indeed, winter moth egg hatching has previously been observed to follow a circadian rhythm (Embree 1970) similar to other insects (Saunders 2002) and responding to photoperiod only at the end of egg development might also explain the small 1.4–2.5 days delay we observed for some of the delayed photoperiod treatments.

Compared to the overriding eight-fold larger temperature effect on egg development time (for arguably a smaller treatment difference of on average 1.36 °C between warm and cold treatments compared to a day length difference of 2 h in mid-March between photoperiod treatments), we would argue that photoperiod received at the egg stage does not play a substantial role in regulating the seasonal timing of egg hatching. In fact, the small photoperiod effects we found were outweighed by both between- and within-clutch variation in egg development time (Fig.S3). Nevertheless, photoperiod received at a different life stage might still play a role, as previous work on the winter moth has found indications of maternal effects of photoperiod (Salis et al. 2017). The first stage of insect embryogenesis critically depends on the maternally set-up environment in the egg (Irvine 2020) and dormancy responses are often maternally regulated in insects (Mousseau and Dingle 1991). For example, egg diapause is maternally induced in the silk moth, B. mori (Kogure 1933). In the winter moth, previous work indicated that photoperiod received by the mother can carry over into the next generation: mothers that received an early season photoperiod as caterpillars laid eggs that took less time to develop and vice versa, with an effect size of up to 10 days depending on the temperature treatment the eggs received (Salis et al. 2017). This trans-generational photoperiod response might be linked to the nutritional status of the mother (i.e. whether she was timed well to budburst as a caterpillar (van Asch et al. 2010), but the causal mechanism behind this maternal effect remains unclear.

Because climate change affects ambient temperature but not photoperiod, the relative importance of temperature and photoperiod as cues has important implications for climate change adaptation. Importantly, phenological traits mostly regulated by temperature are expected to immediately shift under climate change and such temperature-only controls of phenology might be common in moth species (e.g. in at least 34% of 112 analysed Finnish species, (Valtonen et al. 2011). While development rate is temperature dependent in all insect species (Nedved 2009), it might be that insects with obligate diapause (i.e. where diapause does not need to be induced by environmental cues) are more likely to have temperature-only controls of phenology, while species with facultative diapause often also rely on photoperiod as a cue to regulate dormancy (Denlinger 2002). Indeed, photoperiod regulation of diapause induction, maintenance, and termination has mostly been reported for facultative diapausers (e.g. Brunnarius and Dumortier 1984; Wang et al. 2009; Yang et al. 2014), while studies in obligate diapausers tend to focus on the effect of temperature only (e.g. Doherty et al. 2018; Gray et al. 2001). But to properly test this pattern, more experiments investigating the effect of photoperiod on the phenology of obligate diapausers are needed. In addition, it is important to identify where in the life cycle climate change-induced selection acts in order to understand which environmental cues are important for adaptation. This importance is illustrated by the few examples we have of insects evolving under climate change (Merilä and Hendry 2014): so far, genetic changes to the photoperiodic response were involved in pre-dormancy adaptations – e.g. the pitcher plant mosquito, Wyeomyia smithii (Bradshaw and Holzapfel 2001), and speckled wood butterfly, Pararge aegeria (Nielsen et al. 2023), where the photoperiodic response genetically changed to take advantage of the longer growing season. In contrast, genetic adaptation in the winter moth changed post-dormancy seasonal timing, involving changes to the temperature response (van Asch et al. 2007, 2013).

Conclusion

Seasonal timing shifts are one of the most ubiquitous responses to climate change across taxa (Parmesan and Yohe 2003; Root et al. 2003; Thackeray et al. 2010). Elucidating which environmental cues regulate these timing responses is a crucial step in determining how populations can adapt to climate change. We conclude that temperature has an overriding role compared to photoperiod in regulating the seasonal timing of winter moth egg hatching. These relative contributions of temperature and photoperiod could have important implications for climate change adaptation. So far, we know of only few species that have evolved under climate change (Catullo et al. 2019; Merilä and Hendry 2014), but selection often seems to target seasonal timing responses (Bradshaw and Holzapfel 2008; Visser and Gienapp 2019). Future work should take care in determining both the specific selection pressure that climate change exerts on the seasonal timing trait as well as its underlying mechanism, as we might expect different targets of selection depending on the relative contribution of different environmental cues.