Discussion
Oviparous and viviparous squamates are sympatric mainly in tropical and subtropical regions. In these regions, climate warming should cause maternal energy budgets of viviparous species to increase, but those of oviparous species to decrease because of the increment of oviparous reproductive frequency (Fig. 5). Although the offspring survival is predicted to increase in both oviparous and viviparous species, the anticipated increment in oviparous species is small in subtropical regions (Fig. 6). Accordingly, climate warming will benefit viviparous species, but likely lead to negative balance of energy in oviparous species, suggesting oviparous squamates may be more vulnerable to climate warming than viviparous species in the broad sympatric zone of these species.
For oviparous species in most subtropical regions (except for east coast), climate warming enhances reproductive output by increasing clutch numbers (Fig. 3b), but decreases embryonic developmental success (Fig. 4a, b), and leave little time for late-hatching offspring to grow before hibernation (Fig. S1). More specifically, under climate warming, oviparous females in most subtropical regions produce additional clutches because they keep producing new clutches late in the year (around July) (Fig. 6e, f). However, these females fail to provide developmental temperatures that are high enough to ensure developmental success in late reproductive season (Ma et al. 2018). In addition, those late hatchlings may not have enough time before hibernation for energy accumulation to survive the severe winter, and therefore have low survivorship (Sun et al. 2018). These negative impacts on embryos and hatchings will undoubtedly decrease offspring survival, which reduce the reproductive efficiency of oviparous squamates in the subtropical region (i.e. they reproduce more but have fewer surviving offspring).
The reduction of maternal energy budgets for oviparous species under the scenario of climate warming largely results from the increased number of clutches per year (Fig. 5e). Longer activity seasons under warming enable oviparous mothers to produce more clutches, which is in accordance with the facts that nearly half of oviparous species produce multiple clutches per year (see our analysis in the methods) (Mesquitaet al. 2016), and experimental warming can increase clutch numbers in oviparous squamates (Du et al. 2005; Lu et al.2018). It is noteworthy that our analysis assumes that food availability is unlimited for females. If females face food constraints under warming, the negative balance of energy would be even more severe (Raymond B. Huey & Joel G. Kingsolver 2019). Of course, this negative balance can be partly offset by maternal decision on energy allocation, if females select to allocate a constant fraction of assimilated energy to reproduction (Meiri et al. 2012), or dynamically allocate resources to reproduction according to environmental conditions and individual state (Senner et al. 2015). Accordingly, the oviparous species may be able to enhance their own fitness by reducing reproductive frequency and therefore reproductive investment.
Previous studies predict that species in cold regions are more vulnerable to climate warming than oviparous species in warm regions. This “cul-de-sac” hypothesis states progressively warming climates are driving viviparous reptile extinction (Sinervo et al. 2010; Jaraet al. 2019). This prediction is based on the cold climate hypothesis of viviparity evolution, which suggests that the evolution of viviparity has facilitated the radiation of reptiles into cold climates (Shine & Bull 1979). Thus warming climates will favor oviparous species. Extending this previous understanding, our study suggests that oviparous species are more vulnerable than viviparous species in the tropics and subtropics, where both modes are sympatric. The vulnerability of oviparous species is not only due to negative balance of maternal energy budget, but also due to their increment in maternal reproductive investment that does not yield a corresponding increment in offspring survival. Even at high elevations in tropical and subtropical zones (e.g. Mexican Plateau and Colorado Plateau), maternal energy budgets increase for oviparous and viviparous species (Fig. 5b, d), but offspring survival increases only for viviparous species (not for oviparous species) (Fig. 6b, d). These results are contrary to the “cul-de-sac” hypothesis (Pincheira-Donoso et al. 2013). Instead, our results are consistent with the maternal manipulation hypothesis of viviparity evolution (Shine 1995), which suggests that the evolution of viviparity has improved the fitness of reptiles not only in cold regions but also in other conditions where gravid females are able to maintain body temperatures different from those available in external nests (e.g. hot regions and fluctuating environments). Accordingly, we may expect that viviparity enables species to tolerate climate warming (with higher and more variable temperatures) and outcompete oviparous reproduction. This explanation fits well with the observation that viviparity has been a successful strategy among reptiles in warm climates (Packard et al. 1977).
Our analysis focused on the energy budget of mothers and offspring, but did not consider the role of the phenotypic plasticity and evolutionary adaptation of physiological and life-history traits that could occur among species and populations across latitudes and elevations (Sundayet al. 2019). In addition, our study did not take species interactions into account, which may also affect the vulnerability of species to climate warming. For example, the competition between sympatric oviparous and viviparous squamates and/or the food web of squamate-insect-plant may be affected by climate warming, and in turn determine species vulnerability. Unfortunately, those above-mentioned physiological, ecological and evolutionary parameters are currently unavailable for the majority of species, which hampers their inclusion in mechanistic models. Despite these caveats, both our models – biophysical and life-history models – are the first to reveal that oviparous squamates are more vulnerable to climate warming than are oviparous species where both modes are sympatric. This finding suggests that enhanced conservation effort should be directed to oviparous squamates in tropical and subtropical zones. Our results may be relevant to other taxa (e.g., insects, amphibians) that have both viviparous and oviparous modes.